From 0e5dda4cdca527008a86d4522b89dd8c449a88ac Mon Sep 17 00:00:00 2001 From: Zichao <40557101+hxzd5568@users.noreply.github.com> Date: Thu, 17 Jul 2025 14:21:41 +0800 Subject: [PATCH 01/10] Initial commit --- README.md | 1 + 1 file changed, 1 insertion(+) create mode 100644 README.md diff --git a/README.md b/README.md new file mode 100644 index 000000000..95b79d737 --- /dev/null +++ b/README.md @@ -0,0 +1 @@ +# GraphNet \ No newline at end of file From 72ee48c3d9c7b04d38cfb1b7f1265b99341251fb Mon Sep 17 00:00:00 2001 From: hxzd5568 <3257591325@qq.com> Date: Thu, 17 Jul 2025 14:22:56 +0800 Subject: [PATCH 02/10] demo --- .../input_tensor_meta.py | 547 ++++++++ .../model.py | 826 ++++++++++++ samples/torch/resnet18/attribute.json | 3 + .../resnet18/input_tensor_constraints.py | 0 samples/torch/resnet18/model.py | 96 ++ samples/torch/resnet18/source_tensor_meta.py | 1106 +++++++++++++++++ samples/torch/utils.py | 302 +++++ 7 files changed, 2880 insertions(+) create mode 100644 samples/paddle/PaddleX_AutoEncoder_ad_dy2st_train/input_tensor_meta.py create mode 100644 samples/paddle/PaddleX_AutoEncoder_ad_dy2st_train/model.py create mode 100644 samples/torch/resnet18/attribute.json create mode 100644 samples/torch/resnet18/input_tensor_constraints.py create mode 100644 samples/torch/resnet18/model.py create mode 100644 samples/torch/resnet18/source_tensor_meta.py create mode 100644 samples/torch/utils.py diff --git a/samples/paddle/PaddleX_AutoEncoder_ad_dy2st_train/input_tensor_meta.py b/samples/paddle/PaddleX_AutoEncoder_ad_dy2st_train/input_tensor_meta.py new file mode 100644 index 000000000..f19a344a5 --- /dev/null +++ b/samples/paddle/PaddleX_AutoEncoder_ad_dy2st_train/input_tensor_meta.py @@ -0,0 +1,547 @@ +class PirProgram_example_input_tensor_meta_5389940036631769731: + program_id = 880737988565058868 + input_name = "linear_1.b_0" + shape = [16] + data = None + + +class PirProgram_example_input_tensor_meta_43418529140942226: + program_id = 880737988565058868 + input_name = "linear_1.w_0" + shape = [32, 16] + data = None + + +class PirProgram_example_input_tensor_meta_3606508939262698655: + program_id = 880737988565058868 + input_name = "linear_0.b_0" + shape = [32] + data = None + + +class PirProgram_example_input_tensor_meta_3554240975606516965: + program_id = 880737988565058868 + input_name = "linear_0.w_0" + shape = [96, 32] + data = None + + +class PirProgram_example_input_tensor_meta_5179986117148289115: + program_id = 880737988565058868 + input_name = "args_0" + shape = [16, 2, 96] + data = None + + +class PirProgram_example_input_tensor_meta_659966657961177331: + program_id = 5237713637574795923 + input_name = "linear_3.b_0" + shape = [96] + data = None + + +class PirProgram_example_input_tensor_meta_3195965365640632706: + program_id = 5237713637574795923 + input_name = "linear_3.w_0" + shape = [32, 96] + data = None + + +class PirProgram_example_input_tensor_meta_3617487774345626831: + program_id = 5237713637574795923 + input_name = "linear_2.b_0" + shape = [32] + data = None + + +class PirProgram_example_input_tensor_meta_5636200344230617383: + program_id = 5237713637574795923 + input_name = "linear_2.w_0" + shape = [16, 32] + data = None + + +class PirProgram_example_input_tensor_meta_5228671372811070830: + program_id = 5237713637574795923 + input_name = "args_0" + shape = [16, 2, 16] + data = [float("-1.24783"), float("1.35968"), float("-0.0643954"), float("-0.627694"), float("3.57834"), float("-1.40647"), float("-3.06211"), float("-0.852737"), float("1.89044"), float("1.71429"), float("0.873626"), float("2.45266"), float("2.46995"), float("-4.18985"), float("5.24214"), float("-0.769194"), float("-0.432318"), float("1.44443"), float("-0.398696"), float("-1.49099"), float("-0.245603"), float("-1.75011"), float("-1.19734"), float("0.138174"), float("1.28394"), float("-0.147029"), float("1.46563"), float("0.138966"), float("1.49462"), float("-1.24567"), float("1.74074"), float("-0.541634"), float("-0.137766"), float("0.149732"), float("0.813494"), float("0.303299"), float("-0.238571"), float("1.88537"), float("-0.0370909"), float("0.573859"), float("0.349418"), float("0.487316"), float("-1.32582"), float("-0.843882"), float("-0.676632"), float("-0.909084"), float("0.270721"), float("-1.23429"), float("-0.0424248"), float("0.429659"), float("-0.14267"), float("-0.428278"), float("-0.0162147"), float("-0.494324"), float("-0.285505"), float("0.0283108"), float("0.343003"), float("0.0251735"), float("0.319188"), float("0.0917862"), float("0.402113"), float("-0.271283"), float("0.494286"), float("-0.191708"), float("-0.284004"), float("0.298903"), float("1.58414"), float("0.506222"), float("-0.530693"), float("3.27026"), float("-0.117699"), float("1.01495"), float("0.711183"), float("0.811757"), float("-2.06371"), float("-1.2438"), float("-1.13482"), float("-1.38332"), float("0.486574"), float("-2.27285"), float("-0.358158"), float("0.179657"), float("0.702598"), float("0.133837"), float("-0.116537"), float("1.53755"), float("0.00049031"), float("0.551636"), float("0.176739"), float("0.308397"), float("-0.844465"), float("-0.537499"), float("-0.755579"), float("-0.623192"), float("0.182918"), float("-1.02258"), float("-1.34653"), float("0.913991"), float("1.1851"), float("0.400625"), float("0.121387"), float("3.74875"), float("0.507555"), float("1.24397"), float("-0.00871566"), float("0.270786"), float("-2.39195"), float("-1.38387"), float("-2.33894"), float("-0.779432"), float("0.529702"), float("-2.01132"), float("-0.0355457"), float("0.156263"), float("-0.0398702"), float("-0.0941751"), float("0.070695"), float("-0.125185"), float("-0.123143"), float("-0.0328905"), float("0.0242618"), float("0.050551"), float("0.0789997"), float("0.0688375"), float("0.0704337"), float("-0.152359"), float("0.139597"), float("-0.0303556"), float("-0.422165"), float("0.755079"), float("-0.125477"), float("-0.390056"), float("0.129337"), float("-0.44178"), float("-0.83648"), float("-0.363755"), float("-0.396812"), float("-0.217299"), float("1.19051"), float("0.353966"), float("0.100611"), float("-0.841632"), float("1.13851"), float("-0.0570613"), float("-0.321325"), float("1.00824"), float("-0.478592"), float("-1.40818"), float("-0.435232"), float("-1.11203"), float("-0.501488"), float("-0.127583"), float("0.967821"), float("0.100076"), float("0.722625"), float("0.514725"), float("0.67351"), float("-0.719309"), float("0.999483"), float("-0.640918"), float("-1.46169"), float("1.59347"), float("3.25034"), float("0.413147"), float("0.432863"), float("5.4329"), float("0.0917308"), float("1.98376"), float("0.598049"), float("1.64893"), float("-1.92021"), float("-1.70858"), float("-2.384"), float("-1.1942"), float("0.144455"), float("-3.21842"), float("-1.22085"), float("-0.843408"), float("3.91006"), float("0.0781953"), float("-1.01348"), float("6.29112"), float("1.98013"), float("2.26446"), float("0.98046"), float("0.878038"), float("-4.09496"), float("-3.09536"), float("-2.928"), float("-3.87442"), float("-0.000862794"), float("-3.55961"), float("-0.1064"), float("-0.161292"), float("0.187614"), float("0.0255209"), float("-0.162917"), float("0.42479"), float("-0.0453343"), float("0.436514"), float("-0.191061"), float("0.137538"), float("-0.410105"), float("-0.431748"), float("-0.438004"), float("-0.404633"), float("0.218743"), float("-0.427686"), float("-0.424405"), float("0.837761"), float("0.332992"), float("-0.0513256"), float("0.047148"), float("1.35676"), float("-0.0688395"), float("0.506198"), float("0.163514"), float("-0.337938"), float("-1.2557"), float("-0.616636"), float("-1.1297"), float("-0.193176"), float("0.337317"), float("-0.561698"), float("-0.135902"), float("0.393851"), float("0.00424394"), float("-0.196324"), float("0.159519"), float("-0.308631"), float("-0.376123"), float("-0.071268"), float("0.0883756"), float("0.0545766"), float("0.303032"), float("0.111892"), float("0.218541"), float("-0.537371"), float("0.421097"), float("-0.129858"), float("-1.03447"), float("0.513909"), float("2.0756"), float("0.441222"), float("-0.472549"), float("4.29349"), float("0.13518"), float("1.6049"), float("0.482167"), float("0.853731"), float("-2.33474"), float("-1.43174"), float("-2.12045"), float("-1.69272"), float("0.520399"), float("-2.92517"), float("-0.154838"), float("1.12713"), float("-0.729672"), float("-1.66338"), float("-0.657706"), float("-1.18132"), float("-0.874759"), float("0.06689"), float("1.51983"), float("0.351833"), float("0.986195"), float("0.245073"), float("0.796269"), float("-0.622144"), float("0.948545"), float("-1.01152"), float("-1.22342"), float("2.62304"), float("-0.221326"), float("-2.36949"), float("0.0231131"), float("-3.1239"), float("-1.56313"), float("-0.0846508"), float("1.4516"), float("-0.843756"), float("2.11236"), float("0.770368"), float("2.54804"), float("-2.68047"), float("3.21551"), float("-0.423819"), float("-0.392694"), float("0.187697"), float("0.413998"), float("0.057787"), float("-0.13115"), float("1.17413"), float("0.0864602"), float("0.307451"), float("0.0920442"), float("0.143506"), float("-0.635472"), float("-0.255527"), float("-0.542642"), float("-0.322647"), float("0.183446"), float("-0.825342"), float("-0.0203514"), float("0.0198951"), float("0.0865615"), float("0.0272587"), float("-0.00242023"), float("0.195467"), float("0.07148"), float("0.0644732"), float("0.0219086"), float("0.0301208"), float("-0.185134"), float("-0.132513"), float("-0.0794464"), float("-0.100434"), float("0.0120146"), float("-0.08797"), float("-1.29686"), float("1.00538"), float("-0.771612"), float("-1.60371"), float("1.04745"), float("-1.71847"), float("-0.364911"), float("-0.669676"), float("1.52801"), float("-0.0574329"), float("0.59866"), float("1.73576"), float("2.91115"), float("-1.2902"), float("3.56666"), float("0.497791"), float("-0.629735"), float("1.17433"), float("-0.648112"), float("-1.04914"), float("-0.195696"), float("-1.03082"), float("-0.321714"), float("-0.668095"), float("-0.117482"), float("-0.81666"), float("1.54317"), float("0.372422"), float("0.500825"), float("-0.542041"), float("1.91077"), float("0.324363"), float("-0.746501"), float("1.56012"), float("-0.693475"), float("-1.47149"), float("0.417272"), float("-1.5822"), float("-0.614944"), float("-0.337996"), float("0.782261"), float("-0.244622"), float("1.03057"), float("0.317955"), float("1.1072"), float("-1.29199"), float("1.91734"), float("0.0562871"), float("-0.344145"), float("0.582381"), float("0.0177077"), float("-0.394364"), float("0.278082"), float("-0.549242"), float("-0.608972"), float("-0.148099"), float("0.0354738"), float("0.105372"), float("0.54669"), float("0.293604"), float("0.384397"), float("-0.770095"), float("0.530887"), float("-0.0841932"), float("-1.5664"), float("1.47664"), float("-0.0485786"), float("-1.2655"), float("1.12964"), float("-1.11663"), float("-0.127906"), float("-0.158063"), float("0.266339"), float("-0.689339"), float("0.552855"), float("-0.29022"), float("1.41004"), float("-1.61181"), float("1.21825"), float("0.138486"), float("-1.40203"), float("1.25058"), float("0.718333"), float("0.433586"), float("0.213605"), float("3.4107"), float("0.702294"), float("0.844367"), float("-0.269252"), float("0.469749"), float("-2.07015"), float("-0.969774"), float("-1.84929"), float("-0.536334"), float("0.553032"), float("-2.07369"), float("-0.875411"), float("1.6934"), float("-0.485379"), float("-1.53678"), float("0.237551"), float("-1.80035"), float("-0.801288"), float("-0.196758"), float("1.13256"), float("-0.170934"), float("1.21607"), float("0.530888"), float("1.79473"), float("-1.66515"), float("2.21643"), float("-0.170763"), float("-0.242361"), float("0.0761521"), float("0.432139"), float("0.0631636"), float("-0.093372"), float("0.938469"), float("0.0142968"), float("0.313853"), float("0.120945"), float("0.144088"), float("-0.496069"), float("-0.284314"), float("-0.471705"), float("-0.335182"), float("0.114975"), float("-0.613756"), float("-0.133989"), float("0.332343"), float("-0.0773375"), float("-0.386288"), float("-0.0153539"), float("-0.397145"), float("-0.344031"), float("0.0805068"), float("0.348394"), float("0.00255038"), float("0.33544"), float("0.0063985"), float("0.34345"), float("-0.323657"), float("0.404106"), float("-0.197486"), float("-0.0204629"), float("0.165297"), float("0.344726"), float("0.0223876"), float("-0.00523791"), float("0.623413"), float("-0.16317"), float("0.276788"), float("0.216639"), float("0.113306"), float("-0.499786"), float("-0.285747"), float("-0.187579"), float("-0.235846"), float("0.0960729"), float("-0.395305"), float("-0.0911418"), float("0.288873"), float("-0.0312524"), float("-0.338057"), float("-0.0387871"), float("-0.421691"), float("-0.315626"), float("0.125869"), float("0.273582"), float("-0.058278"), float("0.3202"), float("0.015641"), float("0.297726"), float("-0.288548"), float("0.329069"), float("-0.179419"), float("-0.0326695"), float("0.117692"), float("-0.060717"), float("-0.112393"), float("0.0348706"), float("-0.105831"), float("-0.0257303"), float("-0.0271061"), float("0.0484697"), float("-0.00249167"), float("0.0425337"), float("0.0319538"), float("0.0739469"), float("-0.0675583"), float("0.136938"), float("-0.0154109")] + + +class PirProgram_example_input_tensor_meta_6104757736120590527: + program_id = 1242071021843173094 + input_name = "middle_5" + shape = [16, 2, 96] + data = None + + +class PirProgram_example_input_tensor_meta_42754444789929111: + program_id = 1242071021843173094 + input_name = "middle_4" + shape = [16, 2, 32] + data = None + + +class PirProgram_example_input_tensor_meta_3577094903844169474: + program_id = 1242071021843173094 + input_name = "output_grad_0" + shape = [16, 2, 96] + data = None + + +class PirProgram_example_input_tensor_meta_1198995949523510110: + program_id = 1242071021843173094 + input_name = "linear_3.w_0" + shape = [32, 96] + data = None + + +class PirProgram_example_input_tensor_meta_8553907544768773751: + program_id = 1242071021843173094 + input_name = "middle_2" + shape = [1] + data = [float("0.2")] + + +class PirProgram_example_input_tensor_meta_4986101899439820377: + program_id = 1242071021843173094 + input_name = "linear_3.b_0" + shape = [96] + data = [float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0")] + + +class PirProgram_example_input_tensor_meta_8512202976348838334: + program_id = 1242071021843173094 + input_name = "middle_3" + shape = [16, 2, 32] + data = [float("0.0347782"), float("1.16674"), float("0"), float("0.33759"), float("0"), float("0"), float("2.70181"), float("0"), float("0"), float("0"), float("1.46725"), float("1.50891"), float("0"), float("0"), float("0.539994"), float("0"), float("1.30105"), float("0"), float("0"), float("2.44845"), float("0"), float("0"), float("0"), float("0.298854"), float("0"), float("0"), float("0"), float("0.978712"), float("0"), float("0"), float("2.00676"), float("0"), float("0"), float("0.0118907"), float("0"), float("1.28928"), float("0"), float("0"), float("0.902585"), float("0.740692"), float("0"), float("0.553495"), float("0"), float("0"), float("0.295481"), float("0"), float("0"), float("0"), float("0"), float("0.87983"), float("0"), float("1.30724"), float("0"), float("0"), float("0"), float("1.35767"), float("0.412437"), float("0"), float("1.42755"), float("0"), float("0"), float("0"), float("1.31564"), float("0"), float("0.0323608"), float("0.438107"), float("0"), float("0"), float("0"), float("0.759341"), float("0"), float("0"), float("0"), float("0.935668"), float("0"), float("1.4212"), float("0.437199"), float("0"), float("0.36504"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.397975"), float("0"), float("0"), float("0.335801"), float("0.170311"), float("0"), float("0.170158"), float("0"), float("0"), float("0.0497961"), float("0"), float("0"), float("0"), float("0"), float("0.21666"), float("0"), float("0.381272"), float("0"), float("0"), float("0"), float("0.35843"), float("0.109284"), float("0"), float("0.436922"), float("0"), float("0.0547679"), float("0"), float("0.377323"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("1.24158"), float("0"), float("0"), float("2.46088"), float("1.5229"), float("0"), float("2.31923"), float("0.538648"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.258044"), float("0"), float("0"), float("0"), float("0"), float("0"), float("1.01807"), float("0"), float("1.60786"), float("0"), float("0.818661"), float("0"), float("0"), float("0"), float("0.319663"), float("0"), float("0"), float("0"), float("0.711094"), float("0"), float("0"), float("1.17478"), float("0"), float("1.00297"), float("0.975026"), float("0.0937734"), float("0.0181032"), float("0.17601"), float("0"), float("0"), float("0"), float("0"), float("0.160587"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.650962"), float("0"), float("0.506464"), float("0"), float("0.501895"), float("0"), float("0.564878"), float("0"), float("0.826023"), float("0"), float("0"), float("0"), float("2.16452"), float("0"), float("0"), float("3.52679"), float("1.14489"), float("2.1719"), float("0"), float("0.178483"), float("0"), float("0"), float("0.522222"), float("0"), float("0"), float("0.336636"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("1.47996"), float("0"), float("1.47095"), float("0"), float("1.57665"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.0702505"), float("0"), float("0.0277561"), float("0.115806"), float("0.0211074"), float("0"), float("0"), float("0.0265309"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.0315472"), float("0.028493"), float("0.00156406"), float("0.167891"), float("0"), float("0"), float("0"), float("0.0448072"), float("0"), float("0"), float("0.149831"), float("0"), float("0.0563955"), float("0"), float("0"), float("0.010389"), float("0"), float("0.0195092"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.677082"), float("0.256379"), float("0.328802"), float("0.26332"), float("0"), float("0.135194"), float("0"), float("0"), float("0.15623"), float("0"), float("0"), float("0.823846"), float("0"), float("0.437986"), float("0"), float("0.122316"), float("0"), float("0"), float("0"), float("0"), float("0.952759"), float("0"), float("0"), float("0"), float("0.0882541"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.198234"), float("0"), float("0"), float("0"), float("0"), float("0"), float("1.00134"), float("0"), float("0"), float("1.40157"), float("0"), float("0"), float("0"), float("0.800234"), float("0"), float("0"), float("0.18145"), float("0"), float("0"), float("0"), float("2.90421"), float("0"), float("0"), float("5.13843"), float("1.78794"), float("2.42338"), float("3.48572"), float("0"), float("0.827473"), float("0.628247"), float("1.24771"), float("0"), float("0"), float("0"), float("1.20432"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("1.76207"), float("0"), float("2.24063"), float("0"), float("0.8273"), float("0.0147218"), float("0.535274"), float("0"), float("0"), float("0"), float("3.12057"), float("0"), float("0"), float("4.5842"), float("3.66083"), float("3.91709"), float("5.36272"), float("1.2578"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("1.98505"), float("0"), float("0.42635"), float("0"), float("0"), float("0"), float("2.05989"), float("0.0728085"), float("0.442795"), float("0.275162"), float("0.000693649"), float("0"), float("0.136329"), float("0"), float("0"), float("0.208206"), float("0.211367"), float("0"), float("0.226844"), float("0.222011"), float("0"), float("0.138711"), float("0"), float("0"), float("0"), float("0"), float("0.162163"), float("0"), float("0.386705"), float("0"), float("0"), float("0.0113519"), float("0.248948"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.337618"), float("0"), float("0"), float("0.143206"), float("0.973093"), float("0"), float("0"), float("1.5557"), float("0.547614"), float("0.808207"), float("0.809795"), float("0.207993"), float("0.226023"), float("0"), float("0.446442"), float("0"), float("0"), float("0.298011"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.550403"), float("0"), float("0.662424"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.0034696"), float("0.222743"), float("0.020042"), float("0"), float("0"), float("0.0433677"), float("0"), float("0"), float("0"), float("0"), float("0.0376502"), float("0.0464511"), float("0.101323"), float("0.00628365"), float("0.449533"), float("0"), float("0"), float("0"), float("0.0415141"), float("0"), float("0"), float("0.329318"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.871903"), float("0"), float("0"), float("0"), float("2.00545"), float("0"), float("0"), float("0"), float("1.72249"), float("2.65814"), float("2.60701"), float("0.273242"), float("0.0238184"), float("0.55817"), float("0"), float("0"), float("0"), float("0"), float("0.387383"), float("0"), float("0"), float("0"), float("0"), float("0"), float("1.78813"), float("0"), float("1.34975"), float("0"), float("1.48764"), float("0"), float("1.50652"), float("0"), float("0"), float("0"), float("1.50629"), float("0"), float("0"), float("0.539916"), float("0.389776"), float("0"), float("1.01698"), float("0.107516"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.578256"), float("0"), float("1.2537"), float("0"), float("0"), float("0"), float("1.35418"), float("0.705801"), float("0"), float("1.52801"), float("0"), float("0"), float("0"), float("1.11237"), float("0"), float("0"), float("0.0364399"), float("0"), float("1.3821"), float("0"), float("0"), float("1.9378"), float("0.935661"), float("0"), float("0.557419"), float("0"), float("0"), float("0.87944"), float("0"), float("0"), float("0"), float("0"), float("0.81148"), float("0"), float("1.97005"), float("0"), float("0"), float("0"), float("0"), float("0.119622"), float("0"), float("2.45299"), float("0"), float("0.501118"), float("0"), float("1.94722"), float("0"), float("0"), float("0.272278"), float("0"), float("0"), float("0"), float("0.572684"), float("0"), float("0"), float("0.837621"), float("0.405361"), float("0.676569"), float("0.585324"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.456107"), float("0"), float("0.319202"), float("0"), float("0.391309"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.116754"), float("0"), float("0"), float("0.203108"), float("0"), float("0.132567"), float("0.181899"), float("0.087103"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.0459756"), float("0"), float("0"), float("0"), float("0.0294301"), float("0"), float("0.0783622"), float("0"), float("0.0726098"), float("0"), float("0.0563935"), float("0"), float("0.73509"), float("0"), float("0"), float("0"), float("0"), float("2.19167"), float("0.614574"), float("0"), float("0"), float("0"), float("0.149091"), float("0.140033"), float("0"), float("0.175131"), float("0"), float("1.20836"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.0784062"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.0162837"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.737377"), float("0.196769"), float("0"), float("0.265471"), float("0"), float("0.740579"), float("0"), float("0"), float("0.955313"), float("0.0152713"), float("0"), float("0"), float("0"), float("0.576584"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.663437"), float("0"), float("0"), float("1.01948"), float("0.675586"), float("0"), float("0.312976"), float("0"), float("0"), float("0.292665"), float("0"), float("0"), float("0"), float("0"), float("0.0120792"), float("0.0180582"), float("0.879502"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("1.43309"), float("0"), float("0.42474"), float("0"), float("1.11088"), float("0"), float("0"), float("0"), float("0"), float("0.178277"), float("0"), float("0.0671259"), float("0.422192"), float("0.127027"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.294385"), float("0.217661"), float("0.0798928"), float("0.724993"), float("0"), float("0"), float("0"), float("0.182339"), float("0"), float("0"), float("0.597919"), float("0"), float("0.238543"), float("0"), float("0.412386"), float("0"), float("0"), float("0.0349023"), float("0"), float("0.222169"), float("0"), float("0"), float("0.909737"), float("1.39266"), float("0"), float("0.19613"), float("0"), float("0.478526"), float("0"), float("0"), float("0"), float("0.593163"), float("0.518306"), float("0"), float("0.881752"), float("0.491215"), float("0"), float("0"), float("0"), float("0.242513"), float("0"), float("0"), float("0.900461"), float("0"), float("0.303094"), float("0"), float("0.754786"), float("0"), float("0"), float("0.644119"), float("0"), float("0"), float("0"), float("2.18745"), float("0"), float("0"), float("3.23613"), float("1.11745"), float("1.86997"), float("0"), float("0.0982345"), float("0.664672"), float("0"), float("0"), float("0"), float("0"), float("0.438279"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("1.45317"), float("0"), float("1.58249"), float("0"), float("0"), float("0"), float("0.975825"), float("0"), float("0"), float("0"), float("1.01558"), float("0"), float("0"), float("1.36234"), float("0.717539"), float("0"), float("0.324656"), float("0"), float("0"), float("0.4705"), float("0"), float("0"), float("0"), float("0"), float("0.304301"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.198254"), float("0"), float("0"), float("0"), float("0.416886"), float("0"), float("0"), float("0.686297"), float("0.334294"), float("0.576713"), float("0.580936"), float("0"), float("0.00356044"), float("0.0984948"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.387813"), float("0"), float("0.303094"), float("0"), float("0"), float("0"), float("0.336003"), float("0"), float("0.0410064"), float("0"), float("0.342481"), float("0"), float("0"), float("0.198419"), float("0.214347"), float("0"), float("0"), float("0"), float("0"), float("0.0215962"), float("0"), float("0"), float("0"), float("0"), float("0.203783"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.293901"), float("0.111309"), float("0"), float("0.333862"), float("0"), float("0"), float("0"), float("0.320106"), float("0"), float("0"), float("0.193667"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.270118"), float("0"), float("0.590756"), float("0.117252"), float("0"), float("0.000430197"), float("0"), float("0"), float("0"), float("0"), float("0.181376"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.0962679"), float("0"), float("0"), float("0"), float("0.106943"), float("0"), float("0.199012"), float("0"), float("0.0540602"), float("0.0616581"), float("0.305817"), float("0"), float("0"), float("0.204442"), float("0.191712"), float("0"), float("0.109193"), float("0"), float("0"), float("0.0509193"), float("0"), float("0"), float("0"), float("0"), float("0.218017"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.253662"), float("0.134936"), float("0"), float("0.269651"), float("0"), float("0"), float("0"), float("0.295306"), float("0"), float("0"), float("0"), float("0"), float("0.0601464"), float("0"), float("0.00200357"), float("0.0851614"), float("0.0318568"), float("0"), float("0.0348719"), float("0"), float("0"), float("0.0127985"), float("0"), float("0"), float("0"), float("0"), float("0"), float("3.93368e-05"), float("0.0573087"), float("0"), float("0"), float("0"), float("0.0663058"), float("0"), float("0"), float("0.118441"), float("0"), float("0"), float("0"), float("0.072888"), float("0")] + + +class PirProgram_example_input_tensor_meta_7134464969354624939: + program_id = 1242071021843173094 + input_name = "middle_1" + shape = [16, 2, 32] + data = [float("0.0278226"), float("0.933391"), float("0"), float("0.270072"), float("0"), float("0"), float("2.16145"), float("0"), float("0"), float("0"), float("1.1738"), float("1.20713"), float("0"), float("0"), float("0.431996"), float("0"), float("1.04084"), float("0"), float("0"), float("1.95876"), float("0"), float("0"), float("0"), float("0.239083"), float("0"), float("0"), float("1.80388"), float("0.782969"), float("0"), float("0"), float("1.60541"), float("0"), float("0"), float("0.00951254"), float("0"), float("1.03142"), float("0"), float("0"), float("0.722068"), float("0.592554"), float("0"), float("0.442796"), float("0"), float("0"), float("0.236385"), float("0"), float("0"), float("0"), float("0"), float("0.703864"), float("0"), float("1.04579"), float("0"), float("0"), float("0"), float("1.08614"), float("0.32995"), float("0"), float("1.14204"), float("0"), float("0.172047"), float("0"), float("1.05251"), float("0"), float("0.0258886"), float("0.350485"), float("0"), float("0"), float("0"), float("0.607473"), float("0"), float("0"), float("1.18592"), float("0.748534"), float("1.12389"), float("1.13696"), float("0.349759"), float("0"), float("0.292032"), float("0"), float("0"), float("0"), float("0"), float("0.118198"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.452116"), float("0"), float("0.73808"), float("0"), float("0.388192"), float("0"), float("0.583962"), float("0"), float("0"), float("0"), float("0.31838"), float("0"), float("0"), float("0.268641"), float("0.136249"), float("0"), float("0.136126"), float("0"), float("0"), float("0.0398369"), float("0"), float("0"), float("0"), float("0"), float("0.173328"), float("0"), float("0.305018"), float("0"), float("0"), float("0"), float("0.286744"), float("0.0874273"), float("0"), float("0.349538"), float("0"), float("0.0438143"), float("0"), float("0.301858"), float("0"), float("0"), float("0.574676"), float("0"), float("0"), float("0"), float("0.993267"), float("0"), float("0"), float("1.96871"), float("1.21832"), float("1.763"), float("1.85539"), float("0.430918"), float("0"), float("0.511989"), float("0"), float("0"), float("0"), float("0"), float("0.206435"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.814453"), float("0"), float("1.28629"), float("0"), float("0.654929"), float("0"), float("0.880677"), float("0"), float("0.255731"), float("0"), float("0"), float("0"), float("0.568876"), float("0"), float("0"), float("0.939828"), float("0.488117"), float("0.802376"), float("0.78002"), float("0.0750187"), float("0.0144826"), float("0.140808"), float("0"), float("0"), float("0"), float("0"), float("0.12847"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.52077"), float("0"), float("0.405171"), float("0"), float("0.401516"), float("0"), float("0.451902"), float("0"), float("0.660819"), float("0"), float("0"), float("0"), float("1.73161"), float("0"), float("0"), float("2.82143"), float("0.915915"), float("1.73752"), float("1.66533"), float("0.142786"), float("0.414861"), float("0"), float("0.417778"), float("0"), float("0"), float("0.269308"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("1.18397"), float("0"), float("1.17676"), float("0"), float("1.26132"), float("0"), float("1.01996"), float("0"), float("0"), float("0"), float("0.0562004"), float("0"), float("0.0222049"), float("0.0926452"), float("0.0168859"), float("0"), float("0.0163579"), float("0.0212247"), float("0"), float("0.00476881"), float("0"), float("0"), float("0"), float("0.0252377"), float("0.0227944"), float("0.00125125"), float("0.134313"), float("0"), float("0"), float("0"), float("0.0358457"), float("0"), float("0"), float("0.119865"), float("0"), float("0.0451164"), float("0"), float("0.0851875"), float("0.0083112"), float("0"), float("0.0156073"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.0826194"), float("0"), float("0.0455271"), float("0.541666"), float("0.205103"), float("0.263042"), float("0.210656"), float("0.379648"), float("0.108155"), float("0"), float("0"), float("0.124984"), float("0"), float("0"), float("0.659077"), float("0"), float("0.350388"), float("0"), float("0.097853"), float("0.221756"), float("0"), float("0"), float("0"), float("0.762207"), float("0"), float("0"), float("0.52652"), float("0.0706032"), float("0"), float("0.551366"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.158587"), float("0"), float("0.748428"), float("0"), float("0"), float("0"), float("0.801074"), float("0.351115"), float("0"), float("1.12126"), float("0"), float("0"), float("0"), float("0.640187"), float("0"), float("0"), float("0.14516"), float("0"), float("0"), float("0"), float("2.32337"), float("0"), float("0"), float("4.11075"), float("1.43035"), float("1.9387"), float("2.78858"), float("0"), float("0.661978"), float("0.502597"), float("0.998164"), float("0"), float("0"), float("0.55665"), float("0.963454"), float("0"), float("0"), float("0"), float("0"), float("0"), float("2.3835"), float("0"), float("1.40965"), float("0"), float("1.7925"), float("0"), float("0.66184"), float("0.0117774"), float("0.428219"), float("0"), float("0"), float("0"), float("2.49646"), float("0"), float("0"), float("3.66736"), float("2.92867"), float("3.13367"), float("4.29018"), float("1.00624"), float("0"), float("2.21902"), float("0"), float("0"), float("0"), float("0"), float("0.516142"), float("0"), float("1.84779"), float("0"), float("0"), float("0"), float("1.58804"), float("0"), float("0.34108"), float("0"), float("1.56324"), float("0"), float("1.64791"), float("0.0582468"), float("0.354236"), float("0.22013"), float("0.000554919"), float("0"), float("0.109064"), float("0"), float("0"), float("0.166565"), float("0.169093"), float("0.495146"), float("0.181475"), float("0.177609"), float("0"), float("0.110969"), float("0"), float("0"), float("0"), float("0"), float("0.129731"), float("0"), float("0.309364"), float("0"), float("0"), float("0.00908154"), float("0.199158"), float("0"), float("0"), float("0"), float("0.215054"), float("0"), float("0.398292"), float("0"), float("0.270094"), float("0"), float("0"), float("0.114564"), float("0.778475"), float("0"), float("0"), float("1.24456"), float("0.438091"), float("0.646566"), float("0.647836"), float("0.166394"), float("0.180818"), float("0"), float("0.357154"), float("0"), float("0"), float("0.238409"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.0367787"), float("0"), float("0.735905"), float("0"), float("0.440322"), float("0"), float("0.529939"), float("0"), float("0"), float("0"), float("0.0900413"), float("0"), float("0.00277568"), float("0.178194"), float("0.0160336"), float("0"), float("0.0589797"), float("0.0346942"), float("0"), float("0.0824681"), float("0"), float("0"), float("0.0301201"), float("0.0371609"), float("0.0810581"), float("0.00502692"), float("0.359627"), float("0"), float("0"), float("0"), float("0.0332112"), float("0"), float("0"), float("0.263454"), float("0"), float("0.101193"), float("0"), float("0.175708"), float("0"), float("0"), float("0.697523"), float("0"), float("0"), float("0"), float("1.60436"), float("0"), float("0"), float("2.59563"), float("1.37799"), float("2.12651"), float("2.08561"), float("0.218594"), float("0.0190547"), float("0.446536"), float("0"), float("0"), float("0"), float("0"), float("0.309906"), float("0"), float("0"), float("0"), float("0"), float("0"), float("1.4305"), float("0"), float("1.0798"), float("0"), float("1.19011"), float("0"), float("1.20522"), float("0"), float("0"), float("0"), float("1.20503"), float("0"), float("0"), float("0.431932"), float("0.311821"), float("0"), float("0.813583"), float("0.0860126"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.462605"), float("0"), float("1.00296"), float("0"), float("0"), float("0"), float("1.08335"), float("0.564641"), float("0"), float("1.22241"), float("0"), float("0"), float("0"), float("0.889895"), float("0"), float("0"), float("0.0291519"), float("0"), float("1.10568"), float("0"), float("0"), float("1.55024"), float("0.748529"), float("0"), float("0.445935"), float("0"), float("0"), float("0.703552"), float("0"), float("0"), float("0"), float("0.0828259"), float("0.649184"), float("0"), float("1.57604"), float("0"), float("0"), float("0"), float("1.44342"), float("0.0956974"), float("0"), float("1.96239"), float("0"), float("0.400894"), float("0"), float("1.55778"), float("0"), float("0"), float("0.217823"), float("0"), float("0"), float("0"), float("0.458147"), float("0"), float("0"), float("0.670096"), float("0.324289"), float("0.541255"), float("0.468259"), float("0"), float("0.0262447"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.362822"), float("0"), float("0.364886"), float("0"), float("0.255361"), float("0"), float("0.313047"), float("0"), float("0.0198007"), float("0"), float("0"), float("0"), float("0.0934036"), float("0"), float("0"), float("0.162486"), float("0.100014"), float("0.106053"), float("0.145519"), float("0.0696824"), float("0"), float("0.0350842"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.0367805"), float("0"), float("0"), float("0"), float("0.0235441"), float("0"), float("0.0626898"), float("0"), float("0.0580879"), float("0"), float("0.0451148"), float("0"), float("0.588072"), float("0"), float("0.632345"), float("0"), float("0"), float("1.75334"), float("0.491659"), float("0"), float("0"), float("0"), float("0.119273"), float("0.112027"), float("0"), float("0.140105"), float("0"), float("0.966691"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("1.66538"), float("0"), float("0"), float("1.73029"), float("0.062725"), float("0"), float("0"), float("1.35752"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.013027"), float("0"), float("0"), float("0.0148772"), float("0"), float("0"), float("0.220215"), float("0"), float("0"), float("0.589901"), float("0.157416"), float("0.163732"), float("0.212377"), float("0"), float("0.592463"), float("0"), float("0"), float("0.76425"), float("0.012217"), float("0"), float("1.03746"), float("0"), float("0.461267"), float("0"), float("0.246892"), float("0"), float("0"), float("0"), float("0"), float("0.530749"), float("0"), float("0"), float("0.815583"), float("0.540469"), float("0"), float("0.250381"), float("0"), float("0"), float("0.234132"), float("0"), float("0"), float("0"), float("0.000377319"), float("0.00966338"), float("0.0144465"), float("0.703602"), float("0"), float("0"), float("0"), float("0.849869"), float("0"), float("0"), float("1.14647"), float("0"), float("0.339792"), float("0"), float("0.888705"), float("0"), float("0"), float("0"), float("0"), float("0.142622"), float("0"), float("0.0537007"), float("0.337754"), float("0.101622"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.11494"), float("0.235508"), float("0.174129"), float("0.0639142"), float("0.579994"), float("0"), float("0"), float("0"), float("0.145871"), float("0"), float("0"), float("0.478335"), float("0"), float("0.190835"), float("0"), float("0.329909"), float("0"), float("0"), float("0.0279218"), float("0"), float("0.177735"), float("0"), float("0.414542"), float("0.727789"), float("1.11413"), float("0"), float("0.156904"), float("0"), float("0.382821"), float("0.5096"), float("0"), float("0"), float("0.47453"), float("0.414645"), float("0"), float("0.705401"), float("0.392972"), float("0"), float("0"), float("0"), float("0.194011"), float("0"), float("0"), float("0.720369"), float("0"), float("0.242475"), float("0"), float("0.603829"), float("0"), float("0"), float("0.515296"), float("0"), float("0"), float("0"), float("1.74996"), float("0"), float("0"), float("2.5889"), float("0.893962"), float("1.49597"), float("1.12741"), float("0.0785876"), float("0.531738"), float("0"), float("0.2548"), float("0"), float("0"), float("0.350623"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("1.16254"), float("0"), float("1.26599"), float("0"), float("1.17455"), float("0"), float("0.78066"), float("0"), float("0"), float("0"), float("0.812465"), float("0"), float("0"), float("1.08987"), float("0.574031"), float("0"), float("0.259725"), float("0"), float("0"), float("0.3764"), float("0"), float("0"), float("0"), float("0.111538"), float("0.243441"), float("0"), float("0.798357"), float("0"), float("0"), float("0"), float("1.06892"), float("0"), float("0"), float("1.35836"), float("0"), float("0.206873"), float("0"), float("1.11643"), float("0"), float("0"), float("0.158604"), float("0"), float("0"), float("0"), float("0.333509"), float("0"), float("0"), float("0.549038"), float("0.267435"), float("0.46137"), float("0.464749"), float("0.00498525"), float("0.00284835"), float("0.0787959"), float("0.00215647"), float("0"), float("0"), float("0"), float("0.0392426"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.310251"), float("0"), float("0.242476"), float("0"), float("0.223079"), float("0"), float("0.268803"), float("0"), float("0.0328052"), float("0"), float("0.273985"), float("0"), float("0"), float("0.158735"), float("0.171478"), float("0"), float("0.102056"), float("0"), float("0"), float("0.017277"), float("0"), float("0"), float("0"), float("0"), float("0.163026"), float("0"), float("0.306636"), float("0"), float("0"), float("0"), float("0.235121"), float("0.0890468"), float("0"), float("0.267089"), float("0"), float("0.000100172"), float("0"), float("0.256085"), float("0"), float("0"), float("0.154934"), float("0"), float("0"), float("0"), float("0.222688"), float("0"), float("0"), float("0.459038"), float("0.216095"), float("0.341084"), float("0.472604"), float("0.0938015"), float("0"), float("0.000344158"), float("0"), float("0"), float("0"), float("0"), float("0.145101"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.0770143"), float("0"), float("0.334262"), float("0"), float("0.0855547"), float("0"), float("0.159209"), float("0"), float("0.0432482"), float("0.0493264"), float("0.244654"), float("0"), float("0"), float("0.163553"), float("0.153369"), float("0"), float("0.0873547"), float("0"), float("0"), float("0.0407355"), float("0"), float("0"), float("0"), float("0"), float("0.174414"), float("0"), float("0.309304"), float("0"), float("0"), float("0"), float("0.20293"), float("0.107949"), float("0"), float("0.215721"), float("0"), float("0.0288104"), float("0"), float("0.236245"), float("0"), float("0"), float("0"), float("0"), float("0.0481171"), float("0"), float("0.00160286"), float("0.0681292"), float("0.0254854"), float("0"), float("0.0278975"), float("0"), float("0"), float("0.0102388"), float("0"), float("0"), float("0"), float("0"), float("0"), float("3.14694e-05"), float("0.0458469"), float("0"), float("0"), float("0"), float("0.0530446"), float("0"), float("0"), float("0.0947526"), float("0"), float("0.0166749"), float("0"), float("0.0583104"), float("0")] + + +class PirProgram_example_input_tensor_meta_1913256108345190963: + program_id = 1242071021843173094 + input_name = "linear_2.w_0" + shape = [16, 32] + data = [float("0.230423"), float("-0.20309"), float("0.162069"), float("-0.126608"), float("0.180297"), float("-0.20278"), float("0.302492"), float("-0.251559"), float("-0.108057"), float("0.351752"), float("0.0375671"), float("-0.177384"), float("0.304171"), float("-0.156094"), float("0.242355"), float("0.0150476"), float("-0.255715"), float("0.138158"), float("0.0648821"), float("0.26197"), float("-0.0646441"), float("0.197283"), float("-0.0875255"), float("-0.235316"), float("-0.117067"), float("-0.0573046"), float("-0.273241"), float("0.345782"), float("0.0942704"), float("0.0937579"), float("0.057566"), float("0.19286"), float("-0.326869"), float("-0.325651"), float("-0.25319"), float("0.153059"), float("0.00222721"), float("0.326323"), float("0.152709"), float("0.181255"), float("0.32209"), float("0.337139"), float("-0.224294"), float("-0.245217"), float("0.195527"), float("0.20644"), float("-0.279234"), float("0.240308"), float("0.22163"), float("0.179995"), float("0.290756"), float("0.191943"), float("-0.126184"), float("-0.315085"), float("0.00289739"), float("0.0977283"), float("-0.283358"), float("-0.208794"), float("0.136989"), float("0.21537"), float("0.319846"), float("0.221434"), float("0.261604"), float("-0.0966258"), float("-0.191189"), float("-0.101988"), float("-0.0353045"), float("-0.28391"), float("-0.130498"), float("0.177049"), float("0.143925"), float("-0.120895"), float("0.292791"), float("-0.0774804"), float("-0.229526"), float("0.337984"), float("-0.0903337"), float("-0.171707"), float("0.295915"), float("0.344176"), float("0.342363"), float("0.167197"), float("-0.0261476"), float("0.352028"), float("-0.310108"), float("0.28573"), float("-0.231779"), float("-0.349535"), float("0.226729"), float("-0.0431784"), float("-0.332617"), float("-0.170526"), float("-0.114074"), float("0.131645"), float("-0.295109"), float("-0.294845"), float("0.295009"), float("0.0341118"), float("0.212163"), float("0.0298418"), float("0.181503"), float("-0.103151"), float("-0.146327"), float("-0.0275784"), float("0.22929"), float("-0.199881"), float("-0.123251"), float("-0.0388795"), float("0.306805"), float("0.335557"), float("-0.0176815"), float("-0.272611"), float("0.236301"), float("0.0892752"), float("-0.222741"), float("-0.325869"), float("-0.127395"), float("-0.25068"), float("0.0660806"), float("0.0702004"), float("-0.245965"), float("0.096747"), float("-0.124066"), float("0.227929"), float("0.262033"), float("0.353009"), float("0.0637572"), float("-0.30891"), float("0.175011"), float("-0.0126421"), float("-0.350917"), float("-0.219104"), float("0.255108"), float("0.170938"), float("0.265865"), float("0.321288"), float("0.317676"), float("-0.206672"), float("0.0948979"), float("0.339302"), float("-0.104974"), float("0.229254"), float("-0.173685"), float("0.22608"), float("0.182638"), float("-0.206255"), float("0.338822"), float("0.152891"), float("-0.285523"), float("-0.30232"), float("0.197891"), float("-0.214551"), float("-0.305591"), float("0.234063"), float("-0.232026"), float("0.182208"), float("0.0429179"), float("0.0140892"), float("0.192062"), float("-0.344634"), float("0.118818"), float("0.00202157"), float("-0.283396"), float("-0.292219"), float("0.182276"), float("-0.0348539"), float("-0.319024"), float("-0.185221"), float("0.331582"), float("0.0734183"), float("0.0942539"), float("0.333694"), float("0.0249039"), float("0.0709629"), float("0.226096"), float("0.0418429"), float("-0.0366375"), float("-0.0565682"), float("0.144466"), float("-0.345168"), float("-0.074358"), float("-0.307203"), float("0.0179474"), float("-0.138052"), float("-0.112629"), float("0.308599"), float("-0.161632"), float("0.322421"), float("-0.327087"), float("-0.0683733"), float("-0.0945122"), float("0.348189"), float("-0.00324977"), float("-0.340845"), float("-0.0842492"), float("-0.165104"), float("-0.299402"), float("0.161167"), float("0.244718"), float("-0.179583"), float("0.0124633"), float("0.304233"), float("-0.268381"), float("-0.16432"), float("0.157403"), float("-0.0637301"), float("0.270001"), float("-0.044371"), float("-0.120994"), float("-0.345206"), float("0.0607044"), float("-0.267447"), float("-0.308716"), float("0.0326651"), float("0.217073"), float("-0.128421"), float("0.0860339"), float("-0.109498"), float("-0.177799"), float("-0.128405"), float("0.161585"), float("0.116405"), float("-0.122393"), float("-0.272491"), float("-0.225742"), float("0.304446"), float("0.224986"), float("0.256058"), float("0.277367"), float("-0.100766"), float("0.188724"), float("0.212626"), float("-0.21794"), float("-0.126419"), float("-0.224952"), float("-0.0924843"), float("0.206705"), float("0.021292"), float("0.0137566"), float("-0.0360458"), float("-0.234257"), float("0.117413"), float("0.153925"), float("0.191284"), float("-0.253123"), float("0.21299"), float("0.327653"), float("0.0754134"), float("0.193985"), float("-0.0368757"), float("-0.326489"), float("-0.267416"), float("0.0922668"), float("0.105197"), float("0.280179"), float("0.110465"), float("-0.0979321"), float("-0.275342"), float("-0.279135"), float("0.0396403"), float("0.208052"), float("-0.0856857"), float("0.03993"), float("0.299113"), float("0.0134926"), float("0.0971514"), float("-0.0327729"), float("0.324181"), float("-0.162795"), float("0.104321"), float("-0.00222742"), float("-0.314876"), float("-0.186012"), float("-0.141716"), float("-0.305346"), float("0.0716737"), float("-0.315062"), float("-0.305029"), float("0.105486"), float("0.311472"), float("-0.13068"), float("-0.241825"), float("-0.261683"), float("0.344032"), float("-0.211968"), float("-0.30109"), float("0.301081"), float("-0.300443"), float("0.0113317"), float("-0.13623"), float("0.00588745"), float("0.323762"), float("-0.27139"), float("0.128905"), float("0.226326"), float("-0.0567057"), float("0.0513325"), float("-0.00563676"), float("0.292468"), float("-0.00536835"), float("-0.240204"), float("0.0458691"), float("0.258935"), float("-0.259172"), float("0.155908"), float("0.154713"), float("-0.29383"), float("0.345019"), float("-0.302847"), float("-0.102248"), float("-0.224491"), float("0.00236819"), float("0.181067"), float("0.289794"), float("0.118373"), float("-0.00321519"), float("0.166811"), float("-0.0312314"), float("-0.0175973"), float("0.23624"), float("0.196907"), float("-0.292757"), float("0.0645859"), float("-0.1919"), float("0.219958"), float("-0.347093"), float("-0.0693511"), float("-0.0524661"), float("-0.237186"), float("0.0300816"), float("-0.259862"), float("-0.319011"), float("-0.0807318"), float("0.135482"), float("0.307606"), float("0.350905"), float("0.0697996"), float("0.283031"), float("0.237269"), float("0.0447312"), float("-0.131454"), float("-0.28127"), float("0.160253"), float("0.148049"), float("0.243601"), float("0.323207"), float("-0.143284"), float("-0.341898"), float("0.293442"), float("0.063323"), float("0.120295"), float("-0.18036"), float("-0.0975274"), float("-0.0455448"), float("-0.183233"), float("-0.0857107"), float("0.278475"), float("0.174762"), float("0.340955"), float("-0.308548"), float("-0.350232"), float("-0.251843"), float("-0.218846"), float("-0.27189"), float("-0.261468"), float("0.0743435"), float("-0.044729"), float("-0.248535"), float("0.257598"), float("-0.103987"), float("-0.145567"), float("0.023903"), float("-0.311359"), float("-0.0946209"), float("-0.0670924"), float("0.256815"), float("0.0447149"), float("-0.0385811"), float("0.215715"), float("0.0890531"), float("-0.120938"), float("-0.135609"), float("0.295268"), float("0.222618"), float("-0.117695"), float("0.215494"), float("0.14366"), float("0.277732"), float("-0.296978"), float("-0.189063"), float("0.266609"), float("0.243941"), float("-0.106652"), float("-0.0466308"), float("-0.295165"), float("0.223554"), float("0.269018"), float("-0.199119"), float("0.138165"), float("-0.308141"), float("0.35326"), float("0.321212"), float("0.019633"), float("-0.321271"), float("0.0288094"), float("-0.229809"), float("-0.0147547"), float("0.204604"), float("0.184154"), float("-0.259976"), float("0.211335"), float("0.0558831"), float("-0.110034"), float("-0.327218"), float("0.213624"), float("-0.329281"), float("-0.176617"), float("0.163939"), float("-0.111053"), float("0.15307"), float("0.195678"), float("-0.122387"), float("-0.13225"), float("0.175086"), float("0.207095"), float("-0.319538"), float("-0.189705"), float("-0.090235"), float("-0.340103"), float("0.0825443"), float("-0.303053"), float("0.135722"), float("0.323061"), float("0.284381"), float("0.0299432"), float("-0.28277"), float("0.179539"), float("-0.283711"), float("0.303636"), float("0.306957"), float("0.345974"), float("0.0915332"), float("0.0688096"), float("0.333376"), float("-0.0633284"), float("0.0370856"), float("9.25966e-05"), float("-0.239747"), float("7.92782e-05"), float("0.296225"), float("-0.107539"), float("-0.087091"), float("-0.183428"), float("-0.248449"), float("-0.105759"), float("-0.352878"), float("-0.137407"), float("-0.203483"), float("0.0538141"), float("-0.0800397"), float("0.14867"), float("-0.253796"), float("0.169363"), float("-0.103523"), float("-0.105282"), float("-0.102652"), float("-0.298457"), float("-0.199085"), float("-0.013719"), float("-0.323702"), float("-0.154908"), float("-0.0621661"), float("0.0333456"), float("-0.317511"), float("0.0649108"), float("0.152055"), float("-0.113154"), float("-0.337959"), float("-0.261652"), float("0.12889"), float("-0.0705134"), float("-0.162341"), float("-0.21903"), float("-0.195068"), float("0.163776"), float("-0.0572965"), float("0.179957"), float("0.298765"), float("0.0771351"), float("-0.239391"), float("-0.246127"), float("0.232749"), float("0.132148"), float("0.0270761"), float("0.256751"), float("0.0966847"), float("0.338032"), float("0.034458"), float("-0.0380405"), float("0.0232373"), float("0.332713"), float("0.287575"), float("0.328612"), float("0.216737"), float("-0.268043"), float("-0.24701"), float("-0.352788"), float("-0.170355"), float("0.220089"), float("-0.181284"), float("0.178662"), float("0.316528")] + + +class PirProgram_example_input_tensor_meta_560967842253293313: + program_id = 1242071021843173094 + input_name = "middle_0" + shape = [16, 2, 32] + data = None + + +class PirProgram_example_input_tensor_meta_7562718223343614406: + program_id = 1242071021843173094 + input_name = "linear_2.b_0" + shape = [32] + data = [float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0")] + + +class PirProgram_example_input_tensor_meta_602773716366570321: + program_id = 1242071021843173094 + input_name = "args_0" + shape = [16, 2, 16] + data = [float("-1.24783"), float("1.35968"), float("-0.0643954"), float("-0.627694"), float("3.57834"), float("-1.40647"), float("-3.06211"), float("-0.852737"), float("1.89044"), float("1.71429"), float("0.873626"), float("2.45266"), float("2.46995"), float("-4.18985"), float("5.24214"), float("-0.769194"), float("-0.432318"), float("1.44443"), float("-0.398696"), float("-1.49099"), float("-0.245603"), float("-1.75011"), float("-1.19734"), float("0.138174"), float("1.28394"), float("-0.147029"), float("1.46563"), float("0.138966"), float("1.49462"), float("-1.24567"), float("1.74074"), float("-0.541634"), float("-0.137766"), float("0.149732"), float("0.813494"), float("0.303299"), float("-0.238571"), float("1.88537"), float("-0.0370909"), float("0.573859"), float("0.349418"), float("0.487316"), float("-1.32582"), float("-0.843882"), float("-0.676632"), float("-0.909084"), float("0.270721"), float("-1.23429"), float("-0.0424248"), float("0.429659"), float("-0.14267"), float("-0.428278"), float("-0.0162147"), float("-0.494324"), float("-0.285505"), float("0.0283108"), float("0.343003"), float("0.0251735"), float("0.319188"), float("0.0917862"), float("0.402113"), float("-0.271283"), float("0.494286"), float("-0.191708"), float("-0.284004"), float("0.298903"), float("1.58414"), float("0.506222"), float("-0.530693"), float("3.27026"), float("-0.117699"), float("1.01495"), float("0.711183"), float("0.811757"), float("-2.06371"), float("-1.2438"), float("-1.13482"), float("-1.38332"), float("0.486574"), float("-2.27285"), float("-0.358158"), float("0.179657"), float("0.702598"), float("0.133837"), float("-0.116537"), float("1.53755"), float("0.00049031"), float("0.551636"), float("0.176739"), float("0.308397"), float("-0.844465"), float("-0.537499"), float("-0.755579"), float("-0.623192"), float("0.182918"), float("-1.02258"), float("-1.34653"), float("0.913991"), float("1.1851"), float("0.400625"), float("0.121387"), float("3.74875"), float("0.507555"), float("1.24397"), float("-0.00871566"), float("0.270786"), float("-2.39195"), float("-1.38387"), float("-2.33894"), float("-0.779432"), float("0.529702"), float("-2.01132"), float("-0.0355457"), float("0.156263"), float("-0.0398702"), float("-0.0941751"), float("0.070695"), float("-0.125185"), float("-0.123143"), float("-0.0328905"), float("0.0242618"), float("0.050551"), float("0.0789997"), float("0.0688375"), float("0.0704337"), float("-0.152359"), float("0.139597"), float("-0.0303556"), float("-0.422165"), float("0.755079"), float("-0.125477"), float("-0.390056"), float("0.129337"), float("-0.44178"), float("-0.83648"), float("-0.363755"), float("-0.396812"), float("-0.217299"), float("1.19051"), float("0.353966"), float("0.100611"), float("-0.841632"), float("1.13851"), float("-0.0570613"), float("-0.321325"), float("1.00824"), float("-0.478592"), float("-1.40818"), float("-0.435232"), float("-1.11203"), float("-0.501488"), float("-0.127583"), float("0.967821"), float("0.100076"), float("0.722625"), float("0.514725"), float("0.67351"), float("-0.719309"), float("0.999483"), float("-0.640918"), float("-1.46169"), float("1.59347"), float("3.25034"), float("0.413147"), float("0.432863"), float("5.4329"), float("0.0917308"), float("1.98376"), float("0.598049"), float("1.64893"), float("-1.92021"), float("-1.70858"), float("-2.384"), float("-1.1942"), float("0.144455"), float("-3.21842"), float("-1.22085"), float("-0.843408"), float("3.91006"), float("0.0781953"), float("-1.01348"), float("6.29112"), float("1.98013"), float("2.26446"), float("0.98046"), float("0.878038"), float("-4.09496"), float("-3.09536"), float("-2.928"), float("-3.87442"), float("-0.000862794"), float("-3.55961"), float("-0.1064"), float("-0.161292"), float("0.187614"), float("0.0255209"), float("-0.162917"), float("0.42479"), float("-0.0453343"), float("0.436514"), float("-0.191061"), float("0.137538"), float("-0.410105"), float("-0.431748"), float("-0.438004"), float("-0.404633"), float("0.218743"), float("-0.427686"), float("-0.424405"), float("0.837761"), float("0.332992"), float("-0.0513256"), float("0.047148"), float("1.35676"), float("-0.0688395"), float("0.506198"), float("0.163514"), float("-0.337938"), float("-1.2557"), float("-0.616636"), float("-1.1297"), float("-0.193176"), float("0.337317"), float("-0.561698"), float("-0.135902"), float("0.393851"), float("0.00424394"), float("-0.196324"), float("0.159519"), float("-0.308631"), float("-0.376123"), float("-0.071268"), float("0.0883756"), float("0.0545766"), float("0.303032"), float("0.111892"), float("0.218541"), float("-0.537371"), float("0.421097"), float("-0.129858"), float("-1.03447"), float("0.513909"), float("2.0756"), float("0.441222"), float("-0.472549"), float("4.29349"), float("0.13518"), float("1.6049"), float("0.482167"), float("0.853731"), float("-2.33474"), float("-1.43174"), float("-2.12045"), float("-1.69272"), float("0.520399"), float("-2.92517"), float("-0.154838"), float("1.12713"), float("-0.729672"), float("-1.66338"), float("-0.657706"), float("-1.18132"), float("-0.874759"), float("0.06689"), float("1.51983"), float("0.351833"), float("0.986195"), float("0.245073"), float("0.796269"), float("-0.622144"), float("0.948545"), float("-1.01152"), float("-1.22342"), float("2.62304"), float("-0.221326"), float("-2.36949"), float("0.0231131"), float("-3.1239"), float("-1.56313"), float("-0.0846508"), float("1.4516"), float("-0.843756"), float("2.11236"), float("0.770368"), float("2.54804"), float("-2.68047"), float("3.21551"), float("-0.423819"), float("-0.392694"), float("0.187697"), float("0.413998"), float("0.057787"), float("-0.13115"), float("1.17413"), float("0.0864602"), float("0.307451"), float("0.0920442"), float("0.143506"), float("-0.635472"), float("-0.255527"), float("-0.542642"), float("-0.322647"), float("0.183446"), float("-0.825342"), float("-0.0203514"), float("0.0198951"), float("0.0865615"), float("0.0272587"), float("-0.00242023"), float("0.195467"), float("0.07148"), float("0.0644732"), float("0.0219086"), float("0.0301208"), float("-0.185134"), float("-0.132513"), float("-0.0794464"), float("-0.100434"), float("0.0120146"), float("-0.08797"), float("-1.29686"), float("1.00538"), float("-0.771612"), float("-1.60371"), float("1.04745"), float("-1.71847"), float("-0.364911"), float("-0.669676"), float("1.52801"), float("-0.0574329"), float("0.59866"), float("1.73576"), float("2.91115"), float("-1.2902"), float("3.56666"), float("0.497791"), float("-0.629735"), float("1.17433"), float("-0.648112"), float("-1.04914"), float("-0.195696"), float("-1.03082"), float("-0.321714"), float("-0.668095"), float("-0.117482"), float("-0.81666"), float("1.54317"), float("0.372422"), float("0.500825"), float("-0.542041"), float("1.91077"), float("0.324363"), float("-0.746501"), float("1.56012"), float("-0.693475"), float("-1.47149"), float("0.417272"), float("-1.5822"), float("-0.614944"), float("-0.337996"), float("0.782261"), float("-0.244622"), float("1.03057"), float("0.317955"), float("1.1072"), float("-1.29199"), float("1.91734"), float("0.0562871"), float("-0.344145"), float("0.582381"), float("0.0177077"), float("-0.394364"), float("0.278082"), float("-0.549242"), float("-0.608972"), float("-0.148099"), float("0.0354738"), float("0.105372"), float("0.54669"), float("0.293604"), float("0.384397"), float("-0.770095"), float("0.530887"), float("-0.0841932"), float("-1.5664"), float("1.47664"), float("-0.0485786"), float("-1.2655"), float("1.12964"), float("-1.11663"), float("-0.127906"), float("-0.158063"), float("0.266339"), float("-0.689339"), float("0.552855"), float("-0.29022"), float("1.41004"), float("-1.61181"), float("1.21825"), float("0.138486"), float("-1.40203"), float("1.25058"), float("0.718333"), float("0.433586"), float("0.213605"), float("3.4107"), float("0.702294"), float("0.844367"), float("-0.269252"), float("0.469749"), float("-2.07015"), float("-0.969774"), float("-1.84929"), float("-0.536334"), float("0.553032"), float("-2.07369"), float("-0.875411"), float("1.6934"), float("-0.485379"), float("-1.53678"), float("0.237551"), float("-1.80035"), float("-0.801288"), float("-0.196758"), float("1.13256"), float("-0.170934"), float("1.21607"), float("0.530888"), float("1.79473"), float("-1.66515"), float("2.21643"), float("-0.170763"), float("-0.242361"), float("0.0761521"), float("0.432139"), float("0.0631636"), float("-0.093372"), float("0.938469"), float("0.0142968"), float("0.313853"), float("0.120945"), float("0.144088"), float("-0.496069"), float("-0.284314"), float("-0.471705"), float("-0.335182"), float("0.114975"), float("-0.613756"), float("-0.133989"), float("0.332343"), float("-0.0773375"), float("-0.386288"), float("-0.0153539"), float("-0.397145"), float("-0.344031"), float("0.0805068"), float("0.348394"), float("0.00255038"), float("0.33544"), float("0.0063985"), float("0.34345"), float("-0.323657"), float("0.404106"), float("-0.197486"), float("-0.0204629"), float("0.165297"), float("0.344726"), float("0.0223876"), float("-0.00523791"), float("0.623413"), float("-0.16317"), float("0.276788"), float("0.216639"), float("0.113306"), float("-0.499786"), float("-0.285747"), float("-0.187579"), float("-0.235846"), float("0.0960729"), float("-0.395305"), float("-0.0911418"), float("0.288873"), float("-0.0312524"), float("-0.338057"), float("-0.0387871"), float("-0.421691"), float("-0.315626"), float("0.125869"), float("0.273582"), float("-0.058278"), float("0.3202"), float("0.015641"), float("0.297726"), float("-0.288548"), float("0.329069"), float("-0.179419"), float("-0.0326695"), float("0.117692"), float("-0.060717"), float("-0.112393"), float("0.0348706"), float("-0.105831"), float("-0.0257303"), float("-0.0271061"), float("0.0484697"), float("-0.00249167"), float("0.0425337"), float("0.0319538"), float("0.0739469"), float("-0.0675583"), float("0.136938"), float("-0.0154109")] + + +class PirProgram_example_input_tensor_meta_6573343046139926864: + program_id = 692320333968606381 + input_name = "output_grad_0" + shape = [16, 2, 16] + data = [float("0.00107317"), float("-0.00117165"), float("-0.00163671"), float("0.00156766"), float("0.00201485"), float("7.37798e-05"), float("-0.00288858"), float("0.00324427"), float("0.00363222"), float("0.00455793"), float("0.000339677"), float("0.00159846"), float("0.00616652"), float("-0.000968744"), float("-0.00142565"), float("0.00341143"), float("-0.000397156"), float("0.00313079"), float("-0.0038428"), float("-0.000317453"), float("-0.00244509"), float("-0.00321148"), float("-0.00215432"), float("0.0027039"), float("0.000365354"), float("0.00177382"), float("-7.06572e-05"), float("0.00144024"), float("0.00486523"), float("0.00112444"), float("-0.00136742"), float("-0.00121718"), float("-0.000322603"), float("0.000538117"), float("3.002e-05"), float("-0.000153229"), float("0.000238948"), float("-4.99467e-05"), float("-0.000162901"), float("5.13133e-05"), float("0.000178962"), float("-0.00028677"), float("-0.000726859"), float("-0.000173296"), float("8.02252e-05"), float("7.30726e-05"), float("-0.00029218"), float("-0.000122243"), float("-1.10911e-05"), float("0.000961878"), float("-0.000971545"), float("-0.000160983"), float("-0.000659728"), float("-0.000886996"), float("-0.000450224"), float("0.000584157"), float("0.000198652"), float("0.000577065"), float("4.15716e-05"), float("0.000471037"), float("0.00124457"), float("0.000217747"), float("-0.000470511"), float("-0.000316141"), float("0.000342187"), float("0.0018494"), float("0.000947351"), float("0.00101066"), float("0.00129266"), float("0.00142459"), float("0.000266141"), float("-0.00110398"), float("0.000510692"), float("0.000157268"), float("-0.00170386"), float("-0.00100753"), float("-0.00057964"), float("0.000633022"), float("-0.000892884"), float("-0.000346073"), float("-4.86121e-05"), float("0.000248871"), float("-8.16761e-05"), float("-0.000720575"), float("0.000460798"), float("0.000718931"), float("-0.0010086"), float("-0.00063695"), float("-0.000415294"), float("0.000742808"), float("-0.00158288"), float("-0.000315739"), float("-0.0013079"), float("3.30604e-05"), float("-0.000385291"), float("-0.000363103"), float("-0.000218203"), float("0.00222529"), float("0.000970178"), float("0.0010858"), float("0.00236524"), float("0.00202229"), float("-6.98109e-05"), float("-0.00177226"), float("-0.000319509"), float("0.00120065"), float("-0.00222744"), float("-0.00133017"), float("-0.0020776"), float("0.000887785"), float("-0.00186867"), float("-0.0012222"), float("-8.52198e-05"), float("0.000224672"), float("-0.000124413"), float("-0.00012138"), float("-7.69958e-05"), float("-0.000256524"), float("-0.000147351"), float("-1.61188e-05"), float("-0.000101862"), float("0.000245915"), float("-9.56549e-05"), float("0.000121793"), float("0.000120966"), float("2.49652e-05"), float("-0.000102786"), float("-0.000184276"), float("-0.000588331"), float("0.00143901"), float("-0.000900593"), float("-0.000131671"), float("0.000196109"), float("-0.00068363"), float("-0.000781266"), float("0.000212953"), float("-0.000563651"), float("0.000203772"), float("0.000687689"), float("0.000925853"), float("0.000966789"), float("0.000618601"), float("-0.000566369"), float("-0.000386996"), float("-0.00146396"), float("0.00202857"), float("-0.00351244"), float("7.24205e-05"), float("-0.00133867"), float("-0.00197239"), float("-0.00188278"), float("0.000753329"), float("0.000766242"), float("0.00134591"), float("-0.00017611"), float("0.00196395"), float("0.0027534"), float("0.00141287"), float("-0.00103552"), float("-0.000716369"), float("0.000550839"), float("0.00322606"), float("0.000856896"), float("-0.00125687"), float("0.00129017"), float("0.00116238"), float("-0.00187795"), float("-0.00158381"), float("-0.00141739"), float("0.00180403"), float("-0.00490405"), float("-0.000981318"), float("-0.00421745"), float("-0.000114125"), float("-0.00195622"), float("-0.000424439"), float("-0.000280211"), float("0.00171129"), float("0.000688157"), float("-0.00253045"), float("0.00109986"), float("0.00417076"), float("-0.0016617"), float("-0.0023778"), float("-0.000190303"), float("0.00340801"), float("-0.00539941"), float("-0.00185255"), float("-0.00372863"), float("-0.00238725"), float("-0.00133698"), float("0.00134814"), float("0.000206168"), float("-0.000115355"), float("0.000135783"), float("-5.93454e-05"), float("-0.000104663"), float("-0.000363951"), float("-0.000272538"), float("0.000322769"), float("-0.000417613"), float("-0.000232908"), float("-0.000300922"), float("-8.19516e-05"), float("-0.000290464"), float("-0.000267267"), float("-0.000150961"), float("9.07732e-05"), float("0.000208024"), float("0.000412389"), float("0.000340767"), float("-0.000807369"), float("-2.84969e-05"), float("0.000996116"), float("-0.000259763"), float("-0.000458858"), float("-0.00100166"), float("0.000698113"), float("-0.00121583"), float("-0.000378595"), float("-0.00100867"), float("-0.000326933"), float("-0.000104752"), float("0.000252775"), float("-0.000104392"), float("0.000384491"), float("-0.000120818"), float("-0.000332935"), float("-0.00014957"), float("-0.000383103"), float("-0.00033556"), float("-0.000210315"), float("-0.000283282"), float("0.00040653"), float("-0.000114731"), float("0.000345088"), float("0.000101139"), float("-3.16119e-05"), float("-0.000134473"), float("-0.000333086"), float("0.000365319"), float("2.31993e-05"), float("-0.00101835"), float("-0.00198467"), float("0.000607929"), float("0.00105125"), float("-0.00256911"), float("-0.00117818"), float("-0.000836815"), float("0.00171317"), float("-0.00357581"), float("0.000157612"), float("-0.00292617"), float("-0.000337635"), float("-0.000517322"), float("-0.00113363"), float("-0.000858934"), float("0.00275311"), float("-0.00322123"), float("-0.00128946"), float("-0.00257051"), float("-0.00303501"), float("-0.00161937"), float("0.00149826"), float("0.000905219"), float("0.00253491"), float("0.000515937"), float("0.00195283"), float("0.00346017"), float("0.00160003"), float("-0.00186369"), float("-0.00181285"), float("0.000569412"), float("0.00475362"), float("-0.00489154"), float("-0.000602263"), float("-0.00312457"), float("-0.00445273"), float("-0.00261735"), float("0.00360081"), float("-0.00123231"), float("0.00259976"), float("-0.00111133"), float("0.00116551"), float("0.00690931"), float("-8.18758e-05"), float("-0.00153739"), float("-0.00306445"), float("0.00017831"), float("0.000354979"), float("-0.000123217"), float("-0.000102344"), float("0.000186917"), float("0.000809431"), float("-0.000367907"), float("-0.000401029"), float("-0.000246482"), float("0.000556523"), float("-0.000902953"), float("-0.000132123"), float("-0.000778616"), float("-0.000146372"), float("-0.00013007"), float("7.63733e-05"), float("2.0302e-05"), float("3.89685e-05"), float("3.33519e-05"), float("-6.25571e-06"), float("5.20851e-05"), float("0.000157613"), float("-7.67726e-05"), float("-6.23804e-05"), float("-4.95858e-05"), float("8.56801e-05"), float("-0.000203655"), float("-9.11566e-05"), float("-0.000137199"), float("-5.19502e-05"), float("-4.72111e-05"), float("4.96732e-05"), float("0.000994176"), float("-0.00101409"), float("0.00191827"), float("0.000882816"), float("0.000591085"), float("0.00105649"), float("0.000372834"), float("0.000875936"), float("-0.000349729"), float("0.000747718"), float("-0.000399308"), float("-0.000425689"), float("0.00309901"), float("-0.00110301"), float("0.00112329"), float("0.00215364"), float("-0.000519875"), float("0.000449016"), float("-0.000270285"), float("-0.000219273"), float("-0.000324491"), float("-0.000262691"), float("-0.000138284"), float("0.0003396"), float("-0.000301592"), float("-0.000257651"), float("0.00117054"), float("-0.000151151"), float("3.56859e-05"), float("0.000969652"), float("-0.00039694"), float("0.000225646"), float("0.000454721"), float("0.00344821"), float("-0.00299805"), float("-0.000152523"), float("-0.000732639"), float("-0.00203668"), float("-0.000872876"), float("0.00124991"), float("-0.000206255"), float("0.000911826"), float("-0.000406766"), float("0.000470595"), float("0.00296606"), float("-0.000641648"), float("-0.00152614"), float("-0.000957327"), float("-0.000434856"), float("0.00139467"), float("-0.000964782"), float("-0.000269292"), float("-0.000128748"), float("-0.00115347"), float("-0.000888983"), float("0.000360765"), float("1.10851e-06"), float("0.000905672"), float("-7.21941e-05"), float("0.000970206"), float("0.00103436"), float("0.000360162"), float("-0.00084144"), float("-0.000245883"), float("-0.00154833"), float("0.00260887"), float("-0.0028161"), float("-0.000606185"), float("0.000816593"), float("-0.00119184"), float("-0.00179874"), float("0.00128897"), float("0.000705463"), float("0.000194246"), float("-0.000344325"), float("0.000336249"), float("0.00326308"), float("0.00110578"), float("-0.00228927"), float("4.76386e-05"), float("0.000593787"), float("0.000946524"), float("0.000321888"), float("-0.00150594"), float("0.0006321"), float("0.0028326"), float("-0.000986253"), float("-0.0018503"), float("-0.00087765"), float("0.001987"), float("-0.00344485"), float("-0.000441437"), float("-0.00218665"), float("-0.000122176"), float("-0.000163618"), float("-0.000218914"), float("0.00134576"), float("0.00197272"), float("-0.000995735"), float("0.000263282"), float("-0.00112765"), float("-0.00111128"), float("0.000379471"), float("0.00132961"), float("-7.55798e-07"), float("0.000778359"), float("-0.000557666"), float("-0.00122852"), float("0.00195573"), float("-0.000791029"), float("-0.000683356"), float("-0.000497815"), float("-1.17727e-05"), float("0.00016727"), float("-0.000183097"), float("-0.000404088"), float("0.000217476"), float("0.000506551"), float("-0.000556639"), float("-0.000408932"), float("-0.000231675"), float("0.000294711"), float("-0.00109808"), float("-0.000204529"), float("-0.000558736"), float("1.61032e-05"), float("-3.83038e-05"), float("-0.000171005"), float("-0.000396657"), float("0.000533907"), float("-0.00102416"), float("8.19507e-05"), float("-0.000511321"), float("-0.000713884"), float("-0.00065625"), float("0.000688043"), float("7.72698e-05"), float("0.000343118"), float("-3.60998e-05"), float("0.000457628"), float("0.00127803"), float("0.000562512"), float("-0.000228446"), float("-0.000199705"), float("-6.09644e-05"), float("9.9358e-05"), float("-2.25066e-05"), float("-0.000130122"), float("7.57663e-05"), float("-4.82356e-05"), float("-0.000337591"), float("9.73819e-05"), float("9.94797e-06"), float("-4.78918e-05"), float("-0.000385803"), float("2.49486e-05"), float("-0.000145524"), float("-1.2884e-05"), float("-4.45019e-05"), float("0.000119422"), float("-0.000170435"), float("0.000654842"), float("-0.000939046"), float("-1.96642e-05"), float("-0.000560269"), float("-0.000625978"), float("-0.000445959"), float("0.000620287"), float("0.000111282"), float("0.00032595"), float("-5.50135e-06"), float("0.000306219"), float("0.00119186"), float("0.000381669"), float("-0.000291645"), float("-0.00027276"), float("-1.00973e-05"), float("0.000269494"), float("-0.000242654"), float("-3.09646e-05"), float("-5.82016e-05"), float("-0.000165689"), float("-5.35421e-05"), float("8.65551e-05"), float("1.77047e-05"), float("7.4701e-05"), float("-4.16094e-05"), float("0.000107224"), float("0.000217084"), float("-1.98606e-05"), float("-0.000127421"), float("-5.16393e-05")] + + +class PirProgram_example_input_tensor_meta_984001809579706501: + program_id = 692320333968606381 + input_name = "middle_5" + shape = [16, 2, 16] + data = None + + +class PirProgram_example_input_tensor_meta_1407663771573075047: + program_id = 692320333968606381 + input_name = "middle_4" + shape = [16, 2, 32] + data = None + + +class PirProgram_example_input_tensor_meta_7391043412817019354: + program_id = 692320333968606381 + input_name = "middle_2" + shape = [1] + data = [float("0.2")] + + +class PirProgram_example_input_tensor_meta_6623077475521577491: + program_id = 692320333968606381 + input_name = "middle_0" + shape = [16, 2, 32] + data = None + + +class PirProgram_example_input_tensor_meta_3111587029586218519: + program_id = 692320333968606381 + input_name = "linear_1.w_0" + shape = [32, 16] + data = [float("0.344869"), float("-0.262408"), float("0.0439143"), float("0.0156327"), float("0.172872"), float("0.0675123"), float("0.328597"), float("0.28137"), float("0.193017"), float("0.118872"), float("0.0326795"), float("0.0050282"), float("0.0623816"), float("-0.172865"), float("0.280214"), float("-0.221634"), float("0.158068"), float("-0.227596"), float("0.345091"), float("0.0911414"), float("-0.0102415"), float("0.167898"), float("-0.158481"), float("0.289152"), float("0.113772"), float("0.103651"), float("-0.119918"), float("-0.337519"), float("-0.25293"), float("-0.307349"), float("-0.0620606"), float("0.0564728"), float("0.227404"), float("0.139312"), float("0.0545452"), float("-0.304748"), float("0.167032"), float("0.0143448"), float("0.135743"), float("-0.105886"), float("-0.343549"), float("0.290257"), float("-0.2818"), float("-0.160174"), float("0.1422"), float("-0.229487"), float("0.177531"), float("-0.315231"), float("0.331407"), float("0.0724641"), float("-0.265019"), float("-0.103591"), float("-0.30779"), float("-0.196846"), float("-0.240486"), float("-0.0806981"), float("0.246875"), float("0.155696"), float("0.207512"), float("0.245401"), float("-0.135328"), float("0.104992"), float("0.289559"), float("-0.0550811"), float("-0.0135079"), float("0.345409"), float("0.237245"), float("-0.201128"), float("0.14598"), float("0.235113"), float("0.208274"), float("-0.244138"), float("0.348133"), float("0.0957723"), float("-0.00846937"), float("0.128104"), float("0.236483"), float("-0.026668"), float("0.278156"), float("0.122785"), float("-0.148078"), float("-0.251864"), float("0.327847"), float("0.185817"), float("-0.0450313"), float("0.238892"), float("0.252549"), float("0.330687"), float("-0.0880492"), float("0.309633"), float("-0.219191"), float("-0.199132"), float("0.0599504"), float("-0.1848"), float("-0.11876"), float("-0.0490333"), float("0.0584787"), float("-0.0681539"), float("-0.298419"), float("-0.0999435"), float("0.346551"), float("0.129203"), float("-0.332224"), float("-0.147948"), float("0.00135393"), float("0.0816889"), float("-0.131467"), float("-0.224166"), float("-0.0461218"), float("-0.214483"), float("-0.0326581"), float("0.10036"), float("-0.264792"), float("-0.32798"), float("0.271663"), float("0.288337"), float("-0.350532"), float("0.06591"), float("-0.0911039"), float("0.0444793"), float("-0.00501465"), float("-0.213402"), float("0.349463"), float("-0.232356"), float("-0.011135"), float("-0.234883"), float("-0.223632"), float("0.207395"), float("0.242189"), float("-0.34568"), float("0.281336"), float("-0.141172"), float("-0.279639"), float("-0.0623853"), float("-0.31392"), float("0.0430904"), float("0.319302"), float("0.0603828"), float("0.165175"), float("0.0803925"), float("0.241733"), float("-0.31437"), float("-0.0401438"), float("-0.186046"), float("-0.0437005"), float("0.3525"), float("0.00955195"), float("-0.245716"), float("-0.23917"), float("-0.22811"), float("-0.180175"), float("0.0921878"), float("-0.153823"), float("-0.0783991"), float("0.297338"), float("-0.230138"), float("-0.11621"), float("-0.171227"), float("-0.235734"), float("-0.13929"), float("-0.198997"), float("0.16052"), float("0.210111"), float("0.18088"), float("0.242678"), float("-0.276179"), float("-0.183217"), float("0.00408171"), float("-0.230232"), float("-0.305358"), float("0.207404"), float("-0.0695701"), float("0.349871"), float("-0.293597"), float("0.342355"), float("0.349062"), float("-0.0969384"), float("-0.304206"), float("-0.100894"), float("0.218041"), float("-0.238812"), float("0.175587"), float("0.324251"), float("-0.219158"), float("-0.192937"), float("0.17492"), float("0.190574"), float("0.0470951"), float("-0.140387"), float("-0.180639"), float("0.0101866"), float("-0.195554"), float("0.0171249"), float("-0.0556396"), float("0.220385"), float("0.0830055"), float("-0.209667"), float("0.0225179"), float("-0.0834754"), float("0.111096"), float("0.125786"), float("0.0543055"), float("0.0282446"), float("0.0734042"), float("0.313033"), float("-0.0875646"), float("0.0155789"), float("-0.191"), float("-0.174547"), float("0.29949"), float("0.101439"), float("0.167632"), float("0.294102"), float("0.284237"), float("-0.306646"), float("0.123889"), float("0.273155"), float("0.0722914"), float("0.343355"), float("0.200973"), float("-0.0359481"), float("0.295771"), float("-0.0428864"), float("-0.276863"), float("0.219781"), float("-0.321594"), float("-0.193598"), float("0.0766501"), float("-0.0693026"), float("-0.108357"), float("0.239546"), float("-0.0313863"), float("-0.235817"), float("0.312584"), float("0.147659"), float("0.199029"), float("0.0308142"), float("0.319813"), float("0.276646"), float("0.232516"), float("-0.108472"), float("-0.271957"), float("0.214807"), float("0.131503"), float("0.0458386"), float("0.158341"), float("-0.258767"), float("-0.0813985"), float("0.00375828"), float("0.135654"), float("0.265497"), float("0.0394395"), float("-0.234527"), float("0.0994757"), float("-0.342846"), float("0.215965"), float("-0.319046"), float("0.146627"), float("-0.0292457"), float("0.0153329"), float("0.351146"), float("0.082035"), float("0.350166"), float("-0.317657"), float("-0.232845"), float("-0.0578491"), float("-0.266218"), float("0.270595"), float("0.119427"), float("-0.238121"), float("0.319077"), float("0.299425"), float("-0.125333"), float("0.0425065"), float("-0.157729"), float("-0.291869"), float("-0.202124"), float("-0.17027"), float("0.290271"), float("0.0382349"), float("0.322722"), float("-0.215145"), float("-0.0237843"), float("-0.0231641"), float("0.347289"), float("0.217441"), float("0.182893"), float("0.0933436"), float("-0.00364189"), float("0.251967"), float("-0.248983"), float("0.0451025"), float("-0.0902002"), float("0.0672752"), float("-0.33523"), float("0.00100524"), float("-0.244151"), float("0.116821"), float("-0.218376"), float("0.0135987"), float("-0.155867"), float("0.0408046"), float("0.272804"), float("-0.157648"), float("0.178179"), float("0.0977667"), float("-0.119494"), float("-0.184943"), float("-0.330054"), float("-0.00757139"), float("-0.086305"), float("-0.00671808"), float("-0.221499"), float("0.178183"), float("-0.315276"), float("-0.0872174"), float("-0.142795"), float("-0.0392847"), float("0.125988"), float("-0.309111"), float("-0.100147"), float("-0.281519"), float("-0.108678"), float("0.170384"), float("-0.315366"), float("0.262136"), float("-0.0733001"), float("0.127174"), float("0.176889"), float("-0.211834"), float("-0.323871"), float("-0.0813174"), float("-0.263945"), float("-0.287189"), float("0.0900709"), float("-0.149072"), float("-0.100897"), float("-5.45169e-05"), float("0.0549951"), float("-0.167664"), float("0.286603"), float("-0.187299"), float("-0.0942526"), float("0.178792"), float("0.300509"), float("0.353007"), float("-0.32969"), float("0.044596"), float("0.255667"), float("-0.191181"), float("-0.295033"), float("-0.0528022"), float("0.287379"), float("0.158327"), float("0.00928331"), float("0.228015"), float("0.00668824"), float("0.0287411"), float("-0.0175968"), float("0.0713126"), float("0.347188"), float("-0.00186841"), float("-0.0819523"), float("-0.221304"), float("-0.0364172"), float("-0.264244"), float("0.307289"), float("-0.312169"), float("-0.0896873"), float("0.110611"), float("-0.347699"), float("-0.266626"), float("0.151276"), float("0.317222"), float("-0.0747132"), float("0.0141522"), float("-0.18889"), float("-0.282493"), float("-0.114794"), float("-0.19188"), float("-0.189031"), float("0.038257"), float("0.182447"), float("0.338656"), float("-0.185908"), float("0.282292"), float("-0.333012"), float("-0.215363"), float("0.198045"), float("0.10666"), float("0.0252068"), float("-0.186876"), float("0.33483"), float("0.246478"), float("-0.0593159"), float("-0.0243462"), float("0.271037"), float("-0.140778"), float("0.321753"), float("-0.0425401"), float("-0.0232625"), float("0.15166"), float("0.0834763"), float("0.0449372"), float("0.0559293"), float("0.351461"), float("0.0579684"), float("-0.0641627"), float("0.161772"), float("-0.141108"), float("-0.331302"), float("-0.0164107"), float("-0.119114"), float("0.0939414"), float("-0.254304"), float("-0.0356149"), float("-0.342954"), float("0.0774205"), float("0.050752"), float("-0.102107"), float("0.106109"), float("0.099873"), float("0.144805"), float("0.170335"), float("-0.202735"), float("-0.0852358"), float("0.1246"), float("0.0621777"), float("-0.35238"), float("0.0236915"), float("-0.0462948"), float("-0.0346923"), float("-0.0105365"), float("-0.0434867"), float("0.0888052"), float("-0.0298811"), float("0.278204"), float("0.248552"), float("0.0612633"), float("-0.296462"), float("-0.149241"), float("-0.00470171"), float("0.122117"), float("0.00352487"), float("-0.313724"), float("0.157657"), float("0.286055"), float("-0.179478"), float("0.165632"), float("0.200262"), float("-0.115712"), float("0.0705119"), float("-0.313889"), float("-0.285383"), float("0.272408"), float("-0.298484"), float("-0.133519"), float("-0.268654"), float("0.234496"), float("0.17597"), float("0.253718"), float("0.0419651"), float("0.324774"), float("0.332485"), float("-0.28583"), float("0.111105"), float("-0.194049"), float("-0.168264"), float("0.295419"), float("-0.0469957"), float("-0.323113"), float("0.219891"), float("-0.205654"), float("-0.136962"), float("0.0286586"), float("0.247516"), float("0.0750707"), float("-0.286282"), float("-0.0419273"), float("-0.167156"), float("0.338594"), float("0.18547"), float("-0.179599"), float("0.169526"), float("0.214873"), float("-0.340293"), float("-0.149433"), float("0.0603927"), float("-0.08848"), float("0.324922"), float("0.0252319"), float("-0.278085"), float("0.275881"), float("-0.322733"), float("-0.119825"), float("0.178038"), float("-0.117301"), float("0.346557"), float("0.122787"), float("-0.161778"), float("0.114787"), float("0.123685"), float("0.0642729"), float("0.262809"), float("-0.173838"), float("-0.114229"), float("-0.167517"), float("-0.324107"), float("-0.244136"), float("0.172947"), float("0.207802"), float("-0.277253")] + + +class PirProgram_example_input_tensor_meta_81239810641304043: + program_id = 692320333968606381 + input_name = "linear_1.b_0" + shape = [16] + data = [float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0")] + + +class PirProgram_example_input_tensor_meta_1643386415584303293: + program_id = 692320333968606381 + input_name = "middle_3" + shape = [16, 2, 32] + data = [float("0.723589"), float("0"), float("0"), float("6.28983"), float("0"), float("0"), float("0"), float("1.92776"), float("0"), float("0"), float("4.57823"), float("0"), float("3.68971"), float("0"), float("0.607042"), float("0.85038"), float("2.91316"), float("0"), float("0"), float("0.443321"), float("0"), float("6.09899"), float("2.37732"), float("0.4176"), float("2.09232"), float("0"), float("0"), float("5.69163"), float("0"), float("3.48354"), float("0"), float("0"), float("0.271179"), float("0"), float("0"), float("2.45632"), float("0"), float("0"), float("0"), float("0.763683"), float("0"), float("2.4179"), float("1.79379"), float("0"), float("1.44576"), float("0"), float("0.210515"), float("0.347592"), float("0"), float("2.32378"), float("0"), float("0.174633"), float("0"), float("2.37125"), float("0.922107"), float("0.187329"), float("0.81025"), float("0"), float("0"), float("2.25267"), float("0.632394"), float("1.35589"), float("0"), float("0"), float("0"), float("0.664025"), float("0.530738"), float("0"), float("0.843825"), float("1.08603"), float("0.16219"), float("0"), float("0.791713"), float("0"), float("0"), float("0.949193"), float("0"), float("0.600742"), float("0"), float("0"), float("0"), float("0"), float("0.897179"), float("0"), float("1.29137"), float("0"), float("0"), float("0"), float("0"), float("1.61578"), float("0"), float("0"), float("0"), float("0"), float("0.354193"), float("1.23875"), float("0.0676746"), float("0"), float("0"), float("0.646917"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.632321"), float("0.472122"), float("0"), float("0.376537"), float("0"), float("0.0776579"), float("0.0903725"), float("0"), float("0.60918"), float("0"), float("0.0524048"), float("0"), float("0.620071"), float("0.246761"), float("0"), float("0.201337"), float("0"), float("0"), float("0.575339"), float("0.163366"), float("0.340029"), float("0"), float("0"), float("0"), float("1.13115"), float("0.828709"), float("0"), float("1.45559"), float("1.82495"), float("0"), float("0"), float("1.64345"), float("0"), float("0"), float("1.69007"), float("0"), float("1.23017"), float("0"), float("0"), float("0"), float("0"), float("1.66897"), float("0"), float("2.01373"), float("0"), float("0"), float("0"), float("0"), float("2.80322"), float("0"), float("0"), float("0"), float("0"), float("0.433559"), float("2.17109"), float("0"), float("0.499996"), float("0.36502"), float("0"), float("0.64794"), float("0.807835"), float("0.134061"), float("0"), float("0.740266"), float("0"), float("0"), float("0.745333"), float("0"), float("0.544806"), float("0"), float("0"), float("0"), float("0"), float("0.743101"), float("0"), float("0.89026"), float("0"), float("0"), float("0"), float("0"), float("1.241"), float("0.699808"), float("0"), float("0"), float("0"), float("0.186857"), float("0.973017"), float("0"), float("1.19681"), float("0.724072"), float("0"), float("1.45466"), float("1.74084"), float("0.428011"), float("0"), float("0"), float("0"), float("0"), float("1.4446"), float("0"), float("1.06807"), float("0"), float("0"), float("0"), float("0"), float("1.58471"), float("0"), float("2.22805"), float("0"), float("0"), float("0"), float("0"), float("2.90281"), float("1.58111"), float("0"), float("0"), float("0"), float("0"), float("2.41101"), float("0"), float("0"), float("0"), float("0.225791"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.219702"), float("0.13923"), float("0"), float("0.0994024"), float("0"), float("0.0351282"), float("0.0333149"), float("0.115384"), float("0"), float("0"), float("0.00266587"), float("0"), float("0.210125"), float("0.0723203"), float("0.0102157"), float("0.0743553"), float("0"), float("0"), float("0.158375"), float("0.0459132"), float("0.116562"), float("0"), float("0"), float("0.167725"), float("0"), float("0"), float("1.51631"), float("0"), float("0"), float("0"), float("0.458128"), float("0"), float("1.47672"), float("1.08063"), float("0"), float("0.808897"), float("0"), float("0.18597"), float("0.232214"), float("0.753348"), float("0"), float("0"), float("0.00460923"), float("0"), float("0"), float("0"), float("0.0860425"), float("0.529424"), float("0"), float("0"), float("1.25368"), float("0"), float("0.882674"), float("0"), float("0"), float("0"), float("0"), float("0"), float("2.13523"), float("0"), float("0"), float("0"), float("0.66402"), float("0"), float("2.1026"), float("0"), float("0"), float("1.25584"), float("0"), float("0.19961"), float("0.295115"), float("0.982376"), float("2.01711"), float("0"), float("0.152203"), float("0"), float("2.06125"), float("0.806557"), float("0.147271"), float("0"), float("0"), float("0"), float("1.94095"), float("0.547275"), float("1.17657"), float("0"), float("0"), float("0"), float("2.0945"), float("0"), float("0"), float("2.71852"), float("3.39442"), float("0.543131"), float("0"), float("3.05501"), float("0"), float("0"), float("3.12772"), float("0"), float("2.25904"), float("0"), float("0"), float("0"), float("0"), float("3.11314"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("5.23011"), float("2.90551"), float("0"), float("0"), float("0"), float("0.81488"), float("4.04086"), float("0"), float("2.36868"), float("1.72581"), float("0"), float("3.04193"), float("3.81239"), float("0.621253"), float("0"), float("3.48355"), float("0"), float("0"), float("3.57841"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("4.22051"), float("0"), float("0"), float("0"), float("0"), float("5.8696"), float("3.2964"), float("0"), float("0"), float("0"), float("0.907758"), float("4.54258"), float("0"), float("0.315007"), float("0.206424"), float("0"), float("0"), float("0.376567"), float("0.122297"), float("0"), float("0.283101"), float("0"), float("0"), float("0.19817"), float("0"), float("0"), float("0.077143"), float("0"), float("0"), float("0"), float("0.317819"), float("0"), float("0.574079"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.342921"), float("0"), float("0"), float("0"), float("0.0767962"), float("0.601694"), float("0"), float("0.622385"), float("0"), float("0"), float("0.818888"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.948796"), float("0"), float("1.11377"), float("0"), float("0"), float("0"), float("0"), float("1.55717"), float("0.869019"), float("0"), float("0"), float("0"), float("0.238691"), float("1.22154"), float("0.0845757"), float("0"), float("0"), float("0.614002"), float("0"), float("0"), float("0"), float("0.202461"), float("0"), float("0.609987"), float("0.456371"), float("0"), float("0.391232"), float("0"), float("0"), float("0.0745103"), float("0.288329"), float("0"), float("0"), float("0.039889"), float("0"), float("0.59132"), float("0.238821"), float("0.0415231"), float("0.210348"), float("0"), float("0"), float("0.575056"), float("0.155218"), float("0.339351"), float("0"), float("0"), float("0"), float("1.40681"), float("1.02908"), float("0"), float("1.81357"), float("2.29121"), float("0"), float("0"), float("2.07469"), float("0"), float("0"), float("2.14309"), float("0"), float("1.5367"), float("0"), float("0"), float("0"), float("0"), float("2.11169"), float("0"), float("2.52412"), float("0"), float("0"), float("0"), float("0"), float("3.50848"), float("1.97313"), float("0"), float("0"), float("0"), float("0.5445"), float("2.72088"), float("0"), float("0"), float("0"), float("2.24401"), float("0"), float("0"), float("0"), float("0.686096"), float("0"), float("2.21794"), float("0"), float("0"), float("1.30706"), float("0"), float("0.208904"), float("0.309097"), float("0"), float("2.12223"), float("0"), float("0.137407"), float("0"), float("2.17288"), float("0.849765"), float("0.153441"), float("0.754491"), float("0"), float("0"), float("2.01347"), float("0.563138"), float("1.24616"), float("0"), float("0"), float("0.527303"), float("0"), float("0"), float("4.42907"), float("0"), float("0"), float("0"), float("1.36257"), float("0"), float("4.37391"), float("3.2329"), float("0"), float("2.59558"), float("0"), float("0"), float("0.592568"), float("2.05092"), float("4.16704"), float("0"), float("0.312967"), float("0"), float("4.28341"), float("1.67541"), float("0.30598"), float("0"), float("0"), float("0"), float("4.02188"), float("1.12964"), float("2.46205"), float("0"), float("0"), float("0"), float("0"), float("0.2639"), float("0"), float("0.490358"), float("0.634138"), float("0.112012"), float("0"), float("0.54594"), float("0"), float("0"), float("0.577505"), float("0"), float("0.391071"), float("0"), float("0"), float("0"), float("0"), float("0.595201"), float("0"), float("0.710756"), float("0"), float("0"), float("0"), float("0"), float("0.963"), float("0.534581"), float("0"), float("0"), float("0"), float("0"), float("0.746982"), float("0"), float("0.0715721"), float("0.0618026"), float("0"), float("0.102145"), float("0.126783"), float("0.0210766"), float("0"), float("0"), float("0"), float("0"), float("0.0819103"), float("0"), float("0"), float("0.00295741"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.163064"), float("0"), float("0"), float("0"), float("0"), float("0.196886"), float("0"), float("0"), float("0"), float("0"), float("0.060482"), float("0.162623"), float("0.403443"), float("0"), float("0"), float("3.50631"), float("0"), float("0"), float("0"), float("1.05008"), float("0"), float("0"), float("2.5659"), float("0"), float("2.04395"), float("0"), float("0.33452"), float("0"), float("1.64144"), float("3.27211"), float("0"), float("0.21803"), float("0"), float("3.39128"), float("0"), float("0.223745"), float("1.16294"), float("0"), float("0"), float("3.1488"), float("0.884625"), float("1.95163"), float("0"), float("0"), float("0.223519"), float("0"), float("0"), float("1.90329"), float("0"), float("0"), float("0"), float("0.586335"), float("0"), float("1.86716"), float("1.38074"), float("0"), float("0"), float("0"), float("0.179526"), float("0.256382"), float("0.877256"), float("1.78683"), float("0"), float("0.128199"), float("0"), float("0"), float("0"), float("0"), float("0.635889"), float("0"), float("0"), float("1.72533"), float("0.482998"), float("1.05037"), float("0"), float("0"), float("0.262392"), float("0"), float("0"), float("2.26296"), float("0"), float("0"), float("0"), float("0.703865"), float("0"), float("2.25157"), float("1.65439"), float("0"), float("0"), float("0"), float("0.222988"), float("0.301277"), float("1.05197"), float("2.14329"), float("0"), float("0.16106"), float("0"), float("2.20785"), float("0.855672"), float("0.146199"), float("0.753635"), float("0"), float("0"), float("2.05749"), float("0.576322"), float("1.25052"), float("0"), float("0"), float("0.117819"), float("0"), float("0"), float("1.01242"), float("0"), float("0"), float("0"), float("0.314459"), float("0"), float("1.00417"), float("0.740845"), float("0"), float("0.599216"), float("0"), float("0"), float("0.132811"), float("0.470765"), float("0"), float("0"), float("0"), float("0"), float("0.987654"), float("0"), float("0"), float("0.342916"), float("0"), float("0"), float("0.926627"), float("0.257738"), float("0.562952"), float("0"), float("0"), float("0.26498"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.729735"), float("0"), float("2.27784"), float("1.6737"), float("0"), float("0"), float("0"), float("0"), float("0.315342"), float("1.05739"), float("2.1737"), float("0"), float("0.16906"), float("0"), float("2.21541"), float("0.86"), float("0.157351"), float("0.758679"), float("0"), float("0"), float("2.09938"), float("0.591763"), float("1.26801"), float("0"), float("0"), float("0"), float("0"), float("0.828738"), float("0"), float("1.46613"), float("1.81584"), float("0.311517"), float("0"), float("0"), float("0"), float("0"), float("1.69295"), float("0"), float("1.22507"), float("0"), float("0"), float("0"), float("0"), float("1.70447"), float("0"), float("2.02645"), float("0"), float("0"), float("0"), float("0"), float("2.83993"), float("1.60285"), float("0"), float("0"), float("0"), float("0.417853"), float("2.20486"), float("0.324522"), float("0"), float("0"), float("2.62623"), float("0"), float("0"), float("0"), float("0.815165"), float("0"), float("2.57824"), float("1.88527"), float("0"), float("1.4626"), float("0"), float("0.302452"), float("0.349664"), float("1.31955"), float("2.45426"), float("0"), float("0.0641848"), float("0"), float("2.53626"), float("1.00942"), float("0.126232"), float("0.896666"), float("0"), float("0"), float("2.23374"), float("0.599363"), float("1.47756"), float("0"), float("0"), float("0"), float("0.302239"), float("0.212177"), float("0"), float("0.378777"), float("0.468387"), float("0.0813494"), float("0"), float("0.433903"), float("0"), float("0"), float("0.438357"), float("0"), float("0.31956"), float("0"), float("0"), float("0"), float("0"), float("0.426783"), float("0"), float("0.524977"), float("0"), float("0"), float("0"), float("0"), float("0.729943"), float("0.404518"), float("0"), float("0"), float("0"), float("0"), float("0.559602"), float("0.0706485"), float("0"), float("0"), float("0.625332"), float("0"), float("0"), float("0"), float("0.199803"), float("0"), float("0.607633"), float("0.462029"), float("0"), float("0.381851"), float("0"), float("0.0538131"), float("0.0799106"), float("0"), float("0.596368"), float("0"), float("0.050653"), float("0"), float("0.599458"), float("0.238125"), float("0.045"), float("0.204004"), float("0"), float("0"), float("0.58834"), float("0"), float("0.342713"), float("0"), float("0"), float("0"), float("0.235136"), float("0.167431"), float("0"), float("0.303585"), float("0.377351"), float("0.0631602"), float("0"), float("0.344763"), float("0"), float("0"), float("0"), float("0"), float("0.255787"), float("0"), float("0"), float("0"), float("0"), float("0.352881"), float("0"), float("0.412486"), float("0"), float("0"), float("0"), float("0"), float("0.580145"), float("0"), float("0"), float("0"), float("0"), float("0.0895882"), float("0.447694"), float("0.0637865"), float("0"), float("0"), float("0.599056"), float("0"), float("0"), float("0"), float("0.175942"), float("0"), float("0.592536"), float("0.436414"), float("0"), float("0.345927"), float("0"), float("0.0527949"), float("0"), float("0"), float("0.560908"), float("0"), float("0"), float("0"), float("0.578455"), float("0.221061"), float("0.0428592"), float("0"), float("0"), float("0"), float("0.537149"), float("0"), float("0.333177"), float("0"), float("0"), float("0.0157932"), float("0"), float("0"), float("0.141245"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.142142"), float("0.0979528"), float("0"), float("0"), float("0"), float("0.0137536"), float("0"), float("0.0659162"), float("0.138318"), float("0"), float("0.0213317"), float("0"), float("0.138244"), float("0.0449504"), float("0"), float("0.05323"), float("0"), float("0"), float("0.136466"), float("0.0393414"), float("0.0763775"), float("0"), float("0")] + + +class PirProgram_example_input_tensor_meta_6890028235555013676: + program_id = 692320333968606381 + input_name = "middle_1" + shape = [16, 2, 32] + data = [float("0.578871"), float("0"), float("0"), float("5.03186"), float("0"), float("0"), float("0"), float("1.5422"), float("0"), float("4.98103"), float("3.66259"), float("0"), float("2.95177"), float("0"), float("0.485634"), float("0.680304"), float("2.33053"), float("4.73965"), float("0"), float("0.354657"), float("0"), float("4.87919"), float("1.90186"), float("0.33408"), float("1.67386"), float("0"), float("0"), float("4.5533"), float("1.28468"), float("2.78683"), float("0"), float("0"), float("0.216943"), float("0"), float("0"), float("1.96505"), float("0"), float("0"), float("0"), float("0.610946"), float("0"), float("1.93432"), float("1.43503"), float("0"), float("1.1566"), float("0"), float("0.168412"), float("0.278073"), float("0.895773"), float("1.85903"), float("0"), float("0.139707"), float("0"), float("1.897"), float("0.737685"), float("0.149863"), float("0.6482"), float("0"), float("0"), float("1.80214"), float("0.505915"), float("1.08471"), float("0"), float("0"), float("0"), float("0.53122"), float("0.42459"), float("0"), float("0.67506"), float("0.868822"), float("0.129752"), float("0"), float("0.633371"), float("0"), float("0"), float("0.759354"), float("0"), float("0.480594"), float("0"), float("0"), float("0"), float("0"), float("0.717743"), float("0"), float("1.0331"), float("0"), float("0"), float("0"), float("0"), float("1.29263"), float("0.697283"), float("0"), float("0"), float("0"), float("0.283354"), float("0.991004"), float("0.0541397"), float("0"), float("0"), float("0.517533"), float("0"), float("0"), float("0"), float("0.16042"), float("0"), float("0.505857"), float("0.377697"), float("0"), float("0.30123"), float("0"), float("0.0621263"), float("0.072298"), float("0.248242"), float("0.487344"), float("0"), float("0.0419238"), float("0"), float("0.496056"), float("0.197409"), float("0.0200742"), float("0.16107"), float("0"), float("0"), float("0.460271"), float("0.130693"), float("0.272023"), float("0"), float("0"), float("0"), float("0.90492"), float("0.662967"), float("0"), float("1.16447"), float("1.45996"), float("0.227835"), float("0"), float("1.31476"), float("0"), float("0"), float("1.35205"), float("0"), float("0.984135"), float("0"), float("0"), float("0"), float("0"), float("1.33518"), float("0"), float("1.61099"), float("0"), float("0"), float("0"), float("0"), float("2.24257"), float("1.25536"), float("0"), float("0"), float("0"), float("0.346848"), float("1.73687"), float("0"), float("0.399997"), float("0.292016"), float("0"), float("0.518352"), float("0.646268"), float("0.107249"), float("0"), float("0.592213"), float("0"), float("0"), float("0.596267"), float("0"), float("0.435845"), float("0"), float("0"), float("0"), float("0"), float("0.594481"), float("0"), float("0.712208"), float("0"), float("0"), float("0"), float("0"), float("0.992802"), float("0.559846"), float("0"), float("0"), float("0"), float("0.149486"), float("0.778414"), float("0"), float("0.957444"), float("0.579258"), float("0"), float("1.16373"), float("1.39267"), float("0.342409"), float("0"), float("1.2634"), float("0"), float("0"), float("1.15568"), float("0"), float("0.854454"), float("0"), float("0"), float("0"), float("0"), float("1.26777"), float("0"), float("1.78244"), float("0"), float("0"), float("0"), float("0"), float("2.32225"), float("1.26489"), float("0"), float("0"), float("0"), float("0.358627"), float("1.9288"), float("0.0289001"), float("0"), float("0"), float("0.180633"), float("0"), float("0"), float("0"), float("0.0530196"), float("0"), float("0.175761"), float("0.111384"), float("0"), float("0.0795219"), float("0"), float("0.0281026"), float("0.0266519"), float("0.0923073"), float("0.155116"), float("0"), float("0.0021327"), float("0"), float("0.1681"), float("0.0578563"), float("0.00817253"), float("0.0594842"), float("0"), float("0"), float("0.1267"), float("0.0367306"), float("0.0932499"), float("0"), float("0"), float("0.13418"), float("0"), float("0"), float("1.21305"), float("0"), float("0"), float("0"), float("0.366503"), float("0"), float("1.18137"), float("0.864508"), float("0"), float("0.647117"), float("0"), float("0.148776"), float("0.185771"), float("0.602678"), float("1.12672"), float("0"), float("0.00368738"), float("0"), float("1.17025"), float("0.461667"), float("0.068834"), float("0.423539"), float("0"), float("0"), float("1.00295"), float("0.276469"), float("0.706139"), float("0"), float("0"), float("0.196585"), float("0"), float("0"), float("1.70819"), float("0"), float("0"), float("0"), float("0.531216"), float("0"), float("1.68208"), float("1.2358"), float("0"), float("1.00467"), float("0"), float("0.159688"), float("0.236092"), float("0.785901"), float("1.61369"), float("0"), float("0.121763"), float("0"), float("1.649"), float("0.645246"), float("0.117817"), float("0.569933"), float("0"), float("0"), float("1.55276"), float("0.43782"), float("0.941253"), float("0"), float("0"), float("0"), float("1.6756"), float("1.21723"), float("0"), float("2.17482"), float("2.71554"), float("0.434505"), float("0"), float("2.44401"), float("0"), float("0"), float("2.50218"), float("0"), float("1.80723"), float("0"), float("0"), float("0"), float("0"), float("2.49051"), float("0"), float("3.00915"), float("0"), float("0"), float("0"), float("0"), float("4.18409"), float("2.32441"), float("0"), float("0"), float("0"), float("0.651904"), float("3.23269"), float("0"), float("1.89495"), float("1.38065"), float("0"), float("2.43354"), float("3.04991"), float("0.497002"), float("0"), float("2.78684"), float("0"), float("0"), float("2.86273"), float("0"), float("2.06031"), float("0"), float("0"), float("0"), float("0"), float("2.82139"), float("0"), float("3.37641"), float("0"), float("0"), float("0"), float("0"), float("4.69568"), float("2.63712"), float("0"), float("0"), float("0"), float("0.726207"), float("3.63407"), float("0"), float("0.252005"), float("0.165139"), float("0"), float("0.234757"), float("0.301253"), float("0.0978375"), float("0"), float("0.226481"), float("0"), float("0"), float("0.158536"), float("0"), float("0.146847"), float("0.0617144"), float("0"), float("0.0277497"), float("0"), float("0.254255"), float("0"), float("0.459263"), float("0"), float("0"), float("0"), float("0"), float("0.519019"), float("0.274337"), float("0"), float("0"), float("0"), float("0.061437"), float("0.481355"), float("0"), float("0.497908"), float("0.365889"), float("0"), float("0.65511"), float("0.814613"), float("0.128348"), float("0"), float("0.746606"), float("0"), float("0"), float("0.767712"), float("0"), float("0.555277"), float("0"), float("0"), float("0"), float("0"), float("0.759037"), float("0"), float("0.891017"), float("0"), float("0"), float("0"), float("0"), float("1.24574"), float("0.695215"), float("0"), float("0"), float("0"), float("0.190953"), float("0.977231"), float("0.0676606"), float("0"), float("0"), float("0.491202"), float("0"), float("0"), float("0"), float("0.161969"), float("0"), float("0.48799"), float("0.365097"), float("0"), float("0.312986"), float("0"), float("0.0529859"), float("0.0596082"), float("0.230663"), float("0.474683"), float("0"), float("0.0319112"), float("0"), float("0.473056"), float("0.191057"), float("0.0332185"), float("0.168278"), float("0"), float("0"), float("0.460044"), float("0.124175"), float("0.27148"), float("0"), float("0"), float("0"), float("1.12545"), float("0.823264"), float("0"), float("1.45085"), float("1.83297"), float("0.288033"), float("0"), float("1.65976"), float("0"), float("0"), float("1.71447"), float("0"), float("1.22936"), float("0"), float("0"), float("0"), float("0"), float("1.68935"), float("0"), float("2.0193"), float("0"), float("0"), float("0"), float("0"), float("2.80679"), float("1.5785"), float("0"), float("0"), float("0"), float("0.4356"), float("2.17671"), float("0.200711"), float("0"), float("0"), float("1.79521"), float("0"), float("0"), float("0"), float("0.548877"), float("0"), float("1.77435"), float("1.30207"), float("0"), float("1.04565"), float("0"), float("0.167124"), float("0.247277"), float("0.834098"), float("1.69778"), float("0"), float("0.109925"), float("0"), float("1.7383"), float("0.679812"), float("0.122753"), float("0.603593"), float("0"), float("0"), float("1.61078"), float("0.45051"), float("0.996926"), float("0"), float("0"), float("0.421842"), float("0"), float("0"), float("3.54325"), float("0"), float("0"), float("0"), float("1.09006"), float("0"), float("3.49913"), float("2.58632"), float("0"), float("2.07647"), float("0"), float("0.338603"), float("0.474054"), float("1.64073"), float("3.33363"), float("0"), float("0.250374"), float("0"), float("3.42673"), float("1.34033"), float("0.244784"), float("1.18908"), float("0"), float("0"), float("3.2175"), float("0.903709"), float("1.96964"), float("0"), float("0"), float("0"), float("0.314209"), float("0.21112"), float("0"), float("0.392287"), float("0.50731"), float("0.0896099"), float("0"), float("0.436752"), float("0"), float("0"), float("0.462004"), float("0"), float("0.312857"), float("0"), float("0"), float("0"), float("0"), float("0.47616"), float("0"), float("0.568605"), float("0"), float("0"), float("0"), float("0"), float("0.7704"), float("0.427665"), float("0"), float("0"), float("0"), float("0.14075"), float("0.597586"), float("0"), float("0.0572577"), float("0.0494421"), float("0"), float("0.081716"), float("0.101426"), float("0.0168613"), float("0"), float("0.0787916"), float("0"), float("0"), float("0.0655282"), float("0"), float("0.0364558"), float("0.00236592"), float("0"), float("0"), float("0"), float("0.0653458"), float("0"), float("0.130451"), float("0"), float("0"), float("0"), float("0"), float("0.157509"), float("0.0695541"), float("0"), float("0"), float("0"), float("0.0483856"), float("0.130099"), float("0.322755"), float("0"), float("0"), float("2.80505"), float("0"), float("0"), float("0"), float("0.840062"), float("0"), float("2.75153"), float("2.05272"), float("0"), float("1.63516"), float("0"), float("0.267616"), float("0.353298"), float("1.31315"), float("2.61768"), float("0"), float("0.174424"), float("0"), float("2.71303"), float("1.05055"), float("0.178996"), float("0.930353"), float("0"), float("0"), float("2.51904"), float("0.7077"), float("1.5613"), float("0"), float("0"), float("0.178815"), float("0"), float("0"), float("1.52263"), float("0"), float("0"), float("0"), float("0.469068"), float("0"), float("1.49373"), float("1.10459"), float("0"), float("0.891864"), float("0"), float("0.143621"), float("0.205105"), float("0.701805"), float("1.42947"), float("0"), float("0.102559"), float("0"), float("1.46317"), float("0.5694"), float("0.108553"), float("0.508712"), float("0"), float("0"), float("1.38026"), float("0.386398"), float("0.840296"), float("0"), float("0"), float("0.209913"), float("0"), float("0"), float("1.81037"), float("0"), float("0"), float("0"), float("0.563092"), float("0"), float("1.80125"), float("1.32351"), float("0"), float("1.0774"), float("0"), float("0.178391"), float("0.241022"), float("0.841577"), float("1.71463"), float("0"), float("0.128848"), float("0"), float("1.76628"), float("0.684538"), float("0.116959"), float("0.602908"), float("0"), float("0"), float("1.64599"), float("0.461058"), float("1.00041"), float("0"), float("0"), float("0.0942551"), float("0"), float("0"), float("0.809935"), float("0"), float("0"), float("0"), float("0.251567"), float("0"), float("0.803332"), float("0.592676"), float("0"), float("0.479373"), float("0"), float("0.0745486"), float("0.106249"), float("0.376612"), float("0.764936"), float("0"), float("0.0548324"), float("0"), float("0.790123"), float("0.308984"), float("0.0567979"), float("0.274333"), float("0"), float("0"), float("0.741301"), float("0.20619"), float("0.450362"), float("0"), float("0"), float("0.211984"), float("0"), float("0"), float("1.83722"), float("0"), float("0"), float("0"), float("0.583788"), float("0"), float("1.82228"), float("1.33896"), float("0"), float("1.0933"), float("0"), float("0.180633"), float("0.252274"), float("0.845911"), float("1.73896"), float("0"), float("0.135248"), float("0"), float("1.77233"), float("0.688"), float("0.125881"), float("0.606944"), float("0"), float("0"), float("1.67951"), float("0.473411"), float("1.01441"), float("0"), float("0"), float("0"), float("0.925363"), float("0.66299"), float("0"), float("1.1729"), float("1.45267"), float("0.249214"), float("0"), float("1.34885"), float("0"), float("0"), float("1.35436"), float("0"), float("0.980055"), float("0"), float("0"), float("0"), float("0"), float("1.36358"), float("0"), float("1.62116"), float("0"), float("0"), float("0"), float("0"), float("2.27194"), float("1.28228"), float("0"), float("0"), float("0"), float("0.334282"), float("1.76389"), float("0.259617"), float("0"), float("0"), float("2.10098"), float("0"), float("0"), float("0"), float("0.652132"), float("0"), float("2.06259"), float("1.50822"), float("0"), float("1.17008"), float("0"), float("0.241962"), float("0.279731"), float("1.05564"), float("1.96341"), float("0"), float("0.0513479"), float("0"), float("2.02901"), float("0.807535"), float("0.100986"), float("0.717333"), float("0"), float("0"), float("1.78699"), float("0.47949"), float("1.18205"), float("0"), float("0"), float("0"), float("0.241791"), float("0.169742"), float("0"), float("0.303022"), float("0.374709"), float("0.0650795"), float("0"), float("0.347122"), float("0"), float("0"), float("0.350685"), float("0"), float("0.255648"), float("0"), float("0"), float("0"), float("0"), float("0.341427"), float("0"), float("0.419982"), float("0"), float("0"), float("0"), float("0"), float("0.583954"), float("0.323614"), float("0"), float("0"), float("0"), float("0.0932906"), float("0.447681"), float("0.0565188"), float("0"), float("0"), float("0.500265"), float("0"), float("0"), float("0"), float("0.159843"), float("0"), float("0.486106"), float("0.369623"), float("0"), float("0.305481"), float("0"), float("0.0430505"), float("0.0639285"), float("0.224091"), float("0.477095"), float("0"), float("0.0405224"), float("0"), float("0.479566"), float("0.1905"), float("0.036"), float("0.163203"), float("0"), float("0"), float("0.470672"), float("0.133333"), float("0.27417"), float("0"), float("0"), float("0"), float("0.188109"), float("0.133944"), float("0"), float("0.242868"), float("0.301881"), float("0.0505282"), float("0"), float("0.27581"), float("0"), float("0"), float("0.284024"), float("0"), float("0.204629"), float("0"), float("0"), float("0"), float("0"), float("0.282305"), float("0"), float("0.329989"), float("0"), float("0"), float("0"), float("0"), float("0.464116"), float("0.26113"), float("0"), float("0"), float("0"), float("0.0716705"), float("0.358155"), float("0.0510292"), float("0"), float("0"), float("0.479245"), float("0"), float("0"), float("0"), float("0.140753"), float("0"), float("0.474029"), float("0.349131"), float("0"), float("0.276741"), float("0"), float("0.0422359"), float("0.0659246"), float("0.221283"), float("0.448727"), float("0"), float("0.0395124"), float("0"), float("0.462764"), float("0.176849"), float("0.0342874"), float("0.155229"), float("0"), float("0"), float("0.429719"), float("0.123715"), float("0.266542"), float("0"), float("0"), float("0.0126346"), float("0"), float("0"), float("0.112996"), float("0"), float("0"), float("0"), float("0.0360254"), float("0"), float("0.113713"), float("0.0783622"), float("0"), float("0.07168"), float("0"), float("0.0110029"), float("0.0184994"), float("0.052733"), float("0.110654"), float("0"), float("0.0170654"), float("0"), float("0.110596"), float("0.0359603"), float("0.0123128"), float("0.042584"), float("0"), float("0"), float("0.109173"), float("0.0314731"), float("0.061102"), float("0"), float("0")] + + +class PirProgram_example_input_tensor_meta_3560422295928009713: + program_id = 692320333968606381 + input_name = "linear_0.w_0" + shape = [96, 32] + data = None + + +class PirProgram_example_input_tensor_meta_6660463665478598428: + program_id = 692320333968606381 + input_name = "linear_0.b_0" + shape = [32] + data = [float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0")] + + +class PirProgram_example_input_tensor_meta_4787673921159221127: + program_id = 692320333968606381 + input_name = "args_0" + shape = [16, 2, 96] + data = None + + +class PirProgram_example_input_tensor_meta_4218138602689122281: + program_id = 5426476789613677471 + input_name = "linear_1.b_0" + shape = [16] + data = None + + +class PirProgram_example_input_tensor_meta_8039520472926701889: + program_id = 5426476789613677471 + input_name = "linear_1.w_0" + shape = [32, 16] + data = None + + +class PirProgram_example_input_tensor_meta_3987890563716532376: + program_id = 5426476789613677471 + input_name = "linear_0.b_0" + shape = [32] + data = None + + +class PirProgram_example_input_tensor_meta_100547460941153657: + program_id = 5426476789613677471 + input_name = "linear_0.w_0" + shape = [96, 32] + data = None + + +class PirProgram_example_input_tensor_meta_3242029616568894092: + program_id = 5426476789613677471 + input_name = "args_0" + shape = [2, 2, 96] + data = [float("-1.15597"), float("-1.1523"), float("-1.15171"), float("-1.14702"), float("-1.14445"), float("-1.14483"), float("-1.14359"), float("-1.14577"), float("-1.14909"), float("-1.14868"), float("-1.14244"), float("-1.14465"), float("-1.13813"), float("-1.14016"), float("-1.13834"), float("-1.12726"), float("-1.11024"), float("-1.10905"), float("-1.10871"), float("-1.10095"), float("-1.09833"), float("-1.09615"), float("-1.09124"), float("-1.09103"), float("-1.09178"), float("-1.09125"), float("-1.09142"), float("-1.09758"), float("-1.09483"), float("-1.09004"), float("-1.0888"), float("-1.08218"), float("-1.08141"), float("-1.07451"), float("-1.07159"), float("-1.07065"), float("-1.06963"), float("-1.06884"), float("-1.06653"), float("-1.06595"), float("-1.06269"), float("-1.06049"), float("-1.05744"), float("-1.06124"), float("-1.05722"), float("-1.05479"), float("-1.05422"), float("-1.05193"), float("-1.0509"), float("-1.05028"), float("-1.04654"), float("-1.045"), float("-1.046"), float("-1.04667"), float("-1.04904"), float("-1.04042"), float("-1.04003"), float("-1.04359"), float("-1.03823"), float("-1.03557"), float("-1.03127"), float("-1.03583"), float("-1.0344"), float("-1.03144"), float("-1.02952"), float("-1.02593"), float("-1.02701"), float("-1.02957"), float("-1.02561"), float("-1.02673"), float("-1.0235"), float("-1.02171"), float("-1.02023"), float("-1.01943"), float("-1.01996"), float("-1.0124"), float("-1.01049"), float("-1.01181"), float("-1.01104"), float("-1.0073"), float("-1.0079"), float("-1.00004"), float("-0.996691"), float("-0.993982"), float("-0.991675"), float("-0.990566"), float("-0.988743"), float("-0.988268"), float("-0.984146"), float("-0.982912"), float("-0.987096"), float("-0.984222"), float("-0.963696"), float("-0.952533"), float("-0.954977"), float("-0.950661"), float("-0.399858"), float("-0.398499"), float("-0.392917"), float("-0.396343"), float("-0.391605"), float("-0.386859"), float("-0.391783"), float("-0.393174"), float("-0.396282"), float("-0.395804"), float("-0.396976"), float("-0.398565"), float("-0.395854"), float("-0.393964"), float("-0.393872"), float("-0.39339"), float("-0.394667"), float("-0.394263"), float("-0.395912"), float("-0.39461"), float("-0.392164"), float("-0.391116"), float("-0.393774"), float("-0.391869"), float("-0.392982"), float("-0.390163"), float("-0.392436"), float("-0.393987"), float("-0.394744"), float("-0.394476"), float("-0.391628"), float("-0.389037"), float("-0.387299"), float("-0.382867"), float("-0.392128"), float("-0.39309"), float("-0.3956"), float("-0.392959"), float("-0.386446"), float("-0.381103"), float("-0.380557"), float("-0.385733"), float("-0.382358"), float("-0.382535"), float("-0.388842"), float("-0.3847"), float("-0.386946"), float("-0.391374"), float("-0.390959"), float("-0.386347"), float("-0.391026"), float("-0.390428"), float("-0.386892"), float("-0.388793"), float("-0.392527"), float("-0.390458"), float("-0.389178"), float("-0.38852"), float("-0.389764"), float("-0.389545"), float("-0.384279"), float("-0.383312"), float("-0.379087"), float("-0.380502"), float("-0.381453"), float("-0.38131"), float("-0.386445"), float("-0.387503"), float("-0.385849"), float("-0.391521"), float("-0.39044"), float("-0.391326"), float("-0.390687"), float("-0.389549"), float("-0.387867"), float("-0.384638"), float("-0.388114"), float("-0.389713"), float("-0.386914"), float("-0.383812"), float("-0.38576"), float("-0.382273"), float("-0.381319"), float("-0.383497"), float("-0.378321"), float("-0.382726"), float("-0.3803"), float("-0.38556"), float("-0.378433"), float("-0.378988"), float("-0.380473"), float("-0.383853"), float("-0.381981"), float("-0.383138"), float("-0.38374"), float("-0.379838"), float("-0.630672"), float("-0.632913"), float("-0.633336"), float("-0.636795"), float("-0.637253"), float("-0.635754"), float("-0.636042"), float("-0.637148"), float("-0.640553"), float("-0.641178"), float("-0.639793"), float("-0.642557"), float("-0.642459"), float("-0.647011"), float("-0.646651"), float("-0.643922"), float("-0.64385"), float("-0.643057"), float("-0.642857"), float("-0.643463"), float("-0.644947"), float("-0.645577"), float("-0.645093"), float("-0.645523"), float("-0.645906"), float("-0.647578"), float("-0.646427"), float("-0.646111"), float("-0.649196"), float("-0.65109"), float("-0.649286"), float("-0.648321"), float("-0.64727"), float("-0.649662"), float("-0.646437"), float("-0.650501"), float("-0.648313"), float("-0.652139"), float("-0.652555"), float("-0.652121"), float("-0.651807"), float("-0.652984"), float("-0.652824"), float("-0.654689"), float("-0.653612"), float("-0.653592"), float("-0.653823"), float("-0.653214"), float("-0.653469"), float("-0.651772"), float("-0.650067"), float("-0.65063"), float("-0.652344"), float("-0.653487"), float("-0.648193"), float("-0.650301"), float("-0.649187"), float("-0.650017"), float("-0.648082"), float("-0.648181"), float("-0.647695"), float("-0.648911"), float("-0.649349"), float("-0.651401"), float("-0.654608"), float("-0.651767"), float("-0.652217"), float("-0.646641"), float("-0.649977"), float("-0.651891"), float("-0.653053"), float("-0.65444"), float("-0.653964"), float("-0.657094"), float("-0.657665"), float("-0.658948"), float("-0.662083"), float("-0.663417"), float("-0.665791"), float("-0.663225"), float("-0.659105"), float("-0.659393"), float("-0.659387"), float("-0.661882"), float("-0.663708"), float("-0.663021"), float("-0.66179"), float("-0.652634"), float("-0.652383"), float("-0.651621"), float("-0.65448"), float("-0.654242"), float("-0.656716"), float("-0.657377"), float("-0.65797"), float("-0.655658"), float("-0.00436456"), float("-0.00475825"), float("-0.00663337"), float("-0.00523404"), float("-0.00562955"), float("-0.00631295"), float("-0.00424731"), float("-0.00233632"), float("0.00070127"), float("0.00211113"), float("0.000545739"), float("0.00642778"), float("0.0038039"), float("0.000308868"), float("-0.00136501"), float("-0.00140553"), float("-0.00126356"), float("-0.00223812"), float("-0.00588529"), float("-0.00441997"), float("-0.00579612"), float("-0.00591071"), float("-0.00374736"), float("-0.0039841"), float("0.00174819"), float("0.00298575"), float("0.00173817"), float("0.00255231"), float("0.000368191"), float("0.000318813"), float("-0.00178881"), float("-0.00274283"), float("0.000628811"), float("-0.00182678"), float("-7.61417e-05"), float("-0.000906536"), float("-0.00333968"), float("-0.00408479"), float("-0.00325562"), float("-0.00334317"), float("-0.00397552"), float("-0.00359293"), float("-0.0032651"), float("0.00372263"), float("0.00179851"), float("0.00113003"), float("0.00107969"), float("0.000824071"), float("0.00230663"), float("0.00521457"), float("0.00676399"), float("0.00760787"), float("0.00433249"), float("0.0035608"), float("0.000554484"), float("0.00362084"), float("-0.00145232"), float("-0.00354868"), float("-0.00290337"), float("-0.00352887"), float("-0.00181459"), float("-0.00102524"), float("-0.00380784"), float("-0.00352733"), float("-0.00369791"), float("-0.00225879"), float("-0.00303406"), float("0.00243163"), float("0.00128276"), float("0.000701387"), float("0.00297557"), float("0.00318256"), float("0.00367343"), float("0.00317448"), float("0.00563999"), float("0.00400331"), float("-0.00251565"), float("0.00151779"), float("-0.00175092"), float("-0.00389494"), float("-0.0005398"), float("-0.000404088"), float("-0.00223803"), float("-0.00250093"), float("-0.0025293"), float("-0.00180433"), float("0.000374747"), float("0.00324137"), float("0.00465033"), float("0.00285781"), float("0.0061246"), float("0.00211838"), float("0.000508647"), float("-0.00179661"), float("-0.00133032"), float("-0.000782828")] + + +class PirProgram_example_input_tensor_meta_3980394430064143036: + program_id = 7028096434133672773 + input_name = "linear_3.b_0" + shape = [96] + data = None + + +class PirProgram_example_input_tensor_meta_3054776355284465051: + program_id = 7028096434133672773 + input_name = "linear_3.w_0" + shape = [32, 96] + data = None + + +class PirProgram_example_input_tensor_meta_6125757611404651348: + program_id = 7028096434133672773 + input_name = "linear_2.b_0" + shape = [32] + data = None + + +class PirProgram_example_input_tensor_meta_5534661246454474549: + program_id = 7028096434133672773 + input_name = "linear_2.w_0" + shape = [16, 32] + data = None + + +class PirProgram_example_input_tensor_meta_4614797390250714793: + program_id = 7028096434133672773 + input_name = "args_0" + shape = [2, 2, 16] + data = [float("0.63837"), float("-0.484176"), float("1.37562"), float("2.11047"), float("0.00551162"), float("2.62531"), float("-0.372032"), float("2.03205"), float("3.52888"), float("0.886358"), float("-0.225955"), float("-0.324597"), float("0.802664"), float("-2.20063"), float("1.28291"), float("-5.03928"), float("0.139641"), float("0.108348"), float("0.579502"), float("0.525852"), float("0.11746"), float("0.674907"), float("-0.286067"), float("0.566479"), float("0.799654"), float("0.31314"), float("0.146439"), float("-0.275929"), float("0.237963"), float("-0.39452"), float("0.0206265"), float("-1.22811"), float("-0.356625"), float("0.240933"), float("0.447273"), float("0.99085"), float("0.022031"), float("1.57203"), float("-0.341862"), float("0.827176"), float("1.06617"), float("0.33467"), float("0.0470244"), float("-0.0600742"), float("0.227221"), float("-0.507455"), float("-0.00214139"), float("-1.91961"), float("0.133666"), float("0.00946639"), float("0.0763022"), float("-0.017242"), float("0.0861289"), float("0.00317063"), float("0.054149"), float("0.141779"), float("-0.0106162"), float("0.142801"), float("-0.0334722"), float("-0.301938"), float("0.0549173"), float("0.0499354"), float("-0.144601"), float("0.111016")] + + +class PirProgram_example_input_tensor_meta_3371184329773483952: + program_id = 3000443918524221177 + input_name = "middle_5" + shape = [2, 2, 96] + data = None + + +class PirProgram_example_input_tensor_meta_1190819320963532881: + program_id = 3000443918524221177 + input_name = "middle_4" + shape = [2, 2, 32] + data = None + + +class PirProgram_example_input_tensor_meta_6657326147953639809: + program_id = 3000443918524221177 + input_name = "output_grad_0" + shape = [2, 2, 96] + data = [float("0.000443692"), float("0.000333813"), float("0.000337164"), float("0.000376996"), float("0.00031588"), float("0.000140307"), float("0.000155425"), float("0.000254817"), float("0.000186923"), float("0.000238763"), float("0.000174804"), float("0.000121806"), float("0.000140712"), float("0.000191536"), float("0.000127365"), float("-4.77346e-05"), float("-6.26668e-05"), float("-2.02892e-05"), float("1.12926e-05"), float("-9.52116e-05"), float("-8.16081e-05"), float("-3.6167e-05"), float("-3.85133e-06"), float("-2.60981e-05"), float("-4.54684e-05"), float("-4.69554e-05"), float("-0.000101383"), float("-5.73186e-05"), float("-4.71144e-05"), float("-0.000123006"), float("-0.000133709"), float("-0.000119086"), float("-0.000149415"), float("-0.000197483"), float("-0.000172614"), float("-0.000186745"), float("-0.000191804"), float("-0.000183598"), float("-0.000241256"), float("-0.000192098"), float("-0.000170965"), float("-0.000176673"), float("-0.000198307"), float("-0.000156611"), float("-0.000142045"), float("-0.000161603"), float("-0.000160694"), float("-0.000142266"), float("-0.000155407"), float("-0.000171961"), float("-0.000133331"), float("-0.000150841"), float("-0.000125645"), float("-4.11465e-05"), float("1.53606e-05"), float("-6.93277e-05"), float("-2.15477e-05"), float("2.44963e-05"), float("-3.18382e-05"), float("-2.37642e-05"), float("9.24679e-06"), float("4.21846e-05"), float("3.36176e-05"), float("2.54965e-05"), float("2.47645e-05"), float("0.000108346"), float("0.000108482"), float("0.000101298"), float("0.000141421"), float("0.000167558"), float("0.000157186"), float("0.000138657"), float("0.000193282"), float("0.000220386"), float("0.000193898"), float("0.000190735"), float("0.000211619"), float("0.000230261"), float("0.00021528"), float("0.000228187"), float("0.000302432"), float("0.000226292"), float("0.000244117"), float("0.000253747"), float("0.000253201"), float("0.000315778"), float("0.000286737"), float("0.000284055"), float("0.000284057"), float("0.000300978"), float("0.000344142"), float("0.000348519"), float("0.000208636"), float("0.000151769"), float("0.000212618"), float("0.000182387"), float("4.25186e-05"), float("1.55146e-05"), float("-1.29299e-06"), float("2.55707e-05"), float("-8.0474e-06"), float("-5.08373e-05"), float("-4.35299e-05"), float("4.43465e-06"), float("4.81199e-06"), float("1.61176e-05"), float("8.6034e-06"), float("2.54492e-05"), float("2.2127e-05"), float("1.95685e-05"), float("2.03618e-05"), float("-5.84188e-06"), float("2.20123e-05"), float("4.80387e-05"), float("6.19176e-05"), float("4.55705e-05"), float("6.3324e-05"), float("6.17111e-05"), float("0.000102875"), float("0.00010095"), float("9.11544e-05"), float("8.62715e-05"), float("8.6665e-05"), float("0.000107648"), float("0.000111767"), float("0.00010453"), float("9.53272e-05"), float("9.07721e-05"), float("8.25583e-05"), float("5.29328e-05"), float("0.000105609"), float("0.000115843"), float("0.000136405"), float("0.000130128"), float("9.2449e-05"), float("7.78238e-05"), float("9.47646e-05"), float("0.000133679"), float("0.000109147"), float("0.000113715"), float("0.000171358"), float("0.00016322"), float("0.000161961"), float("0.000186543"), float("0.00020642"), float("0.000175628"), float("0.000196759"), float("0.00020202"), float("0.000188262"), float("0.000195771"), float("0.00022144"), float("0.000212061"), float("0.000197066"), float("0.000193823"), float("0.000188813"), float("0.000189964"), float("0.000145314"), float("0.000119696"), float("0.000107136"), float("9.4593e-05"), float("9.97707e-05"), float("9.35298e-05"), float("9.53252e-05"), float("8.86535e-05"), float("9.87882e-05"), float("0.000110338"), float("8.10272e-05"), float("6.78147e-05"), float("5.33696e-05"), float("5.26765e-05"), float("2.62262e-05"), float("-1.23077e-05"), float("2.29062e-05"), float("1.04001e-05"), float("-2.91967e-05"), float("-4.2393e-05"), float("-3.54717e-05"), float("-9.28945e-05"), float("-0.000100431"), float("-7.46758e-05"), float("-0.000122817"), float("-0.000109041"), float("-0.000119485"), float("-0.000103781"), float("-0.000171119"), float("-0.000160994"), float("-0.000149117"), float("-0.00013268"), float("-0.000159193"), float("-0.00015496"), float("-0.000140225"), float("-0.000158241"), float("-0.000762324"), float("-0.000767115"), float("-0.000753975"), float("-0.00072348"), float("-0.000748033"), float("-0.000816843"), float("-0.000808487"), float("-0.000739376"), float("-0.000756383"), float("-0.000731987"), float("-0.000758532"), float("-0.000756136"), float("-0.000741849"), float("-0.000682354"), float("-0.000714943"), float("-0.000763774"), float("-0.000742451"), float("-0.000723821"), float("-0.000704298"), float("-0.0007232"), float("-0.000703063"), float("-0.000678821"), float("-0.000624749"), float("-0.000675047"), float("-0.000674623"), float("-0.000625239"), float("-0.000637009"), float("-0.000641791"), float("-0.000637589"), float("-0.000608231"), float("-0.000597854"), float("-0.000589748"), float("-0.000600376"), float("-0.000582922"), float("-0.000588446"), float("-0.000555943"), float("-0.000551712"), float("-0.000508178"), float("-0.000500373"), float("-0.00049847"), float("-0.000467944"), float("-0.000431473"), float("-0.000442191"), float("-0.000412144"), float("-0.000376516"), float("-0.000374266"), float("-0.000364806"), float("-0.000311562"), float("-0.000328529"), float("-0.000328796"), float("-0.000263592"), float("-0.000287732"), float("-0.000251361"), float("-0.000222262"), float("-0.000213702"), float("-0.00019776"), float("-0.000168036"), float("-0.000169818"), float("-0.000179459"), float("-0.00015092"), float("-0.000140475"), float("-0.000135543"), float("-0.000105994"), float("-9.4315e-05"), float("-6.30614e-05"), float("-2.45928e-05"), float("-1.47304e-05"), float("-8.07969e-05"), float("-4.10341e-05"), float("-9.95025e-06"), float("-8.09847e-06"), float("2.15607e-05"), float("2.36972e-05"), float("4.01298e-05"), float("5.78094e-05"), float("7.7993e-05"), float("7.89144e-05"), float("0.000119707"), float("0.000113879"), float("9.84538e-05"), float("9.42778e-05"), float("6.83578e-05"), float("7.44164e-05"), float("0.000107444"), float("0.000120769"), float("8.44241e-05"), float("7.24259e-05"), float("5.70118e-05"), float("4.95479e-05"), float("3.81417e-05"), float("7.03329e-05"), float("7.46877e-05"), float("6.71179e-05"), float("7.26103e-05"), float("9.57397e-05"), float("7.91801e-05"), float("-0.000149472"), float("-0.00014213"), float("-0.000131744"), float("-0.000140043"), float("-0.000137351"), float("-0.000136615"), float("-0.000151119"), float("-0.000142876"), float("-0.000168719"), float("-0.00017359"), float("-0.000164497"), float("-0.000199011"), float("-0.000174208"), float("-0.000162659"), float("-0.000147246"), float("-0.000139938"), float("-0.000130565"), float("-0.000135166"), float("-9.48119e-05"), float("-9.42852e-05"), float("-7.41053e-05"), float("-7.15872e-05"), float("-7.1455e-05"), float("-6.04486e-05"), float("-8.1519e-05"), float("-7.22456e-05"), float("-6.37999e-05"), float("-5.91884e-05"), float("-4.43363e-05"), float("-3.85257e-05"), float("-3.43068e-06"), float("4.23698e-06"), float("-1.38437e-05"), float("6.94508e-06"), float("2.44123e-06"), float("1.97679e-05"), float("2.92677e-05"), float("3.48e-05"), float("3.98916e-05"), float("3.75445e-05"), float("4.12756e-05"), float("4.7296e-05"), float("3.69121e-05"), float("8.78256e-06"), float("2.21561e-05"), float("2.35112e-05"), float("1.53588e-05"), float("2.79886e-05"), float("1.545e-05"), float("-1.5073e-05"), float("-1.2325e-05"), float("-2.31248e-05"), float("-8.68771e-06"), float("-5.88729e-06"), float("1.19861e-05"), float("2.16062e-06"), float("3.66875e-05"), float("3.44659e-05"), float("3.93272e-05"), float("3.75808e-05"), float("2.52072e-05"), float("2.62236e-05"), float("4.71685e-05"), float("3.84434e-05"), float("3.61974e-05"), float("4.22836e-05"), float("3.47363e-05"), float("1.43028e-05"), float("2.4955e-05"), float("2.90605e-05"), float("5.53731e-06"), float("1.6158e-05"), float("5.64251e-06"), float("1.99968e-05"), float("8.72064e-06"), float("4.71728e-06"), float("3.85144e-05"), float("2.5692e-05"), float("2.56004e-05"), float("3.43408e-05"), float("2.86808e-05"), float("1.72165e-05"), float("2.5901e-05"), float("3.57338e-05"), float("3.10321e-05"), float("1.95796e-05"), float("2.35217e-05"), float("1.14026e-05"), float("-1.22701e-05"), float("6.5969e-06"), float("-1.01829e-05"), float("1.21936e-05"), float("1.92287e-05"), float("1.96028e-05"), float("2.41576e-05"), float("3.20103e-05")] + + +class PirProgram_example_input_tensor_meta_1859405807941431992: + program_id = 3000443918524221177 + input_name = "linear_3.w_0" + shape = [32, 96] + data = None + + +class PirProgram_example_input_tensor_meta_2446845994563353672: + program_id = 3000443918524221177 + input_name = "middle_2" + shape = [1] + data = [float("0.2")] + + +class PirProgram_example_input_tensor_meta_3189608938430428800: + program_id = 3000443918524221177 + input_name = "linear_3.b_0" + shape = [96] + data = [float("0.0485458"), float("0.048584"), float("0.0494509"), float("0.0514775"), float("0.0513561"), float("0.0539435"), float("0.053923"), float("0.053999"), float("0.0532589"), float("0.0548013"), float("0.0545531"), float("0.0531352"), float("0.0552883"), float("0.0543044"), float("0.0539071"), float("0.0559445"), float("0.0566455"), float("0.0558921"), float("0.0559904"), float("0.0579908"), float("0.0594786"), float("0.0587086"), float("0.0591194"), float("0.0593815"), float("0.0592324"), float("0.0608538"), float("0.060729"), float("0.0611746"), float("0.0610663"), float("0.0603559"), float("0.0624431"), float("0.0632785"), float("0.0626797"), float("0.061953"), float("0.0608844"), float("0.061694"), float("0.0620486"), float("0.0616028"), float("0.0614813"), float("0.0607751"), float("0.0605967"), float("0.0610775"), float("0.06023"), float("0.0599733"), float("0.0603637"), float("0.0607957"), float("0.0595258"), float("0.0607369"), float("0.0593041"), float("0.0590547"), float("0.0602343"), float("0.05974"), float("0.0587402"), float("0.0584157"), float("0.059519"), float("0.0591371"), float("0.0598814"), float("0.0598889"), float("0.0592714"), float("0.0588438"), float("0.0606669"), float("0.0606387"), float("0.060385"), float("0.062044"), float("0.0617631"), float("0.0626737"), float("0.0635835"), float("0.063831"), float("0.0633426"), float("0.0645196"), float("0.0646552"), float("0.0664435"), float("0.0664267"), float("0.0671475"), float("0.0679168"), float("0.0685048"), float("0.0680123"), float("0.0694771"), float("0.0680605"), float("0.0687388"), float("0.0703835"), float("0.0694004"), float("0.0691108"), float("0.0696977"), float("0.0700212"), float("0.0698567"), float("0.0703321"), float("0.0709236"), float("0.0696366"), float("0.0705553"), float("0.0702906"), float("0.0700228"), float("0.0695461"), float("0.0689861"), float("0.0691985"), float("0.0688401")] + + +class PirProgram_example_input_tensor_meta_4224543760103808494: + program_id = 3000443918524221177 + input_name = "middle_3" + shape = [2, 2, 32] + data = [float("0.942804"), float("1.70377"), float("1.09786"), float("1.38926"), float("0"), float("0"), float("0"), float("0"), float("1.07601"), float("0"), float("1.56775"), float("1.42995"), float("1.9211"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.443868"), float("0"), float("0"), float("2.23675"), float("0"), float("0"), float("1.42107"), float("0"), float("0.289613"), float("0.480792"), float("0.445812"), float("0.489618"), float("0.107548"), float("0.48011"), float("0"), float("0"), float("0.549541"), float("0.420605"), float("0"), float("0.585187"), float("0.59219"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.321631"), float("0"), float("0"), float("0.639618"), float("0"), float("0.0198829"), float("0.443799"), float("0"), float("0.387322"), float("0.817863"), float("0"), float("0.74025"), float("0.489511"), float("0.669735"), float("0"), float("0"), float("0.759979"), float("0.446918"), float("0.665187"), float("0.711636"), float("0.819093"), float("0.535416"), float("0.548715"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.278116"), float("0"), float("0"), float("1.01575"), float("0"), float("0"), float("0.604064"), float("0"), float("0.17553"), float("0"), float("0.174411"), float("0.232705"), float("0"), float("0.266127"), float("0.32481"), float("0.322138"), float("0"), float("0.19189"), float("0.227142"), float("0.317245"), float("0"), float("0"), float("0.264847"), float("0.188336"), float("0.249843"), float("0.277018"), float("0"), float("0.278298"), float("0.270292"), float("0.274346"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.272329"), float("0.314617"), float("0.0190948"), float("0.226759"), float("0")] + + +class PirProgram_example_input_tensor_meta_1661268218216156872: + program_id = 3000443918524221177 + input_name = "middle_1" + shape = [2, 2, 32] + data = [float("0.754243"), float("1.36301"), float("0.878288"), float("1.11141"), float("0"), float("0.754685"), float("0"), float("0"), float("0.860806"), float("0.719321"), float("1.2542"), float("1.14396"), float("1.53688"), float("0"), float("1.29534"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.355094"), float("0.157341"), float("0"), float("1.7894"), float("0"), float("0"), float("1.13686"), float("0"), float("0.23169"), float("0.384634"), float("0.35665"), float("0.391695"), float("0.0860384"), float("0.384088"), float("0"), float("0"), float("0.439633"), float("0.336484"), float("0.403318"), float("0.46815"), float("0.473752"), float("0.0790446"), float("0.443998"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.257305"), float("0.191734"), float("0"), float("0.511694"), float("0"), float("0.0159064"), float("0.355039"), float("0"), float("0.309857"), float("0.654291"), float("0.311981"), float("0.5922"), float("0.391609"), float("0.535788"), float("0"), float("0"), float("0.607983"), float("0.357534"), float("0.53215"), float("0.569309"), float("0.655275"), float("0.428333"), float("0.438972"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.222493"), float("0.474749"), float("0"), float("0.812597"), float("0"), float("0"), float("0.483251"), float("0"), float("0.140424"), float("0.155835"), float("0.139529"), float("0.186164"), float("0"), float("0.212902"), float("0.259848"), float("0.257711"), float("0.23834"), float("0.153512"), float("0.181713"), float("0.253796"), float("0.198702"), float("0"), float("0.211877"), float("0.150669"), float("0.199874"), float("0.221614"), float("0.180791"), float("0.222638"), float("0.216234"), float("0.219477"), float("0.191163"), float("0.13138"), float("0.153447"), float("0"), float("0"), float("0.217864"), float("0.251694"), float("0.0152759"), float("0.181407"), float("0.198531")] + + +class PirProgram_example_input_tensor_meta_3873387511994723854: + program_id = 3000443918524221177 + input_name = "linear_2.w_0" + shape = [16, 32] + data = [float("0.138689"), float("-0.106322"), float("0.172949"), float("-0.118539"), float("0.170557"), float("-0.0672584"), float("0.204199"), float("-0.206781"), float("0.059026"), float("0.325195"), float("0.0539278"), float("-0.0231287"), float("0.291649"), float("-0.0653538"), float("0.210411"), float("0.0745167"), float("-0.203244"), float("0.00686247"), float("0.10958"), float("0.157849"), float("-0.0889362"), float("0.298802"), float("-0.0925571"), float("-0.246793"), float("-0.11154"), float("-0.027271"), float("-0.196191"), float("0.275181"), float("0.0250688"), float("0.204275"), float("0.198187"), float("0.132799"), float("-0.285483"), float("-0.383002"), float("-0.0782729"), float("0.0404985"), float("0.165563"), float("0.206953"), float("0.1843"), float("0.173485"), float("0.129389"), float("0.375723"), float("-0.235547"), float("-0.253078"), float("0.160617"), float("0.282609"), float("-0.204999"), float("0.137476"), float("0.233969"), float("0.185019"), float("0.205812"), float("0.206644"), float("-0.142799"), float("-0.404361"), float("0.0459731"), float("0.0739177"), float("-0.249473"), float("-0.106103"), float("0.0529779"), float("0.0155141"), float("0.310476"), float("0.196231"), float("0.0647549"), float("-0.0507605"), float("-0.130941"), float("-0.0484275"), float("-0.0269265"), float("-0.18644"), float("-0.118492"), float("0.224969"), float("0.13941"), float("-0.146658"), float("0.276094"), float("-0.0997558"), float("-0.17141"), float("0.292359"), float("-0.0742287"), float("-0.164869"), float("0.268466"), float("0.304443"), float("0.228287"), float("0.0761761"), float("-0.0371925"), float("0.296199"), float("-0.191571"), float("0.247706"), float("-0.273535"), float("-0.263982"), float("0.229751"), float("-0.0848374"), float("-0.260978"), float("-0.119975"), float("-0.162848"), float("0.176309"), float("-0.187869"), float("-0.230427"), float("0.299697"), float("0.103454"), float("0.186159"), float("0.117861"), float("0.214886"), float("-0.0160362"), float("-0.191379"), float("-0.0251439"), float("0.253566"), float("-0.193173"), float("-0.0737695"), float("-0.00845166"), float("0.369554"), float("0.347756"), float("0.0289703"), float("-0.210089"), float("0.15666"), float("-0.0107514"), float("-0.205177"), float("-0.370911"), float("-0.123499"), float("-0.165401"), float("0.0389836"), float("0.0961651"), float("-0.18811"), float("0.09999"), float("-0.0234326"), float("0.195959"), float("0.196091"), float("0.293869"), float("0.119714"), float("-0.35079"), float("0.216349"), float("0.0228813"), float("-0.274141"), float("-0.255975"), float("0.247017"), float("0.0754483"), float("0.229817"), float("0.23657"), float("0.161538"), float("-0.221736"), float("0.120394"), float("0.183345"), float("-0.0264143"), float("0.210679"), float("-0.107138"), float("0.124702"), float("0.118588"), float("-0.0608178"), float("0.233959"), float("0.155236"), float("-0.206555"), float("-0.168823"), float("0.219217"), float("-0.0797468"), float("-0.289087"), float("0.274242"), float("-0.367646"), float("0.00935798"), float("0.10757"), float("-0.131774"), float("0.101516"), float("-0.233815"), float("0.142226"), float("0.0382131"), float("-0.246448"), float("-0.232159"), float("0.183271"), float("-0.0259207"), float("-0.29719"), float("-0.189024"), float("0.286622"), float("0.0378376"), float("0.0687375"), float("0.296513"), float("0.115993"), float("0.0502835"), float("0.281012"), float("0.0420163"), float("-0.0551966"), float("-0.0533368"), float("0.0877916"), float("-0.359907"), float("-0.0330066"), float("-0.264281"), float("-0.00644859"), float("-0.103175"), float("-0.0902519"), float("0.246111"), float("-0.111925"), float("0.270801"), float("-0.294941"), float("-0.147398"), float("-0.0205641"), float("0.302264"), float("0.0983701"), float("-0.216157"), float("-0.0542728"), float("-0.0155529"), float("-0.264782"), float("0.168931"), float("0.0981406"), float("-0.106492"), float("0.0882299"), float("0.216699"), float("-0.211478"), float("-0.0953006"), float("0.0748619"), float("-0.100698"), float("0.108855"), float("0.0412198"), float("-0.119328"), float("-0.172325"), float("0.122055"), float("-0.0801006"), float("-0.18746"), float("0.190164"), float("0.0690614"), float("-0.0622083"), float("0.130328"), float("-0.174397"), float("-0.2734"), float("0.0485907"), float("0.161157"), float("0.154754"), float("0.164222"), float("-0.222811"), float("-0.244355"), float("0.30284"), float("0.22183"), float("0.223936"), float("0.25338"), float("-0.0558757"), float("0.1592"), float("0.11035"), float("-0.229928"), float("-0.111174"), float("-0.202858"), float("-0.109098"), float("0.220881"), float("0.00651073"), float("0.0485816"), float("-0.107969"), float("-0.15052"), float("-0.00632081"), float("0.00609574"), float("0.0133219"), float("-0.208563"), float("0.153815"), float("0.269685"), float("0.0584512"), float("0.177424"), float("-0.0912139"), float("-0.345211"), float("-0.256602"), float("0.102258"), float("-0.01134"), float("0.335877"), float("0.0870043"), float("-0.156129"), float("-0.211225"), float("-0.231017"), float("0.0358312"), float("0.203831"), float("0.0187823"), float("-0.0363091"), float("0.213651"), float("0.0315795"), float("0.0870476"), float("0.0339045"), float("0.257777"), float("-0.128626"), float("0.142826"), float("0.0238258"), float("-0.181365"), float("-0.115817"), float("-0.216976"), float("-0.289225"), float("0.00338232"), float("-0.143151"), float("-0.301035"), float("0.0747563"), float("0.207678"), float("-0.153696"), float("-0.2239"), float("-0.345777"), float("0.244627"), float("-0.125439"), float("-0.307755"), float("0.304526"), float("-0.206704"), float("0.00549386"), float("-0.0245539"), float("-0.0432738"), float("0.291515"), float("-0.172698"), float("0.153061"), float("0.138099"), float("-0.0670694"), float("0.0995701"), float("-0.0270735"), float("0.238724"), float("0.0714559"), float("-0.0813835"), float("0.0283591"), float("0.254933"), float("-0.138964"), float("0.177769"), float("0.097657"), float("-0.237825"), float("0.228541"), float("-0.211276"), float("0.0191841"), float("-0.17721"), float("0.0404337"), float("0.101901"), float("0.153098"), float("0.174044"), float("-0.0329636"), float("0.172645"), float("-0.142116"), float("0.0631262"), float("0.181635"), float("0.162319"), float("-0.297797"), float("0.190953"), float("-0.225768"), float("0.257752"), float("-0.29906"), float("0.000686928"), float("-0.0377532"), float("-0.244496"), float("0.0995946"), float("-0.187728"), float("-0.266484"), float("-0.0836021"), float("0.185958"), float("0.186861"), float("0.310878"), float("0.0343862"), float("0.237059"), float("0.251551"), float("0.0472566"), float("-0.115398"), float("-0.383304"), float("0.101012"), float("0.0714781"), float("0.221221"), float("0.314253"), float("-0.272301"), float("-0.379694"), float("0.220204"), float("0.239134"), float("-0.0663735"), float("-0.080212"), float("-0.16315"), float("-0.00506175"), float("-0.0545347"), float("-0.0465481"), float("0.475722"), float("0.179901"), float("0.161356"), float("-0.17223"), float("-0.338577"), float("-0.0519715"), float("-0.103634"), float("-0.181719"), float("-0.0823647"), float("0.279138"), float("-0.0167731"), float("-0.0912584"), float("0.056401"), float("-0.175043"), float("-0.111419"), float("0.0304177"), float("-0.354337"), float("-0.050518"), float("-0.0543624"), float("0.0608217"), float("-0.0743314"), float("0.0759768"), float("0.236895"), float("-0.0438436"), float("-0.134808"), float("-0.0798067"), float("0.186131"), float("0.115307"), float("-0.155228"), float("0.187059"), float("0.0529251"), float("0.185473"), float("-0.274505"), float("-0.15848"), float("0.0693555"), float("0.103163"), float("-0.0561303"), float("-0.133256"), float("-0.178763"), float("0.0890734"), float("0.17019"), float("-0.163824"), float("0.0706935"), float("-0.173212"), float("0.278386"), float("0.297952"), float("0.0195602"), float("-0.269767"), float("0.0546989"), float("-0.311704"), float("-0.0584766"), float("0.145253"), float("0.0326612"), float("-0.137887"), float("0.0357246"), float("0.0743239"), float("-0.0922011"), float("-0.239939"), float("0.0901577"), float("-0.18774"), float("-0.138902"), float("0.201292"), float("-0.004765"), float("0.160184"), float("0.246065"), float("-0.170404"), float("-0.0220289"), float("0.283042"), float("0.164063"), float("-0.216808"), float("-0.125219"), float("0.0188683"), float("-0.209135"), float("0.172225"), float("-0.197942"), float("0.07705"), float("0.340536"), float("0.23112"), float("0.0997995"), float("-0.14498"), float("0.0878749"), float("-0.232691"), float("0.171985"), float("0.192874"), float("0.349805"), float("0.146311"), float("0.243025"), float("0.266444"), float("-0.0826823"), float("0.0912206"), float("-0.00912116"), float("-0.18689"), float("-0.017315"), float("0.0672263"), float("-0.0861141"), float("-0.189335"), float("-0.287294"), float("-0.287037"), float("-0.101011"), float("-0.288937"), float("-0.110381"), float("-0.372508"), float("-0.166248"), float("-0.236813"), float("-0.0719964"), float("-0.311204"), float("-0.051435"), float("-0.123605"), float("-0.138647"), float("-0.0663331"), float("-0.263123"), float("-0.137314"), float("-0.0564986"), float("-0.368283"), float("-0.135173"), float("-0.0914974"), float("0.00832917"), float("-0.175773"), float("-0.0174043"), float("-0.0604438"), float("-0.115462"), float("-0.193237"), float("-0.333979"), float("0.110844"), float("-0.094741"), float("-0.20917"), float("-0.270162"), float("-0.244399"), float("0.113335"), float("-0.113068"), float("0.237874"), float("0.317679"), float("0.0349081"), float("-0.230409"), float("-0.230016"), float("0.172636"), float("0.0211285"), float("0.00908043"), float("0.191938"), float("0.0610569"), float("0.337628"), float("0.17004"), float("0.0341012"), float("0.15482"), float("0.292289"), float("0.24799"), float("0.333809"), float("0.267258"), float("-0.277675"), float("-0.184417"), float("-0.394628"), float("-0.134093"), float("0.248116"), float("-0.180201"), float("0.0719522"), float("0.389222")] + + +class PirProgram_example_input_tensor_meta_770961235158691829: + program_id = 3000443918524221177 + input_name = "middle_0" + shape = [2, 2, 32] + data = None + + +class PirProgram_example_input_tensor_meta_3897698076100692765: + program_id = 3000443918524221177 + input_name = "linear_2.b_0" + shape = [32] + data = [float("0.127184"), float("0.148213"), float("0.126821"), float("0.128469"), float("-0.0799158"), float("0.205759"), float("0.155048"), float("0.126542"), float("0.0762877"), float("0.116064"), float("0.164428"), float("0.110229"), float("0.1077"), float("-0.069753"), float("0.103557"), float("0.0965401"), float("0.129614"), float("0.111043"), float("0.0933711"), float("0.123984"), float("0.169367"), float("0.0588203"), float("0.114893"), float("0.144424"), float("0.124464"), float("-0.0191884"), float("-0.133152"), float("0.195669"), float("0.13529"), float("-0.0156079"), float("0.0929306"), float("0.208163")] + + +class PirProgram_example_input_tensor_meta_5160725927315730163: + program_id = 3000443918524221177 + input_name = "args_0" + shape = [2, 2, 16] + data = [float("0.63837"), float("-0.484176"), float("1.37562"), float("2.11047"), float("0.00551162"), float("2.62531"), float("-0.372032"), float("2.03205"), float("3.52888"), float("0.886358"), float("-0.225955"), float("-0.324597"), float("0.802664"), float("-2.20063"), float("1.28291"), float("-5.03928"), float("0.139641"), float("0.108348"), float("0.579502"), float("0.525852"), float("0.11746"), float("0.674907"), float("-0.286067"), float("0.566479"), float("0.799654"), float("0.31314"), float("0.146439"), float("-0.275929"), float("0.237963"), float("-0.39452"), float("0.0206265"), float("-1.22811"), float("-0.356625"), float("0.240933"), float("0.447273"), float("0.99085"), float("0.022031"), float("1.57203"), float("-0.341862"), float("0.827176"), float("1.06617"), float("0.33467"), float("0.0470244"), float("-0.0600742"), float("0.227221"), float("-0.507455"), float("-0.00214139"), float("-1.91961"), float("0.133666"), float("0.00946639"), float("0.0763022"), float("-0.017242"), float("0.0861289"), float("0.00317063"), float("0.054149"), float("0.141779"), float("-0.0106162"), float("0.142801"), float("-0.0334722"), float("-0.301938"), float("0.0549173"), float("0.0499354"), float("-0.144601"), float("0.111016")] + + +class PirProgram_example_input_tensor_meta_2653122763850172560: + program_id = 3111236241086672326 + input_name = "output_grad_0" + shape = [2, 2, 16] + data = [float("-0.000259765"), float("0.000502711"), float("-3.34286e-05"), float("-0.000308996"), float("0.000766827"), float("0.00031456"), float("5.42812e-05"), float("-0.000930537"), float("0.000139215"), float("-0.000372329"), float("0.000279334"), float("0.00021588"), float("-0.00051067"), float("-0.00061697"), float("0.000621474"), float("0.000900432"), float("-0.000251851"), float("0.000434745"), float("-1.13758e-05"), float("-0.000917027"), float("-0.000684694"), float("-0.000777882"), float("7.10025e-05"), float("-5.16687e-05"), float("-0.000292353"), float("-0.000278136"), float("0.000741509"), float("8.23935e-05"), float("-2.14872e-05"), float("-0.000322378"), float("0.000605169"), float("0.000113745"), float("0.00375845"), float("0.00106431"), float("-0.000995514"), float("0.00707899"), float("0.00362036"), float("0.00517876"), float("-0.000870925"), float("0.00109624"), float("0.00352798"), float("0.00302349"), float("-0.00333248"), float("0.0019234"), float("-0.00183069"), float("0.00179234"), float("-0.00927467"), float("-0.00196363"), float("0.00013415"), float("-0.000518328"), float("-0.000200835"), float("-0.000149648"), float("-0.000169967"), float("0.000350789"), float("9.03782e-05"), float("0.000154067"), float("0.000613054"), float("0.000406082"), float("-0.000935107"), float("0.00018923"), float("0.000132097"), float("-0.000361182"), float("-0.000315851"), float("-0.000232278")] + + +class PirProgram_example_input_tensor_meta_9055065804403674970: + program_id = 3111236241086672326 + input_name = "middle_5" + shape = [2, 2, 16] + data = None + + +class PirProgram_example_input_tensor_meta_7382231132507793619: + program_id = 3111236241086672326 + input_name = "middle_4" + shape = [2, 2, 32] + data = None + + +class PirProgram_example_input_tensor_meta_3182402915473364379: + program_id = 3111236241086672326 + input_name = "middle_2" + shape = [1] + data = [float("0.2")] + + +class PirProgram_example_input_tensor_meta_1974672370162666358: + program_id = 3111236241086672326 + input_name = "middle_0" + shape = [2, 2, 32] + data = None + + +class PirProgram_example_input_tensor_meta_7940412681263697295: + program_id = 3111236241086672326 + input_name = "linear_1.w_0" + shape = [32, 16] + data = [float("0.235596"), float("-0.166064"), float("-0.00841034"), float("0.0251242"), float("0.121917"), float("0.0115089"), float("0.293106"), float("0.189091"), float("0.228253"), float("0.0750035"), float("-0.0840923"), float("0.0157098"), float("0.0583272"), float("-0.113309"), float("0.256736"), float("-0.266579"), float("0.0569002"), float("-0.148237"), float("0.168182"), float("0.0465327"), float("-0.104699"), float("0.100044"), float("-0.136119"), float("0.217071"), float("0.250393"), float("0.0458567"), float("-0.165222"), float("-0.247985"), float("-0.0402208"), float("-0.265832"), float("-0.00288254"), float("0.0103337"), float("0.221503"), float("0.0711672"), float("-0.0150802"), float("-0.153977"), float("0.232742"), float("0.0707905"), float("0.109877"), float("0.0475617"), float("-0.0662123"), float("0.242932"), float("-0.290769"), float("-0.0773183"), float("0.211586"), float("-0.140735"), float("0.143677"), float("-0.318947"), float("0.309584"), float("0.155288"), float("-0.142704"), float("-0.208616"), float("-0.161949"), float("-0.225615"), float("-0.119309"), float("-0.0151871"), float("0.243764"), float("-0.125401"), float("0.334346"), float("0.0876797"), float("-0.12878"), float("0.155844"), float("0.25528"), float("0.0449498"), float("-0.109217"), float("0.361007"), float("0.0548769"), float("-0.225299"), float("0.129158"), float("0.182534"), float("0.111121"), float("-0.170268"), float("0.26808"), float("0.0713453"), float("0.0276709"), float("0.0309021"), float("0.222045"), float("0.0556532"), float("0.225857"), float("0.165388"), float("-0.214861"), float("-0.189104"), float("0.0905981"), float("0.164752"), float("-0.0745658"), float("0.17254"), float("0.230335"), float("0.172994"), float("-0.0394944"), float("0.162471"), float("-0.178272"), float("-0.121411"), float("0.119618"), float("-0.197991"), float("-0.0220652"), float("-0.0271392"), float("0.0401553"), float("-0.0364811"), float("-0.180565"), float("-0.149028"), float("0.336991"), float("0.0416623"), float("-0.233224"), float("-0.080196"), float("0.00397957"), float("0.0222934"), float("-0.033727"), float("-0.151176"), float("0.0296938"), float("-0.174967"), float("0.0433006"), float("0.146219"), float("-0.295726"), float("-0.200405"), float("0.292424"), float("0.105805"), float("-0.278996"), float("-0.0612777"), float("-0.14825"), float("0.00739621"), float("-0.119451"), float("-0.253996"), float("0.501842"), float("-0.138736"), float("-0.055433"), float("-0.109043"), float("-0.210503"), float("0.283465"), float("0.218386"), float("-0.230348"), float("0.266489"), float("-0.0496932"), float("-0.214893"), float("-0.0331828"), float("-0.199718"), float("-0.0481673"), float("0.382266"), float("-0.000796923"), float("0.0316693"), float("-0.00864441"), float("0.260415"), float("-0.304373"), float("-0.076894"), float("-0.239595"), float("-0.0672012"), float("0.319481"), float("0.0133308"), float("-0.206744"), float("-0.268379"), float("-0.218137"), float("-0.110468"), float("0.120406"), float("0.0381229"), float("-0.171569"), float("0.281952"), float("-0.130553"), float("-0.0789959"), float("-0.175866"), float("-0.148795"), float("-0.123259"), float("-0.18909"), float("0.127911"), float("0.152386"), float("0.112031"), float("0.196341"), float("-0.264504"), float("-0.152368"), float("-0.0758584"), float("-0.202836"), float("-0.304334"), float("0.174291"), float("-0.0152358"), float("0.200058"), float("-0.284122"), float("0.376094"), float("0.310537"), float("-0.116116"), float("-0.245531"), float("-0.0476175"), float("0.201079"), float("-0.16955"), float("0.134718"), float("0.328129"), float("-0.201016"), float("-0.122684"), float("0.0527725"), float("0.014955"), float("-0.0159163"), float("-0.024892"), float("-0.172767"), float("0.00895791"), float("-0.126559"), float("0.017916"), float("-0.0624295"), float("0.248658"), float("0.104764"), float("-0.0995163"), float("0.021314"), float("-0.00632193"), float("-0.0202286"), float("0.0669071"), float("-0.0343595"), float("-0.0058106"), float("-0.0319011"), float("0.211846"), float("-0.143065"), float("0.00494012"), float("-0.178171"), float("-0.142023"), float("0.236462"), float("0.0113232"), float("0.203394"), float("0.213219"), float("0.287939"), float("-0.266193"), float("0.145776"), float("0.277013"), float("0.0349097"), float("0.339889"), float("0.257273"), float("-0.031818"), float("0.106094"), float("-0.016972"), float("-0.341205"), float("0.195945"), float("-0.273676"), float("-0.0920491"), float("-0.00215972"), float("0.0761255"), float("-0.117919"), float("0.292542"), float("-0.0279444"), float("-0.138423"), float("0.120147"), float("0.190762"), float("0.0532155"), float("-0.00462688"), float("0.29658"), float("0.264242"), float("0.279836"), float("-0.0740407"), float("-0.258875"), float("0.204921"), float("0.0259398"), float("-0.0169675"), float("0.074128"), float("-0.164556"), float("-0.0276452"), float("0.0226643"), float("0.0141938"), float("0.249415"), float("0.0347015"), float("-0.165602"), float("0.077402"), float("-0.226793"), float("0.244763"), float("-0.305538"), float("0.167513"), float("-0.026212"), float("-0.0884336"), float("0.323809"), float("0.000747159"), float("0.308381"), float("-0.199611"), float("-0.193336"), float("-0.144533"), float("-0.186432"), float("0.208955"), float("0.0886495"), float("-0.191583"), float("0.349786"), float("0.35748"), float("-0.0242003"), float("0.0790487"), float("-0.0257431"), float("-0.347474"), float("-0.154624"), float("-0.18064"), float("0.292404"), float("-0.0464253"), float("0.238564"), float("-0.238146"), float("0.0556705"), float("0.0297406"), float("0.167295"), float("0.196666"), float("0.233435"), float("0.119447"), float("-0.0764451"), float("0.305884"), float("-0.210522"), float("-0.0604975"), float("-0.101583"), float("-0.0236538"), float("-0.235309"), float("-0.0823719"), float("-0.248695"), float("0.0203888"), float("-0.178912"), float("0.120222"), float("-0.123081"), float("0.0837287"), float("0.264862"), float("-0.137024"), float("0.18452"), float("0.117199"), float("-0.0519129"), float("-0.0791504"), float("-0.250732"), float("0.0108192"), float("-0.0176924"), float("0.00619382"), float("-0.14818"), float("0.0474773"), float("-0.218766"), float("-0.152883"), float("-0.108489"), float("0.0400477"), float("0.0824107"), float("-0.326851"), float("-0.167915"), float("-0.206116"), float("-0.070031"), float("0.186761"), float("-0.299896"), float("0.256832"), float("0.0676038"), float("0.0782485"), float("0.191245"), float("-0.187385"), float("-0.239842"), float("-0.0772858"), float("-0.153211"), float("-0.220496"), float("0.0923377"), float("-0.181751"), float("-0.122754"), float("0.0135214"), float("0.0596246"), float("-0.241126"), float("0.284071"), float("-0.177228"), float("-0.0907132"), float("0.138534"), float("0.194014"), float("0.157129"), float("-0.215676"), float("-0.0922354"), float("0.246938"), float("-0.123517"), float("-0.250844"), float("0.0446277"), float("0.176366"), float("-0.0482041"), float("0.116668"), float("0.210133"), float("-0.0399605"), float("0.0173124"), float("-0.00336163"), float("0.111738"), float("0.267352"), float("0.0239314"), float("-0.12188"), float("-0.200225"), float("-0.0374264"), float("-0.223423"), float("0.28684"), float("-0.316664"), float("-0.0740798"), float("0.101619"), float("-0.15827"), float("-0.242475"), float("0.116453"), float("0.23239"), float("0.0285625"), float("0.0559415"), float("-0.111179"), float("-0.187168"), float("-0.0821701"), float("-0.100173"), float("-0.122167"), float("0.0214796"), float("0.275201"), float("0.326983"), float("-0.157285"), float("0.308136"), float("-0.231156"), float("-0.303319"), float("0.229308"), float("0.0719604"), float("0.0252571"), float("-0.137748"), float("0.196749"), float("0.0238845"), float("0.0721256"), float("-0.111398"), float("0.193114"), float("-0.0786931"), float("0.346398"), float("0.0366481"), float("-0.0774674"), float("-0.0469274"), float("0.0359128"), float("-0.0288154"), float("-0.175977"), float("0.22237"), float("0.186215"), float("-0.00687118"), float("0.151105"), float("-0.156596"), float("-0.342589"), float("-0.0769077"), float("0.140154"), float("-0.0963023"), float("-0.152346"), float("-0.00415342"), float("-0.203184"), float("-0.07618"), float("0.142056"), float("-0.0463533"), float("0.000432739"), float("0.113857"), float("0.239217"), float("0.183889"), float("-0.148585"), float("-0.0760731"), float("-0.0540795"), float("-0.137623"), float("-0.24345"), float("-0.0384814"), float("-0.0854952"), float("-0.151667"), float("0.00554339"), float("-0.106125"), float("0.0946545"), float("-0.0221862"), float("0.257953"), float("0.194255"), float("0.117183"), float("-0.239613"), float("-0.276004"), float("-0.0432311"), float("0.174063"), float("-0.0634706"), float("-0.324576"), float("0.200557"), float("0.27033"), float("-0.141904"), float("0.163621"), float("0.159129"), float("0.0331803"), float("0.0716283"), float("-0.259889"), float("-0.281286"), float("0.311005"), float("-0.192631"), float("-0.025214"), float("-0.335516"), float("0.290179"), float("0.163395"), float("0.0640547"), float("0.0451905"), float("0.3239"), float("0.274364"), float("-0.305999"), float("0.102385"), float("-0.106225"), float("-0.213116"), float("0.344277"), float("-0.0780973"), float("-0.224715"), float("0.241605"), float("-0.151579"), float("-0.210299"), float("0.0823131"), float("0.100404"), float("0.132546"), float("-0.277527"), float("0.0598196"), float("-0.0433582"), float("0.230157"), float("0.123731"), float("-0.100427"), float("0.193713"), float("0.202449"), float("-0.266377"), float("-0.0941569"), float("-0.0321949"), float("-0.0972526"), float("0.167801"), float("0.0583861"), float("-0.270751"), float("0.18621"), float("-0.278426"), float("-0.0386036"), float("0.160098"), float("-0.13865"), float("0.383023"), float("0.051967"), float("-0.180901"), float("0.171269"), float("0.0455311"), float("0.0692674"), float("0.38991"), float("-0.0709482"), float("-0.0925497"), float("-0.125314"), float("-0.28824"), float("-0.101276"), float("0.180632"), float("0.229682"), float("-0.129278")] + + +class PirProgram_example_input_tensor_meta_434379246092404667: + program_id = 3111236241086672326 + input_name = "linear_1.b_0" + shape = [16] + data = [float("0.125644"), float("0.0118724"), float("0.0756429"), float("-0.0202031"), float("0.0825371"), float("0.00196512"), float("0.0474929"), float("0.136103"), float("-0.0232748"), float("0.140402"), float("-0.0294003"), float("-0.296898"), float("0.0529667"), float("0.0581229"), float("-0.159494"), float("0.123836")] + + +class PirProgram_example_input_tensor_meta_8868019711434113141: + program_id = 3111236241086672326 + input_name = "middle_3" + shape = [2, 2, 32] + data = [float("2.3204"), float("1.27108"), float("1.55032"), float("0"), float("0"), float("1.33579"), float("0"), float("0"), float("1.66375"), float("0"), float("0"), float("0.766774"), float("1.97588"), float("5.28458"), float("0"), float("0"), float("0"), float("0"), float("0"), float("1.20052"), float("1.82524"), float("0"), float("2.05893"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.23679"), float("0.317447"), float("0"), float("0"), float("0.289374"), float("0"), float("0"), float("0.508824"), float("0"), float("0"), float("0.178782"), float("0.712137"), float("1.66309"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.357579"), float("0"), float("0"), float("0.756404"), float("0"), float("0"), float("0.0961974"), float("0.327142"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.581249"), float("0.744442"), float("0"), float("0"), float("0.680321"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.493762"), float("1.26236"), float("3.09834"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.783214"), float("1.05384"), float("0"), float("0"), float("0"), float("0"), float("0.346895"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.0223007"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.0226074"), float("0"), float("0.0077469"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0")] + + +class PirProgram_example_input_tensor_meta_7754446314117936854: + program_id = 3111236241086672326 + input_name = "middle_1" + shape = [2, 2, 32] + data = [float("1.85632"), float("1.01686"), float("1.24026"), float("0"), float("0"), float("1.06863"), float("0"), float("0"), float("1.331"), float("0"), float("0"), float("0.613419"), float("1.5807"), float("4.22766"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.960414"), float("1.46019"), float("0"), float("1.64714"), float("0"), float("0"), float("0.822619"), float("1.06868"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.674439"), float("0.189432"), float("0.253958"), float("0"), float("0"), float("0.231499"), float("0"), float("0"), float("0.407059"), float("0"), float("0"), float("0.143025"), float("0.569709"), float("1.33048"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.286063"), float("0.5215"), float("0"), float("0.605123"), float("0"), float("0"), float("0.0769579"), float("0.261714"), float("0"), float("0"), float("0"), float("0"), float("0"), float("1.09256"), float("0.464999"), float("0.595554"), float("0"), float("0"), float("0.544257"), float("0"), float("0"), float("0.735279"), float("0"), float("0"), float("0.39501"), float("1.00989"), float("2.47867"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.626571"), float("0.843074"), float("0"), float("0.986253"), float("0"), float("0"), float("0.277516"), float("0.654774"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.0178405"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.0151772"), float("0"), float("0.0180859"), float("0"), float("0.00619752"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0")] + + +class PirProgram_example_input_tensor_meta_4575370185374825974: + program_id = 3111236241086672326 + input_name = "linear_0.w_0" + shape = [96, 32] + data = None + + +class PirProgram_example_input_tensor_meta_8444399123059722995: + program_id = 3111236241086672326 + input_name = "linear_0.b_0" + shape = [32] + data = [float("0.0201408"), float("-0.266505"), float("-0.278198"), float("-0.00211538"), float("-0.269264"), float("-0.232061"), float("-0.364265"), float("-0.248469"), float("-0.107383"), float("-0.31451"), float("-0.262869"), float("-0.176201"), float("-0.0559743"), float("-0.359228"), float("-0.388085"), float("-0.37184"), float("-0.0378036"), float("-0.0328156"), float("-0.336554"), float("-0.176577"), float("0.0164908"), float("-0.371851"), float("0.0113843"), float("-0.332664"), float("0.0101896"), float("-0.282343"), float("-0.26443"), float("-0.205234"), float("-0.164831"), float("-0.265903"), float("-0.343348"), float("-0.374968")] + + +class PirProgram_example_input_tensor_meta_3925472342396615484: + program_id = 3111236241086672326 + input_name = "args_0" + shape = [2, 2, 96] + data = [float("-1.15597"), float("-1.1523"), float("-1.15171"), float("-1.14702"), float("-1.14445"), float("-1.14483"), float("-1.14359"), float("-1.14577"), float("-1.14909"), float("-1.14868"), float("-1.14244"), float("-1.14465"), float("-1.13813"), float("-1.14016"), float("-1.13834"), float("-1.12726"), float("-1.11024"), float("-1.10905"), float("-1.10871"), float("-1.10095"), float("-1.09833"), float("-1.09615"), float("-1.09124"), float("-1.09103"), float("-1.09178"), float("-1.09125"), float("-1.09142"), float("-1.09758"), float("-1.09483"), float("-1.09004"), float("-1.0888"), float("-1.08218"), float("-1.08141"), float("-1.07451"), float("-1.07159"), float("-1.07065"), float("-1.06963"), float("-1.06884"), float("-1.06653"), float("-1.06595"), float("-1.06269"), float("-1.06049"), float("-1.05744"), float("-1.06124"), float("-1.05722"), float("-1.05479"), float("-1.05422"), float("-1.05193"), float("-1.0509"), float("-1.05028"), float("-1.04654"), float("-1.045"), float("-1.046"), float("-1.04667"), float("-1.04904"), float("-1.04042"), float("-1.04003"), float("-1.04359"), float("-1.03823"), float("-1.03557"), float("-1.03127"), float("-1.03583"), float("-1.0344"), float("-1.03144"), float("-1.02952"), float("-1.02593"), float("-1.02701"), float("-1.02957"), float("-1.02561"), float("-1.02673"), float("-1.0235"), float("-1.02171"), float("-1.02023"), float("-1.01943"), float("-1.01996"), float("-1.0124"), float("-1.01049"), float("-1.01181"), float("-1.01104"), float("-1.0073"), float("-1.0079"), float("-1.00004"), float("-0.996691"), float("-0.993982"), float("-0.991675"), float("-0.990566"), float("-0.988743"), float("-0.988268"), float("-0.984146"), float("-0.982912"), float("-0.987096"), float("-0.984222"), float("-0.963696"), float("-0.952533"), float("-0.954977"), float("-0.950661"), float("-0.399858"), float("-0.398499"), float("-0.392917"), float("-0.396343"), float("-0.391605"), float("-0.386859"), float("-0.391783"), float("-0.393174"), float("-0.396282"), float("-0.395804"), float("-0.396976"), float("-0.398565"), float("-0.395854"), float("-0.393964"), float("-0.393872"), float("-0.39339"), float("-0.394667"), float("-0.394263"), float("-0.395912"), float("-0.39461"), float("-0.392164"), float("-0.391116"), float("-0.393774"), float("-0.391869"), float("-0.392982"), float("-0.390163"), float("-0.392436"), float("-0.393987"), float("-0.394744"), float("-0.394476"), float("-0.391628"), float("-0.389037"), float("-0.387299"), float("-0.382867"), float("-0.392128"), float("-0.39309"), float("-0.3956"), float("-0.392959"), float("-0.386446"), float("-0.381103"), float("-0.380557"), float("-0.385733"), float("-0.382358"), float("-0.382535"), float("-0.388842"), float("-0.3847"), float("-0.386946"), float("-0.391374"), float("-0.390959"), float("-0.386347"), float("-0.391026"), float("-0.390428"), float("-0.386892"), float("-0.388793"), float("-0.392527"), float("-0.390458"), float("-0.389178"), float("-0.38852"), float("-0.389764"), float("-0.389545"), float("-0.384279"), float("-0.383312"), float("-0.379087"), float("-0.380502"), float("-0.381453"), float("-0.38131"), float("-0.386445"), float("-0.387503"), float("-0.385849"), float("-0.391521"), float("-0.39044"), float("-0.391326"), float("-0.390687"), float("-0.389549"), float("-0.387867"), float("-0.384638"), float("-0.388114"), float("-0.389713"), float("-0.386914"), float("-0.383812"), float("-0.38576"), float("-0.382273"), float("-0.381319"), float("-0.383497"), float("-0.378321"), float("-0.382726"), float("-0.3803"), float("-0.38556"), float("-0.378433"), float("-0.378988"), float("-0.380473"), float("-0.383853"), float("-0.381981"), float("-0.383138"), float("-0.38374"), float("-0.379838"), float("-0.630672"), float("-0.632913"), float("-0.633336"), float("-0.636795"), float("-0.637253"), float("-0.635754"), float("-0.636042"), float("-0.637148"), float("-0.640553"), float("-0.641178"), float("-0.639793"), float("-0.642557"), float("-0.642459"), float("-0.647011"), float("-0.646651"), float("-0.643922"), float("-0.64385"), float("-0.643057"), float("-0.642857"), float("-0.643463"), float("-0.644947"), float("-0.645577"), float("-0.645093"), float("-0.645523"), float("-0.645906"), float("-0.647578"), float("-0.646427"), float("-0.646111"), float("-0.649196"), float("-0.65109"), float("-0.649286"), float("-0.648321"), float("-0.64727"), float("-0.649662"), float("-0.646437"), float("-0.650501"), float("-0.648313"), float("-0.652139"), float("-0.652555"), float("-0.652121"), float("-0.651807"), float("-0.652984"), float("-0.652824"), float("-0.654689"), float("-0.653612"), float("-0.653592"), float("-0.653823"), float("-0.653214"), float("-0.653469"), float("-0.651772"), float("-0.650067"), float("-0.65063"), float("-0.652344"), float("-0.653487"), float("-0.648193"), float("-0.650301"), float("-0.649187"), float("-0.650017"), float("-0.648082"), float("-0.648181"), float("-0.647695"), float("-0.648911"), float("-0.649349"), float("-0.651401"), float("-0.654608"), float("-0.651767"), float("-0.652217"), float("-0.646641"), float("-0.649977"), float("-0.651891"), float("-0.653053"), float("-0.65444"), float("-0.653964"), float("-0.657094"), float("-0.657665"), float("-0.658948"), float("-0.662083"), float("-0.663417"), float("-0.665791"), float("-0.663225"), float("-0.659105"), float("-0.659393"), float("-0.659387"), float("-0.661882"), float("-0.663708"), float("-0.663021"), float("-0.66179"), float("-0.652634"), float("-0.652383"), float("-0.651621"), float("-0.65448"), float("-0.654242"), float("-0.656716"), float("-0.657377"), float("-0.65797"), float("-0.655658"), float("-0.00436456"), float("-0.00475825"), float("-0.00663337"), float("-0.00523404"), float("-0.00562955"), float("-0.00631295"), float("-0.00424731"), float("-0.00233632"), float("0.00070127"), float("0.00211113"), float("0.000545739"), float("0.00642778"), float("0.0038039"), float("0.000308868"), float("-0.00136501"), float("-0.00140553"), float("-0.00126356"), float("-0.00223812"), float("-0.00588529"), float("-0.00441997"), float("-0.00579612"), float("-0.00591071"), float("-0.00374736"), float("-0.0039841"), float("0.00174819"), float("0.00298575"), float("0.00173817"), float("0.00255231"), float("0.000368191"), float("0.000318813"), float("-0.00178881"), float("-0.00274283"), float("0.000628811"), float("-0.00182678"), float("-7.61417e-05"), float("-0.000906536"), float("-0.00333968"), float("-0.00408479"), float("-0.00325562"), float("-0.00334317"), float("-0.00397552"), float("-0.00359293"), float("-0.0032651"), float("0.00372263"), float("0.00179851"), float("0.00113003"), float("0.00107969"), float("0.000824071"), float("0.00230663"), float("0.00521457"), float("0.00676399"), float("0.00760787"), float("0.00433249"), float("0.0035608"), float("0.000554484"), float("0.00362084"), float("-0.00145232"), float("-0.00354868"), float("-0.00290337"), float("-0.00352887"), float("-0.00181459"), float("-0.00102524"), float("-0.00380784"), float("-0.00352733"), float("-0.00369791"), float("-0.00225879"), float("-0.00303406"), float("0.00243163"), float("0.00128276"), float("0.000701387"), float("0.00297557"), float("0.00318256"), float("0.00367343"), float("0.00317448"), float("0.00563999"), float("0.00400331"), float("-0.00251565"), float("0.00151779"), float("-0.00175092"), float("-0.00389494"), float("-0.0005398"), float("-0.000404088"), float("-0.00223803"), float("-0.00250093"), float("-0.0025293"), float("-0.00180433"), float("0.000374747"), float("0.00324137"), float("0.00465033"), float("0.00285781"), float("0.0061246"), float("0.00211838"), float("0.000508647"), float("-0.00179661"), float("-0.00133032"), float("-0.000782828")] + + +class PirProgram_example_input_tensor_meta_3211373362553701247: + program_id = 4630264566612914126 + input_name = "linear_1.b_0" + shape = [16] + data = None + + +class PirProgram_example_input_tensor_meta_3038923928244044276: + program_id = 4630264566612914126 + input_name = "linear_1.w_0" + shape = [32, 16] + data = None + + +class PirProgram_example_input_tensor_meta_1110867825722039225: + program_id = 4630264566612914126 + input_name = "linear_0.b_0" + shape = [32] + data = None + + +class PirProgram_example_input_tensor_meta_7373294236970362073: + program_id = 4630264566612914126 + input_name = "linear_0.w_0" + shape = [96, 32] + data = None + + +class PirProgram_example_input_tensor_meta_7822756705787592613: + program_id = 4630264566612914126 + input_name = "args_0" + shape = [16, 2, 96] + data = None + + +class PirProgram_example_input_tensor_meta_1185347860821980671: + program_id = 2498128585497043328 + input_name = "linear_3.b_0" + shape = [96] + data = None + + +class PirProgram_example_input_tensor_meta_8144689379135590626: + program_id = 2498128585497043328 + input_name = "linear_3.w_0" + shape = [32, 96] + data = None + + +class PirProgram_example_input_tensor_meta_5725363569888070898: + program_id = 2498128585497043328 + input_name = "linear_2.b_0" + shape = [32] + data = None + + +class PirProgram_example_input_tensor_meta_8026832391553157584: + program_id = 2498128585497043328 + input_name = "linear_2.w_0" + shape = [16, 32] + data = None + + +class PirProgram_example_input_tensor_meta_4610283840437367128: + program_id = 2498128585497043328 + input_name = "args_0" + shape = [16, 2, 16] + data = [float("-0.15216"), float("0.757773"), float("-0.329263"), float("-0.540604"), float("0.181305"), float("-0.183984"), float("-0.0277555"), float("-0.0668215"), float("-0.164828"), float("-0.152864"), float("-0.041962"), float("-0.119368"), float("0.12759"), float("0.124398"), float("0.762339"), float("0.281282"), float("0.353945"), float("-2.13213"), float("5.45738"), float("6.78211"), float("-0.901334"), float("10.5608"), float("0.638606"), float("7.42863"), float("11.3598"), float("1.95794"), float("-2.3519"), float("-1.56203"), float("1.85749"), float("-8.1808"), float("3.51677"), float("-18.3846"), float("0.175046"), float("-0.0554654"), float("0.142715"), float("0.1279"), float("0.0595285"), float("0.126531"), float("0.0591089"), float("0.250216"), float("0.231589"), float("0.128165"), float("-0.0492941"), float("-0.301204"), float("0.0698216"), float("-0.084365"), float("-0.0207583"), float("-0.169802"), float("0.206708"), float("-0.686835"), float("1.87437"), float("2.33026"), float("-0.205472"), float("3.51035"), float("0.219181"), float("2.54969"), float("3.77282"), float("0.725045"), float("-0.642187"), float("-0.656401"), float("0.567246"), float("-2.6316"), float("1.1654"), float("-6.16229"), float("0.0315424"), float("0.26434"), float("-0.0882699"), float("-0.218807"), float("0.113201"), float("-0.0394262"), float("0.0516854"), float("0.0276483"), float("-0.0290755"), float("0.0801737"), float("-0.0651081"), float("-0.237694"), float("0.0902895"), float("0.0880786"), float("0.151029"), float("0.160605"), float("0.206246"), float("-0.571435"), float("1.51467"), float("1.89243"), float("-0.183881"), float("2.81018"), float("0.168251"), float("2.10894"), float("3.12867"), float("0.59224"), float("-0.562897"), float("-0.582865"), float("0.516875"), float("-2.15694"), float("0.906377"), float("-4.90434"), float("-0.368032"), float("1.16993"), float("-0.514465"), float("-0.850272"), float("0.344572"), float("-0.307727"), float("-0.113579"), float("-0.13008"), float("-0.331641"), float("-0.35746"), float("0.0170699"), float("-0.0903696"), float("0.173143"), float("0.124255"), float("1.31038"), float("0.514586"), float("0.227267"), float("-1.11517"), float("2.94774"), float("3.65704"), float("-0.388165"), float("5.601"), float("0.358289"), float("3.96366"), float("5.94604"), float("1.08666"), float("-1.08567"), float("-0.923208"), float("0.887275"), float("-4.22929"), float("1.879"), float("-9.79137"), float("0.162628"), float("-0.137703"), float("0.503044"), float("0.6243"), float("0.0336264"), float("0.796969"), float("0.0392233"), float("0.718495"), float("0.939775"), float("0.253406"), float("-0.0352974"), float("-0.313293"), float("0.134459"), float("-0.549998"), float("0.238071"), float("-1.44879"), float("0.230935"), float("-1.7091"), float("4.48235"), float("5.54546"), float("-0.611005"), float("8.6001"), float("0.589296"), float("5.91443"), float("8.9119"), float("1.59889"), float("-1.64335"), float("-1.29905"), float("1.25007"), float("-6.4299"), float("2.92819"), float("-15.0178"), float("0.184487"), float("-0.271964"), float("0.839011"), float("1.04533"), float("-0.0367303"), float("1.46748"), float("0.0771252"), float("1.18206"), float("1.66538"), float("0.373778"), float("-0.200558"), float("-0.395642"), float("0.261839"), float("-1.07963"), float("0.462218"), float("-2.62307"), float("0.214251"), float("-0.996448"), float("2.64592"), float("3.29088"), float("-0.336725"), float("5.02743"), float("0.329378"), float("3.56391"), float("5.33463"), float("0.980632"), float("-0.961669"), float("-0.845799"), float("0.793455"), float("-3.77941"), float("1.6815"), float("-8.78449"), float("0.230756"), float("-0.0902852"), float("0.313066"), float("0.373669"), float("0.0560226"), float("0.459538"), float("-0.00736624"), float("0.49768"), float("0.6865"), float("0.191823"), float("-0.057096"), float("-0.286257"), float("0.149784"), float("-0.337131"), float("0.140222"), float("-0.830687"), float("0.208667"), float("-0.436397"), float("1.23756"), float("1.54828"), float("-0.11709"), float("2.26445"), float("0.125928"), float("1.73361"), float("2.52496"), float("0.51826"), float("-0.403326"), float("-0.501762"), float("0.408805"), float("-1.7112"), float("0.734785"), float("-4.00791"), float("0.187755"), float("-0.102415"), float("0.457224"), float("0.545884"), float("0.03179"), float("0.690048"), float("0.0183367"), float("0.648476"), float("0.835108"), float("0.241401"), float("-0.0139194"), float("-0.301483"), float("0.122606"), float("-0.467988"), float("0.20309"), float("-1.30769"), float("0.244079"), float("-1.70861"), float("4.47838"), float("5.55282"), float("-0.619873"), float("8.61168"), float("0.583028"), float("5.91956"), float("8.93747"), float("1.60415"), float("-1.65242"), float("-1.29492"), float("1.26257"), float("-6.44304"), float("2.93482"), float("-15.0458"), float("0.171207"), float("-0.0780683"), float("0.156868"), float("0.230114"), float("0.0477679"), float("0.248112"), float("0.0828771"), float("0.324246"), float("0.362926"), float("0.126835"), float("-0.0397792"), float("-0.276738"), float("0.0498615"), float("-0.144937"), float("0.051245"), float("-0.390421"), float("0.238441"), float("-0.348271"), float("0.995058"), float("1.21009"), float("-0.0690387"), float("1.7613"), float("0.0694272"), float("1.40697"), float("2.07087"), float("0.418077"), float("-0.335243"), float("-0.448782"), float("0.368917"), float("-1.34953"), float("0.568891"), float("-3.1105"), float("0.24836"), float("-0.602299"), float("1.54613"), float("1.93115"), float("-0.169884"), float("2.91118"), float("0.158147"), float("2.18802"), float("3.31096"), float("0.613529"), float("-0.650253"), float("-0.596971"), float("0.594206"), float("-2.2655"), float("0.950745"), float("-5.0319"), float("0.238552"), float("-1.86656"), float("4.77206"), float("5.94475"), float("-0.68505"), float("9.23835"), float("0.640713"), float("6.37201"), float("9.65914"), float("1.70325"), float("-1.87225"), float("-1.38542"), float("1.41939"), float("-6.97447"), float("3.12238"), float("-16.035"), float("0.203786"), float("-0.366669"), float("1.09684"), float("1.35079"), float("-0.0881401"), float("1.95277"), float("0.0972831"), float("1.52543"), float("2.1953"), float("0.459335"), float("-0.31404"), float("-0.459989"), float("0.350135"), float("-1.46276"), float("0.626179"), float("-3.48701"), float("0.199294"), float("-0.480529"), float("1.355"), float("1.69083"), float("-0.132597"), float("2.49271"), float("0.146592"), float("1.88249"), float("2.74644"), float("0.553286"), float("-0.447864"), float("-0.53054"), float("0.437591"), float("-1.87747"), float("0.806791"), float("-4.39344"), float("0.220828"), float("-0.207872"), float("0.573615"), float("0.699591"), float("0.0113688"), float("0.976057"), float("0.0260232"), float("0.855339"), float("1.24417"), float("0.275647"), float("-0.186788"), float("-0.356107"), float("0.243965"), float("-0.749682"), float("0.2994"), float("-1.68212"), float("0.241174"), float("-1.10363"), float("2.90263"), float("3.60878"), float("-0.407001"), float("5.52414"), float("0.349477"), float("3.9248"), float("5.90016"), float("1.07812"), float("-1.10482"), float("-0.917721"), float("0.906962"), float("-4.2014"), float("1.84607"), float("-9.6683"), float("1.36937"), float("-0.201856"), float("3.0553"), float("2.67667"), float("-0.260319"), float("4.74185"), float("-0.375263"), float("3.77967"), float("5.23098"), float("1.77046"), float("-0.285616"), float("-0.768714"), float("0.863539"), float("-3.52976"), float("1.80543"), float("-10.6851"), float("0.502645"), float("1.74433"), float("4.67628"), float("2.66077"), float("-2.54735"), float("4.18144"), float("-1.47307"), float("4.59314"), float("5.66316"), float("-0.827085"), float("4.81157"), float("-0.960794"), float("-1.35232"), float("-3.15221"), float("2.04182"), float("-12.572"), float("0.201046"), float("-0.178521"), float("0.583239"), float("0.723738"), float("0.0144532"), float("0.983445"), float("0.0399271"), float("0.853576"), float("1.18278"), float("0.288476"), float("-0.127038"), float("-0.342783"), float("0.199223"), float("-0.716623"), float("0.309393"), float("-1.76484"), float("0.239322"), float("-1.69879"), float("4.45429"), float("5.51995"), float("-0.617142"), float("8.56166"), float("0.579973"), float("5.89439"), float("8.89391"), float("1.59283"), float("-1.65349"), float("-1.29239"), float("1.26163"), float("-6.41205"), float("2.91166"), float("-14.95"), float("0.205208"), float("-0.0755278"), float("0.319165"), float("0.375614"), float("0.0495638"), float("0.447425"), float("5.57229e-05"), float("0.484023"), float("0.630457"), float("0.186171"), float("-0.0182203"), float("-0.281987"), float("0.116728"), float("-0.306026"), float("0.124836"), float("-0.833024"), float("0.213171"), float("-0.713007"), float("1.94641"), float("2.41507"), float("-0.232947"), float("3.63342"), float("0.21747"), float("2.64788"), float("3.92872"), float("0.746185"), float("-0.676981"), float("-0.674473"), float("0.602595"), float("-2.74229"), float("1.20013"), float("-6.38517"), float("0.0413053"), float("0.302979"), float("-0.114017"), float("-0.241728"), float("0.090609"), float("-0.0588678"), float("0.0378077"), float("0.0154076"), float("-0.0298173"), float("0.0667273"), float("-0.0560023"), float("-0.224117"), float("0.0819501"), float("0.106694"), float("0.195902"), float("0.156805"), float("0.208495"), float("-0.48272"), float("1.36807"), float("1.70364"), float("-0.130933"), float("2.51416"), float("0.147009"), float("1.90362"), float("2.77373"), float("0.562943"), float("-0.454971"), float("-0.535181"), float("0.443537"), float("-1.89619"), float("0.820157"), float("-4.4422")] + + + diff --git a/samples/paddle/PaddleX_AutoEncoder_ad_dy2st_train/model.py b/samples/paddle/PaddleX_AutoEncoder_ad_dy2st_train/model.py new file mode 100644 index 000000000..17f00d164 --- /dev/null +++ b/samples/paddle/PaddleX_AutoEncoder_ad_dy2st_train/model.py @@ -0,0 +1,826 @@ +class PirProgram_880737988565058868: + + def __init__(self): + + self.parameter_105 = self.Op("builtin.parameter", 105, input_types=[], output_types=[self.t_dtensor([16], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_1.b_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.parameter_106 = self.Op("builtin.parameter", 106, input_types=[], output_types=[self.t_dtensor([32, 16], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_1.w_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.parameter_107 = self.Op("builtin.parameter", 107, input_types=[], output_types=[self.t_dtensor([32], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_0.b_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.parameter_108 = self.Op("builtin.parameter", 108, input_types=[], output_types=[self.t_dtensor([96, 32], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "struct_name":self.a_str("/Linear/"), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_0.w_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.data_109 = self.Op("pd_op.data", 109, input_types=[], output_types=[self.t_dtensor([16, 2, 96], self.t_f32())], attrs={"struct_name":self.a_str("/Linear/"), "stop_gradient":self.a_array(self.a_bool(True)), "name":self.a_str("args_0"), "shape":self.a_intarray(16, 2, 96), "dtype":self.a_dtype("float32"), "place":self.a_place("undefined", 0), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.matmul_110 = self.Op("pd_op.matmul", 110, input_types=[self.t_dtensor([16, 2, 96], self.t_f32()), self.t_dtensor([96, 32], self.t_f32())], output_types=[self.t_dtensor([16, 2, 32], self.t_f32())], attrs={"struct_name":self.a_str("/Linear/"), "stop_gradient":self.a_array(self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.add_111 = self.Op("pd_op.add", 111, input_types=[self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([32], self.t_f32())], output_types=[self.t_dtensor([16, 2, 32], self.t_f32())], attrs={"struct_name":self.a_str("/Linear/"), "stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.relu_112 = self.Op("pd_op.relu", 112, input_types=[self.t_dtensor([16, 2, 32], self.t_f32())], output_types=[self.t_dtensor([16, 2, 32], self.t_f32())], attrs={"struct_name":self.a_str("/ReLU/"), "stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.full_113 = self.Op("pd_op.full", 113, input_types=[], output_types=[self.t_dtensor([1], self.t_f32())], attrs={"struct_name":self.a_str("/Dropout/"), "stop_gradient":self.a_array(self.a_bool(True)), "shape":self.a_intarray(1), "value":self.a_f64("0.2"), "dtype":self.a_dtype("float32"), "place":self.a_place("cpu"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.dropout_114 = self.Op("pd_op.dropout", 114, input_types=[self.t_dtensor([16, 2, 32], self.t_f32()), self.t_null(), self.t_dtensor([1], self.t_f32())], output_types=[self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([16, 2, 32], self.t_ui8())], attrs={"struct_name":self.a_str("/Dropout/"), "stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "is_test":self.a_bool(False), "seed":self.a_i32(0), "mode":self.a_str("upscale_in_train"), "fix_seed":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) + + self.matmul_115 = self.Op("pd_op.matmul", 115, input_types=[self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([32, 16], self.t_f32())], output_types=[self.t_dtensor([16, 2, 16], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_1/"), "stop_gradient":self.a_array(self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.add_116 = self.Op("pd_op.add", 116, input_types=[self.t_dtensor([16, 2, 16], self.t_f32()), self.t_dtensor([16], self.t_f32())], output_types=[self.t_dtensor([16, 2, 16], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_1/"), "stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.shadow_output_117 = self.Op("builtin.shadow_output", 117, input_types=[self.t_dtensor([16, 2, 32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_0"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.shadow_output_118 = self.Op("builtin.shadow_output", 118, input_types=[self.t_dtensor([16, 2, 32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_1"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.shadow_output_119 = self.Op("builtin.shadow_output", 119, input_types=[self.t_dtensor([1], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_2"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.shadow_output_120 = self.Op("builtin.shadow_output", 120, input_types=[self.t_dtensor([16, 2, 32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_3"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.shadow_output_121 = self.Op("builtin.shadow_output", 121, input_types=[self.t_dtensor([16, 2, 32], self.t_ui8())], output_types=[], attrs={"output_name":self.a_str("middle_4"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.shadow_output_122 = self.Op("builtin.shadow_output", 122, input_types=[self.t_dtensor([16, 2, 16], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_5"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.shadow_output_123 = self.Op("builtin.shadow_output", 123, input_types=[self.t_dtensor([16, 2, 16], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("output_0"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.module_103 = self.Op("builtin.module", 103, input_types=[], output_types=[], attrs={"program":self.a_pointer("0x584e0a20"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array()}, block_positional_arg_names=[[[]]], block_keyword_arg_names=[[{}]], block_positional_arg_types=[[[]]], block_keyword_arg_types=[[[]]], ) + + + + def module_103_block00(self, call): + + def ret_lambda_module_103_block00(): + + parameter_1050, = call(self.parameter_105) + + parameter_1060, = call(self.parameter_106) + + parameter_1070, = call(self.parameter_107) + + parameter_1080, = call(self.parameter_108) + + data_1090, = call(self.data_109) + + matmul_1100, = call(self.matmul_110, data_1090, parameter_1080) + + add_1110, = call(self.add_111, matmul_1100, parameter_1070) + + relu_1120, = call(self.relu_112, add_1110) + + full_1130, = call(self.full_113) + + dropout_1140, dropout_1141, = call(self.dropout_114, relu_1120, None, full_1130) + + matmul_1150, = call(self.matmul_115, dropout_1140, parameter_1060) + + add_1160, = call(self.add_116, matmul_1150, parameter_1050) + + call(self.shadow_output_117, matmul_1100) + + call(self.shadow_output_118, relu_1120) + + call(self.shadow_output_119, full_1130) + + call(self.shadow_output_120, dropout_1140) + + call(self.shadow_output_121, dropout_1141) + + call(self.shadow_output_122, matmul_1150) + + call(self.shadow_output_123, add_1160) + + return ret_lambda_module_103_block00 + + + + def __call__(self, call, *args, **kwargs): + + self.SetArgs(args) + + self.SetKeywordArgs(kwargs) + + return call(self.module_103, blocks=[[(self.module_103_block00,)]]) + + +class PirProgram_5237713637574795923: + + def __init__(self): + + self.parameter_259 = self.Op("builtin.parameter", 259, input_types=[], output_types=[self.t_dtensor([96], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_3.b_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.parameter_260 = self.Op("builtin.parameter", 260, input_types=[], output_types=[self.t_dtensor([32, 96], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_3.w_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.parameter_261 = self.Op("builtin.parameter", 261, input_types=[], output_types=[self.t_dtensor([32], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_2.b_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.parameter_262 = self.Op("builtin.parameter", 262, input_types=[], output_types=[self.t_dtensor([16, 32], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "struct_name":self.a_str("/Linear_2/"), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_2.w_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.data_263 = self.Op("pd_op.data", 263, input_types=[], output_types=[self.t_dtensor([16, 2, -1], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_2/"), "stop_gradient":self.a_array(self.a_bool(False)), "name":self.a_str("args_0"), "shape":self.a_intarray(16, 2, -1), "dtype":self.a_dtype("float32"), "place":self.a_place("undefined", 0), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.matmul_264 = self.Op("pd_op.matmul", 264, input_types=[self.t_dtensor([16, 2, -1], self.t_f32()), self.t_dtensor([16, 32], self.t_f32())], output_types=[self.t_dtensor([16, 2, 32], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_2/"), "stop_gradient":self.a_array(self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.add_265 = self.Op("pd_op.add", 265, input_types=[self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([32], self.t_f32())], output_types=[self.t_dtensor([16, 2, 32], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_2/"), "stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.relu_266 = self.Op("pd_op.relu", 266, input_types=[self.t_dtensor([16, 2, 32], self.t_f32())], output_types=[self.t_dtensor([16, 2, 32], self.t_f32())], attrs={"struct_name":self.a_str("/ReLU_1/"), "stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.full_267 = self.Op("pd_op.full", 267, input_types=[], output_types=[self.t_dtensor([1], self.t_f32())], attrs={"struct_name":self.a_str("/Dropout_1/"), "stop_gradient":self.a_array(self.a_bool(True)), "shape":self.a_intarray(1), "value":self.a_f64("0.2"), "dtype":self.a_dtype("float32"), "place":self.a_place("cpu"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.dropout_268 = self.Op("pd_op.dropout", 268, input_types=[self.t_dtensor([16, 2, 32], self.t_f32()), self.t_null(), self.t_dtensor([1], self.t_f32())], output_types=[self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([16, 2, 32], self.t_ui8())], attrs={"struct_name":self.a_str("/Dropout_1/"), "stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "is_test":self.a_bool(False), "seed":self.a_i32(0), "mode":self.a_str("upscale_in_train"), "fix_seed":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) + + self.matmul_269 = self.Op("pd_op.matmul", 269, input_types=[self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([32, 96], self.t_f32())], output_types=[self.t_dtensor([16, 2, 96], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_3/"), "stop_gradient":self.a_array(self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.add_270 = self.Op("pd_op.add", 270, input_types=[self.t_dtensor([16, 2, 96], self.t_f32()), self.t_dtensor([96], self.t_f32())], output_types=[self.t_dtensor([16, 2, 96], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_3/"), "stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.shadow_output_271 = self.Op("builtin.shadow_output", 271, input_types=[self.t_dtensor([16, 2, 32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_0"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.shadow_output_272 = self.Op("builtin.shadow_output", 272, input_types=[self.t_dtensor([16, 2, 32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_1"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.shadow_output_273 = self.Op("builtin.shadow_output", 273, input_types=[self.t_dtensor([1], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_2"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.shadow_output_274 = self.Op("builtin.shadow_output", 274, input_types=[self.t_dtensor([16, 2, 32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_3"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.shadow_output_275 = self.Op("builtin.shadow_output", 275, input_types=[self.t_dtensor([16, 2, 32], self.t_ui8())], output_types=[], attrs={"output_name":self.a_str("middle_4"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.shadow_output_276 = self.Op("builtin.shadow_output", 276, input_types=[self.t_dtensor([16, 2, 96], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_5"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.shadow_output_277 = self.Op("builtin.shadow_output", 277, input_types=[self.t_dtensor([16, 2, 96], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("output_0"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.module_257 = self.Op("builtin.module", 257, input_types=[], output_types=[], attrs={"program":self.a_pointer("0x6b30d5a0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array()}, block_positional_arg_names=[[[]]], block_keyword_arg_names=[[{}]], block_positional_arg_types=[[[]]], block_keyword_arg_types=[[[]]], ) + + + + def module_257_block00(self, call): + + def ret_lambda_module_257_block00(): + + parameter_2590, = call(self.parameter_259) + + parameter_2600, = call(self.parameter_260) + + parameter_2610, = call(self.parameter_261) + + parameter_2620, = call(self.parameter_262) + + data_2630, = call(self.data_263) + + matmul_2640, = call(self.matmul_264, data_2630, parameter_2620) + + add_2650, = call(self.add_265, matmul_2640, parameter_2610) + + relu_2660, = call(self.relu_266, add_2650) + + full_2670, = call(self.full_267) + + dropout_2680, dropout_2681, = call(self.dropout_268, relu_2660, None, full_2670) + + matmul_2690, = call(self.matmul_269, dropout_2680, parameter_2600) + + add_2700, = call(self.add_270, matmul_2690, parameter_2590) + + call(self.shadow_output_271, matmul_2640) + + call(self.shadow_output_272, relu_2660) + + call(self.shadow_output_273, full_2670) + + call(self.shadow_output_274, dropout_2680) + + call(self.shadow_output_275, dropout_2681) + + call(self.shadow_output_276, matmul_2690) + + call(self.shadow_output_277, add_2700) + + return ret_lambda_module_257_block00 + + + + def __call__(self, call, *args, **kwargs): + + self.SetArgs(args) + + self.SetKeywordArgs(kwargs) + + return call(self.module_257, blocks=[[(self.module_257_block00,)]]) + + +class PirProgram_1242071021843173094: + + def __init__(self): + + self.add_grad_278 = self.Op("pd_op.add_grad", 278, input_types=[self.t_dtensor([16, 2, 96], self.t_f32()), self.t_dtensor([96], self.t_f32()), self.t_dtensor([16, 2, 96], self.t_f32())], output_types=[self.t_dtensor([16, 2, 96], self.t_f32()), self.t_dtensor([96], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "axis":self.a_i32(-1), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) + + self.matmul_grad_279 = self.Op("pd_op.matmul_grad", 279, input_types=[self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([32, 96], self.t_f32()), self.t_dtensor([16, 2, 96], self.t_f32())], output_types=[self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([32, 96], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) + + self.dropout_grad_280 = self.Op("pd_op.dropout_grad", 280, input_types=[self.t_dtensor([16, 2, 32], self.t_ui8()), self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([1], self.t_f32())], output_types=[self.t_dtensor([16, 2, 32], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False)), "is_test":self.a_bool(False), "mode":self.a_str("upscale_in_train"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.relu_grad_281 = self.Op("pd_op.relu_grad", 281, input_types=[self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([16, 2, 32], self.t_f32())], output_types=[self.t_dtensor([16, 2, 32], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.add_grad_282 = self.Op("pd_op.add_grad", 282, input_types=[self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([32], self.t_f32()), self.t_dtensor([16, 2, 32], self.t_f32())], output_types=[self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([32], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "axis":self.a_i32(-1), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) + + self.matmul_grad_283 = self.Op("pd_op.matmul_grad", 283, input_types=[self.t_dtensor([16, 2, -1], self.t_f32()), self.t_dtensor([16, 32], self.t_f32()), self.t_dtensor([16, 2, 32], self.t_f32())], output_types=[self.t_dtensor([16, 2, -1], self.t_f32()), self.t_dtensor([16, 32], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) + + self.shadow_output_284 = self.Op("builtin.shadow_output", 284, input_types=[self.t_dtensor([16, 2, -1], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("input_grad_0"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.shadow_output_285 = self.Op("builtin.shadow_output", 285, input_types=[self.t_dtensor([32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("param_grad_0"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.shadow_output_286 = self.Op("builtin.shadow_output", 286, input_types=[self.t_dtensor([16, 32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("param_grad_1"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.shadow_output_287 = self.Op("builtin.shadow_output", 287, input_types=[self.t_dtensor([96], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("param_grad_2"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.shadow_output_288 = self.Op("builtin.shadow_output", 288, input_types=[self.t_dtensor([32, 96], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("param_grad_3"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.module_258 = self.Op("builtin.module", 258, input_types=[], output_types=[], attrs={"program":self.a_pointer("0x6b30eee0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array()}, block_positional_arg_names=[[[]]], block_keyword_arg_names=[[{"middle_5": "arg_1798365008", "middle_4": "arg_1798348928", "output_grad_0": "arg_1798365584", "linear_3.w_0": "arg_1481522992", "middle_2": "arg_1798347072", "linear_3.b_0": "arg_1798306704", "middle_3": "arg_1798348720", "middle_1": "arg_1798334304", "linear_2.w_0": "arg_1798084816", "middle_0": "arg_1798334032", "linear_2.b_0": "arg_1798322512", "args_0": "arg_1798315712"}]], block_positional_arg_types=[[[]]], block_keyword_arg_types=[[[self.t_dtensor([16, 2, 96], self.t_f32()), self.t_dtensor([16, 2, 32], self.t_ui8()), self.t_dtensor([16, 2, 96], self.t_f32()), self.t_dtensor([32, 96], self.t_f32()), self.t_dtensor([1], self.t_f32()), self.t_dtensor([96], self.t_f32()), self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([16, 32], self.t_f32()), self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([32], self.t_f32()), self.t_dtensor([16, 2, -1], self.t_f32())]]], ) + + + + def module_258_block00(self, call): + + def ret_lambda_module_258_block00(arg_1798365008, arg_1798348928, arg_1798365584, arg_1481522992, arg_1798347072, arg_1798306704, arg_1798348720, arg_1798334304, arg_1798084816, arg_1798334032, arg_1798322512, arg_1798315712): + + add_grad_2780, add_grad_2781, = call(self.add_grad_278, arg_1798365008, arg_1798306704, arg_1798365584) + + matmul_grad_2790, matmul_grad_2791, = call(self.matmul_grad_279, arg_1798348720, arg_1481522992, add_grad_2780) + + dropout_grad_2800, = call(self.dropout_grad_280, arg_1798348928, matmul_grad_2790, arg_1798347072) + + relu_grad_2810, = call(self.relu_grad_281, arg_1798334304, dropout_grad_2800) + + add_grad_2820, add_grad_2821, = call(self.add_grad_282, arg_1798334032, arg_1798322512, relu_grad_2810) + + matmul_grad_2830, matmul_grad_2831, = call(self.matmul_grad_283, arg_1798315712, arg_1798084816, add_grad_2820) + + call(self.shadow_output_284, matmul_grad_2830) + + call(self.shadow_output_285, add_grad_2821) + + call(self.shadow_output_286, matmul_grad_2831) + + call(self.shadow_output_287, add_grad_2781) + + call(self.shadow_output_288, matmul_grad_2791) + + return ret_lambda_module_258_block00 + + + + def __call__(self, call, *args, **kwargs): + + self.SetArgs(args) + + self.SetKeywordArgs(kwargs) + + return call(self.module_258, blocks=[[(self.module_258_block00,)]]) + + +class PirProgram_692320333968606381: + + def __init__(self): + + self.add_grad_124 = self.Op("pd_op.add_grad", 124, input_types=[self.t_dtensor([16, 2, 16], self.t_f32()), self.t_dtensor([16], self.t_f32()), self.t_dtensor([16, 2, 16], self.t_f32())], output_types=[self.t_dtensor([16, 2, 16], self.t_f32()), self.t_dtensor([16], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "axis":self.a_i32(-1), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) + + self.matmul_grad_125 = self.Op("pd_op.matmul_grad", 125, input_types=[self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([32, 16], self.t_f32()), self.t_dtensor([16, 2, 16], self.t_f32())], output_types=[self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([32, 16], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) + + self.dropout_grad_126 = self.Op("pd_op.dropout_grad", 126, input_types=[self.t_dtensor([16, 2, 32], self.t_ui8()), self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([1], self.t_f32())], output_types=[self.t_dtensor([16, 2, 32], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False)), "is_test":self.a_bool(False), "mode":self.a_str("upscale_in_train"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.relu_grad_127 = self.Op("pd_op.relu_grad", 127, input_types=[self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([16, 2, 32], self.t_f32())], output_types=[self.t_dtensor([16, 2, 32], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.add_grad_128 = self.Op("pd_op.add_grad", 128, input_types=[self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([32], self.t_f32()), self.t_dtensor([16, 2, 32], self.t_f32())], output_types=[self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([32], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "axis":self.a_i32(-1), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) + + self.matmul_grad_129 = self.Op("pd_op.matmul_grad", 129, input_types=[self.t_dtensor([16, 2, 96], self.t_f32()), self.t_dtensor([96, 32], self.t_f32()), self.t_dtensor([16, 2, 32], self.t_f32())], output_types=[self.t_null(), self.t_dtensor([96, 32], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) + + self.shadow_output_130 = self.Op("builtin.shadow_output", 130, input_types=[self.t_dtensor([32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("param_grad_0"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.shadow_output_131 = self.Op("builtin.shadow_output", 131, input_types=[self.t_dtensor([96, 32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("param_grad_1"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.shadow_output_132 = self.Op("builtin.shadow_output", 132, input_types=[self.t_dtensor([16], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("param_grad_2"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.shadow_output_133 = self.Op("builtin.shadow_output", 133, input_types=[self.t_dtensor([32, 16], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("param_grad_3"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.module_104 = self.Op("builtin.module", 104, input_types=[], output_types=[], attrs={"program":self.a_pointer("0x584e2360"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array()}, block_positional_arg_names=[[[]]], block_keyword_arg_names=[[{"output_grad_0": "arg_1481526688", "middle_5": "arg_1497418592", "middle_4": "arg_1497418384", "middle_2": "arg_1481508288", "middle_0": "arg_1481527936", "linear_1.w_0": "arg_1457990560", "linear_1.b_0": "arg_1457989680", "middle_3": "arg_1481508496", "middle_1": "arg_1481528144", "linear_0.w_0": "arg_1497436400", "linear_0.b_0": "arg_1457913760", "args_0": "arg_1497471536"}]], block_positional_arg_types=[[[]]], block_keyword_arg_types=[[[self.t_dtensor([16, 2, 16], self.t_f32()), self.t_dtensor([16, 2, 16], self.t_f32()), self.t_dtensor([16, 2, 32], self.t_ui8()), self.t_dtensor([1], self.t_f32()), self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([32, 16], self.t_f32()), self.t_dtensor([16], self.t_f32()), self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([96, 32], self.t_f32()), self.t_dtensor([32], self.t_f32()), self.t_dtensor([16, 2, 96], self.t_f32())]]], ) + + + + def module_104_block00(self, call): + + def ret_lambda_module_104_block00(arg_1481526688, arg_1497418592, arg_1497418384, arg_1481508288, arg_1481527936, arg_1457990560, arg_1457989680, arg_1481508496, arg_1481528144, arg_1497436400, arg_1457913760, arg_1497471536): + + add_grad_1240, add_grad_1241, = call(self.add_grad_124, arg_1497418592, arg_1457989680, arg_1481526688) + + matmul_grad_1250, matmul_grad_1251, = call(self.matmul_grad_125, arg_1481508496, arg_1457990560, add_grad_1240) + + dropout_grad_1260, = call(self.dropout_grad_126, arg_1497418384, matmul_grad_1250, arg_1481508288) + + relu_grad_1270, = call(self.relu_grad_127, arg_1481528144, dropout_grad_1260) + + add_grad_1280, add_grad_1281, = call(self.add_grad_128, arg_1481527936, arg_1457913760, relu_grad_1270) + + matmul_grad_1290, matmul_grad_1291, = call(self.matmul_grad_129, arg_1497471536, arg_1497436400, add_grad_1280) + + call(self.shadow_output_130, add_grad_1281) + + call(self.shadow_output_131, matmul_grad_1291) + + call(self.shadow_output_132, add_grad_1241) + + call(self.shadow_output_133, matmul_grad_1251) + + return ret_lambda_module_104_block00 + + + + def __call__(self, call, *args, **kwargs): + + self.SetArgs(args) + + self.SetKeywordArgs(kwargs) + + return call(self.module_104, blocks=[[(self.module_104_block00,)]]) + + +class PirProgram_5426476789613677471: + + def __init__(self): + + self.parameter_459 = self.Op("builtin.parameter", 459, input_types=[], output_types=[self.t_dtensor([16], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_1.b_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.parameter_460 = self.Op("builtin.parameter", 460, input_types=[], output_types=[self.t_dtensor([32, 16], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_1.w_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.parameter_461 = self.Op("builtin.parameter", 461, input_types=[], output_types=[self.t_dtensor([32], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_0.b_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.parameter_462 = self.Op("builtin.parameter", 462, input_types=[], output_types=[self.t_dtensor([96, 32], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "struct_name":self.a_str("/Linear_4/"), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_0.w_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.data_463 = self.Op("pd_op.data", 463, input_types=[], output_types=[self.t_dtensor([-1, 2, -1], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_4/"), "stop_gradient":self.a_array(self.a_bool(True)), "name":self.a_str("args_0"), "shape":self.a_intarray(-1, 2, -1), "dtype":self.a_dtype("float32"), "place":self.a_place("undefined", 0), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.matmul_464 = self.Op("pd_op.matmul", 464, input_types=[self.t_dtensor([-1, 2, -1], self.t_f32()), self.t_dtensor([96, 32], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_4/"), "stop_gradient":self.a_array(self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.add_465 = self.Op("pd_op.add", 465, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([32], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_4/"), "stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.relu_466 = self.Op("pd_op.relu", 466, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], attrs={"struct_name":self.a_str("/ReLU_2/"), "stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.full_467 = self.Op("pd_op.full", 467, input_types=[], output_types=[self.t_dtensor([1], self.t_f32())], attrs={"struct_name":self.a_str("/Dropout_2/"), "stop_gradient":self.a_array(self.a_bool(True)), "shape":self.a_intarray(1), "value":self.a_f64("0.2"), "dtype":self.a_dtype("float32"), "place":self.a_place("cpu"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.dropout_468 = self.Op("pd_op.dropout", 468, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_null(), self.t_dtensor([1], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([-1, 2, 32], self.t_ui8())], attrs={"struct_name":self.a_str("/Dropout_2/"), "stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "is_test":self.a_bool(False), "seed":self.a_i32(0), "mode":self.a_str("upscale_in_train"), "fix_seed":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) + + self.matmul_469 = self.Op("pd_op.matmul", 469, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([32, 16], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 16], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_5/"), "stop_gradient":self.a_array(self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.add_470 = self.Op("pd_op.add", 470, input_types=[self.t_dtensor([-1, 2, 16], self.t_f32()), self.t_dtensor([16], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 16], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_5/"), "stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.shadow_output_471 = self.Op("builtin.shadow_output", 471, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_0"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.shadow_output_472 = self.Op("builtin.shadow_output", 472, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_1"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.shadow_output_473 = self.Op("builtin.shadow_output", 473, input_types=[self.t_dtensor([1], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_2"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.shadow_output_474 = self.Op("builtin.shadow_output", 474, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_3"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.shadow_output_475 = self.Op("builtin.shadow_output", 475, input_types=[self.t_dtensor([-1, 2, 32], self.t_ui8())], output_types=[], attrs={"output_name":self.a_str("middle_4"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.shadow_output_476 = self.Op("builtin.shadow_output", 476, input_types=[self.t_dtensor([-1, 2, 16], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_5"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.shadow_output_477 = self.Op("builtin.shadow_output", 477, input_types=[self.t_dtensor([-1, 2, 16], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("output_0"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.module_457 = self.Op("builtin.module", 457, input_types=[], output_types=[], attrs={"program":self.a_pointer("0x6b3cf350"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array()}, block_positional_arg_names=[[[]]], block_keyword_arg_names=[[{}]], block_positional_arg_types=[[[]]], block_keyword_arg_types=[[[]]], ) + + + + def module_457_block00(self, call): + + def ret_lambda_module_457_block00(): + + parameter_4590, = call(self.parameter_459) + + parameter_4600, = call(self.parameter_460) + + parameter_4610, = call(self.parameter_461) + + parameter_4620, = call(self.parameter_462) + + data_4630, = call(self.data_463) + + matmul_4640, = call(self.matmul_464, data_4630, parameter_4620) + + add_4650, = call(self.add_465, matmul_4640, parameter_4610) + + relu_4660, = call(self.relu_466, add_4650) + + full_4670, = call(self.full_467) + + dropout_4680, dropout_4681, = call(self.dropout_468, relu_4660, None, full_4670) + + matmul_4690, = call(self.matmul_469, dropout_4680, parameter_4600) + + add_4700, = call(self.add_470, matmul_4690, parameter_4590) + + call(self.shadow_output_471, matmul_4640) + + call(self.shadow_output_472, relu_4660) + + call(self.shadow_output_473, full_4670) + + call(self.shadow_output_474, dropout_4680) + + call(self.shadow_output_475, dropout_4681) + + call(self.shadow_output_476, matmul_4690) + + call(self.shadow_output_477, add_4700) + + return ret_lambda_module_457_block00 + + + + def __call__(self, call, *args, **kwargs): + + self.SetArgs(args) + + self.SetKeywordArgs(kwargs) + + return call(self.module_457, blocks=[[(self.module_457_block00,)]]) + + +class PirProgram_7028096434133672773: + + def __init__(self): + + self.parameter_613 = self.Op("builtin.parameter", 613, input_types=[], output_types=[self.t_dtensor([96], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_3.b_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.parameter_614 = self.Op("builtin.parameter", 614, input_types=[], output_types=[self.t_dtensor([32, 96], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_3.w_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.parameter_615 = self.Op("builtin.parameter", 615, input_types=[], output_types=[self.t_dtensor([32], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_2.b_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.parameter_616 = self.Op("builtin.parameter", 616, input_types=[], output_types=[self.t_dtensor([16, 32], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "struct_name":self.a_str("/Linear_6/"), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_2.w_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.data_617 = self.Op("pd_op.data", 617, input_types=[], output_types=[self.t_dtensor([-1, 2, -1], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_6/"), "stop_gradient":self.a_array(self.a_bool(False)), "name":self.a_str("args_0"), "shape":self.a_intarray(-1, 2, -1), "dtype":self.a_dtype("float32"), "place":self.a_place("undefined", 0), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.matmul_618 = self.Op("pd_op.matmul", 618, input_types=[self.t_dtensor([-1, 2, -1], self.t_f32()), self.t_dtensor([16, 32], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_6/"), "stop_gradient":self.a_array(self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.add_619 = self.Op("pd_op.add", 619, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([32], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_6/"), "stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.relu_620 = self.Op("pd_op.relu", 620, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], attrs={"struct_name":self.a_str("/ReLU_3/"), "stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.full_621 = self.Op("pd_op.full", 621, input_types=[], output_types=[self.t_dtensor([1], self.t_f32())], attrs={"struct_name":self.a_str("/Dropout_3/"), "stop_gradient":self.a_array(self.a_bool(True)), "shape":self.a_intarray(1), "value":self.a_f64("0.2"), "dtype":self.a_dtype("float32"), "place":self.a_place("cpu"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.dropout_622 = self.Op("pd_op.dropout", 622, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_null(), self.t_dtensor([1], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([-1, 2, 32], self.t_ui8())], attrs={"struct_name":self.a_str("/Dropout_3/"), "stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "is_test":self.a_bool(False), "seed":self.a_i32(0), "mode":self.a_str("upscale_in_train"), "fix_seed":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) + + self.matmul_623 = self.Op("pd_op.matmul", 623, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([32, 96], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 96], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_7/"), "stop_gradient":self.a_array(self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.add_624 = self.Op("pd_op.add", 624, input_types=[self.t_dtensor([-1, 2, 96], self.t_f32()), self.t_dtensor([96], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 96], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_7/"), "stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.shadow_output_625 = self.Op("builtin.shadow_output", 625, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_0"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.shadow_output_626 = self.Op("builtin.shadow_output", 626, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_1"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.shadow_output_627 = self.Op("builtin.shadow_output", 627, input_types=[self.t_dtensor([1], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_2"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.shadow_output_628 = self.Op("builtin.shadow_output", 628, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_3"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.shadow_output_629 = self.Op("builtin.shadow_output", 629, input_types=[self.t_dtensor([-1, 2, 32], self.t_ui8())], output_types=[], attrs={"output_name":self.a_str("middle_4"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.shadow_output_630 = self.Op("builtin.shadow_output", 630, input_types=[self.t_dtensor([-1, 2, 96], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_5"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.shadow_output_631 = self.Op("builtin.shadow_output", 631, input_types=[self.t_dtensor([-1, 2, 96], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("output_0"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.module_611 = self.Op("builtin.module", 611, input_types=[], output_types=[], attrs={"program":self.a_pointer("0x6b400f50"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array()}, block_positional_arg_names=[[[]]], block_keyword_arg_names=[[{}]], block_positional_arg_types=[[[]]], block_keyword_arg_types=[[[]]], ) + + + + def module_611_block00(self, call): + + def ret_lambda_module_611_block00(): + + parameter_6130, = call(self.parameter_613) + + parameter_6140, = call(self.parameter_614) + + parameter_6150, = call(self.parameter_615) + + parameter_6160, = call(self.parameter_616) + + data_6170, = call(self.data_617) + + matmul_6180, = call(self.matmul_618, data_6170, parameter_6160) + + add_6190, = call(self.add_619, matmul_6180, parameter_6150) + + relu_6200, = call(self.relu_620, add_6190) + + full_6210, = call(self.full_621) + + dropout_6220, dropout_6221, = call(self.dropout_622, relu_6200, None, full_6210) + + matmul_6230, = call(self.matmul_623, dropout_6220, parameter_6140) + + add_6240, = call(self.add_624, matmul_6230, parameter_6130) + + call(self.shadow_output_625, matmul_6180) + + call(self.shadow_output_626, relu_6200) + + call(self.shadow_output_627, full_6210) + + call(self.shadow_output_628, dropout_6220) + + call(self.shadow_output_629, dropout_6221) + + call(self.shadow_output_630, matmul_6230) + + call(self.shadow_output_631, add_6240) + + return ret_lambda_module_611_block00 + + + + def __call__(self, call, *args, **kwargs): + + self.SetArgs(args) + + self.SetKeywordArgs(kwargs) + + return call(self.module_611, blocks=[[(self.module_611_block00,)]]) + + +class PirProgram_3000443918524221177: + + def __init__(self): + + self.add_grad_632 = self.Op("pd_op.add_grad", 632, input_types=[self.t_dtensor([-1, 2, 96], self.t_f32()), self.t_dtensor([96], self.t_f32()), self.t_dtensor([-1, 2, 96], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 96], self.t_f32()), self.t_dtensor([96], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "axis":self.a_i32(-1), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) + + self.matmul_grad_633 = self.Op("pd_op.matmul_grad", 633, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([32, 96], self.t_f32()), self.t_dtensor([-1, 2, 96], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([32, 96], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) + + self.dropout_grad_634 = self.Op("pd_op.dropout_grad", 634, input_types=[self.t_dtensor([-1, 2, 32], self.t_ui8()), self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([1], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False)), "is_test":self.a_bool(False), "mode":self.a_str("upscale_in_train"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.relu_grad_635 = self.Op("pd_op.relu_grad", 635, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([-1, 2, 32], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.add_grad_636 = self.Op("pd_op.add_grad", 636, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([32], self.t_f32()), self.t_dtensor([-1, 2, 32], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([32], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "axis":self.a_i32(-1), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) + + self.matmul_grad_637 = self.Op("pd_op.matmul_grad", 637, input_types=[self.t_dtensor([-1, 2, -1], self.t_f32()), self.t_dtensor([16, 32], self.t_f32()), self.t_dtensor([-1, 2, 32], self.t_f32())], output_types=[self.t_dtensor([-1, 2, -1], self.t_f32()), self.t_dtensor([16, 32], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) + + self.shadow_output_638 = self.Op("builtin.shadow_output", 638, input_types=[self.t_dtensor([-1, 2, -1], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("input_grad_0"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.shadow_output_639 = self.Op("builtin.shadow_output", 639, input_types=[self.t_dtensor([32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("param_grad_0"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.shadow_output_640 = self.Op("builtin.shadow_output", 640, input_types=[self.t_dtensor([16, 32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("param_grad_1"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.shadow_output_641 = self.Op("builtin.shadow_output", 641, input_types=[self.t_dtensor([96], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("param_grad_2"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.shadow_output_642 = self.Op("builtin.shadow_output", 642, input_types=[self.t_dtensor([32, 96], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("param_grad_3"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.module_612 = self.Op("builtin.module", 612, input_types=[], output_types=[], attrs={"program":self.a_pointer("0x6b3f4bd0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array()}, block_positional_arg_names=[[[]]], block_keyword_arg_names=[[{"middle_5": "arg_1799310528", "middle_4": "arg_1799317136", "output_grad_0": "arg_1799310976", "linear_3.w_0": "arg_1799293984", "middle_2": "arg_1799316720", "linear_3.b_0": "arg_1799157344", "middle_3": "arg_1799316928", "middle_1": "arg_1799301824", "linear_2.w_0": "arg_1799170736", "middle_0": "arg_1799301552", "linear_2.b_0": "arg_1799157488", "args_0": "arg_1799179680"}]], block_positional_arg_types=[[[]]], block_keyword_arg_types=[[[self.t_dtensor([-1, 2, 96], self.t_f32()), self.t_dtensor([-1, 2, 32], self.t_ui8()), self.t_dtensor([-1, 2, 96], self.t_f32()), self.t_dtensor([32, 96], self.t_f32()), self.t_dtensor([1], self.t_f32()), self.t_dtensor([96], self.t_f32()), self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([16, 32], self.t_f32()), self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([32], self.t_f32()), self.t_dtensor([-1, 2, -1], self.t_f32())]]], ) + + + + def module_612_block00(self, call): + + def ret_lambda_module_612_block00(arg_1799310528, arg_1799317136, arg_1799310976, arg_1799293984, arg_1799316720, arg_1799157344, arg_1799316928, arg_1799301824, arg_1799170736, arg_1799301552, arg_1799157488, arg_1799179680): + + add_grad_6320, add_grad_6321, = call(self.add_grad_632, arg_1799310528, arg_1799157344, arg_1799310976) + + matmul_grad_6330, matmul_grad_6331, = call(self.matmul_grad_633, arg_1799316928, arg_1799293984, add_grad_6320) + + dropout_grad_6340, = call(self.dropout_grad_634, arg_1799317136, matmul_grad_6330, arg_1799316720) + + relu_grad_6350, = call(self.relu_grad_635, arg_1799301824, dropout_grad_6340) + + add_grad_6360, add_grad_6361, = call(self.add_grad_636, arg_1799301552, arg_1799157488, relu_grad_6350) + + matmul_grad_6370, matmul_grad_6371, = call(self.matmul_grad_637, arg_1799179680, arg_1799170736, add_grad_6360) + + call(self.shadow_output_638, matmul_grad_6370) + + call(self.shadow_output_639, add_grad_6361) + + call(self.shadow_output_640, matmul_grad_6371) + + call(self.shadow_output_641, add_grad_6321) + + call(self.shadow_output_642, matmul_grad_6331) + + return ret_lambda_module_612_block00 + + + + def __call__(self, call, *args, **kwargs): + + self.SetArgs(args) + + self.SetKeywordArgs(kwargs) + + return call(self.module_612, blocks=[[(self.module_612_block00,)]]) + + +class PirProgram_3111236241086672326: + + def __init__(self): + + self.add_grad_478 = self.Op("pd_op.add_grad", 478, input_types=[self.t_dtensor([-1, 2, 16], self.t_f32()), self.t_dtensor([16], self.t_f32()), self.t_dtensor([-1, 2, 16], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 16], self.t_f32()), self.t_dtensor([16], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "axis":self.a_i32(-1), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) + + self.matmul_grad_479 = self.Op("pd_op.matmul_grad", 479, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([32, 16], self.t_f32()), self.t_dtensor([-1, 2, 16], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([32, 16], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) + + self.dropout_grad_480 = self.Op("pd_op.dropout_grad", 480, input_types=[self.t_dtensor([-1, 2, 32], self.t_ui8()), self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([1], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False)), "is_test":self.a_bool(False), "mode":self.a_str("upscale_in_train"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.relu_grad_481 = self.Op("pd_op.relu_grad", 481, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([-1, 2, 32], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.add_grad_482 = self.Op("pd_op.add_grad", 482, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([32], self.t_f32()), self.t_dtensor([-1, 2, 32], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([32], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "axis":self.a_i32(-1), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) + + self.matmul_grad_483 = self.Op("pd_op.matmul_grad", 483, input_types=[self.t_dtensor([-1, 2, -1], self.t_f32()), self.t_dtensor([96, 32], self.t_f32()), self.t_dtensor([-1, 2, 32], self.t_f32())], output_types=[self.t_null(), self.t_dtensor([96, 32], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) + + self.shadow_output_484 = self.Op("builtin.shadow_output", 484, input_types=[self.t_dtensor([32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("param_grad_0"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.shadow_output_485 = self.Op("builtin.shadow_output", 485, input_types=[self.t_dtensor([96, 32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("param_grad_1"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.shadow_output_486 = self.Op("builtin.shadow_output", 486, input_types=[self.t_dtensor([16], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("param_grad_2"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.shadow_output_487 = self.Op("builtin.shadow_output", 487, input_types=[self.t_dtensor([32, 16], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("param_grad_3"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.module_458 = self.Op("builtin.module", 458, input_types=[], output_types=[], attrs={"program":self.a_pointer("0x6b3d0ad0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array()}, block_positional_arg_names=[[[]]], block_keyword_arg_names=[[{"output_grad_0": "arg_1799168768", "middle_5": "arg_1799149040", "middle_4": "arg_1799148832", "middle_2": "arg_1799159872", "middle_0": "arg_1799136912", "linear_1.w_0": "arg_1735320288", "linear_1.b_0": "arg_1798932704", "middle_3": "arg_1799148624", "middle_1": "arg_1799137056", "linear_0.w_0": "arg_1457571760", "linear_0.b_0": "arg_1798932016", "args_0": "arg_1798507504"}]], block_positional_arg_types=[[[]]], block_keyword_arg_types=[[[self.t_dtensor([-1, 2, 16], self.t_f32()), self.t_dtensor([-1, 2, 16], self.t_f32()), self.t_dtensor([-1, 2, 32], self.t_ui8()), self.t_dtensor([1], self.t_f32()), self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([32, 16], self.t_f32()), self.t_dtensor([16], self.t_f32()), self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([96, 32], self.t_f32()), self.t_dtensor([32], self.t_f32()), self.t_dtensor([-1, 2, -1], self.t_f32())]]], ) + + + + def module_458_block00(self, call): + + def ret_lambda_module_458_block00(arg_1799168768, arg_1799149040, arg_1799148832, arg_1799159872, arg_1799136912, arg_1735320288, arg_1798932704, arg_1799148624, arg_1799137056, arg_1457571760, arg_1798932016, arg_1798507504): + + add_grad_4780, add_grad_4781, = call(self.add_grad_478, arg_1799149040, arg_1798932704, arg_1799168768) + + matmul_grad_4790, matmul_grad_4791, = call(self.matmul_grad_479, arg_1799148624, arg_1735320288, add_grad_4780) + + dropout_grad_4800, = call(self.dropout_grad_480, arg_1799148832, matmul_grad_4790, arg_1799159872) + + relu_grad_4810, = call(self.relu_grad_481, arg_1799137056, dropout_grad_4800) + + add_grad_4820, add_grad_4821, = call(self.add_grad_482, arg_1799136912, arg_1798932016, relu_grad_4810) + + matmul_grad_4830, matmul_grad_4831, = call(self.matmul_grad_483, arg_1798507504, arg_1457571760, add_grad_4820) + + call(self.shadow_output_484, add_grad_4821) + + call(self.shadow_output_485, matmul_grad_4831) + + call(self.shadow_output_486, add_grad_4781) + + call(self.shadow_output_487, matmul_grad_4791) + + return ret_lambda_module_458_block00 + + + + def __call__(self, call, *args, **kwargs): + + self.SetArgs(args) + + self.SetKeywordArgs(kwargs) + + return call(self.module_458, blocks=[[(self.module_458_block00,)]]) + + +class PirProgram_4630264566612914126: + + def __init__(self): + + self.parameter_750 = self.Op("builtin.parameter", 750, input_types=[], output_types=[self.t_dtensor([16], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_1.b_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.parameter_751 = self.Op("builtin.parameter", 751, input_types=[], output_types=[self.t_dtensor([32, 16], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_1.w_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.parameter_752 = self.Op("builtin.parameter", 752, input_types=[], output_types=[self.t_dtensor([32], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_0.b_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.parameter_753 = self.Op("builtin.parameter", 753, input_types=[], output_types=[self.t_dtensor([96, 32], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "struct_name":self.a_str("/Linear_8/"), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_0.w_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.data_754 = self.Op("pd_op.data", 754, input_types=[], output_types=[self.t_dtensor([-1, 2, -1], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_8/"), "stop_gradient":self.a_array(self.a_bool(True)), "name":self.a_str("args_0"), "shape":self.a_intarray(-1, 2, -1), "dtype":self.a_dtype("float32"), "place":self.a_place("undefined", 0), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.matmul_755 = self.Op("pd_op.matmul", 755, input_types=[self.t_dtensor([-1, 2, -1], self.t_f32()), self.t_dtensor([96, 32], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_8/"), "stop_gradient":self.a_array(self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.add_756 = self.Op("pd_op.add", 756, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([32], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_8/"), "stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.relu_757 = self.Op("pd_op.relu", 757, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], attrs={"struct_name":self.a_str("/ReLU_4/"), "stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.full_758 = self.Op("pd_op.full", 758, input_types=[], output_types=[self.t_dtensor([1], self.t_f32())], attrs={"struct_name":self.a_str("/Dropout_4/"), "stop_gradient":self.a_array(self.a_bool(True)), "shape":self.a_intarray(1), "value":self.a_f64("0.2"), "dtype":self.a_dtype("float32"), "place":self.a_place("cpu"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.dropout_759 = self.Op("pd_op.dropout", 759, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_null(), self.t_dtensor([1], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([-1, 2, 32], self.t_ui8())], attrs={"struct_name":self.a_str("/Dropout_4/"), "stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "is_test":self.a_bool(True), "seed":self.a_i32(0), "mode":self.a_str("upscale_in_train"), "fix_seed":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) + + self.matmul_760 = self.Op("pd_op.matmul", 760, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([32, 16], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 16], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_9/"), "stop_gradient":self.a_array(self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.add_761 = self.Op("pd_op.add", 761, input_types=[self.t_dtensor([-1, 2, 16], self.t_f32()), self.t_dtensor([16], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 16], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_9/"), "stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.shadow_output_762 = self.Op("builtin.shadow_output", 762, input_types=[self.t_dtensor([-1, 2, 16], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("output_0"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.module_748 = self.Op("builtin.module", 748, input_types=[], output_types=[], attrs={"program":self.a_pointer("0x6b3cafd0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array()}, block_positional_arg_names=[[[]]], block_keyword_arg_names=[[{}]], block_positional_arg_types=[[[]]], block_keyword_arg_types=[[[]]], ) + + + + def module_748_block00(self, call): + + def ret_lambda_module_748_block00(): + + parameter_7500, = call(self.parameter_750) + + parameter_7510, = call(self.parameter_751) + + parameter_7520, = call(self.parameter_752) + + parameter_7530, = call(self.parameter_753) + + data_7540, = call(self.data_754) + + matmul_7550, = call(self.matmul_755, data_7540, parameter_7530) + + add_7560, = call(self.add_756, matmul_7550, parameter_7520) + + relu_7570, = call(self.relu_757, add_7560) + + full_7580, = call(self.full_758) + + dropout_7590, dropout_7591, = call(self.dropout_759, relu_7570, None, full_7580) + + matmul_7600, = call(self.matmul_760, dropout_7590, parameter_7510) + + add_7610, = call(self.add_761, matmul_7600, parameter_7500) + + call(self.shadow_output_762, add_7610) + + return ret_lambda_module_748_block00 + + + + def __call__(self, call, *args, **kwargs): + + self.SetArgs(args) + + self.SetKeywordArgs(kwargs) + + return call(self.module_748, blocks=[[(self.module_748_block00,)]]) + + +class PirProgram_2498128585497043328: + + def __init__(self): + + self.parameter_817 = self.Op("builtin.parameter", 817, input_types=[], output_types=[self.t_dtensor([96], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_3.b_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.parameter_818 = self.Op("builtin.parameter", 818, input_types=[], output_types=[self.t_dtensor([32, 96], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_3.w_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.parameter_819 = self.Op("builtin.parameter", 819, input_types=[], output_types=[self.t_dtensor([32], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_2.b_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.parameter_820 = self.Op("builtin.parameter", 820, input_types=[], output_types=[self.t_dtensor([16, 32], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "struct_name":self.a_str("/Linear_10/"), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_2.w_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.data_821 = self.Op("pd_op.data", 821, input_types=[], output_types=[self.t_dtensor([-1, 2, -1], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_10/"), "stop_gradient":self.a_array(self.a_bool(False)), "name":self.a_str("args_0"), "shape":self.a_intarray(-1, 2, -1), "dtype":self.a_dtype("float32"), "place":self.a_place("undefined", 0), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.matmul_822 = self.Op("pd_op.matmul", 822, input_types=[self.t_dtensor([-1, 2, -1], self.t_f32()), self.t_dtensor([16, 32], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_10/"), "stop_gradient":self.a_array(self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.add_823 = self.Op("pd_op.add", 823, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([32], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_10/"), "stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.relu_824 = self.Op("pd_op.relu", 824, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], attrs={"struct_name":self.a_str("/ReLU_5/"), "stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.full_825 = self.Op("pd_op.full", 825, input_types=[], output_types=[self.t_dtensor([1], self.t_f32())], attrs={"struct_name":self.a_str("/Dropout_5/"), "stop_gradient":self.a_array(self.a_bool(True)), "shape":self.a_intarray(1), "value":self.a_f64("0.2"), "dtype":self.a_dtype("float32"), "place":self.a_place("cpu"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.dropout_826 = self.Op("pd_op.dropout", 826, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_null(), self.t_dtensor([1], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([-1, 2, 32], self.t_ui8())], attrs={"struct_name":self.a_str("/Dropout_5/"), "stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "is_test":self.a_bool(True), "seed":self.a_i32(0), "mode":self.a_str("upscale_in_train"), "fix_seed":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) + + self.matmul_827 = self.Op("pd_op.matmul", 827, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([32, 96], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 96], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_11/"), "stop_gradient":self.a_array(self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.add_828 = self.Op("pd_op.add", 828, input_types=[self.t_dtensor([-1, 2, 96], self.t_f32()), self.t_dtensor([96], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 96], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_11/"), "stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) + + self.shadow_output_829 = self.Op("builtin.shadow_output", 829, input_types=[self.t_dtensor([-1, 2, 96], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("output_0"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) + + self.module_815 = self.Op("builtin.module", 815, input_types=[], output_types=[], attrs={"program":self.a_pointer("0x6bc06ee0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array()}, block_positional_arg_names=[[[]]], block_keyword_arg_names=[[{}]], block_positional_arg_types=[[[]]], block_keyword_arg_types=[[[]]], ) + + + + def module_815_block00(self, call): + + def ret_lambda_module_815_block00(): + + parameter_8170, = call(self.parameter_817) + + parameter_8180, = call(self.parameter_818) + + parameter_8190, = call(self.parameter_819) + + parameter_8200, = call(self.parameter_820) + + data_8210, = call(self.data_821) + + matmul_8220, = call(self.matmul_822, data_8210, parameter_8200) + + add_8230, = call(self.add_823, matmul_8220, parameter_8190) + + relu_8240, = call(self.relu_824, add_8230) + + full_8250, = call(self.full_825) + + dropout_8260, dropout_8261, = call(self.dropout_826, relu_8240, None, full_8250) + + matmul_8270, = call(self.matmul_827, dropout_8260, parameter_8180) + + add_8280, = call(self.add_828, matmul_8270, parameter_8170) + + call(self.shadow_output_829, add_8280) + + return ret_lambda_module_815_block00 + + + + def __call__(self, call, *args, **kwargs): + + self.SetArgs(args) + + self.SetKeywordArgs(kwargs) + + return call(self.module_815, blocks=[[(self.module_815_block00,)]]) + + diff --git a/samples/torch/resnet18/attribute.json b/samples/torch/resnet18/attribute.json new file mode 100644 index 000000000..023a79446 --- /dev/null +++ b/samples/torch/resnet18/attribute.json @@ -0,0 +1,3 @@ +{ + "framework": "torch" +} \ No newline at end of file diff --git a/samples/torch/resnet18/input_tensor_constraints.py b/samples/torch/resnet18/input_tensor_constraints.py new file mode 100644 index 000000000..e69de29bb diff --git a/samples/torch/resnet18/model.py b/samples/torch/resnet18/model.py new file mode 100644 index 000000000..7a904d721 --- /dev/null +++ b/samples/torch/resnet18/model.py @@ -0,0 +1,96 @@ + +import torch +from .. import utils +class GraphModule(torch.nn.Module): + + def forward(self, p_conv1_weight, p_bn1_weight, p_bn1_bias, p_getattr_l__self___layer1___0___conv1_weight, p_getattr_l__self___layer1___0___bn1_weight, p_getattr_l__self___layer1___0___bn1_bias, p_getattr_l__self___layer1___0___conv2_weight, p_getattr_l__self___layer1___0___bn2_weight, p_getattr_l__self___layer1___0___bn2_bias, p_getattr_l__self___layer1___1___conv1_weight, p_getattr_l__self___layer1___1___bn1_weight, p_getattr_l__self___layer1___1___bn1_bias, p_getattr_l__self___layer1___1___conv2_weight, p_getattr_l__self___layer1___1___bn2_weight, p_getattr_l__self___layer1___1___bn2_bias, p_getattr_l__self___layer2___0___conv1_weight, p_getattr_l__self___layer2___0___bn1_weight, p_getattr_l__self___layer2___0___bn1_bias, p_getattr_l__self___layer2___0___conv2_weight, p_getattr_l__self___layer2___0___bn2_weight, p_getattr_l__self___layer2___0___bn2_bias, p_getattr_l__self___layer2___0___downsample_0_weight, p_getattr_l__self___layer2___0___downsample_1_weight, p_getattr_l__self___layer2___0___downsample_1_bias, p_getattr_l__self___layer2___1___conv1_weight, p_getattr_l__self___layer2___1___bn1_weight, p_getattr_l__self___layer2___1___bn1_bias, p_getattr_l__self___layer2___1___conv2_weight, p_getattr_l__self___layer2___1___bn2_weight, p_getattr_l__self___layer2___1___bn2_bias, p_getattr_l__self___layer3___0___conv1_weight, p_getattr_l__self___layer3___0___bn1_weight, p_getattr_l__self___layer3___0___bn1_bias, p_getattr_l__self___layer3___0___conv2_weight, p_getattr_l__self___layer3___0___bn2_weight, p_getattr_l__self___layer3___0___bn2_bias, p_getattr_l__self___layer3___0___downsample_0_weight, p_getattr_l__self___layer3___0___downsample_1_weight, p_getattr_l__self___layer3___0___downsample_1_bias, p_getattr_l__self___layer3___1___conv1_weight, p_getattr_l__self___layer3___1___bn1_weight, p_getattr_l__self___layer3___1___bn1_bias, p_getattr_l__self___layer3___1___conv2_weight, p_getattr_l__self___layer3___1___bn2_weight, p_getattr_l__self___layer3___1___bn2_bias, p_getattr_l__self___layer4___0___conv1_weight, p_getattr_l__self___layer4___0___bn1_weight, p_getattr_l__self___layer4___0___bn1_bias, p_getattr_l__self___layer4___0___conv2_weight, p_getattr_l__self___layer4___0___bn2_weight, p_getattr_l__self___layer4___0___bn2_bias, p_getattr_l__self___layer4___0___downsample_0_weight, p_getattr_l__self___layer4___0___downsample_1_weight, p_getattr_l__self___layer4___0___downsample_1_bias, p_getattr_l__self___layer4___1___conv1_weight, p_getattr_l__self___layer4___1___bn1_weight, p_getattr_l__self___layer4___1___bn1_bias, p_getattr_l__self___layer4___1___conv2_weight, p_getattr_l__self___layer4___1___bn2_weight, p_getattr_l__self___layer4___1___bn2_bias, p_fc_weight, p_fc_bias, b_bn1_running_mean, b_bn1_running_var, b_bn1_num_batches_tracked, b_getattr_l__self___layer1___0___bn1_running_mean, b_getattr_l__self___layer1___0___bn1_running_var, b_getattr_l__self___layer1___0___bn1_num_batches_tracked, b_getattr_l__self___layer1___0___bn2_running_mean, b_getattr_l__self___layer1___0___bn2_running_var, b_getattr_l__self___layer1___0___bn2_num_batches_tracked, b_getattr_l__self___layer1___1___bn1_running_mean, b_getattr_l__self___layer1___1___bn1_running_var, b_getattr_l__self___layer1___1___bn1_num_batches_tracked, b_getattr_l__self___layer1___1___bn2_running_mean, b_getattr_l__self___layer1___1___bn2_running_var, b_getattr_l__self___layer1___1___bn2_num_batches_tracked, b_getattr_l__self___layer2___0___bn1_running_mean, b_getattr_l__self___layer2___0___bn1_running_var, b_getattr_l__self___layer2___0___bn1_num_batches_tracked, b_getattr_l__self___layer2___0___bn2_running_mean, b_getattr_l__self___layer2___0___bn2_running_var, b_getattr_l__self___layer2___0___bn2_num_batches_tracked, b_getattr_l__self___layer2___0___downsample_1_running_mean, b_getattr_l__self___layer2___0___downsample_1_running_var, b_getattr_l__self___layer2___0___downsample_1_num_batches_tracked, b_getattr_l__self___layer2___1___bn1_running_mean, b_getattr_l__self___layer2___1___bn1_running_var, b_getattr_l__self___layer2___1___bn1_num_batches_tracked, b_getattr_l__self___layer2___1___bn2_running_mean, b_getattr_l__self___layer2___1___bn2_running_var, b_getattr_l__self___layer2___1___bn2_num_batches_tracked, b_getattr_l__self___layer3___0___bn1_running_mean, b_getattr_l__self___layer3___0___bn1_running_var, b_getattr_l__self___layer3___0___bn1_num_batches_tracked, b_getattr_l__self___layer3___0___bn2_running_mean, b_getattr_l__self___layer3___0___bn2_running_var, b_getattr_l__self___layer3___0___bn2_num_batches_tracked, b_getattr_l__self___layer3___0___downsample_1_running_mean, b_getattr_l__self___layer3___0___downsample_1_running_var, b_getattr_l__self___layer3___0___downsample_1_num_batches_tracked, b_getattr_l__self___layer3___1___bn1_running_mean, b_getattr_l__self___layer3___1___bn1_running_var, b_getattr_l__self___layer3___1___bn1_num_batches_tracked, b_getattr_l__self___layer3___1___bn2_running_mean, b_getattr_l__self___layer3___1___bn2_running_var, b_getattr_l__self___layer3___1___bn2_num_batches_tracked, b_getattr_l__self___layer4___0___bn1_running_mean, b_getattr_l__self___layer4___0___bn1_running_var, b_getattr_l__self___layer4___0___bn1_num_batches_tracked, b_getattr_l__self___layer4___0___bn2_running_mean, b_getattr_l__self___layer4___0___bn2_running_var, b_getattr_l__self___layer4___0___bn2_num_batches_tracked, b_getattr_l__self___layer4___0___downsample_1_running_mean, b_getattr_l__self___layer4___0___downsample_1_running_var, b_getattr_l__self___layer4___0___downsample_1_num_batches_tracked, b_getattr_l__self___layer4___1___bn1_running_mean, b_getattr_l__self___layer4___1___bn1_running_var, b_getattr_l__self___layer4___1___bn1_num_batches_tracked, b_getattr_l__self___layer4___1___bn2_running_mean, b_getattr_l__self___layer4___1___bn2_running_var, b_getattr_l__self___layer4___1___bn2_num_batches_tracked, x): + conv2d = torch.ops.aten.conv2d.default(x, p_conv1_weight, None, [2, 2], [3, 3]); x = p_conv1_weight = None + batch_norm = torch.ops.aten.batch_norm.default(conv2d, p_bn1_weight, p_bn1_bias, b_bn1_running_mean, b_bn1_running_var, False, 0.1, 1e-05, True); conv2d = p_bn1_weight = p_bn1_bias = b_bn1_running_mean = b_bn1_running_var = None + relu_ = torch.ops.aten.relu_.default(batch_norm); batch_norm = None + max_pool2d = torch.ops.aten.max_pool2d.default(relu_, [3, 3], [2, 2], [1, 1]); relu_ = None + conv2d_1 = torch.ops.aten.conv2d.default(max_pool2d, p_getattr_l__self___layer1___0___conv1_weight, None, [1, 1], [1, 1]); p_getattr_l__self___layer1___0___conv1_weight = None + batch_norm_1 = torch.ops.aten.batch_norm.default(conv2d_1, p_getattr_l__self___layer1___0___bn1_weight, p_getattr_l__self___layer1___0___bn1_bias, b_getattr_l__self___layer1___0___bn1_running_mean, b_getattr_l__self___layer1___0___bn1_running_var, False, 0.1, 1e-05, True); conv2d_1 = p_getattr_l__self___layer1___0___bn1_weight = p_getattr_l__self___layer1___0___bn1_bias = b_getattr_l__self___layer1___0___bn1_running_mean = b_getattr_l__self___layer1___0___bn1_running_var = None + relu__1 = torch.ops.aten.relu_.default(batch_norm_1); batch_norm_1 = None + conv2d_2 = torch.ops.aten.conv2d.default(relu__1, p_getattr_l__self___layer1___0___conv2_weight, None, [1, 1], [1, 1]); relu__1 = p_getattr_l__self___layer1___0___conv2_weight = None + batch_norm_2 = torch.ops.aten.batch_norm.default(conv2d_2, p_getattr_l__self___layer1___0___bn2_weight, p_getattr_l__self___layer1___0___bn2_bias, b_getattr_l__self___layer1___0___bn2_running_mean, b_getattr_l__self___layer1___0___bn2_running_var, False, 0.1, 1e-05, True); conv2d_2 = p_getattr_l__self___layer1___0___bn2_weight = p_getattr_l__self___layer1___0___bn2_bias = b_getattr_l__self___layer1___0___bn2_running_mean = b_getattr_l__self___layer1___0___bn2_running_var = None + add_ = torch.ops.aten.add_.Tensor(batch_norm_2, max_pool2d); batch_norm_2 = max_pool2d = None + relu__2 = torch.ops.aten.relu_.default(add_); add_ = None + conv2d_3 = torch.ops.aten.conv2d.default(relu__2, p_getattr_l__self___layer1___1___conv1_weight, None, [1, 1], [1, 1]); p_getattr_l__self___layer1___1___conv1_weight = None + batch_norm_3 = torch.ops.aten.batch_norm.default(conv2d_3, p_getattr_l__self___layer1___1___bn1_weight, p_getattr_l__self___layer1___1___bn1_bias, b_getattr_l__self___layer1___1___bn1_running_mean, b_getattr_l__self___layer1___1___bn1_running_var, False, 0.1, 1e-05, True); conv2d_3 = p_getattr_l__self___layer1___1___bn1_weight = p_getattr_l__self___layer1___1___bn1_bias = b_getattr_l__self___layer1___1___bn1_running_mean = b_getattr_l__self___layer1___1___bn1_running_var = None + relu__3 = torch.ops.aten.relu_.default(batch_norm_3); batch_norm_3 = None + conv2d_4 = torch.ops.aten.conv2d.default(relu__3, p_getattr_l__self___layer1___1___conv2_weight, None, [1, 1], [1, 1]); relu__3 = p_getattr_l__self___layer1___1___conv2_weight = None + batch_norm_4 = torch.ops.aten.batch_norm.default(conv2d_4, p_getattr_l__self___layer1___1___bn2_weight, p_getattr_l__self___layer1___1___bn2_bias, b_getattr_l__self___layer1___1___bn2_running_mean, b_getattr_l__self___layer1___1___bn2_running_var, False, 0.1, 1e-05, True); conv2d_4 = p_getattr_l__self___layer1___1___bn2_weight = p_getattr_l__self___layer1___1___bn2_bias = b_getattr_l__self___layer1___1___bn2_running_mean = b_getattr_l__self___layer1___1___bn2_running_var = None + add__1 = torch.ops.aten.add_.Tensor(batch_norm_4, relu__2); batch_norm_4 = relu__2 = None + relu__4 = torch.ops.aten.relu_.default(add__1); add__1 = None + conv2d_5 = torch.ops.aten.conv2d.default(relu__4, p_getattr_l__self___layer2___0___conv1_weight, None, [2, 2], [1, 1]); p_getattr_l__self___layer2___0___conv1_weight = None + batch_norm_5 = torch.ops.aten.batch_norm.default(conv2d_5, p_getattr_l__self___layer2___0___bn1_weight, p_getattr_l__self___layer2___0___bn1_bias, b_getattr_l__self___layer2___0___bn1_running_mean, b_getattr_l__self___layer2___0___bn1_running_var, False, 0.1, 1e-05, True); conv2d_5 = p_getattr_l__self___layer2___0___bn1_weight = p_getattr_l__self___layer2___0___bn1_bias = b_getattr_l__self___layer2___0___bn1_running_mean = b_getattr_l__self___layer2___0___bn1_running_var = None + relu__5 = torch.ops.aten.relu_.default(batch_norm_5); batch_norm_5 = None + conv2d_6 = torch.ops.aten.conv2d.default(relu__5, p_getattr_l__self___layer2___0___conv2_weight, None, [1, 1], [1, 1]); relu__5 = p_getattr_l__self___layer2___0___conv2_weight = None + batch_norm_6 = torch.ops.aten.batch_norm.default(conv2d_6, p_getattr_l__self___layer2___0___bn2_weight, p_getattr_l__self___layer2___0___bn2_bias, b_getattr_l__self___layer2___0___bn2_running_mean, b_getattr_l__self___layer2___0___bn2_running_var, False, 0.1, 1e-05, True); conv2d_6 = p_getattr_l__self___layer2___0___bn2_weight = p_getattr_l__self___layer2___0___bn2_bias = b_getattr_l__self___layer2___0___bn2_running_mean = b_getattr_l__self___layer2___0___bn2_running_var = None + conv2d_7 = torch.ops.aten.conv2d.default(relu__4, p_getattr_l__self___layer2___0___downsample_0_weight, None, [2, 2]); relu__4 = p_getattr_l__self___layer2___0___downsample_0_weight = None + batch_norm_7 = torch.ops.aten.batch_norm.default(conv2d_7, p_getattr_l__self___layer2___0___downsample_1_weight, p_getattr_l__self___layer2___0___downsample_1_bias, b_getattr_l__self___layer2___0___downsample_1_running_mean, b_getattr_l__self___layer2___0___downsample_1_running_var, False, 0.1, 1e-05, True); conv2d_7 = p_getattr_l__self___layer2___0___downsample_1_weight = p_getattr_l__self___layer2___0___downsample_1_bias = b_getattr_l__self___layer2___0___downsample_1_running_mean = b_getattr_l__self___layer2___0___downsample_1_running_var = None + add__2 = torch.ops.aten.add_.Tensor(batch_norm_6, batch_norm_7); batch_norm_6 = batch_norm_7 = None + relu__6 = torch.ops.aten.relu_.default(add__2); add__2 = None + conv2d_8 = torch.ops.aten.conv2d.default(relu__6, p_getattr_l__self___layer2___1___conv1_weight, None, [1, 1], [1, 1]); p_getattr_l__self___layer2___1___conv1_weight = None + batch_norm_8 = torch.ops.aten.batch_norm.default(conv2d_8, p_getattr_l__self___layer2___1___bn1_weight, p_getattr_l__self___layer2___1___bn1_bias, b_getattr_l__self___layer2___1___bn1_running_mean, b_getattr_l__self___layer2___1___bn1_running_var, False, 0.1, 1e-05, True); conv2d_8 = p_getattr_l__self___layer2___1___bn1_weight = p_getattr_l__self___layer2___1___bn1_bias = b_getattr_l__self___layer2___1___bn1_running_mean = b_getattr_l__self___layer2___1___bn1_running_var = None + relu__7 = torch.ops.aten.relu_.default(batch_norm_8); batch_norm_8 = None + conv2d_9 = torch.ops.aten.conv2d.default(relu__7, p_getattr_l__self___layer2___1___conv2_weight, None, [1, 1], [1, 1]); relu__7 = p_getattr_l__self___layer2___1___conv2_weight = None + batch_norm_9 = torch.ops.aten.batch_norm.default(conv2d_9, p_getattr_l__self___layer2___1___bn2_weight, p_getattr_l__self___layer2___1___bn2_bias, b_getattr_l__self___layer2___1___bn2_running_mean, b_getattr_l__self___layer2___1___bn2_running_var, False, 0.1, 1e-05, True); conv2d_9 = p_getattr_l__self___layer2___1___bn2_weight = p_getattr_l__self___layer2___1___bn2_bias = b_getattr_l__self___layer2___1___bn2_running_mean = b_getattr_l__self___layer2___1___bn2_running_var = None + add__3 = torch.ops.aten.add_.Tensor(batch_norm_9, relu__6); batch_norm_9 = relu__6 = None + relu__8 = torch.ops.aten.relu_.default(add__3); add__3 = None + conv2d_10 = torch.ops.aten.conv2d.default(relu__8, p_getattr_l__self___layer3___0___conv1_weight, None, [2, 2], [1, 1]); p_getattr_l__self___layer3___0___conv1_weight = None + batch_norm_10 = torch.ops.aten.batch_norm.default(conv2d_10, p_getattr_l__self___layer3___0___bn1_weight, p_getattr_l__self___layer3___0___bn1_bias, b_getattr_l__self___layer3___0___bn1_running_mean, b_getattr_l__self___layer3___0___bn1_running_var, False, 0.1, 1e-05, True); conv2d_10 = p_getattr_l__self___layer3___0___bn1_weight = p_getattr_l__self___layer3___0___bn1_bias = b_getattr_l__self___layer3___0___bn1_running_mean = b_getattr_l__self___layer3___0___bn1_running_var = None + relu__9 = torch.ops.aten.relu_.default(batch_norm_10); batch_norm_10 = None + conv2d_11 = torch.ops.aten.conv2d.default(relu__9, p_getattr_l__self___layer3___0___conv2_weight, None, [1, 1], [1, 1]); relu__9 = p_getattr_l__self___layer3___0___conv2_weight = None + batch_norm_11 = torch.ops.aten.batch_norm.default(conv2d_11, p_getattr_l__self___layer3___0___bn2_weight, p_getattr_l__self___layer3___0___bn2_bias, b_getattr_l__self___layer3___0___bn2_running_mean, b_getattr_l__self___layer3___0___bn2_running_var, False, 0.1, 1e-05, True); conv2d_11 = p_getattr_l__self___layer3___0___bn2_weight = p_getattr_l__self___layer3___0___bn2_bias = b_getattr_l__self___layer3___0___bn2_running_mean = b_getattr_l__self___layer3___0___bn2_running_var = None + conv2d_12 = torch.ops.aten.conv2d.default(relu__8, p_getattr_l__self___layer3___0___downsample_0_weight, None, [2, 2]); relu__8 = p_getattr_l__self___layer3___0___downsample_0_weight = None + batch_norm_12 = torch.ops.aten.batch_norm.default(conv2d_12, p_getattr_l__self___layer3___0___downsample_1_weight, p_getattr_l__self___layer3___0___downsample_1_bias, b_getattr_l__self___layer3___0___downsample_1_running_mean, b_getattr_l__self___layer3___0___downsample_1_running_var, False, 0.1, 1e-05, True); conv2d_12 = p_getattr_l__self___layer3___0___downsample_1_weight = p_getattr_l__self___layer3___0___downsample_1_bias = b_getattr_l__self___layer3___0___downsample_1_running_mean = b_getattr_l__self___layer3___0___downsample_1_running_var = None + add__4 = torch.ops.aten.add_.Tensor(batch_norm_11, batch_norm_12); batch_norm_11 = batch_norm_12 = None + relu__10 = torch.ops.aten.relu_.default(add__4); add__4 = None + conv2d_13 = torch.ops.aten.conv2d.default(relu__10, p_getattr_l__self___layer3___1___conv1_weight, None, [1, 1], [1, 1]); p_getattr_l__self___layer3___1___conv1_weight = None + batch_norm_13 = torch.ops.aten.batch_norm.default(conv2d_13, p_getattr_l__self___layer3___1___bn1_weight, p_getattr_l__self___layer3___1___bn1_bias, b_getattr_l__self___layer3___1___bn1_running_mean, b_getattr_l__self___layer3___1___bn1_running_var, False, 0.1, 1e-05, True); conv2d_13 = p_getattr_l__self___layer3___1___bn1_weight = p_getattr_l__self___layer3___1___bn1_bias = b_getattr_l__self___layer3___1___bn1_running_mean = b_getattr_l__self___layer3___1___bn1_running_var = None + relu__11 = torch.ops.aten.relu_.default(batch_norm_13); batch_norm_13 = None + conv2d_14 = torch.ops.aten.conv2d.default(relu__11, p_getattr_l__self___layer3___1___conv2_weight, None, [1, 1], [1, 1]); relu__11 = p_getattr_l__self___layer3___1___conv2_weight = None + batch_norm_14 = torch.ops.aten.batch_norm.default(conv2d_14, p_getattr_l__self___layer3___1___bn2_weight, p_getattr_l__self___layer3___1___bn2_bias, b_getattr_l__self___layer3___1___bn2_running_mean, b_getattr_l__self___layer3___1___bn2_running_var, False, 0.1, 1e-05, True); conv2d_14 = p_getattr_l__self___layer3___1___bn2_weight = p_getattr_l__self___layer3___1___bn2_bias = b_getattr_l__self___layer3___1___bn2_running_mean = b_getattr_l__self___layer3___1___bn2_running_var = None + add__5 = torch.ops.aten.add_.Tensor(batch_norm_14, relu__10); batch_norm_14 = relu__10 = None + relu__12 = torch.ops.aten.relu_.default(add__5); add__5 = None + conv2d_15 = torch.ops.aten.conv2d.default(relu__12, p_getattr_l__self___layer4___0___conv1_weight, None, [2, 2], [1, 1]); p_getattr_l__self___layer4___0___conv1_weight = None + batch_norm_15 = torch.ops.aten.batch_norm.default(conv2d_15, p_getattr_l__self___layer4___0___bn1_weight, p_getattr_l__self___layer4___0___bn1_bias, b_getattr_l__self___layer4___0___bn1_running_mean, b_getattr_l__self___layer4___0___bn1_running_var, False, 0.1, 1e-05, True); conv2d_15 = p_getattr_l__self___layer4___0___bn1_weight = p_getattr_l__self___layer4___0___bn1_bias = b_getattr_l__self___layer4___0___bn1_running_mean = b_getattr_l__self___layer4___0___bn1_running_var = None + relu__13 = torch.ops.aten.relu_.default(batch_norm_15); batch_norm_15 = None + conv2d_16 = torch.ops.aten.conv2d.default(relu__13, p_getattr_l__self___layer4___0___conv2_weight, None, [1, 1], [1, 1]); relu__13 = p_getattr_l__self___layer4___0___conv2_weight = None + batch_norm_16 = torch.ops.aten.batch_norm.default(conv2d_16, p_getattr_l__self___layer4___0___bn2_weight, p_getattr_l__self___layer4___0___bn2_bias, b_getattr_l__self___layer4___0___bn2_running_mean, b_getattr_l__self___layer4___0___bn2_running_var, False, 0.1, 1e-05, True); conv2d_16 = p_getattr_l__self___layer4___0___bn2_weight = p_getattr_l__self___layer4___0___bn2_bias = b_getattr_l__self___layer4___0___bn2_running_mean = b_getattr_l__self___layer4___0___bn2_running_var = None + conv2d_17 = torch.ops.aten.conv2d.default(relu__12, p_getattr_l__self___layer4___0___downsample_0_weight, None, [2, 2]); relu__12 = p_getattr_l__self___layer4___0___downsample_0_weight = None + batch_norm_17 = torch.ops.aten.batch_norm.default(conv2d_17, p_getattr_l__self___layer4___0___downsample_1_weight, p_getattr_l__self___layer4___0___downsample_1_bias, b_getattr_l__self___layer4___0___downsample_1_running_mean, b_getattr_l__self___layer4___0___downsample_1_running_var, False, 0.1, 1e-05, True); conv2d_17 = p_getattr_l__self___layer4___0___downsample_1_weight = p_getattr_l__self___layer4___0___downsample_1_bias = b_getattr_l__self___layer4___0___downsample_1_running_mean = b_getattr_l__self___layer4___0___downsample_1_running_var = None + add__6 = torch.ops.aten.add_.Tensor(batch_norm_16, batch_norm_17); batch_norm_16 = batch_norm_17 = None + relu__14 = torch.ops.aten.relu_.default(add__6); add__6 = None + conv2d_18 = torch.ops.aten.conv2d.default(relu__14, p_getattr_l__self___layer4___1___conv1_weight, None, [1, 1], [1, 1]); p_getattr_l__self___layer4___1___conv1_weight = None + batch_norm_18 = torch.ops.aten.batch_norm.default(conv2d_18, p_getattr_l__self___layer4___1___bn1_weight, p_getattr_l__self___layer4___1___bn1_bias, b_getattr_l__self___layer4___1___bn1_running_mean, b_getattr_l__self___layer4___1___bn1_running_var, False, 0.1, 1e-05, True); conv2d_18 = p_getattr_l__self___layer4___1___bn1_weight = p_getattr_l__self___layer4___1___bn1_bias = b_getattr_l__self___layer4___1___bn1_running_mean = b_getattr_l__self___layer4___1___bn1_running_var = None + relu__15 = torch.ops.aten.relu_.default(batch_norm_18); batch_norm_18 = None + conv2d_19 = torch.ops.aten.conv2d.default(relu__15, p_getattr_l__self___layer4___1___conv2_weight, None, [1, 1], [1, 1]); relu__15 = p_getattr_l__self___layer4___1___conv2_weight = None + batch_norm_19 = torch.ops.aten.batch_norm.default(conv2d_19, p_getattr_l__self___layer4___1___bn2_weight, p_getattr_l__self___layer4___1___bn2_bias, b_getattr_l__self___layer4___1___bn2_running_mean, b_getattr_l__self___layer4___1___bn2_running_var, False, 0.1, 1e-05, True); conv2d_19 = p_getattr_l__self___layer4___1___bn2_weight = p_getattr_l__self___layer4___1___bn2_bias = b_getattr_l__self___layer4___1___bn2_running_mean = b_getattr_l__self___layer4___1___bn2_running_var = None + add__7 = torch.ops.aten.add_.Tensor(batch_norm_19, relu__14); batch_norm_19 = relu__14 = None + relu__16 = torch.ops.aten.relu_.default(add__7); add__7 = None + adaptive_avg_pool2d = torch.ops.aten.adaptive_avg_pool2d.default(relu__16, [1, 1]); relu__16 = None + flatten = torch.ops.aten.flatten.using_ints(adaptive_avg_pool2d, 1); adaptive_avg_pool2d = None + linear = torch.ops.aten.linear.default(flatten, p_fc_weight, p_fc_bias); flatten = p_fc_weight = p_fc_bias = None + return (linear,) + + # To see more debug info, please use `graph_module.print_readable()` + +model = GraphModule() + +inputs_params = utils.load_converted_from_text(f'./source_tensor_meta.py') +inputs = inputs_params["input_info"] +inputs = [utils.replay_tensor(i) for i in inputs] +params = inputs_params["weight_info"] + +state_dict = {} +for k, v in params.items(): + k = utils.convert_param_name(k) + v = utils.replay_tensor(v) + state_dict[k] = v + +y = model(x=inputs[0], **state_dict)[0] +print(torch.argmin(y), torch.argmax(y)) +print(y.shape) + diff --git a/samples/torch/resnet18/source_tensor_meta.py b/samples/torch/resnet18/source_tensor_meta.py new file mode 100644 index 000000000..afcf7a9ca --- /dev/null +++ b/samples/torch/resnet18/source_tensor_meta.py @@ -0,0 +1,1106 @@ +class Program_input_tensor_meta_65ae4c40166d44aaba3e96cd14381f7f: + name = "input_0" + shape = [1, 3, 1, 5] + dtype = "torch.float32" + device = "cpu" + mean = 0.11563243716955185 + std = 1.629511833190918 + data = [1.504339, -1.144876, -1.709763, -1.148033, -0.312331, -1.652560, 2.062543, 2.058685, 1.823935, -1.789209, 0.325380, -0.886801, -1.486652, 2.395117, 1.694714] + +class Program_weight_tensor_meta_85c2f314979746bdb959a1a16a734af5: + name = "conv1.weight" + shape = [64, 3, 7, 7] + dtype = "torch.float32" + device = "cpu" + mean = 2.941965612990316e-05 + std = 0.1296982318162918 + data = None + +class Program_weight_tensor_meta_35de8f66d10f471f9e3cae054124c286: + name = "bn1.weight" + shape = [64] + dtype = "torch.float32" + device = "cpu" + mean = 0.25757724046707153 + std = 0.12313675880432129 + data = None + +class Program_weight_tensor_meta_307b24206f204bed84b5d0165a322199: + name = "bn1.bias" + shape = [64] + dtype = "torch.float32" + device = "cpu" + mean = 0.18112018704414368 + std = 0.29894745349884033 + data = None + +class Program_weight_tensor_meta_62e65985dcfb4520a020ddb9ddf68ddc: + name = "layer1.0.conv1.weight" + shape = [64, 64, 3, 3] + dtype = "torch.float32" + device = "cpu" + mean = -0.0030870491173118353 + std = 0.0533977635204792 + data = None + +class Program_weight_tensor_meta_02dc02adb4f74681afa8e8ebe7e59af2: + name = "layer1.0.bn1.weight" + shape = [64] + dtype = "torch.float32" + device = "cpu" + mean = 0.3396005630493164 + std = 0.12483347952365875 + data = None + +class Program_weight_tensor_meta_4bac88fb70a84774a3aeea8a2a74c9c8: + name = "layer1.0.bn1.bias" + shape = [64] + dtype = "torch.float32" + device = "cpu" + mean = -0.0341365709900856 + std = 0.20955809950828552 + data = None + +class Program_weight_tensor_meta_3a0d8d57a9bc49cdbad3ebef2c730a66: + name = "layer1.0.conv2.weight" + shape = [64, 64, 3, 3] + dtype = "torch.float32" + device = "cpu" + mean = -0.0008893322665244341 + std = 0.04520029574632645 + data = None + +class Program_weight_tensor_meta_0274802f8e004323856635d6209f9de3: + name = "layer1.0.bn2.weight" + shape = [64] + dtype = "torch.float32" + device = "cpu" + mean = 0.33305495977401733 + std = 0.12227492034435272 + data = None + +class Program_weight_tensor_meta_5f5239fca8874dc7b52e4b39cde8db20: + name = "layer1.0.bn2.bias" + shape = [64] + dtype = "torch.float32" + device = "cpu" + mean = 0.003462875261902809 + std = 0.2010938972234726 + data = None + +class Program_weight_tensor_meta_ff8978d5923742c383e5ee38124159e8: + name = "layer1.1.conv1.weight" + shape = [64, 64, 3, 3] + dtype = "torch.float32" + device = "cpu" + mean = -0.00242004101164639 + std = 0.05083491653203964 + data = None + +class Program_weight_tensor_meta_c2ad0dca000642f9a1a1603ce5f51b93: + name = "layer1.1.bn1.weight" + shape = [64] + dtype = "torch.float32" + device = "cpu" + mean = 0.32869213819503784 + std = 0.07058674097061157 + data = None + +class Program_weight_tensor_meta_aa0b9e8eaa8745ba8e5a4c65a76ebd4b: + name = "layer1.1.bn1.bias" + shape = [64] + dtype = "torch.float32" + device = "cpu" + mean = -0.08357392251491547 + std = 0.16554474830627441 + data = None + +class Program_weight_tensor_meta_eabd670247bf402db1ecf5d38d68aade: + name = "layer1.1.conv2.weight" + shape = [64, 64, 3, 3] + dtype = "torch.float32" + device = "cpu" + mean = -0.0012602502247318625 + std = 0.04397609457373619 + data = None + +class Program_weight_tensor_meta_99943b2040f74dbdace2a5f61f4099a0: + name = "layer1.1.bn2.weight" + shape = [64] + dtype = "torch.float32" + device = "cpu" + mean = 0.3924297094345093 + std = 0.15965722501277924 + data = None + +class Program_weight_tensor_meta_e78c6376420d425b901994215fe07cb0: + name = "layer1.1.bn2.bias" + shape = [64] + dtype = "torch.float32" + device = "cpu" + mean = -0.029983950778841972 + std = 0.15416555106639862 + data = None + +class Program_weight_tensor_meta_2abc0ed690b043fd838569b5417f829b: + name = "layer2.0.conv1.weight" + shape = [128, 64, 3, 3] + dtype = "torch.float32" + device = "cpu" + mean = -0.0014537475071847439 + std = 0.041583266109228134 + data = None + +class Program_weight_tensor_meta_f1c806fd6af9414dbdfd1ca6ceb01f24: + name = "layer2.0.bn1.weight" + shape = [128] + dtype = "torch.float32" + device = "cpu" + mean = 0.31641945242881775 + std = 0.04086580500006676 + data = None + +class Program_weight_tensor_meta_261d7c3898f24432a87123eb2f4ec351: + name = "layer2.0.bn1.bias" + shape = [128] + dtype = "torch.float32" + device = "cpu" + mean = -0.06734631955623627 + std = 0.10906781256198883 + data = None + +class Program_weight_tensor_meta_642402a0ed264d0f99f7e4f209ad9989: + name = "layer2.0.conv2.weight" + shape = [128, 128, 3, 3] + dtype = "torch.float32" + device = "cpu" + mean = -0.0012476659612730145 + std = 0.03401258587837219 + data = None + +class Program_weight_tensor_meta_6797a2e364f44f12bf1662e90d24ac04: + name = "layer2.0.bn2.weight" + shape = [128] + dtype = "torch.float32" + device = "cpu" + mean = 0.3275725543498993 + std = 0.1107170432806015 + data = None + +class Program_weight_tensor_meta_b9c98c282b9c4f07975dc69945f86a3e: + name = "layer2.0.bn2.bias" + shape = [128] + dtype = "torch.float32" + device = "cpu" + mean = -0.0035552978515625 + std = 0.09035798162221909 + data = None + +class Program_weight_tensor_meta_fbcc1fee4f864880a766d5b24dedec7a: + name = "layer2.0.downsample.0.weight" + shape = [128, 64, 1, 1] + dtype = "torch.float32" + device = "cpu" + mean = -0.002587598981335759 + std = 0.0706208124756813 + data = None + +class Program_weight_tensor_meta_ab8393fb1e5141d6a834dec7f52eac70: + name = "layer2.0.downsample.1.weight" + shape = [128] + dtype = "torch.float32" + device = "cpu" + mean = 0.19508539140224457 + std = 0.10172920674085617 + data = None + +class Program_weight_tensor_meta_b33c00a94b7e44978f3ca92e72a0bc39: + name = "layer2.0.downsample.1.bias" + shape = [128] + dtype = "torch.float32" + device = "cpu" + mean = -0.0035552978515625 + std = 0.09035798162221909 + data = None + +class Program_weight_tensor_meta_e513d39e62d0439c93b3b6fb5afa0c4c: + name = "layer2.1.conv1.weight" + shape = [128, 128, 3, 3] + dtype = "torch.float32" + device = "cpu" + mean = -0.0015302165411412716 + std = 0.034170422703027725 + data = None + +class Program_weight_tensor_meta_d62e62ab8f0c4947a51852b4979b7421: + name = "layer2.1.bn1.weight" + shape = [128] + dtype = "torch.float32" + device = "cpu" + mean = 0.3212640881538391 + std = 0.047541555017232895 + data = None + +class Program_weight_tensor_meta_3755f606db814a479712a10b20d31365: + name = "layer2.1.bn1.bias" + shape = [128] + dtype = "torch.float32" + device = "cpu" + mean = -0.2102503925561905 + std = 0.10085009783506393 + data = None + +class Program_weight_tensor_meta_af102187c43142d1bd39743868e564b5: + name = "layer2.1.conv2.weight" + shape = [128, 128, 3, 3] + dtype = "torch.float32" + device = "cpu" + mean = -0.0012723812833428383 + std = 0.03005100041627884 + data = None + +class Program_weight_tensor_meta_cc84e336eb354504b67bd02c8a4686ca: + name = "layer2.1.bn2.weight" + shape = [128] + dtype = "torch.float32" + device = "cpu" + mean = 0.28291094303131104 + std = 0.12486880272626877 + data = None + +class Program_weight_tensor_meta_aa0577e7cf6b46e483db923f9c005d51: + name = "layer2.1.bn2.bias" + shape = [128] + dtype = "torch.float32" + device = "cpu" + mean = -0.15128448605537415 + std = 0.13792425394058228 + data = None + +class Program_weight_tensor_meta_fc5d84023c9e489fb836e7415da804a1: + name = "layer3.0.conv1.weight" + shape = [256, 128, 3, 3] + dtype = "torch.float32" + device = "cpu" + mean = -0.0013684284640476108 + std = 0.029047327116131783 + data = None + +class Program_weight_tensor_meta_2cec7afb6d044b1bb998b9d3f5ecf7da: + name = "layer3.0.bn1.weight" + shape = [256] + dtype = "torch.float32" + device = "cpu" + mean = 0.3122752904891968 + std = 0.041269753128290176 + data = None + +class Program_weight_tensor_meta_f408e8d07f704814ac67fe31012bffb0: + name = "layer3.0.bn1.bias" + shape = [256] + dtype = "torch.float32" + device = "cpu" + mean = -0.11478554457426071 + std = 0.09291965514421463 + data = None + +class Program_weight_tensor_meta_bc16ddeffee74369ac83375a44611489: + name = "layer3.0.conv2.weight" + shape = [256, 256, 3, 3] + dtype = "torch.float32" + device = "cpu" + mean = -0.0007866091327741742 + std = 0.02504604496061802 + data = None + +class Program_weight_tensor_meta_5eb6224efe134c9aa9979c2e556b1137: + name = "layer3.0.bn2.weight" + shape = [256] + dtype = "torch.float32" + device = "cpu" + mean = 0.32025015354156494 + std = 0.07570703327655792 + data = None + +class Program_weight_tensor_meta_8fd70e6834324d11b6e5e5ff319a9f85: + name = "layer3.0.bn2.bias" + shape = [256] + dtype = "torch.float32" + device = "cpu" + mean = -0.030766351148486137 + std = 0.08877784758806229 + data = None + +class Program_weight_tensor_meta_bcd023421d364db39eb5773de1b41157: + name = "layer3.0.downsample.0.weight" + shape = [256, 128, 1, 1] + dtype = "torch.float32" + device = "cpu" + mean = -0.0018958767177537084 + std = 0.03294449672102928 + data = None + +class Program_weight_tensor_meta_8e9a2fa165014f3e915b7a2e1781262b: + name = "layer3.0.downsample.1.weight" + shape = [256] + dtype = "torch.float32" + device = "cpu" + mean = 0.08210360258817673 + std = 0.03994247689843178 + data = None + +class Program_weight_tensor_meta_5e9583b2693a44ffb8021c46da4619b2: + name = "layer3.0.downsample.1.bias" + shape = [256] + dtype = "torch.float32" + device = "cpu" + mean = -0.030766351148486137 + std = 0.08877784758806229 + data = None + +class Program_weight_tensor_meta_56bb9f80af6e4735b0eaef9b8039553d: + name = "layer3.1.conv1.weight" + shape = [256, 256, 3, 3] + dtype = "torch.float32" + device = "cpu" + mean = -0.0016618605004623532 + std = 0.022360969334840775 + data = None + +class Program_weight_tensor_meta_d7cbfcd1990d47329240f1a38a6bd2ae: + name = "layer3.1.bn1.weight" + shape = [256] + dtype = "torch.float32" + device = "cpu" + mean = 0.2782110571861267 + std = 0.055360279977321625 + data = None + +class Program_weight_tensor_meta_ae4b0ed1761b443a8ed25587e2f2e3a9: + name = "layer3.1.bn1.bias" + shape = [256] + dtype = "torch.float32" + device = "cpu" + mean = -0.23746797442436218 + std = 0.11999902129173279 + data = None + +class Program_weight_tensor_meta_042a79120ef942beb342b331aa081352: + name = "layer3.1.conv2.weight" + shape = [256, 256, 3, 3] + dtype = "torch.float32" + device = "cpu" + mean = -0.0014413930475711823 + std = 0.020670222118496895 + data = None + +class Program_weight_tensor_meta_088f3b06ef4747259793d885615f3f07: + name = "layer3.1.bn2.weight" + shape = [256] + dtype = "torch.float32" + device = "cpu" + mean = 0.24584954977035522 + std = 0.12051574885845184 + data = None + +class Program_weight_tensor_meta_23ac968c5aea48558ef84b8f87116e64: + name = "layer3.1.bn2.bias" + shape = [256] + dtype = "torch.float32" + device = "cpu" + mean = -0.16372345387935638 + std = 0.14889785647392273 + data = None + +class Program_weight_tensor_meta_f32571f9e92a4c669c4e3a45fea21f7a: + name = "layer4.0.conv1.weight" + shape = [512, 256, 3, 3] + dtype = "torch.float32" + device = "cpu" + mean = -0.001564666861668229 + std = 0.01987350732088089 + data = None + +class Program_weight_tensor_meta_4bbe8bc767a341219352b34779b7fe41: + name = "layer4.0.bn1.weight" + shape = [512] + dtype = "torch.float32" + device = "cpu" + mean = 0.2643176019191742 + std = 0.03648092970252037 + data = None + +class Program_weight_tensor_meta_95e9c0feb9dc4d2098fbeacfcdaaac75: + name = "layer4.0.bn1.bias" + shape = [512] + dtype = "torch.float32" + device = "cpu" + mean = -0.22572243213653564 + std = 0.0785483568906784 + data = None + +class Program_weight_tensor_meta_8bdbf9b235934a68a5d569966be31027: + name = "layer4.0.conv2.weight" + shape = [512, 512, 3, 3] + dtype = "torch.float32" + device = "cpu" + mean = -0.0013029169058427215 + std = 0.017324835062026978 + data = None + +class Program_weight_tensor_meta_b7a92840802d479e9e586481a4c05bde: + name = "layer4.0.bn2.weight" + shape = [512] + dtype = "torch.float32" + device = "cpu" + mean = 0.4243190586566925 + std = 0.06827376037836075 + data = None + +class Program_weight_tensor_meta_187aac42704c4574a0b5198ba946034f: + name = "layer4.0.bn2.bias" + shape = [512] + dtype = "torch.float32" + device = "cpu" + mean = -0.19763392210006714 + std = 0.0654672160744667 + data = None + +class Program_weight_tensor_meta_5ad457ee116045359f343935bef52a04: + name = "layer4.0.downsample.0.weight" + shape = [512, 256, 1, 1] + dtype = "torch.float32" + device = "cpu" + mean = -0.0008430479210801423 + std = 0.032782770693302155 + data = None + +class Program_weight_tensor_meta_1c31df5c29d04650a64799329f0c7f7a: + name = "layer4.0.downsample.1.weight" + shape = [512] + dtype = "torch.float32" + device = "cpu" + mean = 0.25064170360565186 + std = 0.0765230804681778 + data = None + +class Program_weight_tensor_meta_309d941fb6a24483ad2b08096af17fd6: + name = "layer4.0.downsample.1.bias" + shape = [512] + dtype = "torch.float32" + device = "cpu" + mean = -0.19763392210006714 + std = 0.0654672160744667 + data = None + +class Program_weight_tensor_meta_24e0f04bf5ae4d778f1b5d1f2fdb8edd: + name = "layer4.1.conv1.weight" + shape = [512, 512, 3, 3] + dtype = "torch.float32" + device = "cpu" + mean = -0.0022610959131270647 + std = 0.017829064279794693 + data = None + +class Program_weight_tensor_meta_2043c088eff14e5f8af1b143c4883b0a: + name = "layer4.1.bn1.weight" + shape = [512] + dtype = "torch.float32" + device = "cpu" + mean = 0.28860005736351013 + std = 0.050074681639671326 + data = None + +class Program_weight_tensor_meta_e06d42bafc834b9f8ba0619d92d6efe0: + name = "layer4.1.bn1.bias" + shape = [512] + dtype = "torch.float32" + device = "cpu" + mean = -0.24174046516418457 + std = 0.1060953140258789 + data = None + +class Program_weight_tensor_meta_7075acab78824fcf97d83c59a02a937d: + name = "layer4.1.conv2.weight" + shape = [512, 512, 3, 3] + dtype = "torch.float32" + device = "cpu" + mean = -0.0001077137203537859 + std = 0.013201371766626835 + data = None + +class Program_weight_tensor_meta_30e7f949efb146dbad561e12bb16e5b7: + name = "layer4.1.bn2.weight" + shape = [512] + dtype = "torch.float32" + device = "cpu" + mean = 1.8538099527359009 + std = 0.10653527081012726 + data = None + +class Program_weight_tensor_meta_9a58f2359c7b4aa0887159f796037749: + name = "layer4.1.bn2.bias" + shape = [512] + dtype = "torch.float32" + device = "cpu" + mean = 0.27382197976112366 + std = 0.08192502707242966 + data = None + +class Program_weight_tensor_meta_14947e4958504c37b5b27fb41a9dbde2: + name = "fc.weight" + shape = [1000, 512] + dtype = "torch.float32" + device = "cpu" + mean = 5.850196060919188e-08 + std = 0.06945901364088058 + data = None + +class Program_weight_tensor_meta_04b825b2ac3649fc8568b88aea05590f: + name = "fc.bias" + shape = [1000] + dtype = "torch.float32" + device = "cpu" + mean = -5.978345996027201e-08 + std = 0.015928689390420914 + data = None + +class Program_weight_tensor_meta_c0ff88baab9c4687883c2a63297d2a2f: + name = "bn1.running_mean" + shape = [64] + dtype = "torch.float32" + device = "cpu" + mean = 0.0008474350906908512 + std = 0.04888840392231941 + data = None + +class Program_weight_tensor_meta_03d190e2b30c4ef2bcfb9c68a20188c8: + name = "bn1.running_var" + shape = [64] + dtype = "torch.float32" + device = "cpu" + mean = 3.0204856395721436 + std = 2.947047233581543 + data = None + +class Program_weight_tensor_meta_7f6393bd79ff4263a27583b5964adb0a: + name = "bn1.num_batches_tracked" + shape = [] + dtype = "torch.int64" + device = "cpu" + mean = None + std = None + data = [0] + +class Program_weight_tensor_meta_7d1286228ee94baf948fe02e641d9dbf: + name = "layer1.0.bn1.running_mean" + shape = [64] + dtype = "torch.float32" + device = "cpu" + mean = -0.9048561453819275 + std = 0.8941343426704407 + data = None + +class Program_weight_tensor_meta_597da6b86787499ba64cac60b8274cbc: + name = "layer1.0.bn1.running_var" + shape = [64] + dtype = "torch.float32" + device = "cpu" + mean = 0.5320990681648254 + std = 0.387721985578537 + data = None + +class Program_weight_tensor_meta_849b88820c9b4e60973d15eb707c09bd: + name = "layer1.0.bn1.num_batches_tracked" + shape = [] + dtype = "torch.int64" + device = "cpu" + mean = None + std = None + data = [0] + +class Program_weight_tensor_meta_1c3a050c1c184193b9d4fa83cd2fb497: + name = "layer1.0.bn2.running_mean" + shape = [64] + dtype = "torch.float32" + device = "cpu" + mean = -0.06463823467493057 + std = 0.24609768390655518 + data = None + +class Program_weight_tensor_meta_808f9d2ce5d246e9af03e955e99403d9: + name = "layer1.0.bn2.running_var" + shape = [64] + dtype = "torch.float32" + device = "cpu" + mean = 0.09406351298093796 + std = 0.06209423020482063 + data = None + +class Program_weight_tensor_meta_8ef71eaff15c4d36806ad63b0b95ba01: + name = "layer1.0.bn2.num_batches_tracked" + shape = [] + dtype = "torch.int64" + device = "cpu" + mean = None + std = None + data = [0] + +class Program_weight_tensor_meta_091cdabe1ac34d09bf932c8fbfd20ae2: + name = "layer1.1.bn1.running_mean" + shape = [64] + dtype = "torch.float32" + device = "cpu" + mean = -0.6165497303009033 + std = 0.8586963415145874 + data = None + +class Program_weight_tensor_meta_8db6d3fc0b40431d948dd8cf719f5a02: + name = "layer1.1.bn1.running_var" + shape = [64] + dtype = "torch.float32" + device = "cpu" + mean = 0.4359073042869568 + std = 0.17726503312587738 + data = None + +class Program_weight_tensor_meta_e320dc210a644e9983d9c1807228e518: + name = "layer1.1.bn1.num_batches_tracked" + shape = [] + dtype = "torch.int64" + device = "cpu" + mean = None + std = None + data = [0] + +class Program_weight_tensor_meta_855e0b6e1cb24b39ad4c2bcf9a9636f5: + name = "layer1.1.bn2.running_mean" + shape = [64] + dtype = "torch.float32" + device = "cpu" + mean = -0.04923836514353752 + std = 0.21268729865550995 + data = None + +class Program_weight_tensor_meta_de7999462f4647af9ac4eb45191fdc1f: + name = "layer1.1.bn2.running_var" + shape = [64] + dtype = "torch.float32" + device = "cpu" + mean = 0.05337625369429588 + std = 0.020933449268341064 + data = None + +class Program_weight_tensor_meta_8e6bbd68f6dc44da93d6f15307d1b6f1: + name = "layer1.1.bn2.num_batches_tracked" + shape = [] + dtype = "torch.int64" + device = "cpu" + mean = None + std = None + data = [0] + +class Program_weight_tensor_meta_904ac00a390449abbecea3949c6d4caf: + name = "layer2.0.bn1.running_mean" + shape = [128] + dtype = "torch.float32" + device = "cpu" + mean = -0.2939528524875641 + std = 0.43902844190597534 + data = None + +class Program_weight_tensor_meta_f182b7a748644c2087e36ea1d9b62b92: + name = "layer2.0.bn1.running_var" + shape = [128] + dtype = "torch.float32" + device = "cpu" + mean = 0.6451103687286377 + std = 0.17166651785373688 + data = None + +class Program_weight_tensor_meta_fb81a18fad704f648091054d674792e7: + name = "layer2.0.bn1.num_batches_tracked" + shape = [] + dtype = "torch.int64" + device = "cpu" + mean = None + std = None + data = [0] + +class Program_weight_tensor_meta_0c58403de72f4639bc2f681c91d99691: + name = "layer2.0.bn2.running_mean" + shape = [128] + dtype = "torch.float32" + device = "cpu" + mean = -0.19921736419200897 + std = 0.24043412506580353 + data = None + +class Program_weight_tensor_meta_e1970996c15a49c79b3f169798de1830: + name = "layer2.0.bn2.running_var" + shape = [128] + dtype = "torch.float32" + device = "cpu" + mean = 0.07541705667972565 + std = 0.040632884949445724 + data = None + +class Program_weight_tensor_meta_94dd328475ce454395685687e5f0c796: + name = "layer2.0.bn2.num_batches_tracked" + shape = [] + dtype = "torch.int64" + device = "cpu" + mean = None + std = None + data = [0] + +class Program_weight_tensor_meta_c49e3f7d4cdc45a18dca53be240947a2: + name = "layer2.0.downsample.1.running_mean" + shape = [128] + dtype = "torch.float32" + device = "cpu" + mean = -0.06607633084058762 + std = 0.30794447660446167 + data = None + +class Program_weight_tensor_meta_1ce573312db14be18cba92ffcf051691: + name = "layer2.0.downsample.1.running_var" + shape = [128] + dtype = "torch.float32" + device = "cpu" + mean = 0.07285968959331512 + std = 0.05314474180340767 + data = None + +class Program_weight_tensor_meta_16435170b3e842aa80a52273d920e01c: + name = "layer2.0.downsample.1.num_batches_tracked" + shape = [] + dtype = "torch.int64" + device = "cpu" + mean = None + std = None + data = [0] + +class Program_weight_tensor_meta_e8d82e7218614ded82a840b55543a00c: + name = "layer2.1.bn1.running_mean" + shape = [128] + dtype = "torch.float32" + device = "cpu" + mean = -0.2952965795993805 + std = 0.3311306834220886 + data = None + +class Program_weight_tensor_meta_06c754aa9fe94662a614882a88610ccd: + name = "layer2.1.bn1.running_var" + shape = [128] + dtype = "torch.float32" + device = "cpu" + mean = 0.22464340925216675 + std = 0.07565200328826904 + data = None + +class Program_weight_tensor_meta_27d51787a7464d2c9dc07a8298964bbc: + name = "layer2.1.bn1.num_batches_tracked" + shape = [] + dtype = "torch.int64" + device = "cpu" + mean = None + std = None + data = [0] + +class Program_weight_tensor_meta_f5e093e514ef487188201bb1eb240a95: + name = "layer2.1.bn2.running_mean" + shape = [128] + dtype = "torch.float32" + device = "cpu" + mean = -0.06012530252337456 + std = 0.12898705899715424 + data = None + +class Program_weight_tensor_meta_3fa2d273562449b2b14f88306ed7f135: + name = "layer2.1.bn2.running_var" + shape = [128] + dtype = "torch.float32" + device = "cpu" + mean = 0.032846398651599884 + std = 0.013435718603432178 + data = None + +class Program_weight_tensor_meta_3c24722e0bbe4e64a69da9e6d1428a90: + name = "layer2.1.bn2.num_batches_tracked" + shape = [] + dtype = "torch.int64" + device = "cpu" + mean = None + std = None + data = [0] + +class Program_weight_tensor_meta_afd25474ef5040f686f36f84024744cc: + name = "layer3.0.bn1.running_mean" + shape = [256] + dtype = "torch.float32" + device = "cpu" + mean = -0.32490915060043335 + std = 0.32831883430480957 + data = None + +class Program_weight_tensor_meta_c94c040c3e44420197e72bb0401ca858: + name = "layer3.0.bn1.running_var" + shape = [256] + dtype = "torch.float32" + device = "cpu" + mean = 0.2525804042816162 + std = 0.09307771176099777 + data = None + +class Program_weight_tensor_meta_c98b9ced5a8848fcb07d2321c4fc1e16: + name = "layer3.0.bn1.num_batches_tracked" + shape = [] + dtype = "torch.int64" + device = "cpu" + mean = None + std = None + data = [0] + +class Program_weight_tensor_meta_393b1710b04742dfa273d55773d052b8: + name = "layer3.0.bn2.running_mean" + shape = [256] + dtype = "torch.float32" + device = "cpu" + mean = -0.12485643476247787 + std = 0.1548074185848236 + data = None + +class Program_weight_tensor_meta_017c74bd984542ffa26b7ac2824257cd: + name = "layer3.0.bn2.running_var" + shape = [256] + dtype = "torch.float32" + device = "cpu" + mean = 0.10328030586242676 + std = 0.041878342628479004 + data = None + +class Program_weight_tensor_meta_5de6949e01734f9ba16345318e831c1a: + name = "layer3.0.bn2.num_batches_tracked" + shape = [] + dtype = "torch.int64" + device = "cpu" + mean = None + std = None + data = [0] + +class Program_weight_tensor_meta_bd25c63236fb488aa688c3cb2492f17a: + name = "layer3.0.downsample.1.running_mean" + shape = [256] + dtype = "torch.float32" + device = "cpu" + mean = -0.04699009284377098 + std = 0.08458859473466873 + data = None + +class Program_weight_tensor_meta_f8ef3a48d7e545f9b6f0fce5854b7f5a: + name = "layer3.0.downsample.1.running_var" + shape = [256] + dtype = "torch.float32" + device = "cpu" + mean = 0.014185900799930096 + std = 0.006529815495014191 + data = None + +class Program_weight_tensor_meta_15127a4a47364489bc3c79c9720f602a: + name = "layer3.0.downsample.1.num_batches_tracked" + shape = [] + dtype = "torch.int64" + device = "cpu" + mean = None + std = None + data = [0] + +class Program_weight_tensor_meta_0f486c9222584a5f8e45b5ef1df41d26: + name = "layer3.1.bn1.running_mean" + shape = [256] + dtype = "torch.float32" + device = "cpu" + mean = -0.39026427268981934 + std = 0.33702555298805237 + data = None + +class Program_weight_tensor_meta_f9a9c1e6da3a4626a70a415f99a1899a: + name = "layer3.1.bn1.running_var" + shape = [256] + dtype = "torch.float32" + device = "cpu" + mean = 0.14858609437942505 + std = 0.04802856966853142 + data = None + +class Program_weight_tensor_meta_d2aaf095f3fe4121ae2567eb14b749a7: + name = "layer3.1.bn1.num_batches_tracked" + shape = [] + dtype = "torch.int64" + device = "cpu" + mean = None + std = None + data = [0] + +class Program_weight_tensor_meta_0a0cee1d7d7e4fc29790178f04aabee4: + name = "layer3.1.bn2.running_mean" + shape = [256] + dtype = "torch.float32" + device = "cpu" + mean = -0.08337356150150299 + std = 0.10045936703681946 + data = None + +class Program_weight_tensor_meta_28ebeeaa201c4930bc5bf91b00c4013c: + name = "layer3.1.bn2.running_var" + shape = [256] + dtype = "torch.float32" + device = "cpu" + mean = 0.019977394491434097 + std = 0.009728903882205486 + data = None + +class Program_weight_tensor_meta_ed64b870fe0b49b3a50f840c3112b899: + name = "layer3.1.bn2.num_batches_tracked" + shape = [] + dtype = "torch.int64" + device = "cpu" + mean = None + std = None + data = [0] + +class Program_weight_tensor_meta_29430f6c85c342aa8c21867103f36219: + name = "layer4.0.bn1.running_mean" + shape = [512] + dtype = "torch.float32" + device = "cpu" + mean = -0.4088563323020935 + std = 0.2466270923614502 + data = None + +class Program_weight_tensor_meta_95d4338fa96449c98d4d9f3ab4ebc3ae: + name = "layer4.0.bn1.running_var" + shape = [512] + dtype = "torch.float32" + device = "cpu" + mean = 0.12907549738883972 + std = 0.03143839165568352 + data = None + +class Program_weight_tensor_meta_014e26ed28434b9895b25a125e6dc22c: + name = "layer4.0.bn1.num_batches_tracked" + shape = [] + dtype = "torch.int64" + device = "cpu" + mean = None + std = None + data = [0] + +class Program_weight_tensor_meta_80ad5dc9c9dc4df397922f1a14c3b8ba: + name = "layer4.0.bn2.running_mean" + shape = [512] + dtype = "torch.float32" + device = "cpu" + mean = -0.14308327436447144 + std = 0.06397316604852676 + data = None + +class Program_weight_tensor_meta_035b498c07b94dbea69b5150372ff75c: + name = "layer4.0.bn2.running_var" + shape = [512] + dtype = "torch.float32" + device = "cpu" + mean = 0.022369815036654472 + std = 0.0076219867914915085 + data = None + +class Program_weight_tensor_meta_c9879fa3d44a406b91b92c42a9d8b797: + name = "layer4.0.bn2.num_batches_tracked" + shape = [] + dtype = "torch.int64" + device = "cpu" + mean = None + std = None + data = [0] + +class Program_weight_tensor_meta_0fa9d538e74748f8858be7df47d1555c: + name = "layer4.0.downsample.1.running_mean" + shape = [512] + dtype = "torch.float32" + device = "cpu" + mean = -0.04498213902115822 + std = 0.07837548851966858 + data = None + +class Program_weight_tensor_meta_2be2d15264aa411bb81bc43126b7b19f: + name = "layer4.0.downsample.1.running_var" + shape = [512] + dtype = "torch.float32" + device = "cpu" + mean = 0.02036070078611374 + std = 0.008931221440434456 + data = None + +class Program_weight_tensor_meta_67657100d83047f499e962d40b466d29: + name = "layer4.0.downsample.1.num_batches_tracked" + shape = [] + dtype = "torch.int64" + device = "cpu" + mean = None + std = None + data = [0] + +class Program_weight_tensor_meta_7957f25fdfa44f509a13eec0be420db9: + name = "layer4.1.bn1.running_mean" + shape = [512] + dtype = "torch.float32" + device = "cpu" + mean = -0.5950218439102173 + std = 0.13785943388938904 + data = None + +class Program_weight_tensor_meta_4180fb03ddb147df9968212b0d5c23a1: + name = "layer4.1.bn1.running_var" + shape = [512] + dtype = "torch.float32" + device = "cpu" + mean = 0.13789209723472595 + std = 0.024087077006697655 + data = None + +class Program_weight_tensor_meta_08104b52296b421280856c6430d0551e: + name = "layer4.1.bn1.num_batches_tracked" + shape = [] + dtype = "torch.int64" + device = "cpu" + mean = None + std = None + data = [0] + +class Program_weight_tensor_meta_8a896cdc85134f348954a141f37852e3: + name = "layer4.1.bn2.running_mean" + shape = [512] + dtype = "torch.float32" + device = "cpu" + mean = -0.031392887234687805 + std = 0.020217176526784897 + data = None + +class Program_weight_tensor_meta_fcda38469a2a47d5a7cd6720fda3ab34: + name = "layer4.1.bn2.running_var" + shape = [512] + dtype = "torch.float32" + device = "cpu" + mean = 0.012696623802185059 + std = 0.0016607478028163314 + data = None + +class Program_weight_tensor_meta_415c39b7750a4510bbf7fc5090f031ae: + name = "layer4.1.bn2.num_batches_tracked" + shape = [] + dtype = "torch.int64" + device = "cpu" + mean = None + std = None + data = [0] diff --git a/samples/torch/utils.py b/samples/torch/utils.py new file mode 100644 index 000000000..79817ea79 --- /dev/null +++ b/samples/torch/utils.py @@ -0,0 +1,302 @@ +import re +import torch +import torch.nn as nn +from collections import OrderedDict +import uuid +import json +import os + +dyn_template = """ +import torch +from .. import utils +%MODULE + +model = GraphModule() + +inputs_params = utils.load_converted_from_text(f'./source_tensor_meta.py') +inputs = inputs_params["input_info"] +inputs = [utils.replay_tensor(i) for i in inputs] +params = inputs_params["weight_info"] + +state_dict = {} +for k, v in params.items(): + k = utils.convert_param_name(k) + v = utils.replay_tensor(v) + state_dict[k] = v + +y = model(x=inputs[0], **state_dict)[0] +print(torch.argmin(y), torch.argmax(y)) +print(y.shape) +""" + +def save_constraints_text(converted, file_path): + lines = [] + if converted["dynamic_shapes"] is not None: + raise NotImplementedError("Handling constraints is not implemented yet.") + with open(file_path, 'w') as f: + f.write("\n".join(lines)) + +def save_converted_to_text(converted, file_path): + def generate_uid(): + return str(uuid.uuid4()).replace('-', '') + + def format_data(data): + if data is None: + return "None" + elif isinstance(data, torch.Tensor): + if data.dtype.is_floating_point: + return "[{}]".format(", ".join(f'{x:.6f}' for x in data.tolist())) + else: + return "[{}]".format(", ".join(f'{x}' for x in data.tolist())) + else: + return repr(data) + + lines = [] + + def process_tensor_info(tensor_info, name_prefix="example_input"): + data_list = None + if "input_" in tensor_info["name"]: + if tensor_info["type"] in ["small_tensor", "small_int_tensor"]: + data_list = tensor_info["data"].flatten() + elif tensor_info["type"] == "big_int_tensor": + data_list = f'pt-filename:xxx-key' + else: + pass + else: + if tensor_info["type"] == "small_int_tensor": + data_list = tensor_info["data"].flatten() + if tensor_info["type"] == "big_int_tensor": + raise ValueError("Unexpected cases: there are weights in big tensor of int type ") + info = tensor_info.get("info", {}) + dtype = info.get("dtype", "torch.float") + shape = info.get("shape", []) + device = info.get("device", "cpu") + mean = info.get("mean", 0.0) + std = info.get("std", 1.0) + + uid = f"{name_prefix}_tensor_meta_{generate_uid()}" + lines.append(f"class {uid}:") + lines.append(f"\tname = \"{tensor_info.get('name', '')}\"") + lines.append(f"\tshape = {shape}") + lines.append(f"\tdtype = \"{dtype}\"") + lines.append(f"\tdevice = \"{device}\"") + lines.append(f"\tmean = {mean}") + lines.append(f"\tstd = {std}") + lines.append(f"\tdata = {format_data(data_list)}") + lines.append("") + + input_infos = converted["input_info"] + if isinstance(input_infos, dict): + input_infos = [input_infos] + + for idx, input_info in enumerate(input_infos): + input_info["name"] = f"input_{idx}" + process_tensor_info(input_info, name_prefix="Program_input") + + for name, weight_info in converted["weight_info"].items(): + weight_info["name"] = name + process_tensor_info(weight_info, name_prefix="Program_weight") + + with open(file_path, 'w') as f: + f.write("\n".join(lines)) + +def load_converted_from_text(file_path): + + def parse_value(value_str): + value_str = value_str.strip() + if value_str == "None": + return None + if value_str == "[]": + return [] + elif value_str.startswith('"') or value_str.startswith("'"): + return value_str[1:-1] + elif value_str.startswith('['): + elements = value_str[1:-1].split(',') + result = [] + for e in elements: + e = e.strip() + try: + result.append(eval(e)) + except: + result.append(e) + return result + else: + try: + return eval(value_str) + except: + return value_str + + with open(file_path, 'r') as f: + lines = f.readlines() + + classes = [] + current_class = None + + for line in lines: + line = line.strip() + if line.startswith("class "): + if current_class is not None: + classes.append(current_class) + current_class = {} + elif "=" in line: + key, val = line.split("=", 1) + key = key.strip() + val = val.strip() + current_class[key] = parse_value(val) + + if current_class is not None: + classes.append(current_class) + + input_info = [] + weight_info = {} + + for cls in classes: + if 'input_' in cls["name"]: + item = { + "type": "random_tensor", + "info": { + "shape": cls.get("shape", []), + "dtype": getattr(torch, cls.get("dtype", "torch.float").split('.')[-1]), + "device": cls.get("device", "cpu"), + "mean": cls.get("mean", 0.0), + "std": cls.get("std", 1.0), + } + } + if cls.get("data") is not None: + if isinstance(cls.get("data"), str): + pass + else: + item["data"] = torch.tensor(cls["data"], dtype=item["info"]["dtype"]).reshape(cls.get("shape"), []) + input_info.append(item) + else: + data_value = None + data_type = getattr(torch, cls.get("dtype", "torch.float").split('.')[-1]) + if cls.get("data") is not None: + if isinstance(cls.get("data"), str): + raise ValueError("Unimplemented") + else: + data_value = torch.tensor(cls["data"], dtype=data_type).reshape(cls.get("shape"), []) + weight_info[cls["name"]] = { + "info": { + "shape": cls.get("shape", []), + "dtype": data_type, + "device": cls.get("device", "cpu"), + "mean": cls.get("mean", 0.0), + "std": cls.get("std", 1.0), + }, + "data": data_value, + } + + return { + "input_info": input_info if len(input_info) > 0 else None, + "weight_info": weight_info, + "dynamic_shapes": None + } + +def extract_dynamic_shapes(example_inputs): + pass + +def replay_tensor(info): + + device = info["info"]["device"] + dtype = info["info"]["dtype"] + shape = info["info"]["shape"] + + mean = info["info"]["mean"] + std = info["info"]["std"] + if info["data"] is not None: + return info["data"].to(device) + return torch.randn(size=shape).to(dtype).to(device) * std * 0.2 + mean + +def convert_state_and_inputs(state_dict, example_inputs): + def tensor_info(tensor): + is_float = tensor.dtype.is_floating_point + return { + "shape": list(tensor.shape), + "dtype": str(tensor.dtype), + "device": str(tensor.device), + "mean": float(tensor.mean().item()) if is_float else None, + "std": float(tensor.std().item()) if is_float else None, + } + + def process_tensor(tensor): + if not isinstance(tensor, torch.Tensor): + return {"type": "unknown", "value": tensor} + + info = tensor_info(tensor) + if tensor.dtype in [torch.int8, torch.int16, torch.int32, torch.int64]: + if tensor.numel() < 1024: + return {"type": "small_int_tensor", "data": tensor.clone(), "info": info} + else: + return {"type": "big_int_tensor", "data": tensor.clone(), "info": info} + elif tensor.numel() < 1024: + return {"type": "small_tensor", "data": tensor.clone(), "info": info} + else: + return {"type": "random_tensor", "info": info} + + if isinstance(example_inputs, torch.Tensor): + processed_inputs = process_tensor(example_inputs) + elif isinstance(example_inputs, (list, tuple)): + processed_inputs = [process_tensor(t) for t in example_inputs] + else: + processed_inputs = {"type": "unknown", "value": example_inputs} + + processed_weights = {} + for key, tensor in state_dict.items(): + data_value = None + data_type = "random_tensor" + if tensor.dtype in [torch.int8, torch.int16, torch.int32, torch.int64]: + if tensor.numel() < 1024: + data_type = "small_int_tensor" + data_value = tensor.clone() + else: + data_type = "big_int_tensor" + + info = tensor_info(tensor) + processed_weights[key] = {"info": info} + processed_weights[key]["data"] = data_value + processed_weights[key]["type"] = data_type + + # dynamic_shapes = extract_dynamic_shapes(example_inputs) + return { + "input_info": processed_inputs, + "weight_info": processed_weights, + "dynamic_shapes": None + } + +def convert_param_name(original_name): + if original_name.endswith(('.weight', '.bias')): + prefix = 'p_' + base_name = original_name + + elif any(x in original_name for x in ['running_mean', 'running_var', 'num_batches_tracked']): + prefix = 'b_' + base_name = original_name + else: + raise ValueError(f"Unrecognized parameter type: {original_name}") + + if '.' in base_name: + parts = base_name.split('.') + if len(parts) == 2 and not parts[0].startswith('layer'): + return prefix + parts[0] + '_' + parts[1] + else: + # layer1.0 -> layer1___0___ + pattern = r'(layer\d+)\.(\d+)\.' + replacement = r'\1___\2___' + converted = re.sub(pattern, replacement, base_name) + converted = converted.replace('.', '_') + return f"{prefix}getattr_l__self___{converted}" + else: + return prefix + base_name + +def indent_with_tab(code: str) -> str: + lines = code.splitlines() + indented_lines = [f" {line}" for line in lines] + return "\n".join(indented_lines) + +def apply_templates(code: str) -> str: + code = indent_with_tab(code) + code = code.replace(" GraphModule()", "class GraphModule(torch.nn.Module):") + code = code.replace(" \n" * 3, "\n") + py_code = dyn_template.replace('%MODULE', code) + return py_code From 9839ef4e9498db869c92ce8f9f4bb44d406188cf Mon Sep 17 00:00:00 2001 From: hxzd5568 <3257591325@qq.com> Date: Thu, 17 Jul 2025 20:34:10 +0800 Subject: [PATCH 03/10] update dir --- README.md | 21 +- .../input_tensor_meta.py | 547 ------------ .../model.py | 826 ------------------ samples/torch/resnet18/attribute.json | 4 +- samples/torch/resnet18/model.py | 19 - torch_tools/README.md | 13 + torch_tools/execution/runner.py | 50 ++ torch_tools/execution/utils.py | 138 +++ .../torch => torch_tools/generators}/utils.py | 298 ++----- .../generators/vision_model_generator.py | 101 +++ 10 files changed, 420 insertions(+), 1597 deletions(-) delete mode 100644 samples/paddle/PaddleX_AutoEncoder_ad_dy2st_train/input_tensor_meta.py delete mode 100644 samples/paddle/PaddleX_AutoEncoder_ad_dy2st_train/model.py create mode 100644 torch_tools/README.md create mode 100644 torch_tools/execution/runner.py create mode 100644 torch_tools/execution/utils.py rename {samples/torch => torch_tools/generators}/utils.py (66%) create mode 100644 torch_tools/generators/vision_model_generator.py diff --git a/README.md b/README.md index 95b79d737..c9e8490c2 100644 --- a/README.md +++ b/README.md @@ -1 +1,20 @@ -# GraphNet \ No newline at end of file +# GraphNet + +🧠 GraphNet：高性能编译器优化的基准数据集 +近年来，尽管编译器具备强大的通用优化能力，但其在实际工程中的应用却受到结构复杂、学习成本高、优化 Pass 开发门槛高等因素的限制。为了释放编译器的潜力，我们提出了 GraphNet —— 一个面向高性能编译器优化的大规模基准数据集，旨在为研究者和开发者提供一个统一、开放、可扩展的实验平台。 + +📌 项目简介 +GraphNet 包含大量来自真实高性能计算任务的图结构表示（如计算图、IR图、融合模式等），可用于评估编译器Pass的优化效果、训练编译器智能体、实现从示例到优化Pass的自动生成。 + +通过 GraphNet，用户可以： + +快速测试不同优化策略的效果； +训练模型以自动生成编译器优化Pass； +推动编译器在统一硬件接口、广泛适配后端方面的能力； +降低高性能算法集成到编译器中的门槛。 + +🔍 核心特性 + +📊 数据集概览 + +🧰 使用工具与接口 diff --git a/samples/paddle/PaddleX_AutoEncoder_ad_dy2st_train/input_tensor_meta.py b/samples/paddle/PaddleX_AutoEncoder_ad_dy2st_train/input_tensor_meta.py deleted file mode 100644 index f19a344a5..000000000 --- a/samples/paddle/PaddleX_AutoEncoder_ad_dy2st_train/input_tensor_meta.py +++ /dev/null @@ -1,547 +0,0 @@ -class PirProgram_example_input_tensor_meta_5389940036631769731: - program_id = 880737988565058868 - input_name = "linear_1.b_0" - shape = [16] - data = None - - -class PirProgram_example_input_tensor_meta_43418529140942226: - program_id = 880737988565058868 - input_name = "linear_1.w_0" - shape = [32, 16] - data = None - - -class PirProgram_example_input_tensor_meta_3606508939262698655: - program_id = 880737988565058868 - input_name = "linear_0.b_0" - shape = [32] - data = None - - -class PirProgram_example_input_tensor_meta_3554240975606516965: - program_id = 880737988565058868 - input_name = "linear_0.w_0" - shape = [96, 32] - data = None - - -class PirProgram_example_input_tensor_meta_5179986117148289115: - program_id = 880737988565058868 - input_name = "args_0" - shape = [16, 2, 96] - data = None - - -class PirProgram_example_input_tensor_meta_659966657961177331: - program_id = 5237713637574795923 - input_name = "linear_3.b_0" - shape = [96] - data = None - - -class PirProgram_example_input_tensor_meta_3195965365640632706: - program_id = 5237713637574795923 - input_name = "linear_3.w_0" - shape = [32, 96] - data = None - - -class PirProgram_example_input_tensor_meta_3617487774345626831: - program_id = 5237713637574795923 - input_name = "linear_2.b_0" - shape = [32] - data = None - - -class PirProgram_example_input_tensor_meta_5636200344230617383: - program_id = 5237713637574795923 - input_name = "linear_2.w_0" - shape = [16, 32] - data = None - - -class PirProgram_example_input_tensor_meta_5228671372811070830: - program_id = 5237713637574795923 - input_name = "args_0" - shape = [16, 2, 16] - data = [float("-1.24783"), float("1.35968"), float("-0.0643954"), float("-0.627694"), float("3.57834"), float("-1.40647"), float("-3.06211"), float("-0.852737"), float("1.89044"), float("1.71429"), float("0.873626"), float("2.45266"), float("2.46995"), float("-4.18985"), float("5.24214"), float("-0.769194"), float("-0.432318"), float("1.44443"), float("-0.398696"), float("-1.49099"), float("-0.245603"), float("-1.75011"), float("-1.19734"), float("0.138174"), float("1.28394"), float("-0.147029"), float("1.46563"), float("0.138966"), float("1.49462"), float("-1.24567"), float("1.74074"), float("-0.541634"), float("-0.137766"), float("0.149732"), float("0.813494"), float("0.303299"), float("-0.238571"), float("1.88537"), float("-0.0370909"), float("0.573859"), float("0.349418"), float("0.487316"), float("-1.32582"), float("-0.843882"), float("-0.676632"), float("-0.909084"), float("0.270721"), float("-1.23429"), float("-0.0424248"), float("0.429659"), float("-0.14267"), float("-0.428278"), float("-0.0162147"), float("-0.494324"), float("-0.285505"), float("0.0283108"), float("0.343003"), float("0.0251735"), float("0.319188"), float("0.0917862"), float("0.402113"), float("-0.271283"), float("0.494286"), float("-0.191708"), float("-0.284004"), float("0.298903"), float("1.58414"), float("0.506222"), float("-0.530693"), float("3.27026"), float("-0.117699"), float("1.01495"), float("0.711183"), float("0.811757"), float("-2.06371"), float("-1.2438"), float("-1.13482"), float("-1.38332"), float("0.486574"), float("-2.27285"), float("-0.358158"), float("0.179657"), float("0.702598"), float("0.133837"), float("-0.116537"), float("1.53755"), float("0.00049031"), float("0.551636"), float("0.176739"), float("0.308397"), float("-0.844465"), float("-0.537499"), float("-0.755579"), float("-0.623192"), float("0.182918"), float("-1.02258"), float("-1.34653"), float("0.913991"), float("1.1851"), float("0.400625"), float("0.121387"), float("3.74875"), float("0.507555"), float("1.24397"), float("-0.00871566"), float("0.270786"), float("-2.39195"), float("-1.38387"), float("-2.33894"), float("-0.779432"), float("0.529702"), float("-2.01132"), float("-0.0355457"), float("0.156263"), float("-0.0398702"), float("-0.0941751"), float("0.070695"), float("-0.125185"), float("-0.123143"), float("-0.0328905"), float("0.0242618"), float("0.050551"), float("0.0789997"), float("0.0688375"), float("0.0704337"), float("-0.152359"), float("0.139597"), float("-0.0303556"), float("-0.422165"), float("0.755079"), float("-0.125477"), float("-0.390056"), float("0.129337"), float("-0.44178"), float("-0.83648"), float("-0.363755"), float("-0.396812"), float("-0.217299"), float("1.19051"), float("0.353966"), float("0.100611"), float("-0.841632"), float("1.13851"), float("-0.0570613"), float("-0.321325"), float("1.00824"), float("-0.478592"), float("-1.40818"), float("-0.435232"), float("-1.11203"), float("-0.501488"), float("-0.127583"), float("0.967821"), float("0.100076"), float("0.722625"), float("0.514725"), float("0.67351"), float("-0.719309"), float("0.999483"), float("-0.640918"), float("-1.46169"), float("1.59347"), float("3.25034"), float("0.413147"), float("0.432863"), float("5.4329"), float("0.0917308"), float("1.98376"), float("0.598049"), float("1.64893"), float("-1.92021"), float("-1.70858"), float("-2.384"), float("-1.1942"), float("0.144455"), float("-3.21842"), float("-1.22085"), float("-0.843408"), float("3.91006"), float("0.0781953"), float("-1.01348"), float("6.29112"), float("1.98013"), float("2.26446"), float("0.98046"), float("0.878038"), float("-4.09496"), float("-3.09536"), float("-2.928"), float("-3.87442"), float("-0.000862794"), float("-3.55961"), float("-0.1064"), float("-0.161292"), float("0.187614"), float("0.0255209"), float("-0.162917"), float("0.42479"), float("-0.0453343"), float("0.436514"), float("-0.191061"), float("0.137538"), float("-0.410105"), float("-0.431748"), float("-0.438004"), float("-0.404633"), float("0.218743"), float("-0.427686"), float("-0.424405"), float("0.837761"), float("0.332992"), float("-0.0513256"), float("0.047148"), float("1.35676"), float("-0.0688395"), float("0.506198"), float("0.163514"), float("-0.337938"), float("-1.2557"), float("-0.616636"), float("-1.1297"), float("-0.193176"), float("0.337317"), float("-0.561698"), float("-0.135902"), float("0.393851"), float("0.00424394"), float("-0.196324"), float("0.159519"), float("-0.308631"), float("-0.376123"), float("-0.071268"), float("0.0883756"), float("0.0545766"), float("0.303032"), float("0.111892"), float("0.218541"), float("-0.537371"), float("0.421097"), float("-0.129858"), float("-1.03447"), float("0.513909"), float("2.0756"), float("0.441222"), float("-0.472549"), float("4.29349"), float("0.13518"), float("1.6049"), float("0.482167"), float("0.853731"), float("-2.33474"), float("-1.43174"), float("-2.12045"), float("-1.69272"), float("0.520399"), float("-2.92517"), float("-0.154838"), float("1.12713"), float("-0.729672"), float("-1.66338"), float("-0.657706"), float("-1.18132"), float("-0.874759"), float("0.06689"), float("1.51983"), float("0.351833"), float("0.986195"), float("0.245073"), float("0.796269"), float("-0.622144"), float("0.948545"), float("-1.01152"), float("-1.22342"), float("2.62304"), float("-0.221326"), float("-2.36949"), float("0.0231131"), float("-3.1239"), float("-1.56313"), float("-0.0846508"), float("1.4516"), float("-0.843756"), float("2.11236"), float("0.770368"), float("2.54804"), float("-2.68047"), float("3.21551"), float("-0.423819"), float("-0.392694"), float("0.187697"), float("0.413998"), float("0.057787"), float("-0.13115"), float("1.17413"), float("0.0864602"), float("0.307451"), float("0.0920442"), float("0.143506"), float("-0.635472"), float("-0.255527"), float("-0.542642"), float("-0.322647"), float("0.183446"), float("-0.825342"), float("-0.0203514"), float("0.0198951"), float("0.0865615"), float("0.0272587"), float("-0.00242023"), float("0.195467"), float("0.07148"), float("0.0644732"), float("0.0219086"), float("0.0301208"), float("-0.185134"), float("-0.132513"), float("-0.0794464"), float("-0.100434"), float("0.0120146"), float("-0.08797"), float("-1.29686"), float("1.00538"), float("-0.771612"), float("-1.60371"), float("1.04745"), float("-1.71847"), float("-0.364911"), float("-0.669676"), float("1.52801"), float("-0.0574329"), float("0.59866"), float("1.73576"), float("2.91115"), float("-1.2902"), float("3.56666"), float("0.497791"), float("-0.629735"), float("1.17433"), float("-0.648112"), float("-1.04914"), float("-0.195696"), float("-1.03082"), float("-0.321714"), float("-0.668095"), float("-0.117482"), float("-0.81666"), float("1.54317"), float("0.372422"), float("0.500825"), float("-0.542041"), float("1.91077"), float("0.324363"), float("-0.746501"), float("1.56012"), float("-0.693475"), float("-1.47149"), float("0.417272"), float("-1.5822"), float("-0.614944"), float("-0.337996"), float("0.782261"), float("-0.244622"), float("1.03057"), float("0.317955"), float("1.1072"), float("-1.29199"), float("1.91734"), float("0.0562871"), float("-0.344145"), float("0.582381"), float("0.0177077"), float("-0.394364"), float("0.278082"), float("-0.549242"), float("-0.608972"), float("-0.148099"), float("0.0354738"), float("0.105372"), float("0.54669"), float("0.293604"), float("0.384397"), float("-0.770095"), float("0.530887"), float("-0.0841932"), float("-1.5664"), float("1.47664"), float("-0.0485786"), float("-1.2655"), float("1.12964"), float("-1.11663"), float("-0.127906"), float("-0.158063"), float("0.266339"), float("-0.689339"), float("0.552855"), float("-0.29022"), float("1.41004"), float("-1.61181"), float("1.21825"), float("0.138486"), float("-1.40203"), float("1.25058"), float("0.718333"), float("0.433586"), float("0.213605"), float("3.4107"), float("0.702294"), float("0.844367"), float("-0.269252"), float("0.469749"), float("-2.07015"), float("-0.969774"), float("-1.84929"), float("-0.536334"), float("0.553032"), float("-2.07369"), float("-0.875411"), float("1.6934"), float("-0.485379"), float("-1.53678"), float("0.237551"), float("-1.80035"), float("-0.801288"), float("-0.196758"), float("1.13256"), float("-0.170934"), float("1.21607"), float("0.530888"), float("1.79473"), float("-1.66515"), float("2.21643"), float("-0.170763"), float("-0.242361"), float("0.0761521"), float("0.432139"), float("0.0631636"), float("-0.093372"), float("0.938469"), float("0.0142968"), float("0.313853"), float("0.120945"), float("0.144088"), float("-0.496069"), float("-0.284314"), float("-0.471705"), float("-0.335182"), float("0.114975"), float("-0.613756"), float("-0.133989"), float("0.332343"), float("-0.0773375"), float("-0.386288"), float("-0.0153539"), float("-0.397145"), float("-0.344031"), float("0.0805068"), float("0.348394"), float("0.00255038"), float("0.33544"), float("0.0063985"), float("0.34345"), float("-0.323657"), float("0.404106"), float("-0.197486"), float("-0.0204629"), float("0.165297"), float("0.344726"), float("0.0223876"), float("-0.00523791"), float("0.623413"), float("-0.16317"), float("0.276788"), float("0.216639"), float("0.113306"), float("-0.499786"), float("-0.285747"), float("-0.187579"), float("-0.235846"), float("0.0960729"), float("-0.395305"), float("-0.0911418"), float("0.288873"), float("-0.0312524"), float("-0.338057"), float("-0.0387871"), float("-0.421691"), float("-0.315626"), float("0.125869"), float("0.273582"), float("-0.058278"), float("0.3202"), float("0.015641"), float("0.297726"), float("-0.288548"), float("0.329069"), float("-0.179419"), float("-0.0326695"), float("0.117692"), float("-0.060717"), float("-0.112393"), float("0.0348706"), float("-0.105831"), float("-0.0257303"), float("-0.0271061"), float("0.0484697"), float("-0.00249167"), float("0.0425337"), float("0.0319538"), float("0.0739469"), float("-0.0675583"), float("0.136938"), float("-0.0154109")] - - -class PirProgram_example_input_tensor_meta_6104757736120590527: - program_id = 1242071021843173094 - input_name = "middle_5" - shape = [16, 2, 96] - data = None - - -class PirProgram_example_input_tensor_meta_42754444789929111: - program_id = 1242071021843173094 - input_name = "middle_4" - shape = [16, 2, 32] - data = None - - -class PirProgram_example_input_tensor_meta_3577094903844169474: - program_id = 1242071021843173094 - input_name = "output_grad_0" - shape = [16, 2, 96] - data = None - - -class PirProgram_example_input_tensor_meta_1198995949523510110: - program_id = 1242071021843173094 - input_name = "linear_3.w_0" - shape = [32, 96] - data = None - - -class PirProgram_example_input_tensor_meta_8553907544768773751: - program_id = 1242071021843173094 - input_name = "middle_2" - shape = [1] - data = [float("0.2")] - - -class PirProgram_example_input_tensor_meta_4986101899439820377: - program_id = 1242071021843173094 - input_name = "linear_3.b_0" - shape = [96] - data = [float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0")] - - -class PirProgram_example_input_tensor_meta_8512202976348838334: - program_id = 1242071021843173094 - input_name = "middle_3" - shape = [16, 2, 32] - data = [float("0.0347782"), float("1.16674"), float("0"), float("0.33759"), float("0"), float("0"), float("2.70181"), float("0"), float("0"), float("0"), float("1.46725"), float("1.50891"), float("0"), float("0"), float("0.539994"), float("0"), float("1.30105"), float("0"), float("0"), float("2.44845"), float("0"), float("0"), float("0"), float("0.298854"), float("0"), float("0"), float("0"), float("0.978712"), float("0"), float("0"), float("2.00676"), float("0"), float("0"), float("0.0118907"), float("0"), float("1.28928"), float("0"), float("0"), float("0.902585"), float("0.740692"), float("0"), float("0.553495"), float("0"), float("0"), float("0.295481"), float("0"), float("0"), float("0"), float("0"), float("0.87983"), float("0"), float("1.30724"), float("0"), float("0"), float("0"), float("1.35767"), float("0.412437"), float("0"), float("1.42755"), float("0"), float("0"), float("0"), float("1.31564"), float("0"), float("0.0323608"), float("0.438107"), float("0"), float("0"), float("0"), float("0.759341"), float("0"), float("0"), float("0"), float("0.935668"), float("0"), float("1.4212"), float("0.437199"), float("0"), float("0.36504"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.397975"), float("0"), float("0"), float("0.335801"), float("0.170311"), float("0"), float("0.170158"), float("0"), float("0"), float("0.0497961"), float("0"), float("0"), float("0"), float("0"), float("0.21666"), float("0"), float("0.381272"), float("0"), float("0"), float("0"), float("0.35843"), float("0.109284"), float("0"), float("0.436922"), float("0"), float("0.0547679"), float("0"), float("0.377323"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("1.24158"), float("0"), float("0"), float("2.46088"), float("1.5229"), float("0"), float("2.31923"), float("0.538648"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.258044"), float("0"), float("0"), float("0"), float("0"), float("0"), float("1.01807"), float("0"), float("1.60786"), float("0"), float("0.818661"), float("0"), float("0"), float("0"), float("0.319663"), float("0"), float("0"), float("0"), float("0.711094"), float("0"), float("0"), float("1.17478"), float("0"), float("1.00297"), float("0.975026"), float("0.0937734"), float("0.0181032"), float("0.17601"), float("0"), float("0"), float("0"), float("0"), float("0.160587"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.650962"), float("0"), float("0.506464"), float("0"), float("0.501895"), float("0"), float("0.564878"), float("0"), float("0.826023"), float("0"), float("0"), float("0"), float("2.16452"), float("0"), float("0"), float("3.52679"), float("1.14489"), float("2.1719"), float("0"), float("0.178483"), float("0"), float("0"), float("0.522222"), float("0"), float("0"), float("0.336636"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("1.47996"), float("0"), float("1.47095"), float("0"), float("1.57665"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.0702505"), float("0"), float("0.0277561"), float("0.115806"), float("0.0211074"), float("0"), float("0"), float("0.0265309"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.0315472"), float("0.028493"), float("0.00156406"), float("0.167891"), float("0"), float("0"), float("0"), float("0.0448072"), float("0"), float("0"), float("0.149831"), float("0"), float("0.0563955"), float("0"), float("0"), float("0.010389"), float("0"), float("0.0195092"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.677082"), float("0.256379"), float("0.328802"), float("0.26332"), float("0"), float("0.135194"), float("0"), float("0"), float("0.15623"), float("0"), float("0"), float("0.823846"), float("0"), float("0.437986"), float("0"), float("0.122316"), float("0"), float("0"), float("0"), float("0"), float("0.952759"), float("0"), float("0"), float("0"), float("0.0882541"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.198234"), float("0"), float("0"), float("0"), float("0"), float("0"), float("1.00134"), float("0"), float("0"), float("1.40157"), float("0"), float("0"), float("0"), float("0.800234"), float("0"), float("0"), float("0.18145"), float("0"), float("0"), float("0"), float("2.90421"), float("0"), float("0"), float("5.13843"), float("1.78794"), float("2.42338"), float("3.48572"), float("0"), float("0.827473"), float("0.628247"), float("1.24771"), float("0"), float("0"), float("0"), float("1.20432"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("1.76207"), float("0"), float("2.24063"), float("0"), float("0.8273"), float("0.0147218"), float("0.535274"), float("0"), float("0"), float("0"), float("3.12057"), float("0"), float("0"), float("4.5842"), float("3.66083"), float("3.91709"), float("5.36272"), float("1.2578"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("1.98505"), float("0"), float("0.42635"), float("0"), float("0"), float("0"), float("2.05989"), float("0.0728085"), float("0.442795"), float("0.275162"), float("0.000693649"), float("0"), float("0.136329"), float("0"), float("0"), float("0.208206"), float("0.211367"), float("0"), float("0.226844"), float("0.222011"), float("0"), float("0.138711"), float("0"), float("0"), float("0"), float("0"), float("0.162163"), float("0"), float("0.386705"), float("0"), float("0"), float("0.0113519"), float("0.248948"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.337618"), float("0"), float("0"), float("0.143206"), float("0.973093"), float("0"), float("0"), float("1.5557"), float("0.547614"), float("0.808207"), float("0.809795"), float("0.207993"), float("0.226023"), float("0"), float("0.446442"), float("0"), float("0"), float("0.298011"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.550403"), float("0"), float("0.662424"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.0034696"), float("0.222743"), float("0.020042"), float("0"), float("0"), float("0.0433677"), float("0"), float("0"), float("0"), float("0"), float("0.0376502"), float("0.0464511"), float("0.101323"), float("0.00628365"), float("0.449533"), float("0"), float("0"), float("0"), float("0.0415141"), float("0"), float("0"), float("0.329318"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.871903"), float("0"), float("0"), float("0"), float("2.00545"), float("0"), float("0"), float("0"), float("1.72249"), float("2.65814"), float("2.60701"), float("0.273242"), float("0.0238184"), float("0.55817"), float("0"), float("0"), float("0"), float("0"), float("0.387383"), float("0"), float("0"), float("0"), float("0"), float("0"), float("1.78813"), float("0"), float("1.34975"), float("0"), float("1.48764"), float("0"), float("1.50652"), float("0"), float("0"), float("0"), float("1.50629"), float("0"), float("0"), float("0.539916"), float("0.389776"), float("0"), float("1.01698"), float("0.107516"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.578256"), float("0"), float("1.2537"), float("0"), float("0"), float("0"), float("1.35418"), float("0.705801"), float("0"), float("1.52801"), float("0"), float("0"), float("0"), float("1.11237"), float("0"), float("0"), float("0.0364399"), float("0"), float("1.3821"), float("0"), float("0"), float("1.9378"), float("0.935661"), float("0"), float("0.557419"), float("0"), float("0"), float("0.87944"), float("0"), float("0"), float("0"), float("0"), float("0.81148"), float("0"), float("1.97005"), float("0"), float("0"), float("0"), float("0"), float("0.119622"), float("0"), float("2.45299"), float("0"), float("0.501118"), float("0"), float("1.94722"), float("0"), float("0"), float("0.272278"), float("0"), float("0"), float("0"), float("0.572684"), float("0"), float("0"), float("0.837621"), float("0.405361"), float("0.676569"), float("0.585324"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.456107"), float("0"), float("0.319202"), float("0"), float("0.391309"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.116754"), float("0"), float("0"), float("0.203108"), float("0"), float("0.132567"), float("0.181899"), float("0.087103"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.0459756"), float("0"), float("0"), float("0"), float("0.0294301"), float("0"), float("0.0783622"), float("0"), float("0.0726098"), float("0"), float("0.0563935"), float("0"), float("0.73509"), float("0"), float("0"), float("0"), float("0"), float("2.19167"), float("0.614574"), float("0"), float("0"), float("0"), float("0.149091"), float("0.140033"), float("0"), float("0.175131"), float("0"), float("1.20836"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.0784062"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.0162837"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.737377"), float("0.196769"), float("0"), float("0.265471"), float("0"), float("0.740579"), float("0"), float("0"), float("0.955313"), float("0.0152713"), float("0"), float("0"), float("0"), float("0.576584"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.663437"), float("0"), float("0"), float("1.01948"), float("0.675586"), float("0"), float("0.312976"), float("0"), float("0"), float("0.292665"), float("0"), float("0"), float("0"), float("0"), float("0.0120792"), float("0.0180582"), float("0.879502"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("1.43309"), float("0"), float("0.42474"), float("0"), float("1.11088"), float("0"), float("0"), float("0"), float("0"), float("0.178277"), float("0"), float("0.0671259"), float("0.422192"), float("0.127027"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.294385"), float("0.217661"), float("0.0798928"), float("0.724993"), float("0"), float("0"), float("0"), float("0.182339"), float("0"), float("0"), float("0.597919"), float("0"), float("0.238543"), float("0"), float("0.412386"), float("0"), float("0"), float("0.0349023"), float("0"), float("0.222169"), float("0"), float("0"), float("0.909737"), float("1.39266"), float("0"), float("0.19613"), float("0"), float("0.478526"), float("0"), float("0"), float("0"), float("0.593163"), float("0.518306"), float("0"), float("0.881752"), float("0.491215"), float("0"), float("0"), float("0"), float("0.242513"), float("0"), float("0"), float("0.900461"), float("0"), float("0.303094"), float("0"), float("0.754786"), float("0"), float("0"), float("0.644119"), float("0"), float("0"), float("0"), float("2.18745"), float("0"), float("0"), float("3.23613"), float("1.11745"), float("1.86997"), float("0"), float("0.0982345"), float("0.664672"), float("0"), float("0"), float("0"), float("0"), float("0.438279"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("1.45317"), float("0"), float("1.58249"), float("0"), float("0"), float("0"), float("0.975825"), float("0"), float("0"), float("0"), float("1.01558"), float("0"), float("0"), float("1.36234"), float("0.717539"), float("0"), float("0.324656"), float("0"), float("0"), float("0.4705"), float("0"), float("0"), float("0"), float("0"), float("0.304301"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.198254"), float("0"), float("0"), float("0"), float("0.416886"), float("0"), float("0"), float("0.686297"), float("0.334294"), float("0.576713"), float("0.580936"), float("0"), float("0.00356044"), float("0.0984948"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.387813"), float("0"), float("0.303094"), float("0"), float("0"), float("0"), float("0.336003"), float("0"), float("0.0410064"), float("0"), float("0.342481"), float("0"), float("0"), float("0.198419"), float("0.214347"), float("0"), float("0"), float("0"), float("0"), float("0.0215962"), float("0"), float("0"), float("0"), float("0"), float("0.203783"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.293901"), float("0.111309"), float("0"), float("0.333862"), float("0"), float("0"), float("0"), float("0.320106"), float("0"), float("0"), float("0.193667"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.270118"), float("0"), float("0.590756"), float("0.117252"), float("0"), float("0.000430197"), float("0"), float("0"), float("0"), float("0"), float("0.181376"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.0962679"), float("0"), float("0"), float("0"), float("0.106943"), float("0"), float("0.199012"), float("0"), float("0.0540602"), float("0.0616581"), float("0.305817"), float("0"), float("0"), float("0.204442"), float("0.191712"), float("0"), float("0.109193"), float("0"), float("0"), float("0.0509193"), float("0"), float("0"), float("0"), float("0"), float("0.218017"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.253662"), float("0.134936"), float("0"), float("0.269651"), float("0"), float("0"), float("0"), float("0.295306"), float("0"), float("0"), float("0"), float("0"), float("0.0601464"), float("0"), float("0.00200357"), float("0.0851614"), float("0.0318568"), float("0"), float("0.0348719"), float("0"), float("0"), float("0.0127985"), float("0"), float("0"), float("0"), float("0"), float("0"), float("3.93368e-05"), float("0.0573087"), float("0"), float("0"), float("0"), float("0.0663058"), float("0"), float("0"), float("0.118441"), float("0"), float("0"), float("0"), float("0.072888"), float("0")] - - -class PirProgram_example_input_tensor_meta_7134464969354624939: - program_id = 1242071021843173094 - input_name = "middle_1" - shape = [16, 2, 32] - data = [float("0.0278226"), float("0.933391"), float("0"), float("0.270072"), float("0"), float("0"), float("2.16145"), float("0"), float("0"), float("0"), float("1.1738"), float("1.20713"), float("0"), float("0"), float("0.431996"), float("0"), float("1.04084"), float("0"), float("0"), float("1.95876"), float("0"), float("0"), float("0"), float("0.239083"), float("0"), float("0"), float("1.80388"), float("0.782969"), float("0"), float("0"), float("1.60541"), float("0"), float("0"), float("0.00951254"), float("0"), float("1.03142"), float("0"), float("0"), float("0.722068"), float("0.592554"), float("0"), float("0.442796"), float("0"), float("0"), float("0.236385"), float("0"), float("0"), float("0"), float("0"), float("0.703864"), float("0"), float("1.04579"), float("0"), float("0"), float("0"), float("1.08614"), float("0.32995"), float("0"), float("1.14204"), float("0"), float("0.172047"), float("0"), float("1.05251"), float("0"), float("0.0258886"), float("0.350485"), float("0"), float("0"), float("0"), float("0.607473"), float("0"), float("0"), float("1.18592"), float("0.748534"), float("1.12389"), float("1.13696"), float("0.349759"), float("0"), float("0.292032"), float("0"), float("0"), float("0"), float("0"), float("0.118198"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.452116"), float("0"), float("0.73808"), float("0"), float("0.388192"), float("0"), float("0.583962"), float("0"), float("0"), float("0"), float("0.31838"), float("0"), float("0"), float("0.268641"), float("0.136249"), float("0"), float("0.136126"), float("0"), float("0"), float("0.0398369"), float("0"), float("0"), float("0"), float("0"), float("0.173328"), float("0"), float("0.305018"), float("0"), float("0"), float("0"), float("0.286744"), float("0.0874273"), float("0"), float("0.349538"), float("0"), float("0.0438143"), float("0"), float("0.301858"), float("0"), float("0"), float("0.574676"), float("0"), float("0"), float("0"), float("0.993267"), float("0"), float("0"), float("1.96871"), float("1.21832"), float("1.763"), float("1.85539"), float("0.430918"), float("0"), float("0.511989"), float("0"), float("0"), float("0"), float("0"), float("0.206435"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.814453"), float("0"), float("1.28629"), float("0"), float("0.654929"), float("0"), float("0.880677"), float("0"), float("0.255731"), float("0"), float("0"), float("0"), float("0.568876"), float("0"), float("0"), float("0.939828"), float("0.488117"), float("0.802376"), float("0.78002"), float("0.0750187"), float("0.0144826"), float("0.140808"), float("0"), float("0"), float("0"), float("0"), float("0.12847"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.52077"), float("0"), float("0.405171"), float("0"), float("0.401516"), float("0"), float("0.451902"), float("0"), float("0.660819"), float("0"), float("0"), float("0"), float("1.73161"), float("0"), float("0"), float("2.82143"), float("0.915915"), float("1.73752"), float("1.66533"), float("0.142786"), float("0.414861"), float("0"), float("0.417778"), float("0"), float("0"), float("0.269308"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("1.18397"), float("0"), float("1.17676"), float("0"), float("1.26132"), float("0"), float("1.01996"), float("0"), float("0"), float("0"), float("0.0562004"), float("0"), float("0.0222049"), float("0.0926452"), float("0.0168859"), float("0"), float("0.0163579"), float("0.0212247"), float("0"), float("0.00476881"), float("0"), float("0"), float("0"), float("0.0252377"), float("0.0227944"), float("0.00125125"), float("0.134313"), float("0"), float("0"), float("0"), float("0.0358457"), float("0"), float("0"), float("0.119865"), float("0"), float("0.0451164"), float("0"), float("0.0851875"), float("0.0083112"), float("0"), float("0.0156073"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.0826194"), float("0"), float("0.0455271"), float("0.541666"), float("0.205103"), float("0.263042"), float("0.210656"), float("0.379648"), float("0.108155"), float("0"), float("0"), float("0.124984"), float("0"), float("0"), float("0.659077"), float("0"), float("0.350388"), float("0"), float("0.097853"), float("0.221756"), float("0"), float("0"), float("0"), float("0.762207"), float("0"), float("0"), float("0.52652"), float("0.0706032"), float("0"), float("0.551366"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.158587"), float("0"), float("0.748428"), float("0"), float("0"), float("0"), float("0.801074"), float("0.351115"), float("0"), float("1.12126"), float("0"), float("0"), float("0"), float("0.640187"), float("0"), float("0"), float("0.14516"), float("0"), float("0"), float("0"), float("2.32337"), float("0"), float("0"), float("4.11075"), float("1.43035"), float("1.9387"), float("2.78858"), float("0"), float("0.661978"), float("0.502597"), float("0.998164"), float("0"), float("0"), float("0.55665"), float("0.963454"), float("0"), float("0"), float("0"), float("0"), float("0"), float("2.3835"), float("0"), float("1.40965"), float("0"), float("1.7925"), float("0"), float("0.66184"), float("0.0117774"), float("0.428219"), float("0"), float("0"), float("0"), float("2.49646"), float("0"), float("0"), float("3.66736"), float("2.92867"), float("3.13367"), float("4.29018"), float("1.00624"), float("0"), float("2.21902"), float("0"), float("0"), float("0"), float("0"), float("0.516142"), float("0"), float("1.84779"), float("0"), float("0"), float("0"), float("1.58804"), float("0"), float("0.34108"), float("0"), float("1.56324"), float("0"), float("1.64791"), float("0.0582468"), float("0.354236"), float("0.22013"), float("0.000554919"), float("0"), float("0.109064"), float("0"), float("0"), float("0.166565"), float("0.169093"), float("0.495146"), float("0.181475"), float("0.177609"), float("0"), float("0.110969"), float("0"), float("0"), float("0"), float("0"), float("0.129731"), float("0"), float("0.309364"), float("0"), float("0"), float("0.00908154"), float("0.199158"), float("0"), float("0"), float("0"), float("0.215054"), float("0"), float("0.398292"), float("0"), float("0.270094"), float("0"), float("0"), float("0.114564"), float("0.778475"), float("0"), float("0"), float("1.24456"), float("0.438091"), float("0.646566"), float("0.647836"), float("0.166394"), float("0.180818"), float("0"), float("0.357154"), float("0"), float("0"), float("0.238409"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.0367787"), float("0"), float("0.735905"), float("0"), float("0.440322"), float("0"), float("0.529939"), float("0"), float("0"), float("0"), float("0.0900413"), float("0"), float("0.00277568"), float("0.178194"), float("0.0160336"), float("0"), float("0.0589797"), float("0.0346942"), float("0"), float("0.0824681"), float("0"), float("0"), float("0.0301201"), float("0.0371609"), float("0.0810581"), float("0.00502692"), float("0.359627"), float("0"), float("0"), float("0"), float("0.0332112"), float("0"), float("0"), float("0.263454"), float("0"), float("0.101193"), float("0"), float("0.175708"), float("0"), float("0"), float("0.697523"), float("0"), float("0"), float("0"), float("1.60436"), float("0"), float("0"), float("2.59563"), float("1.37799"), float("2.12651"), float("2.08561"), float("0.218594"), float("0.0190547"), float("0.446536"), float("0"), float("0"), float("0"), float("0"), float("0.309906"), float("0"), float("0"), float("0"), float("0"), float("0"), float("1.4305"), float("0"), float("1.0798"), float("0"), float("1.19011"), float("0"), float("1.20522"), float("0"), float("0"), float("0"), float("1.20503"), float("0"), float("0"), float("0.431932"), float("0.311821"), float("0"), float("0.813583"), float("0.0860126"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.462605"), float("0"), float("1.00296"), float("0"), float("0"), float("0"), float("1.08335"), float("0.564641"), float("0"), float("1.22241"), float("0"), float("0"), float("0"), float("0.889895"), float("0"), float("0"), float("0.0291519"), float("0"), float("1.10568"), float("0"), float("0"), float("1.55024"), float("0.748529"), float("0"), float("0.445935"), float("0"), float("0"), float("0.703552"), float("0"), float("0"), float("0"), float("0.0828259"), float("0.649184"), float("0"), float("1.57604"), float("0"), float("0"), float("0"), float("1.44342"), float("0.0956974"), float("0"), float("1.96239"), float("0"), float("0.400894"), float("0"), float("1.55778"), float("0"), float("0"), float("0.217823"), float("0"), float("0"), float("0"), float("0.458147"), float("0"), float("0"), float("0.670096"), float("0.324289"), float("0.541255"), float("0.468259"), float("0"), float("0.0262447"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.362822"), float("0"), float("0.364886"), float("0"), float("0.255361"), float("0"), float("0.313047"), float("0"), float("0.0198007"), float("0"), float("0"), float("0"), float("0.0934036"), float("0"), float("0"), float("0.162486"), float("0.100014"), float("0.106053"), float("0.145519"), float("0.0696824"), float("0"), float("0.0350842"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.0367805"), float("0"), float("0"), float("0"), float("0.0235441"), float("0"), float("0.0626898"), float("0"), float("0.0580879"), float("0"), float("0.0451148"), float("0"), float("0.588072"), float("0"), float("0.632345"), float("0"), float("0"), float("1.75334"), float("0.491659"), float("0"), float("0"), float("0"), float("0.119273"), float("0.112027"), float("0"), float("0.140105"), float("0"), float("0.966691"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("1.66538"), float("0"), float("0"), float("1.73029"), float("0.062725"), float("0"), float("0"), float("1.35752"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.013027"), float("0"), float("0"), float("0.0148772"), float("0"), float("0"), float("0.220215"), float("0"), float("0"), float("0.589901"), float("0.157416"), float("0.163732"), float("0.212377"), float("0"), float("0.592463"), float("0"), float("0"), float("0.76425"), float("0.012217"), float("0"), float("1.03746"), float("0"), float("0.461267"), float("0"), float("0.246892"), float("0"), float("0"), float("0"), float("0"), float("0.530749"), float("0"), float("0"), float("0.815583"), float("0.540469"), float("0"), float("0.250381"), float("0"), float("0"), float("0.234132"), float("0"), float("0"), float("0"), float("0.000377319"), float("0.00966338"), float("0.0144465"), float("0.703602"), float("0"), float("0"), float("0"), float("0.849869"), float("0"), float("0"), float("1.14647"), float("0"), float("0.339792"), float("0"), float("0.888705"), float("0"), float("0"), float("0"), float("0"), float("0.142622"), float("0"), float("0.0537007"), float("0.337754"), float("0.101622"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.11494"), float("0.235508"), float("0.174129"), float("0.0639142"), float("0.579994"), float("0"), float("0"), float("0"), float("0.145871"), float("0"), float("0"), float("0.478335"), float("0"), float("0.190835"), float("0"), float("0.329909"), float("0"), float("0"), float("0.0279218"), float("0"), float("0.177735"), float("0"), float("0.414542"), float("0.727789"), float("1.11413"), float("0"), float("0.156904"), float("0"), float("0.382821"), float("0.5096"), float("0"), float("0"), float("0.47453"), float("0.414645"), float("0"), float("0.705401"), float("0.392972"), float("0"), float("0"), float("0"), float("0.194011"), float("0"), float("0"), float("0.720369"), float("0"), float("0.242475"), float("0"), float("0.603829"), float("0"), float("0"), float("0.515296"), float("0"), float("0"), float("0"), float("1.74996"), float("0"), float("0"), float("2.5889"), float("0.893962"), float("1.49597"), float("1.12741"), float("0.0785876"), float("0.531738"), float("0"), float("0.2548"), float("0"), float("0"), float("0.350623"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("1.16254"), float("0"), float("1.26599"), float("0"), float("1.17455"), float("0"), float("0.78066"), float("0"), float("0"), float("0"), float("0.812465"), float("0"), float("0"), float("1.08987"), float("0.574031"), float("0"), float("0.259725"), float("0"), float("0"), float("0.3764"), float("0"), float("0"), float("0"), float("0.111538"), float("0.243441"), float("0"), float("0.798357"), float("0"), float("0"), float("0"), float("1.06892"), float("0"), float("0"), float("1.35836"), float("0"), float("0.206873"), float("0"), float("1.11643"), float("0"), float("0"), float("0.158604"), float("0"), float("0"), float("0"), float("0.333509"), float("0"), float("0"), float("0.549038"), float("0.267435"), float("0.46137"), float("0.464749"), float("0.00498525"), float("0.00284835"), float("0.0787959"), float("0.00215647"), float("0"), float("0"), float("0"), float("0.0392426"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.310251"), float("0"), float("0.242476"), float("0"), float("0.223079"), float("0"), float("0.268803"), float("0"), float("0.0328052"), float("0"), float("0.273985"), float("0"), float("0"), float("0.158735"), float("0.171478"), float("0"), float("0.102056"), float("0"), float("0"), float("0.017277"), float("0"), float("0"), float("0"), float("0"), float("0.163026"), float("0"), float("0.306636"), float("0"), float("0"), float("0"), float("0.235121"), float("0.0890468"), float("0"), float("0.267089"), float("0"), float("0.000100172"), float("0"), float("0.256085"), float("0"), float("0"), float("0.154934"), float("0"), float("0"), float("0"), float("0.222688"), float("0"), float("0"), float("0.459038"), float("0.216095"), float("0.341084"), float("0.472604"), float("0.0938015"), float("0"), float("0.000344158"), float("0"), float("0"), float("0"), float("0"), float("0.145101"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.0770143"), float("0"), float("0.334262"), float("0"), float("0.0855547"), float("0"), float("0.159209"), float("0"), float("0.0432482"), float("0.0493264"), float("0.244654"), float("0"), float("0"), float("0.163553"), float("0.153369"), float("0"), float("0.0873547"), float("0"), float("0"), float("0.0407355"), float("0"), float("0"), float("0"), float("0"), float("0.174414"), float("0"), float("0.309304"), float("0"), float("0"), float("0"), float("0.20293"), float("0.107949"), float("0"), float("0.215721"), float("0"), float("0.0288104"), float("0"), float("0.236245"), float("0"), float("0"), float("0"), float("0"), float("0.0481171"), float("0"), float("0.00160286"), float("0.0681292"), float("0.0254854"), float("0"), float("0.0278975"), float("0"), float("0"), float("0.0102388"), float("0"), float("0"), float("0"), float("0"), float("0"), float("3.14694e-05"), float("0.0458469"), float("0"), float("0"), float("0"), float("0.0530446"), float("0"), float("0"), float("0.0947526"), float("0"), float("0.0166749"), float("0"), float("0.0583104"), float("0")] - - -class PirProgram_example_input_tensor_meta_1913256108345190963: - program_id = 1242071021843173094 - input_name = "linear_2.w_0" - shape = [16, 32] - data = [float("0.230423"), float("-0.20309"), float("0.162069"), float("-0.126608"), float("0.180297"), float("-0.20278"), float("0.302492"), float("-0.251559"), float("-0.108057"), float("0.351752"), float("0.0375671"), float("-0.177384"), float("0.304171"), float("-0.156094"), float("0.242355"), float("0.0150476"), float("-0.255715"), float("0.138158"), float("0.0648821"), float("0.26197"), float("-0.0646441"), float("0.197283"), float("-0.0875255"), float("-0.235316"), float("-0.117067"), float("-0.0573046"), float("-0.273241"), float("0.345782"), float("0.0942704"), float("0.0937579"), float("0.057566"), float("0.19286"), float("-0.326869"), float("-0.325651"), float("-0.25319"), float("0.153059"), float("0.00222721"), float("0.326323"), float("0.152709"), float("0.181255"), float("0.32209"), float("0.337139"), float("-0.224294"), float("-0.245217"), float("0.195527"), float("0.20644"), float("-0.279234"), float("0.240308"), float("0.22163"), float("0.179995"), float("0.290756"), float("0.191943"), float("-0.126184"), float("-0.315085"), float("0.00289739"), float("0.0977283"), float("-0.283358"), float("-0.208794"), float("0.136989"), float("0.21537"), float("0.319846"), float("0.221434"), float("0.261604"), float("-0.0966258"), float("-0.191189"), float("-0.101988"), float("-0.0353045"), float("-0.28391"), float("-0.130498"), float("0.177049"), float("0.143925"), float("-0.120895"), float("0.292791"), float("-0.0774804"), float("-0.229526"), float("0.337984"), float("-0.0903337"), float("-0.171707"), float("0.295915"), float("0.344176"), float("0.342363"), float("0.167197"), float("-0.0261476"), float("0.352028"), float("-0.310108"), float("0.28573"), float("-0.231779"), float("-0.349535"), float("0.226729"), float("-0.0431784"), float("-0.332617"), float("-0.170526"), float("-0.114074"), float("0.131645"), float("-0.295109"), float("-0.294845"), float("0.295009"), float("0.0341118"), float("0.212163"), float("0.0298418"), float("0.181503"), float("-0.103151"), float("-0.146327"), float("-0.0275784"), float("0.22929"), float("-0.199881"), float("-0.123251"), float("-0.0388795"), float("0.306805"), float("0.335557"), float("-0.0176815"), float("-0.272611"), float("0.236301"), float("0.0892752"), float("-0.222741"), float("-0.325869"), float("-0.127395"), float("-0.25068"), float("0.0660806"), float("0.0702004"), float("-0.245965"), float("0.096747"), float("-0.124066"), float("0.227929"), float("0.262033"), float("0.353009"), float("0.0637572"), float("-0.30891"), float("0.175011"), float("-0.0126421"), float("-0.350917"), float("-0.219104"), float("0.255108"), float("0.170938"), float("0.265865"), float("0.321288"), float("0.317676"), float("-0.206672"), float("0.0948979"), float("0.339302"), float("-0.104974"), float("0.229254"), float("-0.173685"), float("0.22608"), float("0.182638"), float("-0.206255"), float("0.338822"), float("0.152891"), float("-0.285523"), float("-0.30232"), float("0.197891"), float("-0.214551"), float("-0.305591"), float("0.234063"), float("-0.232026"), float("0.182208"), float("0.0429179"), float("0.0140892"), float("0.192062"), float("-0.344634"), float("0.118818"), float("0.00202157"), float("-0.283396"), float("-0.292219"), float("0.182276"), float("-0.0348539"), float("-0.319024"), float("-0.185221"), float("0.331582"), float("0.0734183"), float("0.0942539"), float("0.333694"), float("0.0249039"), float("0.0709629"), float("0.226096"), float("0.0418429"), float("-0.0366375"), float("-0.0565682"), float("0.144466"), float("-0.345168"), float("-0.074358"), float("-0.307203"), float("0.0179474"), float("-0.138052"), float("-0.112629"), float("0.308599"), float("-0.161632"), float("0.322421"), float("-0.327087"), float("-0.0683733"), float("-0.0945122"), float("0.348189"), float("-0.00324977"), float("-0.340845"), float("-0.0842492"), float("-0.165104"), float("-0.299402"), float("0.161167"), float("0.244718"), float("-0.179583"), float("0.0124633"), float("0.304233"), float("-0.268381"), float("-0.16432"), float("0.157403"), float("-0.0637301"), float("0.270001"), float("-0.044371"), float("-0.120994"), float("-0.345206"), float("0.0607044"), float("-0.267447"), float("-0.308716"), float("0.0326651"), float("0.217073"), float("-0.128421"), float("0.0860339"), float("-0.109498"), float("-0.177799"), float("-0.128405"), float("0.161585"), float("0.116405"), float("-0.122393"), float("-0.272491"), float("-0.225742"), float("0.304446"), float("0.224986"), float("0.256058"), float("0.277367"), float("-0.100766"), float("0.188724"), float("0.212626"), float("-0.21794"), float("-0.126419"), float("-0.224952"), float("-0.0924843"), float("0.206705"), float("0.021292"), float("0.0137566"), float("-0.0360458"), float("-0.234257"), float("0.117413"), float("0.153925"), float("0.191284"), float("-0.253123"), float("0.21299"), float("0.327653"), float("0.0754134"), float("0.193985"), float("-0.0368757"), float("-0.326489"), float("-0.267416"), float("0.0922668"), float("0.105197"), float("0.280179"), float("0.110465"), float("-0.0979321"), float("-0.275342"), float("-0.279135"), float("0.0396403"), float("0.208052"), float("-0.0856857"), float("0.03993"), float("0.299113"), float("0.0134926"), float("0.0971514"), float("-0.0327729"), float("0.324181"), float("-0.162795"), float("0.104321"), float("-0.00222742"), float("-0.314876"), float("-0.186012"), float("-0.141716"), float("-0.305346"), float("0.0716737"), float("-0.315062"), float("-0.305029"), float("0.105486"), float("0.311472"), float("-0.13068"), float("-0.241825"), float("-0.261683"), float("0.344032"), float("-0.211968"), float("-0.30109"), float("0.301081"), float("-0.300443"), float("0.0113317"), float("-0.13623"), float("0.00588745"), float("0.323762"), float("-0.27139"), float("0.128905"), float("0.226326"), float("-0.0567057"), float("0.0513325"), float("-0.00563676"), float("0.292468"), float("-0.00536835"), float("-0.240204"), float("0.0458691"), float("0.258935"), float("-0.259172"), float("0.155908"), float("0.154713"), float("-0.29383"), float("0.345019"), float("-0.302847"), float("-0.102248"), float("-0.224491"), float("0.00236819"), float("0.181067"), float("0.289794"), float("0.118373"), float("-0.00321519"), float("0.166811"), float("-0.0312314"), float("-0.0175973"), float("0.23624"), float("0.196907"), float("-0.292757"), float("0.0645859"), float("-0.1919"), float("0.219958"), float("-0.347093"), float("-0.0693511"), float("-0.0524661"), float("-0.237186"), float("0.0300816"), float("-0.259862"), float("-0.319011"), float("-0.0807318"), float("0.135482"), float("0.307606"), float("0.350905"), float("0.0697996"), float("0.283031"), float("0.237269"), float("0.0447312"), float("-0.131454"), float("-0.28127"), float("0.160253"), float("0.148049"), float("0.243601"), float("0.323207"), float("-0.143284"), float("-0.341898"), float("0.293442"), float("0.063323"), float("0.120295"), float("-0.18036"), float("-0.0975274"), float("-0.0455448"), float("-0.183233"), float("-0.0857107"), float("0.278475"), float("0.174762"), float("0.340955"), float("-0.308548"), float("-0.350232"), float("-0.251843"), float("-0.218846"), float("-0.27189"), float("-0.261468"), float("0.0743435"), float("-0.044729"), float("-0.248535"), float("0.257598"), float("-0.103987"), float("-0.145567"), float("0.023903"), float("-0.311359"), float("-0.0946209"), float("-0.0670924"), float("0.256815"), float("0.0447149"), float("-0.0385811"), float("0.215715"), float("0.0890531"), float("-0.120938"), float("-0.135609"), float("0.295268"), float("0.222618"), float("-0.117695"), float("0.215494"), float("0.14366"), float("0.277732"), float("-0.296978"), float("-0.189063"), float("0.266609"), float("0.243941"), float("-0.106652"), float("-0.0466308"), float("-0.295165"), float("0.223554"), float("0.269018"), float("-0.199119"), float("0.138165"), float("-0.308141"), float("0.35326"), float("0.321212"), float("0.019633"), float("-0.321271"), float("0.0288094"), float("-0.229809"), float("-0.0147547"), float("0.204604"), float("0.184154"), float("-0.259976"), float("0.211335"), float("0.0558831"), float("-0.110034"), float("-0.327218"), float("0.213624"), float("-0.329281"), float("-0.176617"), float("0.163939"), float("-0.111053"), float("0.15307"), float("0.195678"), float("-0.122387"), float("-0.13225"), float("0.175086"), float("0.207095"), float("-0.319538"), float("-0.189705"), float("-0.090235"), float("-0.340103"), float("0.0825443"), float("-0.303053"), float("0.135722"), float("0.323061"), float("0.284381"), float("0.0299432"), float("-0.28277"), float("0.179539"), float("-0.283711"), float("0.303636"), float("0.306957"), float("0.345974"), float("0.0915332"), float("0.0688096"), float("0.333376"), float("-0.0633284"), float("0.0370856"), float("9.25966e-05"), float("-0.239747"), float("7.92782e-05"), float("0.296225"), float("-0.107539"), float("-0.087091"), float("-0.183428"), float("-0.248449"), float("-0.105759"), float("-0.352878"), float("-0.137407"), float("-0.203483"), float("0.0538141"), float("-0.0800397"), float("0.14867"), float("-0.253796"), float("0.169363"), float("-0.103523"), float("-0.105282"), float("-0.102652"), float("-0.298457"), float("-0.199085"), float("-0.013719"), float("-0.323702"), float("-0.154908"), float("-0.0621661"), float("0.0333456"), float("-0.317511"), float("0.0649108"), float("0.152055"), float("-0.113154"), float("-0.337959"), float("-0.261652"), float("0.12889"), float("-0.0705134"), float("-0.162341"), float("-0.21903"), float("-0.195068"), float("0.163776"), float("-0.0572965"), float("0.179957"), float("0.298765"), float("0.0771351"), float("-0.239391"), float("-0.246127"), float("0.232749"), float("0.132148"), float("0.0270761"), float("0.256751"), float("0.0966847"), float("0.338032"), float("0.034458"), float("-0.0380405"), float("0.0232373"), float("0.332713"), float("0.287575"), float("0.328612"), float("0.216737"), float("-0.268043"), float("-0.24701"), float("-0.352788"), float("-0.170355"), float("0.220089"), float("-0.181284"), float("0.178662"), float("0.316528")] - - -class PirProgram_example_input_tensor_meta_560967842253293313: - program_id = 1242071021843173094 - input_name = "middle_0" - shape = [16, 2, 32] - data = None - - -class PirProgram_example_input_tensor_meta_7562718223343614406: - program_id = 1242071021843173094 - input_name = "linear_2.b_0" - shape = [32] - data = [float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0")] - - -class PirProgram_example_input_tensor_meta_602773716366570321: - program_id = 1242071021843173094 - input_name = "args_0" - shape = [16, 2, 16] - data = [float("-1.24783"), float("1.35968"), float("-0.0643954"), float("-0.627694"), float("3.57834"), float("-1.40647"), float("-3.06211"), float("-0.852737"), float("1.89044"), float("1.71429"), float("0.873626"), float("2.45266"), float("2.46995"), float("-4.18985"), float("5.24214"), float("-0.769194"), float("-0.432318"), float("1.44443"), float("-0.398696"), float("-1.49099"), float("-0.245603"), float("-1.75011"), float("-1.19734"), float("0.138174"), float("1.28394"), float("-0.147029"), float("1.46563"), float("0.138966"), float("1.49462"), float("-1.24567"), float("1.74074"), float("-0.541634"), float("-0.137766"), float("0.149732"), float("0.813494"), float("0.303299"), float("-0.238571"), float("1.88537"), float("-0.0370909"), float("0.573859"), float("0.349418"), float("0.487316"), float("-1.32582"), float("-0.843882"), float("-0.676632"), float("-0.909084"), float("0.270721"), float("-1.23429"), float("-0.0424248"), float("0.429659"), float("-0.14267"), float("-0.428278"), float("-0.0162147"), float("-0.494324"), float("-0.285505"), float("0.0283108"), float("0.343003"), float("0.0251735"), float("0.319188"), float("0.0917862"), float("0.402113"), float("-0.271283"), float("0.494286"), float("-0.191708"), float("-0.284004"), float("0.298903"), float("1.58414"), float("0.506222"), float("-0.530693"), float("3.27026"), float("-0.117699"), float("1.01495"), float("0.711183"), float("0.811757"), float("-2.06371"), float("-1.2438"), float("-1.13482"), float("-1.38332"), float("0.486574"), float("-2.27285"), float("-0.358158"), float("0.179657"), float("0.702598"), float("0.133837"), float("-0.116537"), float("1.53755"), float("0.00049031"), float("0.551636"), float("0.176739"), float("0.308397"), float("-0.844465"), float("-0.537499"), float("-0.755579"), float("-0.623192"), float("0.182918"), float("-1.02258"), float("-1.34653"), float("0.913991"), float("1.1851"), float("0.400625"), float("0.121387"), float("3.74875"), float("0.507555"), float("1.24397"), float("-0.00871566"), float("0.270786"), float("-2.39195"), float("-1.38387"), float("-2.33894"), float("-0.779432"), float("0.529702"), float("-2.01132"), float("-0.0355457"), float("0.156263"), float("-0.0398702"), float("-0.0941751"), float("0.070695"), float("-0.125185"), float("-0.123143"), float("-0.0328905"), float("0.0242618"), float("0.050551"), float("0.0789997"), float("0.0688375"), float("0.0704337"), float("-0.152359"), float("0.139597"), float("-0.0303556"), float("-0.422165"), float("0.755079"), float("-0.125477"), float("-0.390056"), float("0.129337"), float("-0.44178"), float("-0.83648"), float("-0.363755"), float("-0.396812"), float("-0.217299"), float("1.19051"), float("0.353966"), float("0.100611"), float("-0.841632"), float("1.13851"), float("-0.0570613"), float("-0.321325"), float("1.00824"), float("-0.478592"), float("-1.40818"), float("-0.435232"), float("-1.11203"), float("-0.501488"), float("-0.127583"), float("0.967821"), float("0.100076"), float("0.722625"), float("0.514725"), float("0.67351"), float("-0.719309"), float("0.999483"), float("-0.640918"), float("-1.46169"), float("1.59347"), float("3.25034"), float("0.413147"), float("0.432863"), float("5.4329"), float("0.0917308"), float("1.98376"), float("0.598049"), float("1.64893"), float("-1.92021"), float("-1.70858"), float("-2.384"), float("-1.1942"), float("0.144455"), float("-3.21842"), float("-1.22085"), float("-0.843408"), float("3.91006"), float("0.0781953"), float("-1.01348"), float("6.29112"), float("1.98013"), float("2.26446"), float("0.98046"), float("0.878038"), float("-4.09496"), float("-3.09536"), float("-2.928"), float("-3.87442"), float("-0.000862794"), float("-3.55961"), float("-0.1064"), float("-0.161292"), float("0.187614"), float("0.0255209"), float("-0.162917"), float("0.42479"), float("-0.0453343"), float("0.436514"), float("-0.191061"), float("0.137538"), float("-0.410105"), float("-0.431748"), float("-0.438004"), float("-0.404633"), float("0.218743"), float("-0.427686"), float("-0.424405"), float("0.837761"), float("0.332992"), float("-0.0513256"), float("0.047148"), float("1.35676"), float("-0.0688395"), float("0.506198"), float("0.163514"), float("-0.337938"), float("-1.2557"), float("-0.616636"), float("-1.1297"), float("-0.193176"), float("0.337317"), float("-0.561698"), float("-0.135902"), float("0.393851"), float("0.00424394"), float("-0.196324"), float("0.159519"), float("-0.308631"), float("-0.376123"), float("-0.071268"), float("0.0883756"), float("0.0545766"), float("0.303032"), float("0.111892"), float("0.218541"), float("-0.537371"), float("0.421097"), float("-0.129858"), float("-1.03447"), float("0.513909"), float("2.0756"), float("0.441222"), float("-0.472549"), float("4.29349"), float("0.13518"), float("1.6049"), float("0.482167"), float("0.853731"), float("-2.33474"), float("-1.43174"), float("-2.12045"), float("-1.69272"), float("0.520399"), float("-2.92517"), float("-0.154838"), float("1.12713"), float("-0.729672"), float("-1.66338"), float("-0.657706"), float("-1.18132"), float("-0.874759"), float("0.06689"), float("1.51983"), float("0.351833"), float("0.986195"), float("0.245073"), float("0.796269"), float("-0.622144"), float("0.948545"), float("-1.01152"), float("-1.22342"), float("2.62304"), float("-0.221326"), float("-2.36949"), float("0.0231131"), float("-3.1239"), float("-1.56313"), float("-0.0846508"), float("1.4516"), float("-0.843756"), float("2.11236"), float("0.770368"), float("2.54804"), float("-2.68047"), float("3.21551"), float("-0.423819"), float("-0.392694"), float("0.187697"), float("0.413998"), float("0.057787"), float("-0.13115"), float("1.17413"), float("0.0864602"), float("0.307451"), float("0.0920442"), float("0.143506"), float("-0.635472"), float("-0.255527"), float("-0.542642"), float("-0.322647"), float("0.183446"), float("-0.825342"), float("-0.0203514"), float("0.0198951"), float("0.0865615"), float("0.0272587"), float("-0.00242023"), float("0.195467"), float("0.07148"), float("0.0644732"), float("0.0219086"), float("0.0301208"), float("-0.185134"), float("-0.132513"), float("-0.0794464"), float("-0.100434"), float("0.0120146"), float("-0.08797"), float("-1.29686"), float("1.00538"), float("-0.771612"), float("-1.60371"), float("1.04745"), float("-1.71847"), float("-0.364911"), float("-0.669676"), float("1.52801"), float("-0.0574329"), float("0.59866"), float("1.73576"), float("2.91115"), float("-1.2902"), float("3.56666"), float("0.497791"), float("-0.629735"), float("1.17433"), float("-0.648112"), float("-1.04914"), float("-0.195696"), float("-1.03082"), float("-0.321714"), float("-0.668095"), float("-0.117482"), float("-0.81666"), float("1.54317"), float("0.372422"), float("0.500825"), float("-0.542041"), float("1.91077"), float("0.324363"), float("-0.746501"), float("1.56012"), float("-0.693475"), float("-1.47149"), float("0.417272"), float("-1.5822"), float("-0.614944"), float("-0.337996"), float("0.782261"), float("-0.244622"), float("1.03057"), float("0.317955"), float("1.1072"), float("-1.29199"), float("1.91734"), float("0.0562871"), float("-0.344145"), float("0.582381"), float("0.0177077"), float("-0.394364"), float("0.278082"), float("-0.549242"), float("-0.608972"), float("-0.148099"), float("0.0354738"), float("0.105372"), float("0.54669"), float("0.293604"), float("0.384397"), float("-0.770095"), float("0.530887"), float("-0.0841932"), float("-1.5664"), float("1.47664"), float("-0.0485786"), float("-1.2655"), float("1.12964"), float("-1.11663"), float("-0.127906"), float("-0.158063"), float("0.266339"), float("-0.689339"), float("0.552855"), float("-0.29022"), float("1.41004"), float("-1.61181"), float("1.21825"), float("0.138486"), float("-1.40203"), float("1.25058"), float("0.718333"), float("0.433586"), float("0.213605"), float("3.4107"), float("0.702294"), float("0.844367"), float("-0.269252"), float("0.469749"), float("-2.07015"), float("-0.969774"), float("-1.84929"), float("-0.536334"), float("0.553032"), float("-2.07369"), float("-0.875411"), float("1.6934"), float("-0.485379"), float("-1.53678"), float("0.237551"), float("-1.80035"), float("-0.801288"), float("-0.196758"), float("1.13256"), float("-0.170934"), float("1.21607"), float("0.530888"), float("1.79473"), float("-1.66515"), float("2.21643"), float("-0.170763"), float("-0.242361"), float("0.0761521"), float("0.432139"), float("0.0631636"), float("-0.093372"), float("0.938469"), float("0.0142968"), float("0.313853"), float("0.120945"), float("0.144088"), float("-0.496069"), float("-0.284314"), float("-0.471705"), float("-0.335182"), float("0.114975"), float("-0.613756"), float("-0.133989"), float("0.332343"), float("-0.0773375"), float("-0.386288"), float("-0.0153539"), float("-0.397145"), float("-0.344031"), float("0.0805068"), float("0.348394"), float("0.00255038"), float("0.33544"), float("0.0063985"), float("0.34345"), float("-0.323657"), float("0.404106"), float("-0.197486"), float("-0.0204629"), float("0.165297"), float("0.344726"), float("0.0223876"), float("-0.00523791"), float("0.623413"), float("-0.16317"), float("0.276788"), float("0.216639"), float("0.113306"), float("-0.499786"), float("-0.285747"), float("-0.187579"), float("-0.235846"), float("0.0960729"), float("-0.395305"), float("-0.0911418"), float("0.288873"), float("-0.0312524"), float("-0.338057"), float("-0.0387871"), float("-0.421691"), float("-0.315626"), float("0.125869"), float("0.273582"), float("-0.058278"), float("0.3202"), float("0.015641"), float("0.297726"), float("-0.288548"), float("0.329069"), float("-0.179419"), float("-0.0326695"), float("0.117692"), float("-0.060717"), float("-0.112393"), float("0.0348706"), float("-0.105831"), float("-0.0257303"), float("-0.0271061"), float("0.0484697"), float("-0.00249167"), float("0.0425337"), float("0.0319538"), float("0.0739469"), float("-0.0675583"), float("0.136938"), float("-0.0154109")] - - -class PirProgram_example_input_tensor_meta_6573343046139926864: - program_id = 692320333968606381 - input_name = "output_grad_0" - shape = [16, 2, 16] - data = [float("0.00107317"), float("-0.00117165"), float("-0.00163671"), float("0.00156766"), float("0.00201485"), float("7.37798e-05"), float("-0.00288858"), float("0.00324427"), float("0.00363222"), float("0.00455793"), float("0.000339677"), float("0.00159846"), float("0.00616652"), float("-0.000968744"), float("-0.00142565"), float("0.00341143"), float("-0.000397156"), float("0.00313079"), float("-0.0038428"), float("-0.000317453"), float("-0.00244509"), float("-0.00321148"), float("-0.00215432"), float("0.0027039"), float("0.000365354"), float("0.00177382"), float("-7.06572e-05"), float("0.00144024"), float("0.00486523"), float("0.00112444"), float("-0.00136742"), float("-0.00121718"), float("-0.000322603"), float("0.000538117"), float("3.002e-05"), float("-0.000153229"), float("0.000238948"), float("-4.99467e-05"), float("-0.000162901"), float("5.13133e-05"), float("0.000178962"), float("-0.00028677"), float("-0.000726859"), float("-0.000173296"), float("8.02252e-05"), float("7.30726e-05"), float("-0.00029218"), float("-0.000122243"), float("-1.10911e-05"), float("0.000961878"), float("-0.000971545"), float("-0.000160983"), float("-0.000659728"), float("-0.000886996"), float("-0.000450224"), float("0.000584157"), float("0.000198652"), float("0.000577065"), float("4.15716e-05"), float("0.000471037"), float("0.00124457"), float("0.000217747"), float("-0.000470511"), float("-0.000316141"), float("0.000342187"), float("0.0018494"), float("0.000947351"), float("0.00101066"), float("0.00129266"), float("0.00142459"), float("0.000266141"), float("-0.00110398"), float("0.000510692"), float("0.000157268"), float("-0.00170386"), float("-0.00100753"), float("-0.00057964"), float("0.000633022"), float("-0.000892884"), float("-0.000346073"), float("-4.86121e-05"), float("0.000248871"), float("-8.16761e-05"), float("-0.000720575"), float("0.000460798"), float("0.000718931"), float("-0.0010086"), float("-0.00063695"), float("-0.000415294"), float("0.000742808"), float("-0.00158288"), float("-0.000315739"), float("-0.0013079"), float("3.30604e-05"), float("-0.000385291"), float("-0.000363103"), float("-0.000218203"), float("0.00222529"), float("0.000970178"), float("0.0010858"), float("0.00236524"), float("0.00202229"), float("-6.98109e-05"), float("-0.00177226"), float("-0.000319509"), float("0.00120065"), float("-0.00222744"), float("-0.00133017"), float("-0.0020776"), float("0.000887785"), float("-0.00186867"), float("-0.0012222"), float("-8.52198e-05"), float("0.000224672"), float("-0.000124413"), float("-0.00012138"), float("-7.69958e-05"), float("-0.000256524"), float("-0.000147351"), float("-1.61188e-05"), float("-0.000101862"), float("0.000245915"), float("-9.56549e-05"), float("0.000121793"), float("0.000120966"), float("2.49652e-05"), float("-0.000102786"), float("-0.000184276"), float("-0.000588331"), float("0.00143901"), float("-0.000900593"), float("-0.000131671"), float("0.000196109"), float("-0.00068363"), float("-0.000781266"), float("0.000212953"), float("-0.000563651"), float("0.000203772"), float("0.000687689"), float("0.000925853"), float("0.000966789"), float("0.000618601"), float("-0.000566369"), float("-0.000386996"), float("-0.00146396"), float("0.00202857"), float("-0.00351244"), float("7.24205e-05"), float("-0.00133867"), float("-0.00197239"), float("-0.00188278"), float("0.000753329"), float("0.000766242"), float("0.00134591"), float("-0.00017611"), float("0.00196395"), float("0.0027534"), float("0.00141287"), float("-0.00103552"), float("-0.000716369"), float("0.000550839"), float("0.00322606"), float("0.000856896"), float("-0.00125687"), float("0.00129017"), float("0.00116238"), float("-0.00187795"), float("-0.00158381"), float("-0.00141739"), float("0.00180403"), float("-0.00490405"), float("-0.000981318"), float("-0.00421745"), float("-0.000114125"), float("-0.00195622"), float("-0.000424439"), float("-0.000280211"), float("0.00171129"), float("0.000688157"), float("-0.00253045"), float("0.00109986"), float("0.00417076"), float("-0.0016617"), float("-0.0023778"), float("-0.000190303"), float("0.00340801"), float("-0.00539941"), float("-0.00185255"), float("-0.00372863"), float("-0.00238725"), float("-0.00133698"), float("0.00134814"), float("0.000206168"), float("-0.000115355"), float("0.000135783"), float("-5.93454e-05"), float("-0.000104663"), float("-0.000363951"), float("-0.000272538"), float("0.000322769"), float("-0.000417613"), float("-0.000232908"), float("-0.000300922"), float("-8.19516e-05"), float("-0.000290464"), float("-0.000267267"), float("-0.000150961"), float("9.07732e-05"), float("0.000208024"), float("0.000412389"), float("0.000340767"), float("-0.000807369"), float("-2.84969e-05"), float("0.000996116"), float("-0.000259763"), float("-0.000458858"), float("-0.00100166"), float("0.000698113"), float("-0.00121583"), float("-0.000378595"), float("-0.00100867"), float("-0.000326933"), float("-0.000104752"), float("0.000252775"), float("-0.000104392"), float("0.000384491"), float("-0.000120818"), float("-0.000332935"), float("-0.00014957"), float("-0.000383103"), float("-0.00033556"), float("-0.000210315"), float("-0.000283282"), float("0.00040653"), float("-0.000114731"), float("0.000345088"), float("0.000101139"), float("-3.16119e-05"), float("-0.000134473"), float("-0.000333086"), float("0.000365319"), float("2.31993e-05"), float("-0.00101835"), float("-0.00198467"), float("0.000607929"), float("0.00105125"), float("-0.00256911"), float("-0.00117818"), float("-0.000836815"), float("0.00171317"), float("-0.00357581"), float("0.000157612"), float("-0.00292617"), float("-0.000337635"), float("-0.000517322"), float("-0.00113363"), float("-0.000858934"), float("0.00275311"), float("-0.00322123"), float("-0.00128946"), float("-0.00257051"), float("-0.00303501"), float("-0.00161937"), float("0.00149826"), float("0.000905219"), float("0.00253491"), float("0.000515937"), float("0.00195283"), float("0.00346017"), float("0.00160003"), float("-0.00186369"), float("-0.00181285"), float("0.000569412"), float("0.00475362"), float("-0.00489154"), float("-0.000602263"), float("-0.00312457"), float("-0.00445273"), float("-0.00261735"), float("0.00360081"), float("-0.00123231"), float("0.00259976"), float("-0.00111133"), float("0.00116551"), float("0.00690931"), float("-8.18758e-05"), float("-0.00153739"), float("-0.00306445"), float("0.00017831"), float("0.000354979"), float("-0.000123217"), float("-0.000102344"), float("0.000186917"), float("0.000809431"), float("-0.000367907"), float("-0.000401029"), float("-0.000246482"), float("0.000556523"), float("-0.000902953"), float("-0.000132123"), float("-0.000778616"), float("-0.000146372"), float("-0.00013007"), float("7.63733e-05"), float("2.0302e-05"), float("3.89685e-05"), float("3.33519e-05"), float("-6.25571e-06"), float("5.20851e-05"), float("0.000157613"), float("-7.67726e-05"), float("-6.23804e-05"), float("-4.95858e-05"), float("8.56801e-05"), float("-0.000203655"), float("-9.11566e-05"), float("-0.000137199"), float("-5.19502e-05"), float("-4.72111e-05"), float("4.96732e-05"), float("0.000994176"), float("-0.00101409"), float("0.00191827"), float("0.000882816"), float("0.000591085"), float("0.00105649"), float("0.000372834"), float("0.000875936"), float("-0.000349729"), float("0.000747718"), float("-0.000399308"), float("-0.000425689"), float("0.00309901"), float("-0.00110301"), float("0.00112329"), float("0.00215364"), float("-0.000519875"), float("0.000449016"), float("-0.000270285"), float("-0.000219273"), float("-0.000324491"), float("-0.000262691"), float("-0.000138284"), float("0.0003396"), float("-0.000301592"), float("-0.000257651"), float("0.00117054"), float("-0.000151151"), float("3.56859e-05"), float("0.000969652"), float("-0.00039694"), float("0.000225646"), float("0.000454721"), float("0.00344821"), float("-0.00299805"), float("-0.000152523"), float("-0.000732639"), float("-0.00203668"), float("-0.000872876"), float("0.00124991"), float("-0.000206255"), float("0.000911826"), float("-0.000406766"), float("0.000470595"), float("0.00296606"), float("-0.000641648"), float("-0.00152614"), float("-0.000957327"), float("-0.000434856"), float("0.00139467"), float("-0.000964782"), float("-0.000269292"), float("-0.000128748"), float("-0.00115347"), float("-0.000888983"), float("0.000360765"), float("1.10851e-06"), float("0.000905672"), float("-7.21941e-05"), float("0.000970206"), float("0.00103436"), float("0.000360162"), float("-0.00084144"), float("-0.000245883"), float("-0.00154833"), float("0.00260887"), float("-0.0028161"), float("-0.000606185"), float("0.000816593"), float("-0.00119184"), float("-0.00179874"), float("0.00128897"), float("0.000705463"), float("0.000194246"), float("-0.000344325"), float("0.000336249"), float("0.00326308"), float("0.00110578"), float("-0.00228927"), float("4.76386e-05"), float("0.000593787"), float("0.000946524"), float("0.000321888"), float("-0.00150594"), float("0.0006321"), float("0.0028326"), float("-0.000986253"), float("-0.0018503"), float("-0.00087765"), float("0.001987"), float("-0.00344485"), float("-0.000441437"), float("-0.00218665"), float("-0.000122176"), float("-0.000163618"), float("-0.000218914"), float("0.00134576"), float("0.00197272"), float("-0.000995735"), float("0.000263282"), float("-0.00112765"), float("-0.00111128"), float("0.000379471"), float("0.00132961"), float("-7.55798e-07"), float("0.000778359"), float("-0.000557666"), float("-0.00122852"), float("0.00195573"), float("-0.000791029"), float("-0.000683356"), float("-0.000497815"), float("-1.17727e-05"), float("0.00016727"), float("-0.000183097"), float("-0.000404088"), float("0.000217476"), float("0.000506551"), float("-0.000556639"), float("-0.000408932"), float("-0.000231675"), float("0.000294711"), float("-0.00109808"), float("-0.000204529"), float("-0.000558736"), float("1.61032e-05"), float("-3.83038e-05"), float("-0.000171005"), float("-0.000396657"), float("0.000533907"), float("-0.00102416"), float("8.19507e-05"), float("-0.000511321"), float("-0.000713884"), float("-0.00065625"), float("0.000688043"), float("7.72698e-05"), float("0.000343118"), float("-3.60998e-05"), float("0.000457628"), float("0.00127803"), float("0.000562512"), float("-0.000228446"), float("-0.000199705"), float("-6.09644e-05"), float("9.9358e-05"), float("-2.25066e-05"), float("-0.000130122"), float("7.57663e-05"), float("-4.82356e-05"), float("-0.000337591"), float("9.73819e-05"), float("9.94797e-06"), float("-4.78918e-05"), float("-0.000385803"), float("2.49486e-05"), float("-0.000145524"), float("-1.2884e-05"), float("-4.45019e-05"), float("0.000119422"), float("-0.000170435"), float("0.000654842"), float("-0.000939046"), float("-1.96642e-05"), float("-0.000560269"), float("-0.000625978"), float("-0.000445959"), float("0.000620287"), float("0.000111282"), float("0.00032595"), float("-5.50135e-06"), float("0.000306219"), float("0.00119186"), float("0.000381669"), float("-0.000291645"), float("-0.00027276"), float("-1.00973e-05"), float("0.000269494"), float("-0.000242654"), float("-3.09646e-05"), float("-5.82016e-05"), float("-0.000165689"), float("-5.35421e-05"), float("8.65551e-05"), float("1.77047e-05"), float("7.4701e-05"), float("-4.16094e-05"), float("0.000107224"), float("0.000217084"), float("-1.98606e-05"), float("-0.000127421"), float("-5.16393e-05")] - - -class PirProgram_example_input_tensor_meta_984001809579706501: - program_id = 692320333968606381 - input_name = "middle_5" - shape = [16, 2, 16] - data = None - - -class PirProgram_example_input_tensor_meta_1407663771573075047: - program_id = 692320333968606381 - input_name = "middle_4" - shape = [16, 2, 32] - data = None - - -class PirProgram_example_input_tensor_meta_7391043412817019354: - program_id = 692320333968606381 - input_name = "middle_2" - shape = [1] - data = [float("0.2")] - - -class PirProgram_example_input_tensor_meta_6623077475521577491: - program_id = 692320333968606381 - input_name = "middle_0" - shape = [16, 2, 32] - data = None - - -class PirProgram_example_input_tensor_meta_3111587029586218519: - program_id = 692320333968606381 - input_name = "linear_1.w_0" - shape = [32, 16] - data = [float("0.344869"), float("-0.262408"), float("0.0439143"), float("0.0156327"), float("0.172872"), float("0.0675123"), float("0.328597"), float("0.28137"), float("0.193017"), float("0.118872"), float("0.0326795"), float("0.0050282"), float("0.0623816"), float("-0.172865"), float("0.280214"), float("-0.221634"), float("0.158068"), float("-0.227596"), float("0.345091"), float("0.0911414"), float("-0.0102415"), float("0.167898"), float("-0.158481"), float("0.289152"), float("0.113772"), float("0.103651"), float("-0.119918"), float("-0.337519"), float("-0.25293"), float("-0.307349"), float("-0.0620606"), float("0.0564728"), float("0.227404"), float("0.139312"), float("0.0545452"), float("-0.304748"), float("0.167032"), float("0.0143448"), float("0.135743"), float("-0.105886"), float("-0.343549"), float("0.290257"), float("-0.2818"), float("-0.160174"), float("0.1422"), float("-0.229487"), float("0.177531"), float("-0.315231"), float("0.331407"), float("0.0724641"), float("-0.265019"), float("-0.103591"), float("-0.30779"), float("-0.196846"), float("-0.240486"), float("-0.0806981"), float("0.246875"), float("0.155696"), float("0.207512"), float("0.245401"), float("-0.135328"), float("0.104992"), float("0.289559"), float("-0.0550811"), float("-0.0135079"), float("0.345409"), float("0.237245"), float("-0.201128"), float("0.14598"), float("0.235113"), float("0.208274"), float("-0.244138"), float("0.348133"), float("0.0957723"), float("-0.00846937"), float("0.128104"), float("0.236483"), float("-0.026668"), float("0.278156"), float("0.122785"), float("-0.148078"), float("-0.251864"), float("0.327847"), float("0.185817"), float("-0.0450313"), float("0.238892"), float("0.252549"), float("0.330687"), float("-0.0880492"), float("0.309633"), float("-0.219191"), float("-0.199132"), float("0.0599504"), float("-0.1848"), float("-0.11876"), float("-0.0490333"), float("0.0584787"), float("-0.0681539"), float("-0.298419"), float("-0.0999435"), float("0.346551"), float("0.129203"), float("-0.332224"), float("-0.147948"), float("0.00135393"), float("0.0816889"), float("-0.131467"), float("-0.224166"), float("-0.0461218"), float("-0.214483"), float("-0.0326581"), float("0.10036"), float("-0.264792"), float("-0.32798"), float("0.271663"), float("0.288337"), float("-0.350532"), float("0.06591"), float("-0.0911039"), float("0.0444793"), float("-0.00501465"), float("-0.213402"), float("0.349463"), float("-0.232356"), float("-0.011135"), float("-0.234883"), float("-0.223632"), float("0.207395"), float("0.242189"), float("-0.34568"), float("0.281336"), float("-0.141172"), float("-0.279639"), float("-0.0623853"), float("-0.31392"), float("0.0430904"), float("0.319302"), float("0.0603828"), float("0.165175"), float("0.0803925"), float("0.241733"), float("-0.31437"), float("-0.0401438"), float("-0.186046"), float("-0.0437005"), float("0.3525"), float("0.00955195"), float("-0.245716"), float("-0.23917"), float("-0.22811"), float("-0.180175"), float("0.0921878"), float("-0.153823"), float("-0.0783991"), float("0.297338"), float("-0.230138"), float("-0.11621"), float("-0.171227"), float("-0.235734"), float("-0.13929"), float("-0.198997"), float("0.16052"), float("0.210111"), float("0.18088"), float("0.242678"), float("-0.276179"), float("-0.183217"), float("0.00408171"), float("-0.230232"), float("-0.305358"), float("0.207404"), float("-0.0695701"), float("0.349871"), float("-0.293597"), float("0.342355"), float("0.349062"), float("-0.0969384"), float("-0.304206"), float("-0.100894"), float("0.218041"), float("-0.238812"), float("0.175587"), float("0.324251"), float("-0.219158"), float("-0.192937"), float("0.17492"), float("0.190574"), float("0.0470951"), float("-0.140387"), float("-0.180639"), float("0.0101866"), float("-0.195554"), float("0.0171249"), float("-0.0556396"), float("0.220385"), float("0.0830055"), float("-0.209667"), float("0.0225179"), float("-0.0834754"), float("0.111096"), float("0.125786"), float("0.0543055"), float("0.0282446"), float("0.0734042"), float("0.313033"), float("-0.0875646"), float("0.0155789"), float("-0.191"), float("-0.174547"), float("0.29949"), float("0.101439"), float("0.167632"), float("0.294102"), float("0.284237"), float("-0.306646"), float("0.123889"), float("0.273155"), float("0.0722914"), float("0.343355"), float("0.200973"), float("-0.0359481"), float("0.295771"), float("-0.0428864"), float("-0.276863"), float("0.219781"), float("-0.321594"), float("-0.193598"), float("0.0766501"), float("-0.0693026"), float("-0.108357"), float("0.239546"), float("-0.0313863"), float("-0.235817"), float("0.312584"), float("0.147659"), float("0.199029"), float("0.0308142"), float("0.319813"), float("0.276646"), float("0.232516"), float("-0.108472"), float("-0.271957"), float("0.214807"), float("0.131503"), float("0.0458386"), float("0.158341"), float("-0.258767"), float("-0.0813985"), float("0.00375828"), float("0.135654"), float("0.265497"), float("0.0394395"), float("-0.234527"), float("0.0994757"), float("-0.342846"), float("0.215965"), float("-0.319046"), float("0.146627"), float("-0.0292457"), float("0.0153329"), float("0.351146"), float("0.082035"), float("0.350166"), float("-0.317657"), float("-0.232845"), float("-0.0578491"), float("-0.266218"), float("0.270595"), float("0.119427"), float("-0.238121"), float("0.319077"), float("0.299425"), float("-0.125333"), float("0.0425065"), float("-0.157729"), float("-0.291869"), float("-0.202124"), float("-0.17027"), float("0.290271"), float("0.0382349"), float("0.322722"), float("-0.215145"), float("-0.0237843"), float("-0.0231641"), float("0.347289"), float("0.217441"), float("0.182893"), float("0.0933436"), float("-0.00364189"), float("0.251967"), float("-0.248983"), float("0.0451025"), float("-0.0902002"), float("0.0672752"), float("-0.33523"), float("0.00100524"), float("-0.244151"), float("0.116821"), float("-0.218376"), float("0.0135987"), float("-0.155867"), float("0.0408046"), float("0.272804"), float("-0.157648"), float("0.178179"), float("0.0977667"), float("-0.119494"), float("-0.184943"), float("-0.330054"), float("-0.00757139"), float("-0.086305"), float("-0.00671808"), float("-0.221499"), float("0.178183"), float("-0.315276"), float("-0.0872174"), float("-0.142795"), float("-0.0392847"), float("0.125988"), float("-0.309111"), float("-0.100147"), float("-0.281519"), float("-0.108678"), float("0.170384"), float("-0.315366"), float("0.262136"), float("-0.0733001"), float("0.127174"), float("0.176889"), float("-0.211834"), float("-0.323871"), float("-0.0813174"), float("-0.263945"), float("-0.287189"), float("0.0900709"), float("-0.149072"), float("-0.100897"), float("-5.45169e-05"), float("0.0549951"), float("-0.167664"), float("0.286603"), float("-0.187299"), float("-0.0942526"), float("0.178792"), float("0.300509"), float("0.353007"), float("-0.32969"), float("0.044596"), float("0.255667"), float("-0.191181"), float("-0.295033"), float("-0.0528022"), float("0.287379"), float("0.158327"), float("0.00928331"), float("0.228015"), float("0.00668824"), float("0.0287411"), float("-0.0175968"), float("0.0713126"), float("0.347188"), float("-0.00186841"), float("-0.0819523"), float("-0.221304"), float("-0.0364172"), float("-0.264244"), float("0.307289"), float("-0.312169"), float("-0.0896873"), float("0.110611"), float("-0.347699"), float("-0.266626"), float("0.151276"), float("0.317222"), float("-0.0747132"), float("0.0141522"), float("-0.18889"), float("-0.282493"), float("-0.114794"), float("-0.19188"), float("-0.189031"), float("0.038257"), float("0.182447"), float("0.338656"), float("-0.185908"), float("0.282292"), float("-0.333012"), float("-0.215363"), float("0.198045"), float("0.10666"), float("0.0252068"), float("-0.186876"), float("0.33483"), float("0.246478"), float("-0.0593159"), float("-0.0243462"), float("0.271037"), float("-0.140778"), float("0.321753"), float("-0.0425401"), float("-0.0232625"), float("0.15166"), float("0.0834763"), float("0.0449372"), float("0.0559293"), float("0.351461"), float("0.0579684"), float("-0.0641627"), float("0.161772"), float("-0.141108"), float("-0.331302"), float("-0.0164107"), float("-0.119114"), float("0.0939414"), float("-0.254304"), float("-0.0356149"), float("-0.342954"), float("0.0774205"), float("0.050752"), float("-0.102107"), float("0.106109"), float("0.099873"), float("0.144805"), float("0.170335"), float("-0.202735"), float("-0.0852358"), float("0.1246"), float("0.0621777"), float("-0.35238"), float("0.0236915"), float("-0.0462948"), float("-0.0346923"), float("-0.0105365"), float("-0.0434867"), float("0.0888052"), float("-0.0298811"), float("0.278204"), float("0.248552"), float("0.0612633"), float("-0.296462"), float("-0.149241"), float("-0.00470171"), float("0.122117"), float("0.00352487"), float("-0.313724"), float("0.157657"), float("0.286055"), float("-0.179478"), float("0.165632"), float("0.200262"), float("-0.115712"), float("0.0705119"), float("-0.313889"), float("-0.285383"), float("0.272408"), float("-0.298484"), float("-0.133519"), float("-0.268654"), float("0.234496"), float("0.17597"), float("0.253718"), float("0.0419651"), float("0.324774"), float("0.332485"), float("-0.28583"), float("0.111105"), float("-0.194049"), float("-0.168264"), float("0.295419"), float("-0.0469957"), float("-0.323113"), float("0.219891"), float("-0.205654"), float("-0.136962"), float("0.0286586"), float("0.247516"), float("0.0750707"), float("-0.286282"), float("-0.0419273"), float("-0.167156"), float("0.338594"), float("0.18547"), float("-0.179599"), float("0.169526"), float("0.214873"), float("-0.340293"), float("-0.149433"), float("0.0603927"), float("-0.08848"), float("0.324922"), float("0.0252319"), float("-0.278085"), float("0.275881"), float("-0.322733"), float("-0.119825"), float("0.178038"), float("-0.117301"), float("0.346557"), float("0.122787"), float("-0.161778"), float("0.114787"), float("0.123685"), float("0.0642729"), float("0.262809"), float("-0.173838"), float("-0.114229"), float("-0.167517"), float("-0.324107"), float("-0.244136"), float("0.172947"), float("0.207802"), float("-0.277253")] - - -class PirProgram_example_input_tensor_meta_81239810641304043: - program_id = 692320333968606381 - input_name = "linear_1.b_0" - shape = [16] - data = [float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0")] - - -class PirProgram_example_input_tensor_meta_1643386415584303293: - program_id = 692320333968606381 - input_name = "middle_3" - shape = [16, 2, 32] - data = [float("0.723589"), float("0"), float("0"), float("6.28983"), float("0"), float("0"), float("0"), float("1.92776"), float("0"), float("0"), float("4.57823"), float("0"), float("3.68971"), float("0"), float("0.607042"), float("0.85038"), float("2.91316"), float("0"), float("0"), float("0.443321"), float("0"), float("6.09899"), float("2.37732"), float("0.4176"), float("2.09232"), float("0"), float("0"), float("5.69163"), float("0"), float("3.48354"), float("0"), float("0"), float("0.271179"), float("0"), float("0"), float("2.45632"), float("0"), float("0"), float("0"), float("0.763683"), float("0"), float("2.4179"), float("1.79379"), float("0"), float("1.44576"), float("0"), float("0.210515"), float("0.347592"), float("0"), float("2.32378"), float("0"), float("0.174633"), float("0"), float("2.37125"), float("0.922107"), float("0.187329"), float("0.81025"), float("0"), float("0"), float("2.25267"), float("0.632394"), float("1.35589"), float("0"), float("0"), float("0"), float("0.664025"), float("0.530738"), float("0"), float("0.843825"), float("1.08603"), float("0.16219"), float("0"), float("0.791713"), float("0"), float("0"), float("0.949193"), float("0"), float("0.600742"), float("0"), float("0"), float("0"), float("0"), float("0.897179"), float("0"), float("1.29137"), float("0"), float("0"), float("0"), float("0"), float("1.61578"), float("0"), float("0"), float("0"), float("0"), float("0.354193"), float("1.23875"), float("0.0676746"), float("0"), float("0"), float("0.646917"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.632321"), float("0.472122"), float("0"), float("0.376537"), float("0"), float("0.0776579"), float("0.0903725"), float("0"), float("0.60918"), float("0"), float("0.0524048"), float("0"), float("0.620071"), float("0.246761"), float("0"), float("0.201337"), float("0"), float("0"), float("0.575339"), float("0.163366"), float("0.340029"), float("0"), float("0"), float("0"), float("1.13115"), float("0.828709"), float("0"), float("1.45559"), float("1.82495"), float("0"), float("0"), float("1.64345"), float("0"), float("0"), float("1.69007"), float("0"), float("1.23017"), float("0"), float("0"), float("0"), float("0"), float("1.66897"), float("0"), float("2.01373"), float("0"), float("0"), float("0"), float("0"), float("2.80322"), float("0"), float("0"), float("0"), float("0"), float("0.433559"), float("2.17109"), float("0"), float("0.499996"), float("0.36502"), float("0"), float("0.64794"), float("0.807835"), float("0.134061"), float("0"), float("0.740266"), float("0"), float("0"), float("0.745333"), float("0"), float("0.544806"), float("0"), float("0"), float("0"), float("0"), float("0.743101"), float("0"), float("0.89026"), float("0"), float("0"), float("0"), float("0"), float("1.241"), float("0.699808"), float("0"), float("0"), float("0"), float("0.186857"), float("0.973017"), float("0"), float("1.19681"), float("0.724072"), float("0"), float("1.45466"), float("1.74084"), float("0.428011"), float("0"), float("0"), float("0"), float("0"), float("1.4446"), float("0"), float("1.06807"), float("0"), float("0"), float("0"), float("0"), float("1.58471"), float("0"), float("2.22805"), float("0"), float("0"), float("0"), float("0"), float("2.90281"), float("1.58111"), float("0"), float("0"), float("0"), float("0"), float("2.41101"), float("0"), float("0"), float("0"), float("0.225791"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.219702"), float("0.13923"), float("0"), float("0.0994024"), float("0"), float("0.0351282"), float("0.0333149"), float("0.115384"), float("0"), float("0"), float("0.00266587"), float("0"), float("0.210125"), float("0.0723203"), float("0.0102157"), float("0.0743553"), float("0"), float("0"), float("0.158375"), float("0.0459132"), float("0.116562"), float("0"), float("0"), float("0.167725"), float("0"), float("0"), float("1.51631"), float("0"), float("0"), float("0"), float("0.458128"), float("0"), float("1.47672"), float("1.08063"), float("0"), float("0.808897"), float("0"), float("0.18597"), float("0.232214"), float("0.753348"), float("0"), float("0"), float("0.00460923"), float("0"), float("0"), float("0"), float("0.0860425"), float("0.529424"), float("0"), float("0"), float("1.25368"), float("0"), float("0.882674"), float("0"), float("0"), float("0"), float("0"), float("0"), float("2.13523"), float("0"), float("0"), float("0"), float("0.66402"), float("0"), float("2.1026"), float("0"), float("0"), float("1.25584"), float("0"), float("0.19961"), float("0.295115"), float("0.982376"), float("2.01711"), float("0"), float("0.152203"), float("0"), float("2.06125"), float("0.806557"), float("0.147271"), float("0"), float("0"), float("0"), float("1.94095"), float("0.547275"), float("1.17657"), float("0"), float("0"), float("0"), float("2.0945"), float("0"), float("0"), float("2.71852"), float("3.39442"), float("0.543131"), float("0"), float("3.05501"), float("0"), float("0"), float("3.12772"), float("0"), float("2.25904"), float("0"), float("0"), float("0"), float("0"), float("3.11314"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("5.23011"), float("2.90551"), float("0"), float("0"), float("0"), float("0.81488"), float("4.04086"), float("0"), float("2.36868"), float("1.72581"), float("0"), float("3.04193"), float("3.81239"), float("0.621253"), float("0"), float("3.48355"), float("0"), float("0"), float("3.57841"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("4.22051"), float("0"), float("0"), float("0"), float("0"), float("5.8696"), float("3.2964"), float("0"), float("0"), float("0"), float("0.907758"), float("4.54258"), float("0"), float("0.315007"), float("0.206424"), float("0"), float("0"), float("0.376567"), float("0.122297"), float("0"), float("0.283101"), float("0"), float("0"), float("0.19817"), float("0"), float("0"), float("0.077143"), float("0"), float("0"), float("0"), float("0.317819"), float("0"), float("0.574079"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.342921"), float("0"), float("0"), float("0"), float("0.0767962"), float("0.601694"), float("0"), float("0.622385"), float("0"), float("0"), float("0.818888"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.948796"), float("0"), float("1.11377"), float("0"), float("0"), float("0"), float("0"), float("1.55717"), float("0.869019"), float("0"), float("0"), float("0"), float("0.238691"), float("1.22154"), float("0.0845757"), float("0"), float("0"), float("0.614002"), float("0"), float("0"), float("0"), float("0.202461"), float("0"), float("0.609987"), float("0.456371"), float("0"), float("0.391232"), float("0"), float("0"), float("0.0745103"), float("0.288329"), float("0"), float("0"), float("0.039889"), float("0"), float("0.59132"), float("0.238821"), float("0.0415231"), float("0.210348"), float("0"), float("0"), float("0.575056"), float("0.155218"), float("0.339351"), float("0"), float("0"), float("0"), float("1.40681"), float("1.02908"), float("0"), float("1.81357"), float("2.29121"), float("0"), float("0"), float("2.07469"), float("0"), float("0"), float("2.14309"), float("0"), float("1.5367"), float("0"), float("0"), float("0"), float("0"), float("2.11169"), float("0"), float("2.52412"), float("0"), float("0"), float("0"), float("0"), float("3.50848"), float("1.97313"), float("0"), float("0"), float("0"), float("0.5445"), float("2.72088"), float("0"), float("0"), float("0"), float("2.24401"), float("0"), float("0"), float("0"), float("0.686096"), float("0"), float("2.21794"), float("0"), float("0"), float("1.30706"), float("0"), float("0.208904"), float("0.309097"), float("0"), float("2.12223"), float("0"), float("0.137407"), float("0"), float("2.17288"), float("0.849765"), float("0.153441"), float("0.754491"), float("0"), float("0"), float("2.01347"), float("0.563138"), float("1.24616"), float("0"), float("0"), float("0.527303"), float("0"), float("0"), float("4.42907"), float("0"), float("0"), float("0"), float("1.36257"), float("0"), float("4.37391"), float("3.2329"), float("0"), float("2.59558"), float("0"), float("0"), float("0.592568"), float("2.05092"), float("4.16704"), float("0"), float("0.312967"), float("0"), float("4.28341"), float("1.67541"), float("0.30598"), float("0"), float("0"), float("0"), float("4.02188"), float("1.12964"), float("2.46205"), float("0"), float("0"), float("0"), float("0"), float("0.2639"), float("0"), float("0.490358"), float("0.634138"), float("0.112012"), float("0"), float("0.54594"), float("0"), float("0"), float("0.577505"), float("0"), float("0.391071"), float("0"), float("0"), float("0"), float("0"), float("0.595201"), float("0"), float("0.710756"), float("0"), float("0"), float("0"), float("0"), float("0.963"), float("0.534581"), float("0"), float("0"), float("0"), float("0"), float("0.746982"), float("0"), float("0.0715721"), float("0.0618026"), float("0"), float("0.102145"), float("0.126783"), float("0.0210766"), float("0"), float("0"), float("0"), float("0"), float("0.0819103"), float("0"), float("0"), float("0.00295741"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.163064"), float("0"), float("0"), float("0"), float("0"), float("0.196886"), float("0"), float("0"), float("0"), float("0"), float("0.060482"), float("0.162623"), float("0.403443"), float("0"), float("0"), float("3.50631"), float("0"), float("0"), float("0"), float("1.05008"), float("0"), float("0"), float("2.5659"), float("0"), float("2.04395"), float("0"), float("0.33452"), float("0"), float("1.64144"), float("3.27211"), float("0"), float("0.21803"), float("0"), float("3.39128"), float("0"), float("0.223745"), float("1.16294"), float("0"), float("0"), float("3.1488"), float("0.884625"), float("1.95163"), float("0"), float("0"), float("0.223519"), float("0"), float("0"), float("1.90329"), float("0"), float("0"), float("0"), float("0.586335"), float("0"), float("1.86716"), float("1.38074"), float("0"), float("0"), float("0"), float("0.179526"), float("0.256382"), float("0.877256"), float("1.78683"), float("0"), float("0.128199"), float("0"), float("0"), float("0"), float("0"), float("0.635889"), float("0"), float("0"), float("1.72533"), float("0.482998"), float("1.05037"), float("0"), float("0"), float("0.262392"), float("0"), float("0"), float("2.26296"), float("0"), float("0"), float("0"), float("0.703865"), float("0"), float("2.25157"), float("1.65439"), float("0"), float("0"), float("0"), float("0.222988"), float("0.301277"), float("1.05197"), float("2.14329"), float("0"), float("0.16106"), float("0"), float("2.20785"), float("0.855672"), float("0.146199"), float("0.753635"), float("0"), float("0"), float("2.05749"), float("0.576322"), float("1.25052"), float("0"), float("0"), float("0.117819"), float("0"), float("0"), float("1.01242"), float("0"), float("0"), float("0"), float("0.314459"), float("0"), float("1.00417"), float("0.740845"), float("0"), float("0.599216"), float("0"), float("0"), float("0.132811"), float("0.470765"), float("0"), float("0"), float("0"), float("0"), float("0.987654"), float("0"), float("0"), float("0.342916"), float("0"), float("0"), float("0.926627"), float("0.257738"), float("0.562952"), float("0"), float("0"), float("0.26498"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.729735"), float("0"), float("2.27784"), float("1.6737"), float("0"), float("0"), float("0"), float("0"), float("0.315342"), float("1.05739"), float("2.1737"), float("0"), float("0.16906"), float("0"), float("2.21541"), float("0.86"), float("0.157351"), float("0.758679"), float("0"), float("0"), float("2.09938"), float("0.591763"), float("1.26801"), float("0"), float("0"), float("0"), float("0"), float("0.828738"), float("0"), float("1.46613"), float("1.81584"), float("0.311517"), float("0"), float("0"), float("0"), float("0"), float("1.69295"), float("0"), float("1.22507"), float("0"), float("0"), float("0"), float("0"), float("1.70447"), float("0"), float("2.02645"), float("0"), float("0"), float("0"), float("0"), float("2.83993"), float("1.60285"), float("0"), float("0"), float("0"), float("0.417853"), float("2.20486"), float("0.324522"), float("0"), float("0"), float("2.62623"), float("0"), float("0"), float("0"), float("0.815165"), float("0"), float("2.57824"), float("1.88527"), float("0"), float("1.4626"), float("0"), float("0.302452"), float("0.349664"), float("1.31955"), float("2.45426"), float("0"), float("0.0641848"), float("0"), float("2.53626"), float("1.00942"), float("0.126232"), float("0.896666"), float("0"), float("0"), float("2.23374"), float("0.599363"), float("1.47756"), float("0"), float("0"), float("0"), float("0.302239"), float("0.212177"), float("0"), float("0.378777"), float("0.468387"), float("0.0813494"), float("0"), float("0.433903"), float("0"), float("0"), float("0.438357"), float("0"), float("0.31956"), float("0"), float("0"), float("0"), float("0"), float("0.426783"), float("0"), float("0.524977"), float("0"), float("0"), float("0"), float("0"), float("0.729943"), float("0.404518"), float("0"), float("0"), float("0"), float("0"), float("0.559602"), float("0.0706485"), float("0"), float("0"), float("0.625332"), float("0"), float("0"), float("0"), float("0.199803"), float("0"), float("0.607633"), float("0.462029"), float("0"), float("0.381851"), float("0"), float("0.0538131"), float("0.0799106"), float("0"), float("0.596368"), float("0"), float("0.050653"), float("0"), float("0.599458"), float("0.238125"), float("0.045"), float("0.204004"), float("0"), float("0"), float("0.58834"), float("0"), float("0.342713"), float("0"), float("0"), float("0"), float("0.235136"), float("0.167431"), float("0"), float("0.303585"), float("0.377351"), float("0.0631602"), float("0"), float("0.344763"), float("0"), float("0"), float("0"), float("0"), float("0.255787"), float("0"), float("0"), float("0"), float("0"), float("0.352881"), float("0"), float("0.412486"), float("0"), float("0"), float("0"), float("0"), float("0.580145"), float("0"), float("0"), float("0"), float("0"), float("0.0895882"), float("0.447694"), float("0.0637865"), float("0"), float("0"), float("0.599056"), float("0"), float("0"), float("0"), float("0.175942"), float("0"), float("0.592536"), float("0.436414"), float("0"), float("0.345927"), float("0"), float("0.0527949"), float("0"), float("0"), float("0.560908"), float("0"), float("0"), float("0"), float("0.578455"), float("0.221061"), float("0.0428592"), float("0"), float("0"), float("0"), float("0.537149"), float("0"), float("0.333177"), float("0"), float("0"), float("0.0157932"), float("0"), float("0"), float("0.141245"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.142142"), float("0.0979528"), float("0"), float("0"), float("0"), float("0.0137536"), float("0"), float("0.0659162"), float("0.138318"), float("0"), float("0.0213317"), float("0"), float("0.138244"), float("0.0449504"), float("0"), float("0.05323"), float("0"), float("0"), float("0.136466"), float("0.0393414"), float("0.0763775"), float("0"), float("0")] - - -class PirProgram_example_input_tensor_meta_6890028235555013676: - program_id = 692320333968606381 - input_name = "middle_1" - shape = [16, 2, 32] - data = [float("0.578871"), float("0"), float("0"), float("5.03186"), float("0"), float("0"), float("0"), float("1.5422"), float("0"), float("4.98103"), float("3.66259"), float("0"), float("2.95177"), float("0"), float("0.485634"), float("0.680304"), float("2.33053"), float("4.73965"), float("0"), float("0.354657"), float("0"), float("4.87919"), float("1.90186"), float("0.33408"), float("1.67386"), float("0"), float("0"), float("4.5533"), float("1.28468"), float("2.78683"), float("0"), float("0"), float("0.216943"), float("0"), float("0"), float("1.96505"), float("0"), float("0"), float("0"), float("0.610946"), float("0"), float("1.93432"), float("1.43503"), float("0"), float("1.1566"), float("0"), float("0.168412"), float("0.278073"), float("0.895773"), float("1.85903"), float("0"), float("0.139707"), float("0"), float("1.897"), float("0.737685"), float("0.149863"), float("0.6482"), float("0"), float("0"), float("1.80214"), float("0.505915"), float("1.08471"), float("0"), float("0"), float("0"), float("0.53122"), float("0.42459"), float("0"), float("0.67506"), float("0.868822"), float("0.129752"), float("0"), float("0.633371"), float("0"), float("0"), float("0.759354"), float("0"), float("0.480594"), float("0"), float("0"), float("0"), float("0"), float("0.717743"), float("0"), float("1.0331"), float("0"), float("0"), float("0"), float("0"), float("1.29263"), float("0.697283"), float("0"), float("0"), float("0"), float("0.283354"), float("0.991004"), float("0.0541397"), float("0"), float("0"), float("0.517533"), float("0"), float("0"), float("0"), float("0.16042"), float("0"), float("0.505857"), float("0.377697"), float("0"), float("0.30123"), float("0"), float("0.0621263"), float("0.072298"), float("0.248242"), float("0.487344"), float("0"), float("0.0419238"), float("0"), float("0.496056"), float("0.197409"), float("0.0200742"), float("0.16107"), float("0"), float("0"), float("0.460271"), float("0.130693"), float("0.272023"), float("0"), float("0"), float("0"), float("0.90492"), float("0.662967"), float("0"), float("1.16447"), float("1.45996"), float("0.227835"), float("0"), float("1.31476"), float("0"), float("0"), float("1.35205"), float("0"), float("0.984135"), float("0"), float("0"), float("0"), float("0"), float("1.33518"), float("0"), float("1.61099"), float("0"), float("0"), float("0"), float("0"), float("2.24257"), float("1.25536"), float("0"), float("0"), float("0"), float("0.346848"), float("1.73687"), float("0"), float("0.399997"), float("0.292016"), float("0"), float("0.518352"), float("0.646268"), float("0.107249"), float("0"), float("0.592213"), float("0"), float("0"), float("0.596267"), float("0"), float("0.435845"), float("0"), float("0"), float("0"), float("0"), float("0.594481"), float("0"), float("0.712208"), float("0"), float("0"), float("0"), float("0"), float("0.992802"), float("0.559846"), float("0"), float("0"), float("0"), float("0.149486"), float("0.778414"), float("0"), float("0.957444"), float("0.579258"), float("0"), float("1.16373"), float("1.39267"), float("0.342409"), float("0"), float("1.2634"), float("0"), float("0"), float("1.15568"), float("0"), float("0.854454"), float("0"), float("0"), float("0"), float("0"), float("1.26777"), float("0"), float("1.78244"), float("0"), float("0"), float("0"), float("0"), float("2.32225"), float("1.26489"), float("0"), float("0"), float("0"), float("0.358627"), float("1.9288"), float("0.0289001"), float("0"), float("0"), float("0.180633"), float("0"), float("0"), float("0"), float("0.0530196"), float("0"), float("0.175761"), float("0.111384"), float("0"), float("0.0795219"), float("0"), float("0.0281026"), float("0.0266519"), float("0.0923073"), float("0.155116"), float("0"), float("0.0021327"), float("0"), float("0.1681"), float("0.0578563"), float("0.00817253"), float("0.0594842"), float("0"), float("0"), float("0.1267"), float("0.0367306"), float("0.0932499"), float("0"), float("0"), float("0.13418"), float("0"), float("0"), float("1.21305"), float("0"), float("0"), float("0"), float("0.366503"), float("0"), float("1.18137"), float("0.864508"), float("0"), float("0.647117"), float("0"), float("0.148776"), float("0.185771"), float("0.602678"), float("1.12672"), float("0"), float("0.00368738"), float("0"), float("1.17025"), float("0.461667"), float("0.068834"), float("0.423539"), float("0"), float("0"), float("1.00295"), float("0.276469"), float("0.706139"), float("0"), float("0"), float("0.196585"), float("0"), float("0"), float("1.70819"), float("0"), float("0"), float("0"), float("0.531216"), float("0"), float("1.68208"), float("1.2358"), float("0"), float("1.00467"), float("0"), float("0.159688"), float("0.236092"), float("0.785901"), float("1.61369"), float("0"), float("0.121763"), float("0"), float("1.649"), float("0.645246"), float("0.117817"), float("0.569933"), float("0"), float("0"), float("1.55276"), float("0.43782"), float("0.941253"), float("0"), float("0"), float("0"), float("1.6756"), float("1.21723"), float("0"), float("2.17482"), float("2.71554"), float("0.434505"), float("0"), float("2.44401"), float("0"), float("0"), float("2.50218"), float("0"), float("1.80723"), float("0"), float("0"), float("0"), float("0"), float("2.49051"), float("0"), float("3.00915"), float("0"), float("0"), float("0"), float("0"), float("4.18409"), float("2.32441"), float("0"), float("0"), float("0"), float("0.651904"), float("3.23269"), float("0"), float("1.89495"), float("1.38065"), float("0"), float("2.43354"), float("3.04991"), float("0.497002"), float("0"), float("2.78684"), float("0"), float("0"), float("2.86273"), float("0"), float("2.06031"), float("0"), float("0"), float("0"), float("0"), float("2.82139"), float("0"), float("3.37641"), float("0"), float("0"), float("0"), float("0"), float("4.69568"), float("2.63712"), float("0"), float("0"), float("0"), float("0.726207"), float("3.63407"), float("0"), float("0.252005"), float("0.165139"), float("0"), float("0.234757"), float("0.301253"), float("0.0978375"), float("0"), float("0.226481"), float("0"), float("0"), float("0.158536"), float("0"), float("0.146847"), float("0.0617144"), float("0"), float("0.0277497"), float("0"), float("0.254255"), float("0"), float("0.459263"), float("0"), float("0"), float("0"), float("0"), float("0.519019"), float("0.274337"), float("0"), float("0"), float("0"), float("0.061437"), float("0.481355"), float("0"), float("0.497908"), float("0.365889"), float("0"), float("0.65511"), float("0.814613"), float("0.128348"), float("0"), float("0.746606"), float("0"), float("0"), float("0.767712"), float("0"), float("0.555277"), float("0"), float("0"), float("0"), float("0"), float("0.759037"), float("0"), float("0.891017"), float("0"), float("0"), float("0"), float("0"), float("1.24574"), float("0.695215"), float("0"), float("0"), float("0"), float("0.190953"), float("0.977231"), float("0.0676606"), float("0"), float("0"), float("0.491202"), float("0"), float("0"), float("0"), float("0.161969"), float("0"), float("0.48799"), float("0.365097"), float("0"), float("0.312986"), float("0"), float("0.0529859"), float("0.0596082"), float("0.230663"), float("0.474683"), float("0"), float("0.0319112"), float("0"), float("0.473056"), float("0.191057"), float("0.0332185"), float("0.168278"), float("0"), float("0"), float("0.460044"), float("0.124175"), float("0.27148"), float("0"), float("0"), float("0"), float("1.12545"), float("0.823264"), float("0"), float("1.45085"), float("1.83297"), float("0.288033"), float("0"), float("1.65976"), float("0"), float("0"), float("1.71447"), float("0"), float("1.22936"), float("0"), float("0"), float("0"), float("0"), float("1.68935"), float("0"), float("2.0193"), float("0"), float("0"), float("0"), float("0"), float("2.80679"), float("1.5785"), float("0"), float("0"), float("0"), float("0.4356"), float("2.17671"), float("0.200711"), float("0"), float("0"), float("1.79521"), float("0"), float("0"), float("0"), float("0.548877"), float("0"), float("1.77435"), float("1.30207"), float("0"), float("1.04565"), float("0"), float("0.167124"), float("0.247277"), float("0.834098"), float("1.69778"), float("0"), float("0.109925"), float("0"), float("1.7383"), float("0.679812"), float("0.122753"), float("0.603593"), float("0"), float("0"), float("1.61078"), float("0.45051"), float("0.996926"), float("0"), float("0"), float("0.421842"), float("0"), float("0"), float("3.54325"), float("0"), float("0"), float("0"), float("1.09006"), float("0"), float("3.49913"), float("2.58632"), float("0"), float("2.07647"), float("0"), float("0.338603"), float("0.474054"), float("1.64073"), float("3.33363"), float("0"), float("0.250374"), float("0"), float("3.42673"), float("1.34033"), float("0.244784"), float("1.18908"), float("0"), float("0"), float("3.2175"), float("0.903709"), float("1.96964"), float("0"), float("0"), float("0"), float("0.314209"), float("0.21112"), float("0"), float("0.392287"), float("0.50731"), float("0.0896099"), float("0"), float("0.436752"), float("0"), float("0"), float("0.462004"), float("0"), float("0.312857"), float("0"), float("0"), float("0"), float("0"), float("0.47616"), float("0"), float("0.568605"), float("0"), float("0"), float("0"), float("0"), float("0.7704"), float("0.427665"), float("0"), float("0"), float("0"), float("0.14075"), float("0.597586"), float("0"), float("0.0572577"), float("0.0494421"), float("0"), float("0.081716"), float("0.101426"), float("0.0168613"), float("0"), float("0.0787916"), float("0"), float("0"), float("0.0655282"), float("0"), float("0.0364558"), float("0.00236592"), float("0"), float("0"), float("0"), float("0.0653458"), float("0"), float("0.130451"), float("0"), float("0"), float("0"), float("0"), float("0.157509"), float("0.0695541"), float("0"), float("0"), float("0"), float("0.0483856"), float("0.130099"), float("0.322755"), float("0"), float("0"), float("2.80505"), float("0"), float("0"), float("0"), float("0.840062"), float("0"), float("2.75153"), float("2.05272"), float("0"), float("1.63516"), float("0"), float("0.267616"), float("0.353298"), float("1.31315"), float("2.61768"), float("0"), float("0.174424"), float("0"), float("2.71303"), float("1.05055"), float("0.178996"), float("0.930353"), float("0"), float("0"), float("2.51904"), float("0.7077"), float("1.5613"), float("0"), float("0"), float("0.178815"), float("0"), float("0"), float("1.52263"), float("0"), float("0"), float("0"), float("0.469068"), float("0"), float("1.49373"), float("1.10459"), float("0"), float("0.891864"), float("0"), float("0.143621"), float("0.205105"), float("0.701805"), float("1.42947"), float("0"), float("0.102559"), float("0"), float("1.46317"), float("0.5694"), float("0.108553"), float("0.508712"), float("0"), float("0"), float("1.38026"), float("0.386398"), float("0.840296"), float("0"), float("0"), float("0.209913"), float("0"), float("0"), float("1.81037"), float("0"), float("0"), float("0"), float("0.563092"), float("0"), float("1.80125"), float("1.32351"), float("0"), float("1.0774"), float("0"), float("0.178391"), float("0.241022"), float("0.841577"), float("1.71463"), float("0"), float("0.128848"), float("0"), float("1.76628"), float("0.684538"), float("0.116959"), float("0.602908"), float("0"), float("0"), float("1.64599"), float("0.461058"), float("1.00041"), float("0"), float("0"), float("0.0942551"), float("0"), float("0"), float("0.809935"), float("0"), float("0"), float("0"), float("0.251567"), float("0"), float("0.803332"), float("0.592676"), float("0"), float("0.479373"), float("0"), float("0.0745486"), float("0.106249"), float("0.376612"), float("0.764936"), float("0"), float("0.0548324"), float("0"), float("0.790123"), float("0.308984"), float("0.0567979"), float("0.274333"), float("0"), float("0"), float("0.741301"), float("0.20619"), float("0.450362"), float("0"), float("0"), float("0.211984"), float("0"), float("0"), float("1.83722"), float("0"), float("0"), float("0"), float("0.583788"), float("0"), float("1.82228"), float("1.33896"), float("0"), float("1.0933"), float("0"), float("0.180633"), float("0.252274"), float("0.845911"), float("1.73896"), float("0"), float("0.135248"), float("0"), float("1.77233"), float("0.688"), float("0.125881"), float("0.606944"), float("0"), float("0"), float("1.67951"), float("0.473411"), float("1.01441"), float("0"), float("0"), float("0"), float("0.925363"), float("0.66299"), float("0"), float("1.1729"), float("1.45267"), float("0.249214"), float("0"), float("1.34885"), float("0"), float("0"), float("1.35436"), float("0"), float("0.980055"), float("0"), float("0"), float("0"), float("0"), float("1.36358"), float("0"), float("1.62116"), float("0"), float("0"), float("0"), float("0"), float("2.27194"), float("1.28228"), float("0"), float("0"), float("0"), float("0.334282"), float("1.76389"), float("0.259617"), float("0"), float("0"), float("2.10098"), float("0"), float("0"), float("0"), float("0.652132"), float("0"), float("2.06259"), float("1.50822"), float("0"), float("1.17008"), float("0"), float("0.241962"), float("0.279731"), float("1.05564"), float("1.96341"), float("0"), float("0.0513479"), float("0"), float("2.02901"), float("0.807535"), float("0.100986"), float("0.717333"), float("0"), float("0"), float("1.78699"), float("0.47949"), float("1.18205"), float("0"), float("0"), float("0"), float("0.241791"), float("0.169742"), float("0"), float("0.303022"), float("0.374709"), float("0.0650795"), float("0"), float("0.347122"), float("0"), float("0"), float("0.350685"), float("0"), float("0.255648"), float("0"), float("0"), float("0"), float("0"), float("0.341427"), float("0"), float("0.419982"), float("0"), float("0"), float("0"), float("0"), float("0.583954"), float("0.323614"), float("0"), float("0"), float("0"), float("0.0932906"), float("0.447681"), float("0.0565188"), float("0"), float("0"), float("0.500265"), float("0"), float("0"), float("0"), float("0.159843"), float("0"), float("0.486106"), float("0.369623"), float("0"), float("0.305481"), float("0"), float("0.0430505"), float("0.0639285"), float("0.224091"), float("0.477095"), float("0"), float("0.0405224"), float("0"), float("0.479566"), float("0.1905"), float("0.036"), float("0.163203"), float("0"), float("0"), float("0.470672"), float("0.133333"), float("0.27417"), float("0"), float("0"), float("0"), float("0.188109"), float("0.133944"), float("0"), float("0.242868"), float("0.301881"), float("0.0505282"), float("0"), float("0.27581"), float("0"), float("0"), float("0.284024"), float("0"), float("0.204629"), float("0"), float("0"), float("0"), float("0"), float("0.282305"), float("0"), float("0.329989"), float("0"), float("0"), float("0"), float("0"), float("0.464116"), float("0.26113"), float("0"), float("0"), float("0"), float("0.0716705"), float("0.358155"), float("0.0510292"), float("0"), float("0"), float("0.479245"), float("0"), float("0"), float("0"), float("0.140753"), float("0"), float("0.474029"), float("0.349131"), float("0"), float("0.276741"), float("0"), float("0.0422359"), float("0.0659246"), float("0.221283"), float("0.448727"), float("0"), float("0.0395124"), float("0"), float("0.462764"), float("0.176849"), float("0.0342874"), float("0.155229"), float("0"), float("0"), float("0.429719"), float("0.123715"), float("0.266542"), float("0"), float("0"), float("0.0126346"), float("0"), float("0"), float("0.112996"), float("0"), float("0"), float("0"), float("0.0360254"), float("0"), float("0.113713"), float("0.0783622"), float("0"), float("0.07168"), float("0"), float("0.0110029"), float("0.0184994"), float("0.052733"), float("0.110654"), float("0"), float("0.0170654"), float("0"), float("0.110596"), float("0.0359603"), float("0.0123128"), float("0.042584"), float("0"), float("0"), float("0.109173"), float("0.0314731"), float("0.061102"), float("0"), float("0")] - - -class PirProgram_example_input_tensor_meta_3560422295928009713: - program_id = 692320333968606381 - input_name = "linear_0.w_0" - shape = [96, 32] - data = None - - -class PirProgram_example_input_tensor_meta_6660463665478598428: - program_id = 692320333968606381 - input_name = "linear_0.b_0" - shape = [32] - data = [float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0")] - - -class PirProgram_example_input_tensor_meta_4787673921159221127: - program_id = 692320333968606381 - input_name = "args_0" - shape = [16, 2, 96] - data = None - - -class PirProgram_example_input_tensor_meta_4218138602689122281: - program_id = 5426476789613677471 - input_name = "linear_1.b_0" - shape = [16] - data = None - - -class PirProgram_example_input_tensor_meta_8039520472926701889: - program_id = 5426476789613677471 - input_name = "linear_1.w_0" - shape = [32, 16] - data = None - - -class PirProgram_example_input_tensor_meta_3987890563716532376: - program_id = 5426476789613677471 - input_name = "linear_0.b_0" - shape = [32] - data = None - - -class PirProgram_example_input_tensor_meta_100547460941153657: - program_id = 5426476789613677471 - input_name = "linear_0.w_0" - shape = [96, 32] - data = None - - -class PirProgram_example_input_tensor_meta_3242029616568894092: - program_id = 5426476789613677471 - input_name = "args_0" - shape = [2, 2, 96] - data = [float("-1.15597"), float("-1.1523"), float("-1.15171"), float("-1.14702"), float("-1.14445"), float("-1.14483"), float("-1.14359"), float("-1.14577"), float("-1.14909"), float("-1.14868"), float("-1.14244"), float("-1.14465"), float("-1.13813"), float("-1.14016"), float("-1.13834"), float("-1.12726"), float("-1.11024"), float("-1.10905"), float("-1.10871"), float("-1.10095"), float("-1.09833"), float("-1.09615"), float("-1.09124"), float("-1.09103"), float("-1.09178"), float("-1.09125"), float("-1.09142"), float("-1.09758"), float("-1.09483"), float("-1.09004"), float("-1.0888"), float("-1.08218"), float("-1.08141"), float("-1.07451"), float("-1.07159"), float("-1.07065"), float("-1.06963"), float("-1.06884"), float("-1.06653"), float("-1.06595"), float("-1.06269"), float("-1.06049"), float("-1.05744"), float("-1.06124"), float("-1.05722"), float("-1.05479"), float("-1.05422"), float("-1.05193"), float("-1.0509"), float("-1.05028"), float("-1.04654"), float("-1.045"), float("-1.046"), float("-1.04667"), float("-1.04904"), float("-1.04042"), float("-1.04003"), float("-1.04359"), float("-1.03823"), float("-1.03557"), float("-1.03127"), float("-1.03583"), float("-1.0344"), float("-1.03144"), float("-1.02952"), float("-1.02593"), float("-1.02701"), float("-1.02957"), float("-1.02561"), float("-1.02673"), float("-1.0235"), float("-1.02171"), float("-1.02023"), float("-1.01943"), float("-1.01996"), float("-1.0124"), float("-1.01049"), float("-1.01181"), float("-1.01104"), float("-1.0073"), float("-1.0079"), float("-1.00004"), float("-0.996691"), float("-0.993982"), float("-0.991675"), float("-0.990566"), float("-0.988743"), float("-0.988268"), float("-0.984146"), float("-0.982912"), float("-0.987096"), float("-0.984222"), float("-0.963696"), float("-0.952533"), float("-0.954977"), float("-0.950661"), float("-0.399858"), float("-0.398499"), float("-0.392917"), float("-0.396343"), float("-0.391605"), float("-0.386859"), float("-0.391783"), float("-0.393174"), float("-0.396282"), float("-0.395804"), float("-0.396976"), float("-0.398565"), float("-0.395854"), float("-0.393964"), float("-0.393872"), float("-0.39339"), float("-0.394667"), float("-0.394263"), float("-0.395912"), float("-0.39461"), float("-0.392164"), float("-0.391116"), float("-0.393774"), float("-0.391869"), float("-0.392982"), float("-0.390163"), float("-0.392436"), float("-0.393987"), float("-0.394744"), float("-0.394476"), float("-0.391628"), float("-0.389037"), float("-0.387299"), float("-0.382867"), float("-0.392128"), float("-0.39309"), float("-0.3956"), float("-0.392959"), float("-0.386446"), float("-0.381103"), float("-0.380557"), float("-0.385733"), float("-0.382358"), float("-0.382535"), float("-0.388842"), float("-0.3847"), float("-0.386946"), float("-0.391374"), float("-0.390959"), float("-0.386347"), float("-0.391026"), float("-0.390428"), float("-0.386892"), float("-0.388793"), float("-0.392527"), float("-0.390458"), float("-0.389178"), float("-0.38852"), float("-0.389764"), float("-0.389545"), float("-0.384279"), float("-0.383312"), float("-0.379087"), float("-0.380502"), float("-0.381453"), float("-0.38131"), float("-0.386445"), float("-0.387503"), float("-0.385849"), float("-0.391521"), float("-0.39044"), float("-0.391326"), float("-0.390687"), float("-0.389549"), float("-0.387867"), float("-0.384638"), float("-0.388114"), float("-0.389713"), float("-0.386914"), float("-0.383812"), float("-0.38576"), float("-0.382273"), float("-0.381319"), float("-0.383497"), float("-0.378321"), float("-0.382726"), float("-0.3803"), float("-0.38556"), float("-0.378433"), float("-0.378988"), float("-0.380473"), float("-0.383853"), float("-0.381981"), float("-0.383138"), float("-0.38374"), float("-0.379838"), float("-0.630672"), float("-0.632913"), float("-0.633336"), float("-0.636795"), float("-0.637253"), float("-0.635754"), float("-0.636042"), float("-0.637148"), float("-0.640553"), float("-0.641178"), float("-0.639793"), float("-0.642557"), float("-0.642459"), float("-0.647011"), float("-0.646651"), float("-0.643922"), float("-0.64385"), float("-0.643057"), float("-0.642857"), float("-0.643463"), float("-0.644947"), float("-0.645577"), float("-0.645093"), float("-0.645523"), float("-0.645906"), float("-0.647578"), float("-0.646427"), float("-0.646111"), float("-0.649196"), float("-0.65109"), float("-0.649286"), float("-0.648321"), float("-0.64727"), float("-0.649662"), float("-0.646437"), float("-0.650501"), float("-0.648313"), float("-0.652139"), float("-0.652555"), float("-0.652121"), float("-0.651807"), float("-0.652984"), float("-0.652824"), float("-0.654689"), float("-0.653612"), float("-0.653592"), float("-0.653823"), float("-0.653214"), float("-0.653469"), float("-0.651772"), float("-0.650067"), float("-0.65063"), float("-0.652344"), float("-0.653487"), float("-0.648193"), float("-0.650301"), float("-0.649187"), float("-0.650017"), float("-0.648082"), float("-0.648181"), float("-0.647695"), float("-0.648911"), float("-0.649349"), float("-0.651401"), float("-0.654608"), float("-0.651767"), float("-0.652217"), float("-0.646641"), float("-0.649977"), float("-0.651891"), float("-0.653053"), float("-0.65444"), float("-0.653964"), float("-0.657094"), float("-0.657665"), float("-0.658948"), float("-0.662083"), float("-0.663417"), float("-0.665791"), float("-0.663225"), float("-0.659105"), float("-0.659393"), float("-0.659387"), float("-0.661882"), float("-0.663708"), float("-0.663021"), float("-0.66179"), float("-0.652634"), float("-0.652383"), float("-0.651621"), float("-0.65448"), float("-0.654242"), float("-0.656716"), float("-0.657377"), float("-0.65797"), float("-0.655658"), float("-0.00436456"), float("-0.00475825"), float("-0.00663337"), float("-0.00523404"), float("-0.00562955"), float("-0.00631295"), float("-0.00424731"), float("-0.00233632"), float("0.00070127"), float("0.00211113"), float("0.000545739"), float("0.00642778"), float("0.0038039"), float("0.000308868"), float("-0.00136501"), float("-0.00140553"), float("-0.00126356"), float("-0.00223812"), float("-0.00588529"), float("-0.00441997"), float("-0.00579612"), float("-0.00591071"), float("-0.00374736"), float("-0.0039841"), float("0.00174819"), float("0.00298575"), float("0.00173817"), float("0.00255231"), float("0.000368191"), float("0.000318813"), float("-0.00178881"), float("-0.00274283"), float("0.000628811"), float("-0.00182678"), float("-7.61417e-05"), float("-0.000906536"), float("-0.00333968"), float("-0.00408479"), float("-0.00325562"), float("-0.00334317"), float("-0.00397552"), float("-0.00359293"), float("-0.0032651"), float("0.00372263"), float("0.00179851"), float("0.00113003"), float("0.00107969"), float("0.000824071"), float("0.00230663"), float("0.00521457"), float("0.00676399"), float("0.00760787"), float("0.00433249"), float("0.0035608"), float("0.000554484"), float("0.00362084"), float("-0.00145232"), float("-0.00354868"), float("-0.00290337"), float("-0.00352887"), float("-0.00181459"), float("-0.00102524"), float("-0.00380784"), float("-0.00352733"), float("-0.00369791"), float("-0.00225879"), float("-0.00303406"), float("0.00243163"), float("0.00128276"), float("0.000701387"), float("0.00297557"), float("0.00318256"), float("0.00367343"), float("0.00317448"), float("0.00563999"), float("0.00400331"), float("-0.00251565"), float("0.00151779"), float("-0.00175092"), float("-0.00389494"), float("-0.0005398"), float("-0.000404088"), float("-0.00223803"), float("-0.00250093"), float("-0.0025293"), float("-0.00180433"), float("0.000374747"), float("0.00324137"), float("0.00465033"), float("0.00285781"), float("0.0061246"), float("0.00211838"), float("0.000508647"), float("-0.00179661"), float("-0.00133032"), float("-0.000782828")] - - -class PirProgram_example_input_tensor_meta_3980394430064143036: - program_id = 7028096434133672773 - input_name = "linear_3.b_0" - shape = [96] - data = None - - -class PirProgram_example_input_tensor_meta_3054776355284465051: - program_id = 7028096434133672773 - input_name = "linear_3.w_0" - shape = [32, 96] - data = None - - -class PirProgram_example_input_tensor_meta_6125757611404651348: - program_id = 7028096434133672773 - input_name = "linear_2.b_0" - shape = [32] - data = None - - -class PirProgram_example_input_tensor_meta_5534661246454474549: - program_id = 7028096434133672773 - input_name = "linear_2.w_0" - shape = [16, 32] - data = None - - -class PirProgram_example_input_tensor_meta_4614797390250714793: - program_id = 7028096434133672773 - input_name = "args_0" - shape = [2, 2, 16] - data = [float("0.63837"), float("-0.484176"), float("1.37562"), float("2.11047"), float("0.00551162"), float("2.62531"), float("-0.372032"), float("2.03205"), float("3.52888"), float("0.886358"), float("-0.225955"), float("-0.324597"), float("0.802664"), float("-2.20063"), float("1.28291"), float("-5.03928"), float("0.139641"), float("0.108348"), float("0.579502"), float("0.525852"), float("0.11746"), float("0.674907"), float("-0.286067"), float("0.566479"), float("0.799654"), float("0.31314"), float("0.146439"), float("-0.275929"), float("0.237963"), float("-0.39452"), float("0.0206265"), float("-1.22811"), float("-0.356625"), float("0.240933"), float("0.447273"), float("0.99085"), float("0.022031"), float("1.57203"), float("-0.341862"), float("0.827176"), float("1.06617"), float("0.33467"), float("0.0470244"), float("-0.0600742"), float("0.227221"), float("-0.507455"), float("-0.00214139"), float("-1.91961"), float("0.133666"), float("0.00946639"), float("0.0763022"), float("-0.017242"), float("0.0861289"), float("0.00317063"), float("0.054149"), float("0.141779"), float("-0.0106162"), float("0.142801"), float("-0.0334722"), float("-0.301938"), float("0.0549173"), float("0.0499354"), float("-0.144601"), float("0.111016")] - - -class PirProgram_example_input_tensor_meta_3371184329773483952: - program_id = 3000443918524221177 - input_name = "middle_5" - shape = [2, 2, 96] - data = None - - -class PirProgram_example_input_tensor_meta_1190819320963532881: - program_id = 3000443918524221177 - input_name = "middle_4" - shape = [2, 2, 32] - data = None - - -class PirProgram_example_input_tensor_meta_6657326147953639809: - program_id = 3000443918524221177 - input_name = "output_grad_0" - shape = [2, 2, 96] - data = [float("0.000443692"), float("0.000333813"), float("0.000337164"), float("0.000376996"), float("0.00031588"), float("0.000140307"), float("0.000155425"), float("0.000254817"), float("0.000186923"), float("0.000238763"), float("0.000174804"), float("0.000121806"), float("0.000140712"), float("0.000191536"), float("0.000127365"), float("-4.77346e-05"), float("-6.26668e-05"), float("-2.02892e-05"), float("1.12926e-05"), float("-9.52116e-05"), float("-8.16081e-05"), float("-3.6167e-05"), float("-3.85133e-06"), float("-2.60981e-05"), float("-4.54684e-05"), float("-4.69554e-05"), float("-0.000101383"), float("-5.73186e-05"), float("-4.71144e-05"), float("-0.000123006"), float("-0.000133709"), float("-0.000119086"), float("-0.000149415"), float("-0.000197483"), float("-0.000172614"), float("-0.000186745"), float("-0.000191804"), float("-0.000183598"), float("-0.000241256"), float("-0.000192098"), float("-0.000170965"), float("-0.000176673"), float("-0.000198307"), float("-0.000156611"), float("-0.000142045"), float("-0.000161603"), float("-0.000160694"), float("-0.000142266"), float("-0.000155407"), float("-0.000171961"), float("-0.000133331"), float("-0.000150841"), float("-0.000125645"), float("-4.11465e-05"), float("1.53606e-05"), float("-6.93277e-05"), float("-2.15477e-05"), float("2.44963e-05"), float("-3.18382e-05"), float("-2.37642e-05"), float("9.24679e-06"), float("4.21846e-05"), float("3.36176e-05"), float("2.54965e-05"), float("2.47645e-05"), float("0.000108346"), float("0.000108482"), float("0.000101298"), float("0.000141421"), float("0.000167558"), float("0.000157186"), float("0.000138657"), float("0.000193282"), float("0.000220386"), float("0.000193898"), float("0.000190735"), float("0.000211619"), float("0.000230261"), float("0.00021528"), float("0.000228187"), float("0.000302432"), float("0.000226292"), float("0.000244117"), float("0.000253747"), float("0.000253201"), float("0.000315778"), float("0.000286737"), float("0.000284055"), float("0.000284057"), float("0.000300978"), float("0.000344142"), float("0.000348519"), float("0.000208636"), float("0.000151769"), float("0.000212618"), float("0.000182387"), float("4.25186e-05"), float("1.55146e-05"), float("-1.29299e-06"), float("2.55707e-05"), float("-8.0474e-06"), float("-5.08373e-05"), float("-4.35299e-05"), float("4.43465e-06"), float("4.81199e-06"), float("1.61176e-05"), float("8.6034e-06"), float("2.54492e-05"), float("2.2127e-05"), float("1.95685e-05"), float("2.03618e-05"), float("-5.84188e-06"), float("2.20123e-05"), float("4.80387e-05"), float("6.19176e-05"), float("4.55705e-05"), float("6.3324e-05"), float("6.17111e-05"), float("0.000102875"), float("0.00010095"), float("9.11544e-05"), float("8.62715e-05"), float("8.6665e-05"), float("0.000107648"), float("0.000111767"), float("0.00010453"), float("9.53272e-05"), float("9.07721e-05"), float("8.25583e-05"), float("5.29328e-05"), float("0.000105609"), float("0.000115843"), float("0.000136405"), float("0.000130128"), float("9.2449e-05"), float("7.78238e-05"), float("9.47646e-05"), float("0.000133679"), float("0.000109147"), float("0.000113715"), float("0.000171358"), float("0.00016322"), float("0.000161961"), float("0.000186543"), float("0.00020642"), float("0.000175628"), float("0.000196759"), float("0.00020202"), float("0.000188262"), float("0.000195771"), float("0.00022144"), float("0.000212061"), float("0.000197066"), float("0.000193823"), float("0.000188813"), float("0.000189964"), float("0.000145314"), float("0.000119696"), float("0.000107136"), float("9.4593e-05"), float("9.97707e-05"), float("9.35298e-05"), float("9.53252e-05"), float("8.86535e-05"), float("9.87882e-05"), float("0.000110338"), float("8.10272e-05"), float("6.78147e-05"), float("5.33696e-05"), float("5.26765e-05"), float("2.62262e-05"), float("-1.23077e-05"), float("2.29062e-05"), float("1.04001e-05"), float("-2.91967e-05"), float("-4.2393e-05"), float("-3.54717e-05"), float("-9.28945e-05"), float("-0.000100431"), float("-7.46758e-05"), float("-0.000122817"), float("-0.000109041"), float("-0.000119485"), float("-0.000103781"), float("-0.000171119"), float("-0.000160994"), float("-0.000149117"), float("-0.00013268"), float("-0.000159193"), float("-0.00015496"), float("-0.000140225"), float("-0.000158241"), float("-0.000762324"), float("-0.000767115"), float("-0.000753975"), float("-0.00072348"), float("-0.000748033"), float("-0.000816843"), float("-0.000808487"), float("-0.000739376"), float("-0.000756383"), float("-0.000731987"), float("-0.000758532"), float("-0.000756136"), float("-0.000741849"), float("-0.000682354"), float("-0.000714943"), float("-0.000763774"), float("-0.000742451"), float("-0.000723821"), float("-0.000704298"), float("-0.0007232"), float("-0.000703063"), float("-0.000678821"), float("-0.000624749"), float("-0.000675047"), float("-0.000674623"), float("-0.000625239"), float("-0.000637009"), float("-0.000641791"), float("-0.000637589"), float("-0.000608231"), float("-0.000597854"), float("-0.000589748"), float("-0.000600376"), float("-0.000582922"), float("-0.000588446"), float("-0.000555943"), float("-0.000551712"), float("-0.000508178"), float("-0.000500373"), float("-0.00049847"), float("-0.000467944"), float("-0.000431473"), float("-0.000442191"), float("-0.000412144"), float("-0.000376516"), float("-0.000374266"), float("-0.000364806"), float("-0.000311562"), float("-0.000328529"), float("-0.000328796"), float("-0.000263592"), float("-0.000287732"), float("-0.000251361"), float("-0.000222262"), float("-0.000213702"), float("-0.00019776"), float("-0.000168036"), float("-0.000169818"), float("-0.000179459"), float("-0.00015092"), float("-0.000140475"), float("-0.000135543"), float("-0.000105994"), float("-9.4315e-05"), float("-6.30614e-05"), float("-2.45928e-05"), float("-1.47304e-05"), float("-8.07969e-05"), float("-4.10341e-05"), float("-9.95025e-06"), float("-8.09847e-06"), float("2.15607e-05"), float("2.36972e-05"), float("4.01298e-05"), float("5.78094e-05"), float("7.7993e-05"), float("7.89144e-05"), float("0.000119707"), float("0.000113879"), float("9.84538e-05"), float("9.42778e-05"), float("6.83578e-05"), float("7.44164e-05"), float("0.000107444"), float("0.000120769"), float("8.44241e-05"), float("7.24259e-05"), float("5.70118e-05"), float("4.95479e-05"), float("3.81417e-05"), float("7.03329e-05"), float("7.46877e-05"), float("6.71179e-05"), float("7.26103e-05"), float("9.57397e-05"), float("7.91801e-05"), float("-0.000149472"), float("-0.00014213"), float("-0.000131744"), float("-0.000140043"), float("-0.000137351"), float("-0.000136615"), float("-0.000151119"), float("-0.000142876"), float("-0.000168719"), float("-0.00017359"), float("-0.000164497"), float("-0.000199011"), float("-0.000174208"), float("-0.000162659"), float("-0.000147246"), float("-0.000139938"), float("-0.000130565"), float("-0.000135166"), float("-9.48119e-05"), float("-9.42852e-05"), float("-7.41053e-05"), float("-7.15872e-05"), float("-7.1455e-05"), float("-6.04486e-05"), float("-8.1519e-05"), float("-7.22456e-05"), float("-6.37999e-05"), float("-5.91884e-05"), float("-4.43363e-05"), float("-3.85257e-05"), float("-3.43068e-06"), float("4.23698e-06"), float("-1.38437e-05"), float("6.94508e-06"), float("2.44123e-06"), float("1.97679e-05"), float("2.92677e-05"), float("3.48e-05"), float("3.98916e-05"), float("3.75445e-05"), float("4.12756e-05"), float("4.7296e-05"), float("3.69121e-05"), float("8.78256e-06"), float("2.21561e-05"), float("2.35112e-05"), float("1.53588e-05"), float("2.79886e-05"), float("1.545e-05"), float("-1.5073e-05"), float("-1.2325e-05"), float("-2.31248e-05"), float("-8.68771e-06"), float("-5.88729e-06"), float("1.19861e-05"), float("2.16062e-06"), float("3.66875e-05"), float("3.44659e-05"), float("3.93272e-05"), float("3.75808e-05"), float("2.52072e-05"), float("2.62236e-05"), float("4.71685e-05"), float("3.84434e-05"), float("3.61974e-05"), float("4.22836e-05"), float("3.47363e-05"), float("1.43028e-05"), float("2.4955e-05"), float("2.90605e-05"), float("5.53731e-06"), float("1.6158e-05"), float("5.64251e-06"), float("1.99968e-05"), float("8.72064e-06"), float("4.71728e-06"), float("3.85144e-05"), float("2.5692e-05"), float("2.56004e-05"), float("3.43408e-05"), float("2.86808e-05"), float("1.72165e-05"), float("2.5901e-05"), float("3.57338e-05"), float("3.10321e-05"), float("1.95796e-05"), float("2.35217e-05"), float("1.14026e-05"), float("-1.22701e-05"), float("6.5969e-06"), float("-1.01829e-05"), float("1.21936e-05"), float("1.92287e-05"), float("1.96028e-05"), float("2.41576e-05"), float("3.20103e-05")] - - -class PirProgram_example_input_tensor_meta_1859405807941431992: - program_id = 3000443918524221177 - input_name = "linear_3.w_0" - shape = [32, 96] - data = None - - -class PirProgram_example_input_tensor_meta_2446845994563353672: - program_id = 3000443918524221177 - input_name = "middle_2" - shape = [1] - data = [float("0.2")] - - -class PirProgram_example_input_tensor_meta_3189608938430428800: - program_id = 3000443918524221177 - input_name = "linear_3.b_0" - shape = [96] - data = [float("0.0485458"), float("0.048584"), float("0.0494509"), float("0.0514775"), float("0.0513561"), float("0.0539435"), float("0.053923"), float("0.053999"), float("0.0532589"), float("0.0548013"), float("0.0545531"), float("0.0531352"), float("0.0552883"), float("0.0543044"), float("0.0539071"), float("0.0559445"), float("0.0566455"), float("0.0558921"), float("0.0559904"), float("0.0579908"), float("0.0594786"), float("0.0587086"), float("0.0591194"), float("0.0593815"), float("0.0592324"), float("0.0608538"), float("0.060729"), float("0.0611746"), float("0.0610663"), float("0.0603559"), float("0.0624431"), float("0.0632785"), float("0.0626797"), float("0.061953"), float("0.0608844"), float("0.061694"), float("0.0620486"), float("0.0616028"), float("0.0614813"), float("0.0607751"), float("0.0605967"), float("0.0610775"), float("0.06023"), float("0.0599733"), float("0.0603637"), float("0.0607957"), float("0.0595258"), float("0.0607369"), float("0.0593041"), float("0.0590547"), float("0.0602343"), float("0.05974"), float("0.0587402"), float("0.0584157"), float("0.059519"), float("0.0591371"), float("0.0598814"), float("0.0598889"), float("0.0592714"), float("0.0588438"), float("0.0606669"), float("0.0606387"), float("0.060385"), float("0.062044"), float("0.0617631"), float("0.0626737"), float("0.0635835"), float("0.063831"), float("0.0633426"), float("0.0645196"), float("0.0646552"), float("0.0664435"), float("0.0664267"), float("0.0671475"), float("0.0679168"), float("0.0685048"), float("0.0680123"), float("0.0694771"), float("0.0680605"), float("0.0687388"), float("0.0703835"), float("0.0694004"), float("0.0691108"), float("0.0696977"), float("0.0700212"), float("0.0698567"), float("0.0703321"), float("0.0709236"), float("0.0696366"), float("0.0705553"), float("0.0702906"), float("0.0700228"), float("0.0695461"), float("0.0689861"), float("0.0691985"), float("0.0688401")] - - -class PirProgram_example_input_tensor_meta_4224543760103808494: - program_id = 3000443918524221177 - input_name = "middle_3" - shape = [2, 2, 32] - data = [float("0.942804"), float("1.70377"), float("1.09786"), float("1.38926"), float("0"), float("0"), float("0"), float("0"), float("1.07601"), float("0"), float("1.56775"), float("1.42995"), float("1.9211"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.443868"), float("0"), float("0"), float("2.23675"), float("0"), float("0"), float("1.42107"), float("0"), float("0.289613"), float("0.480792"), float("0.445812"), float("0.489618"), float("0.107548"), float("0.48011"), float("0"), float("0"), float("0.549541"), float("0.420605"), float("0"), float("0.585187"), float("0.59219"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.321631"), float("0"), float("0"), float("0.639618"), float("0"), float("0.0198829"), float("0.443799"), float("0"), float("0.387322"), float("0.817863"), float("0"), float("0.74025"), float("0.489511"), float("0.669735"), float("0"), float("0"), float("0.759979"), float("0.446918"), float("0.665187"), float("0.711636"), float("0.819093"), float("0.535416"), float("0.548715"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.278116"), float("0"), float("0"), float("1.01575"), float("0"), float("0"), float("0.604064"), float("0"), float("0.17553"), float("0"), float("0.174411"), float("0.232705"), float("0"), float("0.266127"), float("0.32481"), float("0.322138"), float("0"), float("0.19189"), float("0.227142"), float("0.317245"), float("0"), float("0"), float("0.264847"), float("0.188336"), float("0.249843"), float("0.277018"), float("0"), float("0.278298"), float("0.270292"), float("0.274346"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.272329"), float("0.314617"), float("0.0190948"), float("0.226759"), float("0")] - - -class PirProgram_example_input_tensor_meta_1661268218216156872: - program_id = 3000443918524221177 - input_name = "middle_1" - shape = [2, 2, 32] - data = [float("0.754243"), float("1.36301"), float("0.878288"), float("1.11141"), float("0"), float("0.754685"), float("0"), float("0"), float("0.860806"), float("0.719321"), float("1.2542"), float("1.14396"), float("1.53688"), float("0"), float("1.29534"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.355094"), float("0.157341"), float("0"), float("1.7894"), float("0"), float("0"), float("1.13686"), float("0"), float("0.23169"), float("0.384634"), float("0.35665"), float("0.391695"), float("0.0860384"), float("0.384088"), float("0"), float("0"), float("0.439633"), float("0.336484"), float("0.403318"), float("0.46815"), float("0.473752"), float("0.0790446"), float("0.443998"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.257305"), float("0.191734"), float("0"), float("0.511694"), float("0"), float("0.0159064"), float("0.355039"), float("0"), float("0.309857"), float("0.654291"), float("0.311981"), float("0.5922"), float("0.391609"), float("0.535788"), float("0"), float("0"), float("0.607983"), float("0.357534"), float("0.53215"), float("0.569309"), float("0.655275"), float("0.428333"), float("0.438972"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.222493"), float("0.474749"), float("0"), float("0.812597"), float("0"), float("0"), float("0.483251"), float("0"), float("0.140424"), float("0.155835"), float("0.139529"), float("0.186164"), float("0"), float("0.212902"), float("0.259848"), float("0.257711"), float("0.23834"), float("0.153512"), float("0.181713"), float("0.253796"), float("0.198702"), float("0"), float("0.211877"), float("0.150669"), float("0.199874"), float("0.221614"), float("0.180791"), float("0.222638"), float("0.216234"), float("0.219477"), float("0.191163"), float("0.13138"), float("0.153447"), float("0"), float("0"), float("0.217864"), float("0.251694"), float("0.0152759"), float("0.181407"), float("0.198531")] - - -class PirProgram_example_input_tensor_meta_3873387511994723854: - program_id = 3000443918524221177 - input_name = "linear_2.w_0" - shape = [16, 32] - data = [float("0.138689"), float("-0.106322"), float("0.172949"), float("-0.118539"), float("0.170557"), float("-0.0672584"), float("0.204199"), float("-0.206781"), float("0.059026"), float("0.325195"), float("0.0539278"), float("-0.0231287"), float("0.291649"), float("-0.0653538"), float("0.210411"), float("0.0745167"), float("-0.203244"), float("0.00686247"), float("0.10958"), float("0.157849"), float("-0.0889362"), float("0.298802"), float("-0.0925571"), float("-0.246793"), float("-0.11154"), float("-0.027271"), float("-0.196191"), float("0.275181"), float("0.0250688"), float("0.204275"), float("0.198187"), float("0.132799"), float("-0.285483"), float("-0.383002"), float("-0.0782729"), float("0.0404985"), float("0.165563"), float("0.206953"), float("0.1843"), float("0.173485"), float("0.129389"), float("0.375723"), float("-0.235547"), float("-0.253078"), float("0.160617"), float("0.282609"), float("-0.204999"), float("0.137476"), float("0.233969"), float("0.185019"), float("0.205812"), float("0.206644"), float("-0.142799"), float("-0.404361"), float("0.0459731"), float("0.0739177"), float("-0.249473"), float("-0.106103"), float("0.0529779"), float("0.0155141"), float("0.310476"), float("0.196231"), float("0.0647549"), float("-0.0507605"), float("-0.130941"), float("-0.0484275"), float("-0.0269265"), float("-0.18644"), float("-0.118492"), float("0.224969"), float("0.13941"), float("-0.146658"), float("0.276094"), float("-0.0997558"), float("-0.17141"), float("0.292359"), float("-0.0742287"), float("-0.164869"), float("0.268466"), float("0.304443"), float("0.228287"), float("0.0761761"), float("-0.0371925"), float("0.296199"), float("-0.191571"), float("0.247706"), float("-0.273535"), float("-0.263982"), float("0.229751"), float("-0.0848374"), float("-0.260978"), float("-0.119975"), float("-0.162848"), float("0.176309"), float("-0.187869"), float("-0.230427"), float("0.299697"), float("0.103454"), float("0.186159"), float("0.117861"), float("0.214886"), float("-0.0160362"), float("-0.191379"), float("-0.0251439"), float("0.253566"), float("-0.193173"), float("-0.0737695"), float("-0.00845166"), float("0.369554"), float("0.347756"), float("0.0289703"), float("-0.210089"), float("0.15666"), float("-0.0107514"), float("-0.205177"), float("-0.370911"), float("-0.123499"), float("-0.165401"), float("0.0389836"), float("0.0961651"), float("-0.18811"), float("0.09999"), float("-0.0234326"), float("0.195959"), float("0.196091"), float("0.293869"), float("0.119714"), float("-0.35079"), float("0.216349"), float("0.0228813"), float("-0.274141"), float("-0.255975"), float("0.247017"), float("0.0754483"), float("0.229817"), float("0.23657"), float("0.161538"), float("-0.221736"), float("0.120394"), float("0.183345"), float("-0.0264143"), float("0.210679"), float("-0.107138"), float("0.124702"), float("0.118588"), float("-0.0608178"), float("0.233959"), float("0.155236"), float("-0.206555"), float("-0.168823"), float("0.219217"), float("-0.0797468"), float("-0.289087"), float("0.274242"), float("-0.367646"), float("0.00935798"), float("0.10757"), float("-0.131774"), float("0.101516"), float("-0.233815"), float("0.142226"), float("0.0382131"), float("-0.246448"), float("-0.232159"), float("0.183271"), float("-0.0259207"), float("-0.29719"), float("-0.189024"), float("0.286622"), float("0.0378376"), float("0.0687375"), float("0.296513"), float("0.115993"), float("0.0502835"), float("0.281012"), float("0.0420163"), float("-0.0551966"), float("-0.0533368"), float("0.0877916"), float("-0.359907"), float("-0.0330066"), float("-0.264281"), float("-0.00644859"), float("-0.103175"), float("-0.0902519"), float("0.246111"), float("-0.111925"), float("0.270801"), float("-0.294941"), float("-0.147398"), float("-0.0205641"), float("0.302264"), float("0.0983701"), float("-0.216157"), float("-0.0542728"), float("-0.0155529"), float("-0.264782"), float("0.168931"), float("0.0981406"), float("-0.106492"), float("0.0882299"), float("0.216699"), float("-0.211478"), float("-0.0953006"), float("0.0748619"), float("-0.100698"), float("0.108855"), float("0.0412198"), float("-0.119328"), float("-0.172325"), float("0.122055"), float("-0.0801006"), float("-0.18746"), float("0.190164"), float("0.0690614"), float("-0.0622083"), float("0.130328"), float("-0.174397"), float("-0.2734"), float("0.0485907"), float("0.161157"), float("0.154754"), float("0.164222"), float("-0.222811"), float("-0.244355"), float("0.30284"), float("0.22183"), float("0.223936"), float("0.25338"), float("-0.0558757"), float("0.1592"), float("0.11035"), float("-0.229928"), float("-0.111174"), float("-0.202858"), float("-0.109098"), float("0.220881"), float("0.00651073"), float("0.0485816"), float("-0.107969"), float("-0.15052"), float("-0.00632081"), float("0.00609574"), float("0.0133219"), float("-0.208563"), float("0.153815"), float("0.269685"), float("0.0584512"), float("0.177424"), float("-0.0912139"), float("-0.345211"), float("-0.256602"), float("0.102258"), float("-0.01134"), float("0.335877"), float("0.0870043"), float("-0.156129"), float("-0.211225"), float("-0.231017"), float("0.0358312"), float("0.203831"), float("0.0187823"), float("-0.0363091"), float("0.213651"), float("0.0315795"), float("0.0870476"), float("0.0339045"), float("0.257777"), float("-0.128626"), float("0.142826"), float("0.0238258"), float("-0.181365"), float("-0.115817"), float("-0.216976"), float("-0.289225"), float("0.00338232"), float("-0.143151"), float("-0.301035"), float("0.0747563"), float("0.207678"), float("-0.153696"), float("-0.2239"), float("-0.345777"), float("0.244627"), float("-0.125439"), float("-0.307755"), float("0.304526"), float("-0.206704"), float("0.00549386"), float("-0.0245539"), float("-0.0432738"), float("0.291515"), float("-0.172698"), float("0.153061"), float("0.138099"), float("-0.0670694"), float("0.0995701"), float("-0.0270735"), float("0.238724"), float("0.0714559"), float("-0.0813835"), float("0.0283591"), float("0.254933"), float("-0.138964"), float("0.177769"), float("0.097657"), float("-0.237825"), float("0.228541"), float("-0.211276"), float("0.0191841"), float("-0.17721"), float("0.0404337"), float("0.101901"), float("0.153098"), float("0.174044"), float("-0.0329636"), float("0.172645"), float("-0.142116"), float("0.0631262"), float("0.181635"), float("0.162319"), float("-0.297797"), float("0.190953"), float("-0.225768"), float("0.257752"), float("-0.29906"), float("0.000686928"), float("-0.0377532"), float("-0.244496"), float("0.0995946"), float("-0.187728"), float("-0.266484"), float("-0.0836021"), float("0.185958"), float("0.186861"), float("0.310878"), float("0.0343862"), float("0.237059"), float("0.251551"), float("0.0472566"), float("-0.115398"), float("-0.383304"), float("0.101012"), float("0.0714781"), float("0.221221"), float("0.314253"), float("-0.272301"), float("-0.379694"), float("0.220204"), float("0.239134"), float("-0.0663735"), float("-0.080212"), float("-0.16315"), float("-0.00506175"), float("-0.0545347"), float("-0.0465481"), float("0.475722"), float("0.179901"), float("0.161356"), float("-0.17223"), float("-0.338577"), float("-0.0519715"), float("-0.103634"), float("-0.181719"), float("-0.0823647"), float("0.279138"), float("-0.0167731"), float("-0.0912584"), float("0.056401"), float("-0.175043"), float("-0.111419"), float("0.0304177"), float("-0.354337"), float("-0.050518"), float("-0.0543624"), float("0.0608217"), float("-0.0743314"), float("0.0759768"), float("0.236895"), float("-0.0438436"), float("-0.134808"), float("-0.0798067"), float("0.186131"), float("0.115307"), float("-0.155228"), float("0.187059"), float("0.0529251"), float("0.185473"), float("-0.274505"), float("-0.15848"), float("0.0693555"), float("0.103163"), float("-0.0561303"), float("-0.133256"), float("-0.178763"), float("0.0890734"), float("0.17019"), float("-0.163824"), float("0.0706935"), float("-0.173212"), float("0.278386"), float("0.297952"), float("0.0195602"), float("-0.269767"), float("0.0546989"), float("-0.311704"), float("-0.0584766"), float("0.145253"), float("0.0326612"), float("-0.137887"), float("0.0357246"), float("0.0743239"), float("-0.0922011"), float("-0.239939"), float("0.0901577"), float("-0.18774"), float("-0.138902"), float("0.201292"), float("-0.004765"), float("0.160184"), float("0.246065"), float("-0.170404"), float("-0.0220289"), float("0.283042"), float("0.164063"), float("-0.216808"), float("-0.125219"), float("0.0188683"), float("-0.209135"), float("0.172225"), float("-0.197942"), float("0.07705"), float("0.340536"), float("0.23112"), float("0.0997995"), float("-0.14498"), float("0.0878749"), float("-0.232691"), float("0.171985"), float("0.192874"), float("0.349805"), float("0.146311"), float("0.243025"), float("0.266444"), float("-0.0826823"), float("0.0912206"), float("-0.00912116"), float("-0.18689"), float("-0.017315"), float("0.0672263"), float("-0.0861141"), float("-0.189335"), float("-0.287294"), float("-0.287037"), float("-0.101011"), float("-0.288937"), float("-0.110381"), float("-0.372508"), float("-0.166248"), float("-0.236813"), float("-0.0719964"), float("-0.311204"), float("-0.051435"), float("-0.123605"), float("-0.138647"), float("-0.0663331"), float("-0.263123"), float("-0.137314"), float("-0.0564986"), float("-0.368283"), float("-0.135173"), float("-0.0914974"), float("0.00832917"), float("-0.175773"), float("-0.0174043"), float("-0.0604438"), float("-0.115462"), float("-0.193237"), float("-0.333979"), float("0.110844"), float("-0.094741"), float("-0.20917"), float("-0.270162"), float("-0.244399"), float("0.113335"), float("-0.113068"), float("0.237874"), float("0.317679"), float("0.0349081"), float("-0.230409"), float("-0.230016"), float("0.172636"), float("0.0211285"), float("0.00908043"), float("0.191938"), float("0.0610569"), float("0.337628"), float("0.17004"), float("0.0341012"), float("0.15482"), float("0.292289"), float("0.24799"), float("0.333809"), float("0.267258"), float("-0.277675"), float("-0.184417"), float("-0.394628"), float("-0.134093"), float("0.248116"), float("-0.180201"), float("0.0719522"), float("0.389222")] - - -class PirProgram_example_input_tensor_meta_770961235158691829: - program_id = 3000443918524221177 - input_name = "middle_0" - shape = [2, 2, 32] - data = None - - -class PirProgram_example_input_tensor_meta_3897698076100692765: - program_id = 3000443918524221177 - input_name = "linear_2.b_0" - shape = [32] - data = [float("0.127184"), float("0.148213"), float("0.126821"), float("0.128469"), float("-0.0799158"), float("0.205759"), float("0.155048"), float("0.126542"), float("0.0762877"), float("0.116064"), float("0.164428"), float("0.110229"), float("0.1077"), float("-0.069753"), float("0.103557"), float("0.0965401"), float("0.129614"), float("0.111043"), float("0.0933711"), float("0.123984"), float("0.169367"), float("0.0588203"), float("0.114893"), float("0.144424"), float("0.124464"), float("-0.0191884"), float("-0.133152"), float("0.195669"), float("0.13529"), float("-0.0156079"), float("0.0929306"), float("0.208163")] - - -class PirProgram_example_input_tensor_meta_5160725927315730163: - program_id = 3000443918524221177 - input_name = "args_0" - shape = [2, 2, 16] - data = [float("0.63837"), float("-0.484176"), float("1.37562"), float("2.11047"), float("0.00551162"), float("2.62531"), float("-0.372032"), float("2.03205"), float("3.52888"), float("0.886358"), float("-0.225955"), float("-0.324597"), float("0.802664"), float("-2.20063"), float("1.28291"), float("-5.03928"), float("0.139641"), float("0.108348"), float("0.579502"), float("0.525852"), float("0.11746"), float("0.674907"), float("-0.286067"), float("0.566479"), float("0.799654"), float("0.31314"), float("0.146439"), float("-0.275929"), float("0.237963"), float("-0.39452"), float("0.0206265"), float("-1.22811"), float("-0.356625"), float("0.240933"), float("0.447273"), float("0.99085"), float("0.022031"), float("1.57203"), float("-0.341862"), float("0.827176"), float("1.06617"), float("0.33467"), float("0.0470244"), float("-0.0600742"), float("0.227221"), float("-0.507455"), float("-0.00214139"), float("-1.91961"), float("0.133666"), float("0.00946639"), float("0.0763022"), float("-0.017242"), float("0.0861289"), float("0.00317063"), float("0.054149"), float("0.141779"), float("-0.0106162"), float("0.142801"), float("-0.0334722"), float("-0.301938"), float("0.0549173"), float("0.0499354"), float("-0.144601"), float("0.111016")] - - -class PirProgram_example_input_tensor_meta_2653122763850172560: - program_id = 3111236241086672326 - input_name = "output_grad_0" - shape = [2, 2, 16] - data = [float("-0.000259765"), float("0.000502711"), float("-3.34286e-05"), float("-0.000308996"), float("0.000766827"), float("0.00031456"), float("5.42812e-05"), float("-0.000930537"), float("0.000139215"), float("-0.000372329"), float("0.000279334"), float("0.00021588"), float("-0.00051067"), float("-0.00061697"), float("0.000621474"), float("0.000900432"), float("-0.000251851"), float("0.000434745"), float("-1.13758e-05"), float("-0.000917027"), float("-0.000684694"), float("-0.000777882"), float("7.10025e-05"), float("-5.16687e-05"), float("-0.000292353"), float("-0.000278136"), float("0.000741509"), float("8.23935e-05"), float("-2.14872e-05"), float("-0.000322378"), float("0.000605169"), float("0.000113745"), float("0.00375845"), float("0.00106431"), float("-0.000995514"), float("0.00707899"), float("0.00362036"), float("0.00517876"), float("-0.000870925"), float("0.00109624"), float("0.00352798"), float("0.00302349"), float("-0.00333248"), float("0.0019234"), float("-0.00183069"), float("0.00179234"), float("-0.00927467"), float("-0.00196363"), float("0.00013415"), float("-0.000518328"), float("-0.000200835"), float("-0.000149648"), float("-0.000169967"), float("0.000350789"), float("9.03782e-05"), float("0.000154067"), float("0.000613054"), float("0.000406082"), float("-0.000935107"), float("0.00018923"), float("0.000132097"), float("-0.000361182"), float("-0.000315851"), float("-0.000232278")] - - -class PirProgram_example_input_tensor_meta_9055065804403674970: - program_id = 3111236241086672326 - input_name = "middle_5" - shape = [2, 2, 16] - data = None - - -class PirProgram_example_input_tensor_meta_7382231132507793619: - program_id = 3111236241086672326 - input_name = "middle_4" - shape = [2, 2, 32] - data = None - - -class PirProgram_example_input_tensor_meta_3182402915473364379: - program_id = 3111236241086672326 - input_name = "middle_2" - shape = [1] - data = [float("0.2")] - - -class PirProgram_example_input_tensor_meta_1974672370162666358: - program_id = 3111236241086672326 - input_name = "middle_0" - shape = [2, 2, 32] - data = None - - -class PirProgram_example_input_tensor_meta_7940412681263697295: - program_id = 3111236241086672326 - input_name = "linear_1.w_0" - shape = [32, 16] - data = [float("0.235596"), float("-0.166064"), float("-0.00841034"), float("0.0251242"), float("0.121917"), float("0.0115089"), float("0.293106"), float("0.189091"), float("0.228253"), float("0.0750035"), float("-0.0840923"), float("0.0157098"), float("0.0583272"), float("-0.113309"), float("0.256736"), float("-0.266579"), float("0.0569002"), float("-0.148237"), float("0.168182"), float("0.0465327"), float("-0.104699"), float("0.100044"), float("-0.136119"), float("0.217071"), float("0.250393"), float("0.0458567"), float("-0.165222"), float("-0.247985"), float("-0.0402208"), float("-0.265832"), float("-0.00288254"), float("0.0103337"), float("0.221503"), float("0.0711672"), float("-0.0150802"), float("-0.153977"), float("0.232742"), float("0.0707905"), float("0.109877"), float("0.0475617"), float("-0.0662123"), float("0.242932"), float("-0.290769"), float("-0.0773183"), float("0.211586"), float("-0.140735"), float("0.143677"), float("-0.318947"), float("0.309584"), float("0.155288"), float("-0.142704"), float("-0.208616"), float("-0.161949"), float("-0.225615"), float("-0.119309"), float("-0.0151871"), float("0.243764"), float("-0.125401"), float("0.334346"), float("0.0876797"), float("-0.12878"), float("0.155844"), float("0.25528"), float("0.0449498"), float("-0.109217"), float("0.361007"), float("0.0548769"), float("-0.225299"), float("0.129158"), float("0.182534"), float("0.111121"), float("-0.170268"), float("0.26808"), float("0.0713453"), float("0.0276709"), float("0.0309021"), float("0.222045"), float("0.0556532"), float("0.225857"), float("0.165388"), float("-0.214861"), float("-0.189104"), float("0.0905981"), float("0.164752"), float("-0.0745658"), float("0.17254"), float("0.230335"), float("0.172994"), float("-0.0394944"), float("0.162471"), float("-0.178272"), float("-0.121411"), float("0.119618"), float("-0.197991"), float("-0.0220652"), float("-0.0271392"), float("0.0401553"), float("-0.0364811"), float("-0.180565"), float("-0.149028"), float("0.336991"), float("0.0416623"), float("-0.233224"), float("-0.080196"), float("0.00397957"), float("0.0222934"), float("-0.033727"), float("-0.151176"), float("0.0296938"), float("-0.174967"), float("0.0433006"), float("0.146219"), float("-0.295726"), float("-0.200405"), float("0.292424"), float("0.105805"), float("-0.278996"), float("-0.0612777"), float("-0.14825"), float("0.00739621"), float("-0.119451"), float("-0.253996"), float("0.501842"), float("-0.138736"), float("-0.055433"), float("-0.109043"), float("-0.210503"), float("0.283465"), float("0.218386"), float("-0.230348"), float("0.266489"), float("-0.0496932"), float("-0.214893"), float("-0.0331828"), float("-0.199718"), float("-0.0481673"), float("0.382266"), float("-0.000796923"), float("0.0316693"), float("-0.00864441"), float("0.260415"), float("-0.304373"), float("-0.076894"), float("-0.239595"), float("-0.0672012"), float("0.319481"), float("0.0133308"), float("-0.206744"), float("-0.268379"), float("-0.218137"), float("-0.110468"), float("0.120406"), float("0.0381229"), float("-0.171569"), float("0.281952"), float("-0.130553"), float("-0.0789959"), float("-0.175866"), float("-0.148795"), float("-0.123259"), float("-0.18909"), float("0.127911"), float("0.152386"), float("0.112031"), float("0.196341"), float("-0.264504"), float("-0.152368"), float("-0.0758584"), float("-0.202836"), float("-0.304334"), float("0.174291"), float("-0.0152358"), float("0.200058"), float("-0.284122"), float("0.376094"), float("0.310537"), float("-0.116116"), float("-0.245531"), float("-0.0476175"), float("0.201079"), float("-0.16955"), float("0.134718"), float("0.328129"), float("-0.201016"), float("-0.122684"), float("0.0527725"), float("0.014955"), float("-0.0159163"), float("-0.024892"), float("-0.172767"), float("0.00895791"), float("-0.126559"), float("0.017916"), float("-0.0624295"), float("0.248658"), float("0.104764"), float("-0.0995163"), float("0.021314"), float("-0.00632193"), float("-0.0202286"), float("0.0669071"), float("-0.0343595"), float("-0.0058106"), float("-0.0319011"), float("0.211846"), float("-0.143065"), float("0.00494012"), float("-0.178171"), float("-0.142023"), float("0.236462"), float("0.0113232"), float("0.203394"), float("0.213219"), float("0.287939"), float("-0.266193"), float("0.145776"), float("0.277013"), float("0.0349097"), float("0.339889"), float("0.257273"), float("-0.031818"), float("0.106094"), float("-0.016972"), float("-0.341205"), float("0.195945"), float("-0.273676"), float("-0.0920491"), float("-0.00215972"), float("0.0761255"), float("-0.117919"), float("0.292542"), float("-0.0279444"), float("-0.138423"), float("0.120147"), float("0.190762"), float("0.0532155"), float("-0.00462688"), float("0.29658"), float("0.264242"), float("0.279836"), float("-0.0740407"), float("-0.258875"), float("0.204921"), float("0.0259398"), float("-0.0169675"), float("0.074128"), float("-0.164556"), float("-0.0276452"), float("0.0226643"), float("0.0141938"), float("0.249415"), float("0.0347015"), float("-0.165602"), float("0.077402"), float("-0.226793"), float("0.244763"), float("-0.305538"), float("0.167513"), float("-0.026212"), float("-0.0884336"), float("0.323809"), float("0.000747159"), float("0.308381"), float("-0.199611"), float("-0.193336"), float("-0.144533"), float("-0.186432"), float("0.208955"), float("0.0886495"), float("-0.191583"), float("0.349786"), float("0.35748"), float("-0.0242003"), float("0.0790487"), float("-0.0257431"), float("-0.347474"), float("-0.154624"), float("-0.18064"), float("0.292404"), float("-0.0464253"), float("0.238564"), float("-0.238146"), float("0.0556705"), float("0.0297406"), float("0.167295"), float("0.196666"), float("0.233435"), float("0.119447"), float("-0.0764451"), float("0.305884"), float("-0.210522"), float("-0.0604975"), float("-0.101583"), float("-0.0236538"), float("-0.235309"), float("-0.0823719"), float("-0.248695"), float("0.0203888"), float("-0.178912"), float("0.120222"), float("-0.123081"), float("0.0837287"), float("0.264862"), float("-0.137024"), float("0.18452"), float("0.117199"), float("-0.0519129"), float("-0.0791504"), float("-0.250732"), float("0.0108192"), float("-0.0176924"), float("0.00619382"), float("-0.14818"), float("0.0474773"), float("-0.218766"), float("-0.152883"), float("-0.108489"), float("0.0400477"), float("0.0824107"), float("-0.326851"), float("-0.167915"), float("-0.206116"), float("-0.070031"), float("0.186761"), float("-0.299896"), float("0.256832"), float("0.0676038"), float("0.0782485"), float("0.191245"), float("-0.187385"), float("-0.239842"), float("-0.0772858"), float("-0.153211"), float("-0.220496"), float("0.0923377"), float("-0.181751"), float("-0.122754"), float("0.0135214"), float("0.0596246"), float("-0.241126"), float("0.284071"), float("-0.177228"), float("-0.0907132"), float("0.138534"), float("0.194014"), float("0.157129"), float("-0.215676"), float("-0.0922354"), float("0.246938"), float("-0.123517"), float("-0.250844"), float("0.0446277"), float("0.176366"), float("-0.0482041"), float("0.116668"), float("0.210133"), float("-0.0399605"), float("0.0173124"), float("-0.00336163"), float("0.111738"), float("0.267352"), float("0.0239314"), float("-0.12188"), float("-0.200225"), float("-0.0374264"), float("-0.223423"), float("0.28684"), float("-0.316664"), float("-0.0740798"), float("0.101619"), float("-0.15827"), float("-0.242475"), float("0.116453"), float("0.23239"), float("0.0285625"), float("0.0559415"), float("-0.111179"), float("-0.187168"), float("-0.0821701"), float("-0.100173"), float("-0.122167"), float("0.0214796"), float("0.275201"), float("0.326983"), float("-0.157285"), float("0.308136"), float("-0.231156"), float("-0.303319"), float("0.229308"), float("0.0719604"), float("0.0252571"), float("-0.137748"), float("0.196749"), float("0.0238845"), float("0.0721256"), float("-0.111398"), float("0.193114"), float("-0.0786931"), float("0.346398"), float("0.0366481"), float("-0.0774674"), float("-0.0469274"), float("0.0359128"), float("-0.0288154"), float("-0.175977"), float("0.22237"), float("0.186215"), float("-0.00687118"), float("0.151105"), float("-0.156596"), float("-0.342589"), float("-0.0769077"), float("0.140154"), float("-0.0963023"), float("-0.152346"), float("-0.00415342"), float("-0.203184"), float("-0.07618"), float("0.142056"), float("-0.0463533"), float("0.000432739"), float("0.113857"), float("0.239217"), float("0.183889"), float("-0.148585"), float("-0.0760731"), float("-0.0540795"), float("-0.137623"), float("-0.24345"), float("-0.0384814"), float("-0.0854952"), float("-0.151667"), float("0.00554339"), float("-0.106125"), float("0.0946545"), float("-0.0221862"), float("0.257953"), float("0.194255"), float("0.117183"), float("-0.239613"), float("-0.276004"), float("-0.0432311"), float("0.174063"), float("-0.0634706"), float("-0.324576"), float("0.200557"), float("0.27033"), float("-0.141904"), float("0.163621"), float("0.159129"), float("0.0331803"), float("0.0716283"), float("-0.259889"), float("-0.281286"), float("0.311005"), float("-0.192631"), float("-0.025214"), float("-0.335516"), float("0.290179"), float("0.163395"), float("0.0640547"), float("0.0451905"), float("0.3239"), float("0.274364"), float("-0.305999"), float("0.102385"), float("-0.106225"), float("-0.213116"), float("0.344277"), float("-0.0780973"), float("-0.224715"), float("0.241605"), float("-0.151579"), float("-0.210299"), float("0.0823131"), float("0.100404"), float("0.132546"), float("-0.277527"), float("0.0598196"), float("-0.0433582"), float("0.230157"), float("0.123731"), float("-0.100427"), float("0.193713"), float("0.202449"), float("-0.266377"), float("-0.0941569"), float("-0.0321949"), float("-0.0972526"), float("0.167801"), float("0.0583861"), float("-0.270751"), float("0.18621"), float("-0.278426"), float("-0.0386036"), float("0.160098"), float("-0.13865"), float("0.383023"), float("0.051967"), float("-0.180901"), float("0.171269"), float("0.0455311"), float("0.0692674"), float("0.38991"), float("-0.0709482"), float("-0.0925497"), float("-0.125314"), float("-0.28824"), float("-0.101276"), float("0.180632"), float("0.229682"), float("-0.129278")] - - -class PirProgram_example_input_tensor_meta_434379246092404667: - program_id = 3111236241086672326 - input_name = "linear_1.b_0" - shape = [16] - data = [float("0.125644"), float("0.0118724"), float("0.0756429"), float("-0.0202031"), float("0.0825371"), float("0.00196512"), float("0.0474929"), float("0.136103"), float("-0.0232748"), float("0.140402"), float("-0.0294003"), float("-0.296898"), float("0.0529667"), float("0.0581229"), float("-0.159494"), float("0.123836")] - - -class PirProgram_example_input_tensor_meta_8868019711434113141: - program_id = 3111236241086672326 - input_name = "middle_3" - shape = [2, 2, 32] - data = [float("2.3204"), float("1.27108"), float("1.55032"), float("0"), float("0"), float("1.33579"), float("0"), float("0"), float("1.66375"), float("0"), float("0"), float("0.766774"), float("1.97588"), float("5.28458"), float("0"), float("0"), float("0"), float("0"), float("0"), float("1.20052"), float("1.82524"), float("0"), float("2.05893"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.23679"), float("0.317447"), float("0"), float("0"), float("0.289374"), float("0"), float("0"), float("0.508824"), float("0"), float("0"), float("0.178782"), float("0.712137"), float("1.66309"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.357579"), float("0"), float("0"), float("0.756404"), float("0"), float("0"), float("0.0961974"), float("0.327142"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.581249"), float("0.744442"), float("0"), float("0"), float("0.680321"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.493762"), float("1.26236"), float("3.09834"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.783214"), float("1.05384"), float("0"), float("0"), float("0"), float("0"), float("0.346895"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.0223007"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.0226074"), float("0"), float("0.0077469"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0")] - - -class PirProgram_example_input_tensor_meta_7754446314117936854: - program_id = 3111236241086672326 - input_name = "middle_1" - shape = [2, 2, 32] - data = [float("1.85632"), float("1.01686"), float("1.24026"), float("0"), float("0"), float("1.06863"), float("0"), float("0"), float("1.331"), float("0"), float("0"), float("0.613419"), float("1.5807"), float("4.22766"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.960414"), float("1.46019"), float("0"), float("1.64714"), float("0"), float("0"), float("0.822619"), float("1.06868"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.674439"), float("0.189432"), float("0.253958"), float("0"), float("0"), float("0.231499"), float("0"), float("0"), float("0.407059"), float("0"), float("0"), float("0.143025"), float("0.569709"), float("1.33048"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.286063"), float("0.5215"), float("0"), float("0.605123"), float("0"), float("0"), float("0.0769579"), float("0.261714"), float("0"), float("0"), float("0"), float("0"), float("0"), float("1.09256"), float("0.464999"), float("0.595554"), float("0"), float("0"), float("0.544257"), float("0"), float("0"), float("0.735279"), float("0"), float("0"), float("0.39501"), float("1.00989"), float("2.47867"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.626571"), float("0.843074"), float("0"), float("0.986253"), float("0"), float("0"), float("0.277516"), float("0.654774"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.0178405"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0.0151772"), float("0"), float("0.0180859"), float("0"), float("0.00619752"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0"), float("0")] - - -class PirProgram_example_input_tensor_meta_4575370185374825974: - program_id = 3111236241086672326 - input_name = "linear_0.w_0" - shape = [96, 32] - data = None - - -class PirProgram_example_input_tensor_meta_8444399123059722995: - program_id = 3111236241086672326 - input_name = "linear_0.b_0" - shape = [32] - data = [float("0.0201408"), float("-0.266505"), float("-0.278198"), float("-0.00211538"), float("-0.269264"), float("-0.232061"), float("-0.364265"), float("-0.248469"), float("-0.107383"), float("-0.31451"), float("-0.262869"), float("-0.176201"), float("-0.0559743"), float("-0.359228"), float("-0.388085"), float("-0.37184"), float("-0.0378036"), float("-0.0328156"), float("-0.336554"), float("-0.176577"), float("0.0164908"), float("-0.371851"), float("0.0113843"), float("-0.332664"), float("0.0101896"), float("-0.282343"), float("-0.26443"), float("-0.205234"), float("-0.164831"), float("-0.265903"), float("-0.343348"), float("-0.374968")] - - -class PirProgram_example_input_tensor_meta_3925472342396615484: - program_id = 3111236241086672326 - input_name = "args_0" - shape = [2, 2, 96] - data = [float("-1.15597"), float("-1.1523"), float("-1.15171"), float("-1.14702"), float("-1.14445"), float("-1.14483"), float("-1.14359"), float("-1.14577"), float("-1.14909"), float("-1.14868"), float("-1.14244"), float("-1.14465"), float("-1.13813"), float("-1.14016"), float("-1.13834"), float("-1.12726"), float("-1.11024"), float("-1.10905"), float("-1.10871"), float("-1.10095"), float("-1.09833"), float("-1.09615"), float("-1.09124"), float("-1.09103"), float("-1.09178"), float("-1.09125"), float("-1.09142"), float("-1.09758"), float("-1.09483"), float("-1.09004"), float("-1.0888"), float("-1.08218"), float("-1.08141"), float("-1.07451"), float("-1.07159"), float("-1.07065"), float("-1.06963"), float("-1.06884"), float("-1.06653"), float("-1.06595"), float("-1.06269"), float("-1.06049"), float("-1.05744"), float("-1.06124"), float("-1.05722"), float("-1.05479"), float("-1.05422"), float("-1.05193"), float("-1.0509"), float("-1.05028"), float("-1.04654"), float("-1.045"), float("-1.046"), float("-1.04667"), float("-1.04904"), float("-1.04042"), float("-1.04003"), float("-1.04359"), float("-1.03823"), float("-1.03557"), float("-1.03127"), float("-1.03583"), float("-1.0344"), float("-1.03144"), float("-1.02952"), float("-1.02593"), float("-1.02701"), float("-1.02957"), float("-1.02561"), float("-1.02673"), float("-1.0235"), float("-1.02171"), float("-1.02023"), float("-1.01943"), float("-1.01996"), float("-1.0124"), float("-1.01049"), float("-1.01181"), float("-1.01104"), float("-1.0073"), float("-1.0079"), float("-1.00004"), float("-0.996691"), float("-0.993982"), float("-0.991675"), float("-0.990566"), float("-0.988743"), float("-0.988268"), float("-0.984146"), float("-0.982912"), float("-0.987096"), float("-0.984222"), float("-0.963696"), float("-0.952533"), float("-0.954977"), float("-0.950661"), float("-0.399858"), float("-0.398499"), float("-0.392917"), float("-0.396343"), float("-0.391605"), float("-0.386859"), float("-0.391783"), float("-0.393174"), float("-0.396282"), float("-0.395804"), float("-0.396976"), float("-0.398565"), float("-0.395854"), float("-0.393964"), float("-0.393872"), float("-0.39339"), float("-0.394667"), float("-0.394263"), float("-0.395912"), float("-0.39461"), float("-0.392164"), float("-0.391116"), float("-0.393774"), float("-0.391869"), float("-0.392982"), float("-0.390163"), float("-0.392436"), float("-0.393987"), float("-0.394744"), float("-0.394476"), float("-0.391628"), float("-0.389037"), float("-0.387299"), float("-0.382867"), float("-0.392128"), float("-0.39309"), float("-0.3956"), float("-0.392959"), float("-0.386446"), float("-0.381103"), float("-0.380557"), float("-0.385733"), float("-0.382358"), float("-0.382535"), float("-0.388842"), float("-0.3847"), float("-0.386946"), float("-0.391374"), float("-0.390959"), float("-0.386347"), float("-0.391026"), float("-0.390428"), float("-0.386892"), float("-0.388793"), float("-0.392527"), float("-0.390458"), float("-0.389178"), float("-0.38852"), float("-0.389764"), float("-0.389545"), float("-0.384279"), float("-0.383312"), float("-0.379087"), float("-0.380502"), float("-0.381453"), float("-0.38131"), float("-0.386445"), float("-0.387503"), float("-0.385849"), float("-0.391521"), float("-0.39044"), float("-0.391326"), float("-0.390687"), float("-0.389549"), float("-0.387867"), float("-0.384638"), float("-0.388114"), float("-0.389713"), float("-0.386914"), float("-0.383812"), float("-0.38576"), float("-0.382273"), float("-0.381319"), float("-0.383497"), float("-0.378321"), float("-0.382726"), float("-0.3803"), float("-0.38556"), float("-0.378433"), float("-0.378988"), float("-0.380473"), float("-0.383853"), float("-0.381981"), float("-0.383138"), float("-0.38374"), float("-0.379838"), float("-0.630672"), float("-0.632913"), float("-0.633336"), float("-0.636795"), float("-0.637253"), float("-0.635754"), float("-0.636042"), float("-0.637148"), float("-0.640553"), float("-0.641178"), float("-0.639793"), float("-0.642557"), float("-0.642459"), float("-0.647011"), float("-0.646651"), float("-0.643922"), float("-0.64385"), float("-0.643057"), float("-0.642857"), float("-0.643463"), float("-0.644947"), float("-0.645577"), float("-0.645093"), float("-0.645523"), float("-0.645906"), float("-0.647578"), float("-0.646427"), float("-0.646111"), float("-0.649196"), float("-0.65109"), float("-0.649286"), float("-0.648321"), float("-0.64727"), float("-0.649662"), float("-0.646437"), float("-0.650501"), float("-0.648313"), float("-0.652139"), float("-0.652555"), float("-0.652121"), float("-0.651807"), float("-0.652984"), float("-0.652824"), float("-0.654689"), float("-0.653612"), float("-0.653592"), float("-0.653823"), float("-0.653214"), float("-0.653469"), float("-0.651772"), float("-0.650067"), float("-0.65063"), float("-0.652344"), float("-0.653487"), float("-0.648193"), float("-0.650301"), float("-0.649187"), float("-0.650017"), float("-0.648082"), float("-0.648181"), float("-0.647695"), float("-0.648911"), float("-0.649349"), float("-0.651401"), float("-0.654608"), float("-0.651767"), float("-0.652217"), float("-0.646641"), float("-0.649977"), float("-0.651891"), float("-0.653053"), float("-0.65444"), float("-0.653964"), float("-0.657094"), float("-0.657665"), float("-0.658948"), float("-0.662083"), float("-0.663417"), float("-0.665791"), float("-0.663225"), float("-0.659105"), float("-0.659393"), float("-0.659387"), float("-0.661882"), float("-0.663708"), float("-0.663021"), float("-0.66179"), float("-0.652634"), float("-0.652383"), float("-0.651621"), float("-0.65448"), float("-0.654242"), float("-0.656716"), float("-0.657377"), float("-0.65797"), float("-0.655658"), float("-0.00436456"), float("-0.00475825"), float("-0.00663337"), float("-0.00523404"), float("-0.00562955"), float("-0.00631295"), float("-0.00424731"), float("-0.00233632"), float("0.00070127"), float("0.00211113"), float("0.000545739"), float("0.00642778"), float("0.0038039"), float("0.000308868"), float("-0.00136501"), float("-0.00140553"), float("-0.00126356"), float("-0.00223812"), float("-0.00588529"), float("-0.00441997"), float("-0.00579612"), float("-0.00591071"), float("-0.00374736"), float("-0.0039841"), float("0.00174819"), float("0.00298575"), float("0.00173817"), float("0.00255231"), float("0.000368191"), float("0.000318813"), float("-0.00178881"), float("-0.00274283"), float("0.000628811"), float("-0.00182678"), float("-7.61417e-05"), float("-0.000906536"), float("-0.00333968"), float("-0.00408479"), float("-0.00325562"), float("-0.00334317"), float("-0.00397552"), float("-0.00359293"), float("-0.0032651"), float("0.00372263"), float("0.00179851"), float("0.00113003"), float("0.00107969"), float("0.000824071"), float("0.00230663"), float("0.00521457"), float("0.00676399"), float("0.00760787"), float("0.00433249"), float("0.0035608"), float("0.000554484"), float("0.00362084"), float("-0.00145232"), float("-0.00354868"), float("-0.00290337"), float("-0.00352887"), float("-0.00181459"), float("-0.00102524"), float("-0.00380784"), float("-0.00352733"), float("-0.00369791"), float("-0.00225879"), float("-0.00303406"), float("0.00243163"), float("0.00128276"), float("0.000701387"), float("0.00297557"), float("0.00318256"), float("0.00367343"), float("0.00317448"), float("0.00563999"), float("0.00400331"), float("-0.00251565"), float("0.00151779"), float("-0.00175092"), float("-0.00389494"), float("-0.0005398"), float("-0.000404088"), float("-0.00223803"), float("-0.00250093"), float("-0.0025293"), float("-0.00180433"), float("0.000374747"), float("0.00324137"), float("0.00465033"), float("0.00285781"), float("0.0061246"), float("0.00211838"), float("0.000508647"), float("-0.00179661"), float("-0.00133032"), float("-0.000782828")] - - -class PirProgram_example_input_tensor_meta_3211373362553701247: - program_id = 4630264566612914126 - input_name = "linear_1.b_0" - shape = [16] - data = None - - -class PirProgram_example_input_tensor_meta_3038923928244044276: - program_id = 4630264566612914126 - input_name = "linear_1.w_0" - shape = [32, 16] - data = None - - -class PirProgram_example_input_tensor_meta_1110867825722039225: - program_id = 4630264566612914126 - input_name = "linear_0.b_0" - shape = [32] - data = None - - -class PirProgram_example_input_tensor_meta_7373294236970362073: - program_id = 4630264566612914126 - input_name = "linear_0.w_0" - shape = [96, 32] - data = None - - -class PirProgram_example_input_tensor_meta_7822756705787592613: - program_id = 4630264566612914126 - input_name = "args_0" - shape = [16, 2, 96] - data = None - - -class PirProgram_example_input_tensor_meta_1185347860821980671: - program_id = 2498128585497043328 - input_name = "linear_3.b_0" - shape = [96] - data = None - - -class PirProgram_example_input_tensor_meta_8144689379135590626: - program_id = 2498128585497043328 - input_name = "linear_3.w_0" - shape = [32, 96] - data = None - - -class PirProgram_example_input_tensor_meta_5725363569888070898: - program_id = 2498128585497043328 - input_name = "linear_2.b_0" - shape = [32] - data = None - - -class PirProgram_example_input_tensor_meta_8026832391553157584: - program_id = 2498128585497043328 - input_name = "linear_2.w_0" - shape = [16, 32] - data = None - - -class PirProgram_example_input_tensor_meta_4610283840437367128: - program_id = 2498128585497043328 - input_name = "args_0" - shape = [16, 2, 16] - data = [float("-0.15216"), float("0.757773"), float("-0.329263"), float("-0.540604"), float("0.181305"), float("-0.183984"), float("-0.0277555"), float("-0.0668215"), float("-0.164828"), float("-0.152864"), float("-0.041962"), float("-0.119368"), float("0.12759"), float("0.124398"), float("0.762339"), float("0.281282"), float("0.353945"), float("-2.13213"), float("5.45738"), float("6.78211"), float("-0.901334"), float("10.5608"), float("0.638606"), float("7.42863"), float("11.3598"), float("1.95794"), float("-2.3519"), float("-1.56203"), float("1.85749"), float("-8.1808"), float("3.51677"), float("-18.3846"), float("0.175046"), float("-0.0554654"), float("0.142715"), float("0.1279"), float("0.0595285"), float("0.126531"), float("0.0591089"), float("0.250216"), float("0.231589"), float("0.128165"), float("-0.0492941"), float("-0.301204"), float("0.0698216"), float("-0.084365"), float("-0.0207583"), float("-0.169802"), float("0.206708"), float("-0.686835"), float("1.87437"), float("2.33026"), float("-0.205472"), float("3.51035"), float("0.219181"), float("2.54969"), float("3.77282"), float("0.725045"), float("-0.642187"), float("-0.656401"), float("0.567246"), float("-2.6316"), float("1.1654"), float("-6.16229"), float("0.0315424"), float("0.26434"), float("-0.0882699"), float("-0.218807"), float("0.113201"), float("-0.0394262"), float("0.0516854"), float("0.0276483"), float("-0.0290755"), float("0.0801737"), float("-0.0651081"), float("-0.237694"), float("0.0902895"), float("0.0880786"), float("0.151029"), float("0.160605"), float("0.206246"), float("-0.571435"), float("1.51467"), float("1.89243"), float("-0.183881"), float("2.81018"), float("0.168251"), float("2.10894"), float("3.12867"), float("0.59224"), float("-0.562897"), float("-0.582865"), float("0.516875"), float("-2.15694"), float("0.906377"), float("-4.90434"), float("-0.368032"), float("1.16993"), float("-0.514465"), float("-0.850272"), float("0.344572"), float("-0.307727"), float("-0.113579"), float("-0.13008"), float("-0.331641"), float("-0.35746"), float("0.0170699"), float("-0.0903696"), float("0.173143"), float("0.124255"), float("1.31038"), float("0.514586"), float("0.227267"), float("-1.11517"), float("2.94774"), float("3.65704"), float("-0.388165"), float("5.601"), float("0.358289"), float("3.96366"), float("5.94604"), float("1.08666"), float("-1.08567"), float("-0.923208"), float("0.887275"), float("-4.22929"), float("1.879"), float("-9.79137"), float("0.162628"), float("-0.137703"), float("0.503044"), float("0.6243"), float("0.0336264"), float("0.796969"), float("0.0392233"), float("0.718495"), float("0.939775"), float("0.253406"), float("-0.0352974"), float("-0.313293"), float("0.134459"), float("-0.549998"), float("0.238071"), float("-1.44879"), float("0.230935"), float("-1.7091"), float("4.48235"), float("5.54546"), float("-0.611005"), float("8.6001"), float("0.589296"), float("5.91443"), float("8.9119"), float("1.59889"), float("-1.64335"), float("-1.29905"), float("1.25007"), float("-6.4299"), float("2.92819"), float("-15.0178"), float("0.184487"), float("-0.271964"), float("0.839011"), float("1.04533"), float("-0.0367303"), float("1.46748"), float("0.0771252"), float("1.18206"), float("1.66538"), float("0.373778"), float("-0.200558"), float("-0.395642"), float("0.261839"), float("-1.07963"), float("0.462218"), float("-2.62307"), float("0.214251"), float("-0.996448"), float("2.64592"), float("3.29088"), float("-0.336725"), float("5.02743"), float("0.329378"), float("3.56391"), float("5.33463"), float("0.980632"), float("-0.961669"), float("-0.845799"), float("0.793455"), float("-3.77941"), float("1.6815"), float("-8.78449"), float("0.230756"), float("-0.0902852"), float("0.313066"), float("0.373669"), float("0.0560226"), float("0.459538"), float("-0.00736624"), float("0.49768"), float("0.6865"), float("0.191823"), float("-0.057096"), float("-0.286257"), float("0.149784"), float("-0.337131"), float("0.140222"), float("-0.830687"), float("0.208667"), float("-0.436397"), float("1.23756"), float("1.54828"), float("-0.11709"), float("2.26445"), float("0.125928"), float("1.73361"), float("2.52496"), float("0.51826"), float("-0.403326"), float("-0.501762"), float("0.408805"), float("-1.7112"), float("0.734785"), float("-4.00791"), float("0.187755"), float("-0.102415"), float("0.457224"), float("0.545884"), float("0.03179"), float("0.690048"), float("0.0183367"), float("0.648476"), float("0.835108"), float("0.241401"), float("-0.0139194"), float("-0.301483"), float("0.122606"), float("-0.467988"), float("0.20309"), float("-1.30769"), float("0.244079"), float("-1.70861"), float("4.47838"), float("5.55282"), float("-0.619873"), float("8.61168"), float("0.583028"), float("5.91956"), float("8.93747"), float("1.60415"), float("-1.65242"), float("-1.29492"), float("1.26257"), float("-6.44304"), float("2.93482"), float("-15.0458"), float("0.171207"), float("-0.0780683"), float("0.156868"), float("0.230114"), float("0.0477679"), float("0.248112"), float("0.0828771"), float("0.324246"), float("0.362926"), float("0.126835"), float("-0.0397792"), float("-0.276738"), float("0.0498615"), float("-0.144937"), float("0.051245"), float("-0.390421"), float("0.238441"), float("-0.348271"), float("0.995058"), float("1.21009"), float("-0.0690387"), float("1.7613"), float("0.0694272"), float("1.40697"), float("2.07087"), float("0.418077"), float("-0.335243"), float("-0.448782"), float("0.368917"), float("-1.34953"), float("0.568891"), float("-3.1105"), float("0.24836"), float("-0.602299"), float("1.54613"), float("1.93115"), float("-0.169884"), float("2.91118"), float("0.158147"), float("2.18802"), float("3.31096"), float("0.613529"), float("-0.650253"), float("-0.596971"), float("0.594206"), float("-2.2655"), float("0.950745"), float("-5.0319"), float("0.238552"), float("-1.86656"), float("4.77206"), float("5.94475"), float("-0.68505"), float("9.23835"), float("0.640713"), float("6.37201"), float("9.65914"), float("1.70325"), float("-1.87225"), float("-1.38542"), float("1.41939"), float("-6.97447"), float("3.12238"), float("-16.035"), float("0.203786"), float("-0.366669"), float("1.09684"), float("1.35079"), float("-0.0881401"), float("1.95277"), float("0.0972831"), float("1.52543"), float("2.1953"), float("0.459335"), float("-0.31404"), float("-0.459989"), float("0.350135"), float("-1.46276"), float("0.626179"), float("-3.48701"), float("0.199294"), float("-0.480529"), float("1.355"), float("1.69083"), float("-0.132597"), float("2.49271"), float("0.146592"), float("1.88249"), float("2.74644"), float("0.553286"), float("-0.447864"), float("-0.53054"), float("0.437591"), float("-1.87747"), float("0.806791"), float("-4.39344"), float("0.220828"), float("-0.207872"), float("0.573615"), float("0.699591"), float("0.0113688"), float("0.976057"), float("0.0260232"), float("0.855339"), float("1.24417"), float("0.275647"), float("-0.186788"), float("-0.356107"), float("0.243965"), float("-0.749682"), float("0.2994"), float("-1.68212"), float("0.241174"), float("-1.10363"), float("2.90263"), float("3.60878"), float("-0.407001"), float("5.52414"), float("0.349477"), float("3.9248"), float("5.90016"), float("1.07812"), float("-1.10482"), float("-0.917721"), float("0.906962"), float("-4.2014"), float("1.84607"), float("-9.6683"), float("1.36937"), float("-0.201856"), float("3.0553"), float("2.67667"), float("-0.260319"), float("4.74185"), float("-0.375263"), float("3.77967"), float("5.23098"), float("1.77046"), float("-0.285616"), float("-0.768714"), float("0.863539"), float("-3.52976"), float("1.80543"), float("-10.6851"), float("0.502645"), float("1.74433"), float("4.67628"), float("2.66077"), float("-2.54735"), float("4.18144"), float("-1.47307"), float("4.59314"), float("5.66316"), float("-0.827085"), float("4.81157"), float("-0.960794"), float("-1.35232"), float("-3.15221"), float("2.04182"), float("-12.572"), float("0.201046"), float("-0.178521"), float("0.583239"), float("0.723738"), float("0.0144532"), float("0.983445"), float("0.0399271"), float("0.853576"), float("1.18278"), float("0.288476"), float("-0.127038"), float("-0.342783"), float("0.199223"), float("-0.716623"), float("0.309393"), float("-1.76484"), float("0.239322"), float("-1.69879"), float("4.45429"), float("5.51995"), float("-0.617142"), float("8.56166"), float("0.579973"), float("5.89439"), float("8.89391"), float("1.59283"), float("-1.65349"), float("-1.29239"), float("1.26163"), float("-6.41205"), float("2.91166"), float("-14.95"), float("0.205208"), float("-0.0755278"), float("0.319165"), float("0.375614"), float("0.0495638"), float("0.447425"), float("5.57229e-05"), float("0.484023"), float("0.630457"), float("0.186171"), float("-0.0182203"), float("-0.281987"), float("0.116728"), float("-0.306026"), float("0.124836"), float("-0.833024"), float("0.213171"), float("-0.713007"), float("1.94641"), float("2.41507"), float("-0.232947"), float("3.63342"), float("0.21747"), float("2.64788"), float("3.92872"), float("0.746185"), float("-0.676981"), float("-0.674473"), float("0.602595"), float("-2.74229"), float("1.20013"), float("-6.38517"), float("0.0413053"), float("0.302979"), float("-0.114017"), float("-0.241728"), float("0.090609"), float("-0.0588678"), float("0.0378077"), float("0.0154076"), float("-0.0298173"), float("0.0667273"), float("-0.0560023"), float("-0.224117"), float("0.0819501"), float("0.106694"), float("0.195902"), float("0.156805"), float("0.208495"), float("-0.48272"), float("1.36807"), float("1.70364"), float("-0.130933"), float("2.51416"), float("0.147009"), float("1.90362"), float("2.77373"), float("0.562943"), float("-0.454971"), float("-0.535181"), float("0.443537"), float("-1.89619"), float("0.820157"), float("-4.4422")] - - - diff --git a/samples/paddle/PaddleX_AutoEncoder_ad_dy2st_train/model.py b/samples/paddle/PaddleX_AutoEncoder_ad_dy2st_train/model.py deleted file mode 100644 index 17f00d164..000000000 --- a/samples/paddle/PaddleX_AutoEncoder_ad_dy2st_train/model.py +++ /dev/null @@ -1,826 +0,0 @@ -class PirProgram_880737988565058868: - - def __init__(self): - - self.parameter_105 = self.Op("builtin.parameter", 105, input_types=[], output_types=[self.t_dtensor([16], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_1.b_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.parameter_106 = self.Op("builtin.parameter", 106, input_types=[], output_types=[self.t_dtensor([32, 16], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_1.w_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.parameter_107 = self.Op("builtin.parameter", 107, input_types=[], output_types=[self.t_dtensor([32], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_0.b_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.parameter_108 = self.Op("builtin.parameter", 108, input_types=[], output_types=[self.t_dtensor([96, 32], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "struct_name":self.a_str("/Linear/"), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_0.w_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.data_109 = self.Op("pd_op.data", 109, input_types=[], output_types=[self.t_dtensor([16, 2, 96], self.t_f32())], attrs={"struct_name":self.a_str("/Linear/"), "stop_gradient":self.a_array(self.a_bool(True)), "name":self.a_str("args_0"), "shape":self.a_intarray(16, 2, 96), "dtype":self.a_dtype("float32"), "place":self.a_place("undefined", 0), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.matmul_110 = self.Op("pd_op.matmul", 110, input_types=[self.t_dtensor([16, 2, 96], self.t_f32()), self.t_dtensor([96, 32], self.t_f32())], output_types=[self.t_dtensor([16, 2, 32], self.t_f32())], attrs={"struct_name":self.a_str("/Linear/"), "stop_gradient":self.a_array(self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.add_111 = self.Op("pd_op.add", 111, input_types=[self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([32], self.t_f32())], output_types=[self.t_dtensor([16, 2, 32], self.t_f32())], attrs={"struct_name":self.a_str("/Linear/"), "stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.relu_112 = self.Op("pd_op.relu", 112, input_types=[self.t_dtensor([16, 2, 32], self.t_f32())], output_types=[self.t_dtensor([16, 2, 32], self.t_f32())], attrs={"struct_name":self.a_str("/ReLU/"), "stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.full_113 = self.Op("pd_op.full", 113, input_types=[], output_types=[self.t_dtensor([1], self.t_f32())], attrs={"struct_name":self.a_str("/Dropout/"), "stop_gradient":self.a_array(self.a_bool(True)), "shape":self.a_intarray(1), "value":self.a_f64("0.2"), "dtype":self.a_dtype("float32"), "place":self.a_place("cpu"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.dropout_114 = self.Op("pd_op.dropout", 114, input_types=[self.t_dtensor([16, 2, 32], self.t_f32()), self.t_null(), self.t_dtensor([1], self.t_f32())], output_types=[self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([16, 2, 32], self.t_ui8())], attrs={"struct_name":self.a_str("/Dropout/"), "stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "is_test":self.a_bool(False), "seed":self.a_i32(0), "mode":self.a_str("upscale_in_train"), "fix_seed":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) - - self.matmul_115 = self.Op("pd_op.matmul", 115, input_types=[self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([32, 16], self.t_f32())], output_types=[self.t_dtensor([16, 2, 16], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_1/"), "stop_gradient":self.a_array(self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.add_116 = self.Op("pd_op.add", 116, input_types=[self.t_dtensor([16, 2, 16], self.t_f32()), self.t_dtensor([16], self.t_f32())], output_types=[self.t_dtensor([16, 2, 16], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_1/"), "stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.shadow_output_117 = self.Op("builtin.shadow_output", 117, input_types=[self.t_dtensor([16, 2, 32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_0"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.shadow_output_118 = self.Op("builtin.shadow_output", 118, input_types=[self.t_dtensor([16, 2, 32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_1"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.shadow_output_119 = self.Op("builtin.shadow_output", 119, input_types=[self.t_dtensor([1], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_2"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.shadow_output_120 = self.Op("builtin.shadow_output", 120, input_types=[self.t_dtensor([16, 2, 32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_3"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.shadow_output_121 = self.Op("builtin.shadow_output", 121, input_types=[self.t_dtensor([16, 2, 32], self.t_ui8())], output_types=[], attrs={"output_name":self.a_str("middle_4"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.shadow_output_122 = self.Op("builtin.shadow_output", 122, input_types=[self.t_dtensor([16, 2, 16], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_5"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.shadow_output_123 = self.Op("builtin.shadow_output", 123, input_types=[self.t_dtensor([16, 2, 16], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("output_0"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.module_103 = self.Op("builtin.module", 103, input_types=[], output_types=[], attrs={"program":self.a_pointer("0x584e0a20"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array()}, block_positional_arg_names=[[[]]], block_keyword_arg_names=[[{}]], block_positional_arg_types=[[[]]], block_keyword_arg_types=[[[]]], ) - - - - def module_103_block00(self, call): - - def ret_lambda_module_103_block00(): - - parameter_1050, = call(self.parameter_105) - - parameter_1060, = call(self.parameter_106) - - parameter_1070, = call(self.parameter_107) - - parameter_1080, = call(self.parameter_108) - - data_1090, = call(self.data_109) - - matmul_1100, = call(self.matmul_110, data_1090, parameter_1080) - - add_1110, = call(self.add_111, matmul_1100, parameter_1070) - - relu_1120, = call(self.relu_112, add_1110) - - full_1130, = call(self.full_113) - - dropout_1140, dropout_1141, = call(self.dropout_114, relu_1120, None, full_1130) - - matmul_1150, = call(self.matmul_115, dropout_1140, parameter_1060) - - add_1160, = call(self.add_116, matmul_1150, parameter_1050) - - call(self.shadow_output_117, matmul_1100) - - call(self.shadow_output_118, relu_1120) - - call(self.shadow_output_119, full_1130) - - call(self.shadow_output_120, dropout_1140) - - call(self.shadow_output_121, dropout_1141) - - call(self.shadow_output_122, matmul_1150) - - call(self.shadow_output_123, add_1160) - - return ret_lambda_module_103_block00 - - - - def __call__(self, call, *args, **kwargs): - - self.SetArgs(args) - - self.SetKeywordArgs(kwargs) - - return call(self.module_103, blocks=[[(self.module_103_block00,)]]) - - -class PirProgram_5237713637574795923: - - def __init__(self): - - self.parameter_259 = self.Op("builtin.parameter", 259, input_types=[], output_types=[self.t_dtensor([96], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_3.b_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.parameter_260 = self.Op("builtin.parameter", 260, input_types=[], output_types=[self.t_dtensor([32, 96], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_3.w_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.parameter_261 = self.Op("builtin.parameter", 261, input_types=[], output_types=[self.t_dtensor([32], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_2.b_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.parameter_262 = self.Op("builtin.parameter", 262, input_types=[], output_types=[self.t_dtensor([16, 32], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "struct_name":self.a_str("/Linear_2/"), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_2.w_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.data_263 = self.Op("pd_op.data", 263, input_types=[], output_types=[self.t_dtensor([16, 2, -1], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_2/"), "stop_gradient":self.a_array(self.a_bool(False)), "name":self.a_str("args_0"), "shape":self.a_intarray(16, 2, -1), "dtype":self.a_dtype("float32"), "place":self.a_place("undefined", 0), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.matmul_264 = self.Op("pd_op.matmul", 264, input_types=[self.t_dtensor([16, 2, -1], self.t_f32()), self.t_dtensor([16, 32], self.t_f32())], output_types=[self.t_dtensor([16, 2, 32], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_2/"), "stop_gradient":self.a_array(self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.add_265 = self.Op("pd_op.add", 265, input_types=[self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([32], self.t_f32())], output_types=[self.t_dtensor([16, 2, 32], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_2/"), "stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.relu_266 = self.Op("pd_op.relu", 266, input_types=[self.t_dtensor([16, 2, 32], self.t_f32())], output_types=[self.t_dtensor([16, 2, 32], self.t_f32())], attrs={"struct_name":self.a_str("/ReLU_1/"), "stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.full_267 = self.Op("pd_op.full", 267, input_types=[], output_types=[self.t_dtensor([1], self.t_f32())], attrs={"struct_name":self.a_str("/Dropout_1/"), "stop_gradient":self.a_array(self.a_bool(True)), "shape":self.a_intarray(1), "value":self.a_f64("0.2"), "dtype":self.a_dtype("float32"), "place":self.a_place("cpu"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.dropout_268 = self.Op("pd_op.dropout", 268, input_types=[self.t_dtensor([16, 2, 32], self.t_f32()), self.t_null(), self.t_dtensor([1], self.t_f32())], output_types=[self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([16, 2, 32], self.t_ui8())], attrs={"struct_name":self.a_str("/Dropout_1/"), "stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "is_test":self.a_bool(False), "seed":self.a_i32(0), "mode":self.a_str("upscale_in_train"), "fix_seed":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) - - self.matmul_269 = self.Op("pd_op.matmul", 269, input_types=[self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([32, 96], self.t_f32())], output_types=[self.t_dtensor([16, 2, 96], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_3/"), "stop_gradient":self.a_array(self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.add_270 = self.Op("pd_op.add", 270, input_types=[self.t_dtensor([16, 2, 96], self.t_f32()), self.t_dtensor([96], self.t_f32())], output_types=[self.t_dtensor([16, 2, 96], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_3/"), "stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.shadow_output_271 = self.Op("builtin.shadow_output", 271, input_types=[self.t_dtensor([16, 2, 32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_0"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.shadow_output_272 = self.Op("builtin.shadow_output", 272, input_types=[self.t_dtensor([16, 2, 32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_1"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.shadow_output_273 = self.Op("builtin.shadow_output", 273, input_types=[self.t_dtensor([1], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_2"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.shadow_output_274 = self.Op("builtin.shadow_output", 274, input_types=[self.t_dtensor([16, 2, 32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_3"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.shadow_output_275 = self.Op("builtin.shadow_output", 275, input_types=[self.t_dtensor([16, 2, 32], self.t_ui8())], output_types=[], attrs={"output_name":self.a_str("middle_4"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.shadow_output_276 = self.Op("builtin.shadow_output", 276, input_types=[self.t_dtensor([16, 2, 96], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_5"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.shadow_output_277 = self.Op("builtin.shadow_output", 277, input_types=[self.t_dtensor([16, 2, 96], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("output_0"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.module_257 = self.Op("builtin.module", 257, input_types=[], output_types=[], attrs={"program":self.a_pointer("0x6b30d5a0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array()}, block_positional_arg_names=[[[]]], block_keyword_arg_names=[[{}]], block_positional_arg_types=[[[]]], block_keyword_arg_types=[[[]]], ) - - - - def module_257_block00(self, call): - - def ret_lambda_module_257_block00(): - - parameter_2590, = call(self.parameter_259) - - parameter_2600, = call(self.parameter_260) - - parameter_2610, = call(self.parameter_261) - - parameter_2620, = call(self.parameter_262) - - data_2630, = call(self.data_263) - - matmul_2640, = call(self.matmul_264, data_2630, parameter_2620) - - add_2650, = call(self.add_265, matmul_2640, parameter_2610) - - relu_2660, = call(self.relu_266, add_2650) - - full_2670, = call(self.full_267) - - dropout_2680, dropout_2681, = call(self.dropout_268, relu_2660, None, full_2670) - - matmul_2690, = call(self.matmul_269, dropout_2680, parameter_2600) - - add_2700, = call(self.add_270, matmul_2690, parameter_2590) - - call(self.shadow_output_271, matmul_2640) - - call(self.shadow_output_272, relu_2660) - - call(self.shadow_output_273, full_2670) - - call(self.shadow_output_274, dropout_2680) - - call(self.shadow_output_275, dropout_2681) - - call(self.shadow_output_276, matmul_2690) - - call(self.shadow_output_277, add_2700) - - return ret_lambda_module_257_block00 - - - - def __call__(self, call, *args, **kwargs): - - self.SetArgs(args) - - self.SetKeywordArgs(kwargs) - - return call(self.module_257, blocks=[[(self.module_257_block00,)]]) - - -class PirProgram_1242071021843173094: - - def __init__(self): - - self.add_grad_278 = self.Op("pd_op.add_grad", 278, input_types=[self.t_dtensor([16, 2, 96], self.t_f32()), self.t_dtensor([96], self.t_f32()), self.t_dtensor([16, 2, 96], self.t_f32())], output_types=[self.t_dtensor([16, 2, 96], self.t_f32()), self.t_dtensor([96], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "axis":self.a_i32(-1), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) - - self.matmul_grad_279 = self.Op("pd_op.matmul_grad", 279, input_types=[self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([32, 96], self.t_f32()), self.t_dtensor([16, 2, 96], self.t_f32())], output_types=[self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([32, 96], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) - - self.dropout_grad_280 = self.Op("pd_op.dropout_grad", 280, input_types=[self.t_dtensor([16, 2, 32], self.t_ui8()), self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([1], self.t_f32())], output_types=[self.t_dtensor([16, 2, 32], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False)), "is_test":self.a_bool(False), "mode":self.a_str("upscale_in_train"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.relu_grad_281 = self.Op("pd_op.relu_grad", 281, input_types=[self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([16, 2, 32], self.t_f32())], output_types=[self.t_dtensor([16, 2, 32], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.add_grad_282 = self.Op("pd_op.add_grad", 282, input_types=[self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([32], self.t_f32()), self.t_dtensor([16, 2, 32], self.t_f32())], output_types=[self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([32], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "axis":self.a_i32(-1), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) - - self.matmul_grad_283 = self.Op("pd_op.matmul_grad", 283, input_types=[self.t_dtensor([16, 2, -1], self.t_f32()), self.t_dtensor([16, 32], self.t_f32()), self.t_dtensor([16, 2, 32], self.t_f32())], output_types=[self.t_dtensor([16, 2, -1], self.t_f32()), self.t_dtensor([16, 32], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) - - self.shadow_output_284 = self.Op("builtin.shadow_output", 284, input_types=[self.t_dtensor([16, 2, -1], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("input_grad_0"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.shadow_output_285 = self.Op("builtin.shadow_output", 285, input_types=[self.t_dtensor([32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("param_grad_0"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.shadow_output_286 = self.Op("builtin.shadow_output", 286, input_types=[self.t_dtensor([16, 32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("param_grad_1"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.shadow_output_287 = self.Op("builtin.shadow_output", 287, input_types=[self.t_dtensor([96], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("param_grad_2"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.shadow_output_288 = self.Op("builtin.shadow_output", 288, input_types=[self.t_dtensor([32, 96], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("param_grad_3"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.module_258 = self.Op("builtin.module", 258, input_types=[], output_types=[], attrs={"program":self.a_pointer("0x6b30eee0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array()}, block_positional_arg_names=[[[]]], block_keyword_arg_names=[[{"middle_5": "arg_1798365008", "middle_4": "arg_1798348928", "output_grad_0": "arg_1798365584", "linear_3.w_0": "arg_1481522992", "middle_2": "arg_1798347072", "linear_3.b_0": "arg_1798306704", "middle_3": "arg_1798348720", "middle_1": "arg_1798334304", "linear_2.w_0": "arg_1798084816", "middle_0": "arg_1798334032", "linear_2.b_0": "arg_1798322512", "args_0": "arg_1798315712"}]], block_positional_arg_types=[[[]]], block_keyword_arg_types=[[[self.t_dtensor([16, 2, 96], self.t_f32()), self.t_dtensor([16, 2, 32], self.t_ui8()), self.t_dtensor([16, 2, 96], self.t_f32()), self.t_dtensor([32, 96], self.t_f32()), self.t_dtensor([1], self.t_f32()), self.t_dtensor([96], self.t_f32()), self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([16, 32], self.t_f32()), self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([32], self.t_f32()), self.t_dtensor([16, 2, -1], self.t_f32())]]], ) - - - - def module_258_block00(self, call): - - def ret_lambda_module_258_block00(arg_1798365008, arg_1798348928, arg_1798365584, arg_1481522992, arg_1798347072, arg_1798306704, arg_1798348720, arg_1798334304, arg_1798084816, arg_1798334032, arg_1798322512, arg_1798315712): - - add_grad_2780, add_grad_2781, = call(self.add_grad_278, arg_1798365008, arg_1798306704, arg_1798365584) - - matmul_grad_2790, matmul_grad_2791, = call(self.matmul_grad_279, arg_1798348720, arg_1481522992, add_grad_2780) - - dropout_grad_2800, = call(self.dropout_grad_280, arg_1798348928, matmul_grad_2790, arg_1798347072) - - relu_grad_2810, = call(self.relu_grad_281, arg_1798334304, dropout_grad_2800) - - add_grad_2820, add_grad_2821, = call(self.add_grad_282, arg_1798334032, arg_1798322512, relu_grad_2810) - - matmul_grad_2830, matmul_grad_2831, = call(self.matmul_grad_283, arg_1798315712, arg_1798084816, add_grad_2820) - - call(self.shadow_output_284, matmul_grad_2830) - - call(self.shadow_output_285, add_grad_2821) - - call(self.shadow_output_286, matmul_grad_2831) - - call(self.shadow_output_287, add_grad_2781) - - call(self.shadow_output_288, matmul_grad_2791) - - return ret_lambda_module_258_block00 - - - - def __call__(self, call, *args, **kwargs): - - self.SetArgs(args) - - self.SetKeywordArgs(kwargs) - - return call(self.module_258, blocks=[[(self.module_258_block00,)]]) - - -class PirProgram_692320333968606381: - - def __init__(self): - - self.add_grad_124 = self.Op("pd_op.add_grad", 124, input_types=[self.t_dtensor([16, 2, 16], self.t_f32()), self.t_dtensor([16], self.t_f32()), self.t_dtensor([16, 2, 16], self.t_f32())], output_types=[self.t_dtensor([16, 2, 16], self.t_f32()), self.t_dtensor([16], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "axis":self.a_i32(-1), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) - - self.matmul_grad_125 = self.Op("pd_op.matmul_grad", 125, input_types=[self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([32, 16], self.t_f32()), self.t_dtensor([16, 2, 16], self.t_f32())], output_types=[self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([32, 16], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) - - self.dropout_grad_126 = self.Op("pd_op.dropout_grad", 126, input_types=[self.t_dtensor([16, 2, 32], self.t_ui8()), self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([1], self.t_f32())], output_types=[self.t_dtensor([16, 2, 32], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False)), "is_test":self.a_bool(False), "mode":self.a_str("upscale_in_train"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.relu_grad_127 = self.Op("pd_op.relu_grad", 127, input_types=[self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([16, 2, 32], self.t_f32())], output_types=[self.t_dtensor([16, 2, 32], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.add_grad_128 = self.Op("pd_op.add_grad", 128, input_types=[self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([32], self.t_f32()), self.t_dtensor([16, 2, 32], self.t_f32())], output_types=[self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([32], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "axis":self.a_i32(-1), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) - - self.matmul_grad_129 = self.Op("pd_op.matmul_grad", 129, input_types=[self.t_dtensor([16, 2, 96], self.t_f32()), self.t_dtensor([96, 32], self.t_f32()), self.t_dtensor([16, 2, 32], self.t_f32())], output_types=[self.t_null(), self.t_dtensor([96, 32], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) - - self.shadow_output_130 = self.Op("builtin.shadow_output", 130, input_types=[self.t_dtensor([32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("param_grad_0"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.shadow_output_131 = self.Op("builtin.shadow_output", 131, input_types=[self.t_dtensor([96, 32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("param_grad_1"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.shadow_output_132 = self.Op("builtin.shadow_output", 132, input_types=[self.t_dtensor([16], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("param_grad_2"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.shadow_output_133 = self.Op("builtin.shadow_output", 133, input_types=[self.t_dtensor([32, 16], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("param_grad_3"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.module_104 = self.Op("builtin.module", 104, input_types=[], output_types=[], attrs={"program":self.a_pointer("0x584e2360"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array()}, block_positional_arg_names=[[[]]], block_keyword_arg_names=[[{"output_grad_0": "arg_1481526688", "middle_5": "arg_1497418592", "middle_4": "arg_1497418384", "middle_2": "arg_1481508288", "middle_0": "arg_1481527936", "linear_1.w_0": "arg_1457990560", "linear_1.b_0": "arg_1457989680", "middle_3": "arg_1481508496", "middle_1": "arg_1481528144", "linear_0.w_0": "arg_1497436400", "linear_0.b_0": "arg_1457913760", "args_0": "arg_1497471536"}]], block_positional_arg_types=[[[]]], block_keyword_arg_types=[[[self.t_dtensor([16, 2, 16], self.t_f32()), self.t_dtensor([16, 2, 16], self.t_f32()), self.t_dtensor([16, 2, 32], self.t_ui8()), self.t_dtensor([1], self.t_f32()), self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([32, 16], self.t_f32()), self.t_dtensor([16], self.t_f32()), self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([16, 2, 32], self.t_f32()), self.t_dtensor([96, 32], self.t_f32()), self.t_dtensor([32], self.t_f32()), self.t_dtensor([16, 2, 96], self.t_f32())]]], ) - - - - def module_104_block00(self, call): - - def ret_lambda_module_104_block00(arg_1481526688, arg_1497418592, arg_1497418384, arg_1481508288, arg_1481527936, arg_1457990560, arg_1457989680, arg_1481508496, arg_1481528144, arg_1497436400, arg_1457913760, arg_1497471536): - - add_grad_1240, add_grad_1241, = call(self.add_grad_124, arg_1497418592, arg_1457989680, arg_1481526688) - - matmul_grad_1250, matmul_grad_1251, = call(self.matmul_grad_125, arg_1481508496, arg_1457990560, add_grad_1240) - - dropout_grad_1260, = call(self.dropout_grad_126, arg_1497418384, matmul_grad_1250, arg_1481508288) - - relu_grad_1270, = call(self.relu_grad_127, arg_1481528144, dropout_grad_1260) - - add_grad_1280, add_grad_1281, = call(self.add_grad_128, arg_1481527936, arg_1457913760, relu_grad_1270) - - matmul_grad_1290, matmul_grad_1291, = call(self.matmul_grad_129, arg_1497471536, arg_1497436400, add_grad_1280) - - call(self.shadow_output_130, add_grad_1281) - - call(self.shadow_output_131, matmul_grad_1291) - - call(self.shadow_output_132, add_grad_1241) - - call(self.shadow_output_133, matmul_grad_1251) - - return ret_lambda_module_104_block00 - - - - def __call__(self, call, *args, **kwargs): - - self.SetArgs(args) - - self.SetKeywordArgs(kwargs) - - return call(self.module_104, blocks=[[(self.module_104_block00,)]]) - - -class PirProgram_5426476789613677471: - - def __init__(self): - - self.parameter_459 = self.Op("builtin.parameter", 459, input_types=[], output_types=[self.t_dtensor([16], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_1.b_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.parameter_460 = self.Op("builtin.parameter", 460, input_types=[], output_types=[self.t_dtensor([32, 16], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_1.w_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.parameter_461 = self.Op("builtin.parameter", 461, input_types=[], output_types=[self.t_dtensor([32], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_0.b_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.parameter_462 = self.Op("builtin.parameter", 462, input_types=[], output_types=[self.t_dtensor([96, 32], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "struct_name":self.a_str("/Linear_4/"), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_0.w_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.data_463 = self.Op("pd_op.data", 463, input_types=[], output_types=[self.t_dtensor([-1, 2, -1], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_4/"), "stop_gradient":self.a_array(self.a_bool(True)), "name":self.a_str("args_0"), "shape":self.a_intarray(-1, 2, -1), "dtype":self.a_dtype("float32"), "place":self.a_place("undefined", 0), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.matmul_464 = self.Op("pd_op.matmul", 464, input_types=[self.t_dtensor([-1, 2, -1], self.t_f32()), self.t_dtensor([96, 32], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_4/"), "stop_gradient":self.a_array(self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.add_465 = self.Op("pd_op.add", 465, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([32], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_4/"), "stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.relu_466 = self.Op("pd_op.relu", 466, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], attrs={"struct_name":self.a_str("/ReLU_2/"), "stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.full_467 = self.Op("pd_op.full", 467, input_types=[], output_types=[self.t_dtensor([1], self.t_f32())], attrs={"struct_name":self.a_str("/Dropout_2/"), "stop_gradient":self.a_array(self.a_bool(True)), "shape":self.a_intarray(1), "value":self.a_f64("0.2"), "dtype":self.a_dtype("float32"), "place":self.a_place("cpu"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.dropout_468 = self.Op("pd_op.dropout", 468, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_null(), self.t_dtensor([1], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([-1, 2, 32], self.t_ui8())], attrs={"struct_name":self.a_str("/Dropout_2/"), "stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "is_test":self.a_bool(False), "seed":self.a_i32(0), "mode":self.a_str("upscale_in_train"), "fix_seed":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) - - self.matmul_469 = self.Op("pd_op.matmul", 469, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([32, 16], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 16], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_5/"), "stop_gradient":self.a_array(self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.add_470 = self.Op("pd_op.add", 470, input_types=[self.t_dtensor([-1, 2, 16], self.t_f32()), self.t_dtensor([16], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 16], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_5/"), "stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.shadow_output_471 = self.Op("builtin.shadow_output", 471, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_0"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.shadow_output_472 = self.Op("builtin.shadow_output", 472, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_1"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.shadow_output_473 = self.Op("builtin.shadow_output", 473, input_types=[self.t_dtensor([1], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_2"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.shadow_output_474 = self.Op("builtin.shadow_output", 474, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_3"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.shadow_output_475 = self.Op("builtin.shadow_output", 475, input_types=[self.t_dtensor([-1, 2, 32], self.t_ui8())], output_types=[], attrs={"output_name":self.a_str("middle_4"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.shadow_output_476 = self.Op("builtin.shadow_output", 476, input_types=[self.t_dtensor([-1, 2, 16], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_5"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.shadow_output_477 = self.Op("builtin.shadow_output", 477, input_types=[self.t_dtensor([-1, 2, 16], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("output_0"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.module_457 = self.Op("builtin.module", 457, input_types=[], output_types=[], attrs={"program":self.a_pointer("0x6b3cf350"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array()}, block_positional_arg_names=[[[]]], block_keyword_arg_names=[[{}]], block_positional_arg_types=[[[]]], block_keyword_arg_types=[[[]]], ) - - - - def module_457_block00(self, call): - - def ret_lambda_module_457_block00(): - - parameter_4590, = call(self.parameter_459) - - parameter_4600, = call(self.parameter_460) - - parameter_4610, = call(self.parameter_461) - - parameter_4620, = call(self.parameter_462) - - data_4630, = call(self.data_463) - - matmul_4640, = call(self.matmul_464, data_4630, parameter_4620) - - add_4650, = call(self.add_465, matmul_4640, parameter_4610) - - relu_4660, = call(self.relu_466, add_4650) - - full_4670, = call(self.full_467) - - dropout_4680, dropout_4681, = call(self.dropout_468, relu_4660, None, full_4670) - - matmul_4690, = call(self.matmul_469, dropout_4680, parameter_4600) - - add_4700, = call(self.add_470, matmul_4690, parameter_4590) - - call(self.shadow_output_471, matmul_4640) - - call(self.shadow_output_472, relu_4660) - - call(self.shadow_output_473, full_4670) - - call(self.shadow_output_474, dropout_4680) - - call(self.shadow_output_475, dropout_4681) - - call(self.shadow_output_476, matmul_4690) - - call(self.shadow_output_477, add_4700) - - return ret_lambda_module_457_block00 - - - - def __call__(self, call, *args, **kwargs): - - self.SetArgs(args) - - self.SetKeywordArgs(kwargs) - - return call(self.module_457, blocks=[[(self.module_457_block00,)]]) - - -class PirProgram_7028096434133672773: - - def __init__(self): - - self.parameter_613 = self.Op("builtin.parameter", 613, input_types=[], output_types=[self.t_dtensor([96], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_3.b_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.parameter_614 = self.Op("builtin.parameter", 614, input_types=[], output_types=[self.t_dtensor([32, 96], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_3.w_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.parameter_615 = self.Op("builtin.parameter", 615, input_types=[], output_types=[self.t_dtensor([32], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_2.b_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.parameter_616 = self.Op("builtin.parameter", 616, input_types=[], output_types=[self.t_dtensor([16, 32], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "struct_name":self.a_str("/Linear_6/"), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_2.w_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.data_617 = self.Op("pd_op.data", 617, input_types=[], output_types=[self.t_dtensor([-1, 2, -1], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_6/"), "stop_gradient":self.a_array(self.a_bool(False)), "name":self.a_str("args_0"), "shape":self.a_intarray(-1, 2, -1), "dtype":self.a_dtype("float32"), "place":self.a_place("undefined", 0), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.matmul_618 = self.Op("pd_op.matmul", 618, input_types=[self.t_dtensor([-1, 2, -1], self.t_f32()), self.t_dtensor([16, 32], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_6/"), "stop_gradient":self.a_array(self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.add_619 = self.Op("pd_op.add", 619, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([32], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_6/"), "stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.relu_620 = self.Op("pd_op.relu", 620, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], attrs={"struct_name":self.a_str("/ReLU_3/"), "stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.full_621 = self.Op("pd_op.full", 621, input_types=[], output_types=[self.t_dtensor([1], self.t_f32())], attrs={"struct_name":self.a_str("/Dropout_3/"), "stop_gradient":self.a_array(self.a_bool(True)), "shape":self.a_intarray(1), "value":self.a_f64("0.2"), "dtype":self.a_dtype("float32"), "place":self.a_place("cpu"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.dropout_622 = self.Op("pd_op.dropout", 622, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_null(), self.t_dtensor([1], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([-1, 2, 32], self.t_ui8())], attrs={"struct_name":self.a_str("/Dropout_3/"), "stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "is_test":self.a_bool(False), "seed":self.a_i32(0), "mode":self.a_str("upscale_in_train"), "fix_seed":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) - - self.matmul_623 = self.Op("pd_op.matmul", 623, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([32, 96], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 96], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_7/"), "stop_gradient":self.a_array(self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.add_624 = self.Op("pd_op.add", 624, input_types=[self.t_dtensor([-1, 2, 96], self.t_f32()), self.t_dtensor([96], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 96], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_7/"), "stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.shadow_output_625 = self.Op("builtin.shadow_output", 625, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_0"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.shadow_output_626 = self.Op("builtin.shadow_output", 626, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_1"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.shadow_output_627 = self.Op("builtin.shadow_output", 627, input_types=[self.t_dtensor([1], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_2"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.shadow_output_628 = self.Op("builtin.shadow_output", 628, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_3"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.shadow_output_629 = self.Op("builtin.shadow_output", 629, input_types=[self.t_dtensor([-1, 2, 32], self.t_ui8())], output_types=[], attrs={"output_name":self.a_str("middle_4"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.shadow_output_630 = self.Op("builtin.shadow_output", 630, input_types=[self.t_dtensor([-1, 2, 96], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("middle_5"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.shadow_output_631 = self.Op("builtin.shadow_output", 631, input_types=[self.t_dtensor([-1, 2, 96], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("output_0"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.module_611 = self.Op("builtin.module", 611, input_types=[], output_types=[], attrs={"program":self.a_pointer("0x6b400f50"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array()}, block_positional_arg_names=[[[]]], block_keyword_arg_names=[[{}]], block_positional_arg_types=[[[]]], block_keyword_arg_types=[[[]]], ) - - - - def module_611_block00(self, call): - - def ret_lambda_module_611_block00(): - - parameter_6130, = call(self.parameter_613) - - parameter_6140, = call(self.parameter_614) - - parameter_6150, = call(self.parameter_615) - - parameter_6160, = call(self.parameter_616) - - data_6170, = call(self.data_617) - - matmul_6180, = call(self.matmul_618, data_6170, parameter_6160) - - add_6190, = call(self.add_619, matmul_6180, parameter_6150) - - relu_6200, = call(self.relu_620, add_6190) - - full_6210, = call(self.full_621) - - dropout_6220, dropout_6221, = call(self.dropout_622, relu_6200, None, full_6210) - - matmul_6230, = call(self.matmul_623, dropout_6220, parameter_6140) - - add_6240, = call(self.add_624, matmul_6230, parameter_6130) - - call(self.shadow_output_625, matmul_6180) - - call(self.shadow_output_626, relu_6200) - - call(self.shadow_output_627, full_6210) - - call(self.shadow_output_628, dropout_6220) - - call(self.shadow_output_629, dropout_6221) - - call(self.shadow_output_630, matmul_6230) - - call(self.shadow_output_631, add_6240) - - return ret_lambda_module_611_block00 - - - - def __call__(self, call, *args, **kwargs): - - self.SetArgs(args) - - self.SetKeywordArgs(kwargs) - - return call(self.module_611, blocks=[[(self.module_611_block00,)]]) - - -class PirProgram_3000443918524221177: - - def __init__(self): - - self.add_grad_632 = self.Op("pd_op.add_grad", 632, input_types=[self.t_dtensor([-1, 2, 96], self.t_f32()), self.t_dtensor([96], self.t_f32()), self.t_dtensor([-1, 2, 96], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 96], self.t_f32()), self.t_dtensor([96], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "axis":self.a_i32(-1), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) - - self.matmul_grad_633 = self.Op("pd_op.matmul_grad", 633, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([32, 96], self.t_f32()), self.t_dtensor([-1, 2, 96], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([32, 96], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) - - self.dropout_grad_634 = self.Op("pd_op.dropout_grad", 634, input_types=[self.t_dtensor([-1, 2, 32], self.t_ui8()), self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([1], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False)), "is_test":self.a_bool(False), "mode":self.a_str("upscale_in_train"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.relu_grad_635 = self.Op("pd_op.relu_grad", 635, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([-1, 2, 32], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.add_grad_636 = self.Op("pd_op.add_grad", 636, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([32], self.t_f32()), self.t_dtensor([-1, 2, 32], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([32], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "axis":self.a_i32(-1), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) - - self.matmul_grad_637 = self.Op("pd_op.matmul_grad", 637, input_types=[self.t_dtensor([-1, 2, -1], self.t_f32()), self.t_dtensor([16, 32], self.t_f32()), self.t_dtensor([-1, 2, 32], self.t_f32())], output_types=[self.t_dtensor([-1, 2, -1], self.t_f32()), self.t_dtensor([16, 32], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) - - self.shadow_output_638 = self.Op("builtin.shadow_output", 638, input_types=[self.t_dtensor([-1, 2, -1], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("input_grad_0"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.shadow_output_639 = self.Op("builtin.shadow_output", 639, input_types=[self.t_dtensor([32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("param_grad_0"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.shadow_output_640 = self.Op("builtin.shadow_output", 640, input_types=[self.t_dtensor([16, 32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("param_grad_1"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.shadow_output_641 = self.Op("builtin.shadow_output", 641, input_types=[self.t_dtensor([96], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("param_grad_2"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.shadow_output_642 = self.Op("builtin.shadow_output", 642, input_types=[self.t_dtensor([32, 96], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("param_grad_3"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.module_612 = self.Op("builtin.module", 612, input_types=[], output_types=[], attrs={"program":self.a_pointer("0x6b3f4bd0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array()}, block_positional_arg_names=[[[]]], block_keyword_arg_names=[[{"middle_5": "arg_1799310528", "middle_4": "arg_1799317136", "output_grad_0": "arg_1799310976", "linear_3.w_0": "arg_1799293984", "middle_2": "arg_1799316720", "linear_3.b_0": "arg_1799157344", "middle_3": "arg_1799316928", "middle_1": "arg_1799301824", "linear_2.w_0": "arg_1799170736", "middle_0": "arg_1799301552", "linear_2.b_0": "arg_1799157488", "args_0": "arg_1799179680"}]], block_positional_arg_types=[[[]]], block_keyword_arg_types=[[[self.t_dtensor([-1, 2, 96], self.t_f32()), self.t_dtensor([-1, 2, 32], self.t_ui8()), self.t_dtensor([-1, 2, 96], self.t_f32()), self.t_dtensor([32, 96], self.t_f32()), self.t_dtensor([1], self.t_f32()), self.t_dtensor([96], self.t_f32()), self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([16, 32], self.t_f32()), self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([32], self.t_f32()), self.t_dtensor([-1, 2, -1], self.t_f32())]]], ) - - - - def module_612_block00(self, call): - - def ret_lambda_module_612_block00(arg_1799310528, arg_1799317136, arg_1799310976, arg_1799293984, arg_1799316720, arg_1799157344, arg_1799316928, arg_1799301824, arg_1799170736, arg_1799301552, arg_1799157488, arg_1799179680): - - add_grad_6320, add_grad_6321, = call(self.add_grad_632, arg_1799310528, arg_1799157344, arg_1799310976) - - matmul_grad_6330, matmul_grad_6331, = call(self.matmul_grad_633, arg_1799316928, arg_1799293984, add_grad_6320) - - dropout_grad_6340, = call(self.dropout_grad_634, arg_1799317136, matmul_grad_6330, arg_1799316720) - - relu_grad_6350, = call(self.relu_grad_635, arg_1799301824, dropout_grad_6340) - - add_grad_6360, add_grad_6361, = call(self.add_grad_636, arg_1799301552, arg_1799157488, relu_grad_6350) - - matmul_grad_6370, matmul_grad_6371, = call(self.matmul_grad_637, arg_1799179680, arg_1799170736, add_grad_6360) - - call(self.shadow_output_638, matmul_grad_6370) - - call(self.shadow_output_639, add_grad_6361) - - call(self.shadow_output_640, matmul_grad_6371) - - call(self.shadow_output_641, add_grad_6321) - - call(self.shadow_output_642, matmul_grad_6331) - - return ret_lambda_module_612_block00 - - - - def __call__(self, call, *args, **kwargs): - - self.SetArgs(args) - - self.SetKeywordArgs(kwargs) - - return call(self.module_612, blocks=[[(self.module_612_block00,)]]) - - -class PirProgram_3111236241086672326: - - def __init__(self): - - self.add_grad_478 = self.Op("pd_op.add_grad", 478, input_types=[self.t_dtensor([-1, 2, 16], self.t_f32()), self.t_dtensor([16], self.t_f32()), self.t_dtensor([-1, 2, 16], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 16], self.t_f32()), self.t_dtensor([16], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "axis":self.a_i32(-1), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) - - self.matmul_grad_479 = self.Op("pd_op.matmul_grad", 479, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([32, 16], self.t_f32()), self.t_dtensor([-1, 2, 16], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([32, 16], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) - - self.dropout_grad_480 = self.Op("pd_op.dropout_grad", 480, input_types=[self.t_dtensor([-1, 2, 32], self.t_ui8()), self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([1], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False)), "is_test":self.a_bool(False), "mode":self.a_str("upscale_in_train"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.relu_grad_481 = self.Op("pd_op.relu_grad", 481, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([-1, 2, 32], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.add_grad_482 = self.Op("pd_op.add_grad", 482, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([32], self.t_f32()), self.t_dtensor([-1, 2, 32], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([32], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "axis":self.a_i32(-1), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) - - self.matmul_grad_483 = self.Op("pd_op.matmul_grad", 483, input_types=[self.t_dtensor([-1, 2, -1], self.t_f32()), self.t_dtensor([96, 32], self.t_f32()), self.t_dtensor([-1, 2, 32], self.t_f32())], output_types=[self.t_null(), self.t_dtensor([96, 32], self.t_f32())], attrs={"stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) - - self.shadow_output_484 = self.Op("builtin.shadow_output", 484, input_types=[self.t_dtensor([32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("param_grad_0"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.shadow_output_485 = self.Op("builtin.shadow_output", 485, input_types=[self.t_dtensor([96, 32], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("param_grad_1"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.shadow_output_486 = self.Op("builtin.shadow_output", 486, input_types=[self.t_dtensor([16], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("param_grad_2"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.shadow_output_487 = self.Op("builtin.shadow_output", 487, input_types=[self.t_dtensor([32, 16], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("param_grad_3"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.module_458 = self.Op("builtin.module", 458, input_types=[], output_types=[], attrs={"program":self.a_pointer("0x6b3d0ad0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array()}, block_positional_arg_names=[[[]]], block_keyword_arg_names=[[{"output_grad_0": "arg_1799168768", "middle_5": "arg_1799149040", "middle_4": "arg_1799148832", "middle_2": "arg_1799159872", "middle_0": "arg_1799136912", "linear_1.w_0": "arg_1735320288", "linear_1.b_0": "arg_1798932704", "middle_3": "arg_1799148624", "middle_1": "arg_1799137056", "linear_0.w_0": "arg_1457571760", "linear_0.b_0": "arg_1798932016", "args_0": "arg_1798507504"}]], block_positional_arg_types=[[[]]], block_keyword_arg_types=[[[self.t_dtensor([-1, 2, 16], self.t_f32()), self.t_dtensor([-1, 2, 16], self.t_f32()), self.t_dtensor([-1, 2, 32], self.t_ui8()), self.t_dtensor([1], self.t_f32()), self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([32, 16], self.t_f32()), self.t_dtensor([16], self.t_f32()), self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([96, 32], self.t_f32()), self.t_dtensor([32], self.t_f32()), self.t_dtensor([-1, 2, -1], self.t_f32())]]], ) - - - - def module_458_block00(self, call): - - def ret_lambda_module_458_block00(arg_1799168768, arg_1799149040, arg_1799148832, arg_1799159872, arg_1799136912, arg_1735320288, arg_1798932704, arg_1799148624, arg_1799137056, arg_1457571760, arg_1798932016, arg_1798507504): - - add_grad_4780, add_grad_4781, = call(self.add_grad_478, arg_1799149040, arg_1798932704, arg_1799168768) - - matmul_grad_4790, matmul_grad_4791, = call(self.matmul_grad_479, arg_1799148624, arg_1735320288, add_grad_4780) - - dropout_grad_4800, = call(self.dropout_grad_480, arg_1799148832, matmul_grad_4790, arg_1799159872) - - relu_grad_4810, = call(self.relu_grad_481, arg_1799137056, dropout_grad_4800) - - add_grad_4820, add_grad_4821, = call(self.add_grad_482, arg_1799136912, arg_1798932016, relu_grad_4810) - - matmul_grad_4830, matmul_grad_4831, = call(self.matmul_grad_483, arg_1798507504, arg_1457571760, add_grad_4820) - - call(self.shadow_output_484, add_grad_4821) - - call(self.shadow_output_485, matmul_grad_4831) - - call(self.shadow_output_486, add_grad_4781) - - call(self.shadow_output_487, matmul_grad_4791) - - return ret_lambda_module_458_block00 - - - - def __call__(self, call, *args, **kwargs): - - self.SetArgs(args) - - self.SetKeywordArgs(kwargs) - - return call(self.module_458, blocks=[[(self.module_458_block00,)]]) - - -class PirProgram_4630264566612914126: - - def __init__(self): - - self.parameter_750 = self.Op("builtin.parameter", 750, input_types=[], output_types=[self.t_dtensor([16], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_1.b_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.parameter_751 = self.Op("builtin.parameter", 751, input_types=[], output_types=[self.t_dtensor([32, 16], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_1.w_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.parameter_752 = self.Op("builtin.parameter", 752, input_types=[], output_types=[self.t_dtensor([32], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_0.b_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.parameter_753 = self.Op("builtin.parameter", 753, input_types=[], output_types=[self.t_dtensor([96, 32], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "struct_name":self.a_str("/Linear_8/"), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_0.w_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.data_754 = self.Op("pd_op.data", 754, input_types=[], output_types=[self.t_dtensor([-1, 2, -1], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_8/"), "stop_gradient":self.a_array(self.a_bool(True)), "name":self.a_str("args_0"), "shape":self.a_intarray(-1, 2, -1), "dtype":self.a_dtype("float32"), "place":self.a_place("undefined", 0), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.matmul_755 = self.Op("pd_op.matmul", 755, input_types=[self.t_dtensor([-1, 2, -1], self.t_f32()), self.t_dtensor([96, 32], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_8/"), "stop_gradient":self.a_array(self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.add_756 = self.Op("pd_op.add", 756, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([32], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_8/"), "stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.relu_757 = self.Op("pd_op.relu", 757, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], attrs={"struct_name":self.a_str("/ReLU_4/"), "stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.full_758 = self.Op("pd_op.full", 758, input_types=[], output_types=[self.t_dtensor([1], self.t_f32())], attrs={"struct_name":self.a_str("/Dropout_4/"), "stop_gradient":self.a_array(self.a_bool(True)), "shape":self.a_intarray(1), "value":self.a_f64("0.2"), "dtype":self.a_dtype("float32"), "place":self.a_place("cpu"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.dropout_759 = self.Op("pd_op.dropout", 759, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_null(), self.t_dtensor([1], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([-1, 2, 32], self.t_ui8())], attrs={"struct_name":self.a_str("/Dropout_4/"), "stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "is_test":self.a_bool(True), "seed":self.a_i32(0), "mode":self.a_str("upscale_in_train"), "fix_seed":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) - - self.matmul_760 = self.Op("pd_op.matmul", 760, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([32, 16], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 16], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_9/"), "stop_gradient":self.a_array(self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.add_761 = self.Op("pd_op.add", 761, input_types=[self.t_dtensor([-1, 2, 16], self.t_f32()), self.t_dtensor([16], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 16], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_9/"), "stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.shadow_output_762 = self.Op("builtin.shadow_output", 762, input_types=[self.t_dtensor([-1, 2, 16], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("output_0"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.module_748 = self.Op("builtin.module", 748, input_types=[], output_types=[], attrs={"program":self.a_pointer("0x6b3cafd0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array()}, block_positional_arg_names=[[[]]], block_keyword_arg_names=[[{}]], block_positional_arg_types=[[[]]], block_keyword_arg_types=[[[]]], ) - - - - def module_748_block00(self, call): - - def ret_lambda_module_748_block00(): - - parameter_7500, = call(self.parameter_750) - - parameter_7510, = call(self.parameter_751) - - parameter_7520, = call(self.parameter_752) - - parameter_7530, = call(self.parameter_753) - - data_7540, = call(self.data_754) - - matmul_7550, = call(self.matmul_755, data_7540, parameter_7530) - - add_7560, = call(self.add_756, matmul_7550, parameter_7520) - - relu_7570, = call(self.relu_757, add_7560) - - full_7580, = call(self.full_758) - - dropout_7590, dropout_7591, = call(self.dropout_759, relu_7570, None, full_7580) - - matmul_7600, = call(self.matmul_760, dropout_7590, parameter_7510) - - add_7610, = call(self.add_761, matmul_7600, parameter_7500) - - call(self.shadow_output_762, add_7610) - - return ret_lambda_module_748_block00 - - - - def __call__(self, call, *args, **kwargs): - - self.SetArgs(args) - - self.SetKeywordArgs(kwargs) - - return call(self.module_748, blocks=[[(self.module_748_block00,)]]) - - -class PirProgram_2498128585497043328: - - def __init__(self): - - self.parameter_817 = self.Op("builtin.parameter", 817, input_types=[], output_types=[self.t_dtensor([96], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_3.b_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.parameter_818 = self.Op("builtin.parameter", 818, input_types=[], output_types=[self.t_dtensor([32, 96], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_3.w_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.parameter_819 = self.Op("builtin.parameter", 819, input_types=[], output_types=[self.t_dtensor([32], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_2.b_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.parameter_820 = self.Op("builtin.parameter", 820, input_types=[], output_types=[self.t_dtensor([16, 32], self.t_f32())], attrs={"is_parameter":self.a_array(self.a_bool(True)), "is_distributed":self.a_array(self.a_bool(False)), "need_clip":self.a_array(self.a_bool(True)), "struct_name":self.a_str("/Linear_10/"), "persistable":self.a_array(self.a_bool(True)), "stop_gradient":self.a_array(self.a_bool(False)), "trainable":self.a_array(self.a_bool(True)), "parameter_name":self.a_str("linear_2.w_0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.data_821 = self.Op("pd_op.data", 821, input_types=[], output_types=[self.t_dtensor([-1, 2, -1], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_10/"), "stop_gradient":self.a_array(self.a_bool(False)), "name":self.a_str("args_0"), "shape":self.a_intarray(-1, 2, -1), "dtype":self.a_dtype("float32"), "place":self.a_place("undefined", 0), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.matmul_822 = self.Op("pd_op.matmul", 822, input_types=[self.t_dtensor([-1, 2, -1], self.t_f32()), self.t_dtensor([16, 32], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_10/"), "stop_gradient":self.a_array(self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.add_823 = self.Op("pd_op.add", 823, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([32], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_10/"), "stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.relu_824 = self.Op("pd_op.relu", 824, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32())], attrs={"struct_name":self.a_str("/ReLU_5/"), "stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.full_825 = self.Op("pd_op.full", 825, input_types=[], output_types=[self.t_dtensor([1], self.t_f32())], attrs={"struct_name":self.a_str("/Dropout_5/"), "stop_gradient":self.a_array(self.a_bool(True)), "shape":self.a_intarray(1), "value":self.a_f64("0.2"), "dtype":self.a_dtype("float32"), "place":self.a_place("cpu"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.dropout_826 = self.Op("pd_op.dropout", 826, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_null(), self.t_dtensor([1], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([-1, 2, 32], self.t_ui8())], attrs={"struct_name":self.a_str("/Dropout_5/"), "stop_gradient":self.a_array(self.a_bool(False), self.a_bool(False)), "is_test":self.a_bool(True), "seed":self.a_i32(0), "mode":self.a_str("upscale_in_train"), "fix_seed":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null()))}) - - self.matmul_827 = self.Op("pd_op.matmul", 827, input_types=[self.t_dtensor([-1, 2, 32], self.t_f32()), self.t_dtensor([32, 96], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 96], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_11/"), "stop_gradient":self.a_array(self.a_bool(False)), "transpose_x":self.a_bool(False), "transpose_y":self.a_bool(False), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.add_828 = self.Op("pd_op.add", 828, input_types=[self.t_dtensor([-1, 2, 96], self.t_f32()), self.t_dtensor([96], self.t_f32())], output_types=[self.t_dtensor([-1, 2, 96], self.t_f32())], attrs={"struct_name":self.a_str("/Linear_11/"), "stop_gradient":self.a_array(self.a_bool(False)), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null()), self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array(self.a_symbol(self.s_null()))}) - - self.shadow_output_829 = self.Op("builtin.shadow_output", 829, input_types=[self.t_dtensor([-1, 2, 96], self.t_f32())], output_types=[], attrs={"output_name":self.a_str("output_0"), "__operands_symbols_signature__":self.a_array(self.a_symbol(self.s_null())), "__results_symbols_signature__":self.a_array()}) - - self.module_815 = self.Op("builtin.module", 815, input_types=[], output_types=[], attrs={"program":self.a_pointer("0x6bc06ee0"), "__operands_symbols_signature__":self.a_array(), "__results_symbols_signature__":self.a_array()}, block_positional_arg_names=[[[]]], block_keyword_arg_names=[[{}]], block_positional_arg_types=[[[]]], block_keyword_arg_types=[[[]]], ) - - - - def module_815_block00(self, call): - - def ret_lambda_module_815_block00(): - - parameter_8170, = call(self.parameter_817) - - parameter_8180, = call(self.parameter_818) - - parameter_8190, = call(self.parameter_819) - - parameter_8200, = call(self.parameter_820) - - data_8210, = call(self.data_821) - - matmul_8220, = call(self.matmul_822, data_8210, parameter_8200) - - add_8230, = call(self.add_823, matmul_8220, parameter_8190) - - relu_8240, = call(self.relu_824, add_8230) - - full_8250, = call(self.full_825) - - dropout_8260, dropout_8261, = call(self.dropout_826, relu_8240, None, full_8250) - - matmul_8270, = call(self.matmul_827, dropout_8260, parameter_8180) - - add_8280, = call(self.add_828, matmul_8270, parameter_8170) - - call(self.shadow_output_829, add_8280) - - return ret_lambda_module_815_block00 - - - - def __call__(self, call, *args, **kwargs): - - self.SetArgs(args) - - self.SetKeywordArgs(kwargs) - - return call(self.module_815, blocks=[[(self.module_815_block00,)]]) - - diff --git a/samples/torch/resnet18/attribute.json b/samples/torch/resnet18/attribute.json index 023a79446..1373fe3b5 100644 --- a/samples/torch/resnet18/attribute.json +++ b/samples/torch/resnet18/attribute.json @@ -1,3 +1,5 @@ { - "framework": "torch" + "framework": "torch", + "num_devices_required": 1, + "num_nodes_required": 1 } \ No newline at end of file diff --git a/samples/torch/resnet18/model.py b/samples/torch/resnet18/model.py index 7a904d721..5d585cc76 100644 --- a/samples/torch/resnet18/model.py +++ b/samples/torch/resnet18/model.py @@ -1,6 +1,3 @@ - -import torch -from .. import utils class GraphModule(torch.nn.Module): def forward(self, p_conv1_weight, p_bn1_weight, p_bn1_bias, p_getattr_l__self___layer1___0___conv1_weight, p_getattr_l__self___layer1___0___bn1_weight, p_getattr_l__self___layer1___0___bn1_bias, p_getattr_l__self___layer1___0___conv2_weight, p_getattr_l__self___layer1___0___bn2_weight, p_getattr_l__self___layer1___0___bn2_bias, p_getattr_l__self___layer1___1___conv1_weight, p_getattr_l__self___layer1___1___bn1_weight, p_getattr_l__self___layer1___1___bn1_bias, p_getattr_l__self___layer1___1___conv2_weight, p_getattr_l__self___layer1___1___bn2_weight, p_getattr_l__self___layer1___1___bn2_bias, p_getattr_l__self___layer2___0___conv1_weight, p_getattr_l__self___layer2___0___bn1_weight, p_getattr_l__self___layer2___0___bn1_bias, p_getattr_l__self___layer2___0___conv2_weight, p_getattr_l__self___layer2___0___bn2_weight, p_getattr_l__self___layer2___0___bn2_bias, p_getattr_l__self___layer2___0___downsample_0_weight, p_getattr_l__self___layer2___0___downsample_1_weight, p_getattr_l__self___layer2___0___downsample_1_bias, p_getattr_l__self___layer2___1___conv1_weight, p_getattr_l__self___layer2___1___bn1_weight, p_getattr_l__self___layer2___1___bn1_bias, p_getattr_l__self___layer2___1___conv2_weight, p_getattr_l__self___layer2___1___bn2_weight, p_getattr_l__self___layer2___1___bn2_bias, p_getattr_l__self___layer3___0___conv1_weight, p_getattr_l__self___layer3___0___bn1_weight, p_getattr_l__self___layer3___0___bn1_bias, p_getattr_l__self___layer3___0___conv2_weight, p_getattr_l__self___layer3___0___bn2_weight, p_getattr_l__self___layer3___0___bn2_bias, p_getattr_l__self___layer3___0___downsample_0_weight, p_getattr_l__self___layer3___0___downsample_1_weight, p_getattr_l__self___layer3___0___downsample_1_bias, p_getattr_l__self___layer3___1___conv1_weight, p_getattr_l__self___layer3___1___bn1_weight, p_getattr_l__self___layer3___1___bn1_bias, p_getattr_l__self___layer3___1___conv2_weight, p_getattr_l__self___layer3___1___bn2_weight, p_getattr_l__self___layer3___1___bn2_bias, p_getattr_l__self___layer4___0___conv1_weight, p_getattr_l__self___layer4___0___bn1_weight, p_getattr_l__self___layer4___0___bn1_bias, p_getattr_l__self___layer4___0___conv2_weight, p_getattr_l__self___layer4___0___bn2_weight, p_getattr_l__self___layer4___0___bn2_bias, p_getattr_l__self___layer4___0___downsample_0_weight, p_getattr_l__self___layer4___0___downsample_1_weight, p_getattr_l__self___layer4___0___downsample_1_bias, p_getattr_l__self___layer4___1___conv1_weight, p_getattr_l__self___layer4___1___bn1_weight, p_getattr_l__self___layer4___1___bn1_bias, p_getattr_l__self___layer4___1___conv2_weight, p_getattr_l__self___layer4___1___bn2_weight, p_getattr_l__self___layer4___1___bn2_bias, p_fc_weight, p_fc_bias, b_bn1_running_mean, b_bn1_running_var, b_bn1_num_batches_tracked, b_getattr_l__self___layer1___0___bn1_running_mean, b_getattr_l__self___layer1___0___bn1_running_var, b_getattr_l__self___layer1___0___bn1_num_batches_tracked, b_getattr_l__self___layer1___0___bn2_running_mean, b_getattr_l__self___layer1___0___bn2_running_var, b_getattr_l__self___layer1___0___bn2_num_batches_tracked, b_getattr_l__self___layer1___1___bn1_running_mean, b_getattr_l__self___layer1___1___bn1_running_var, b_getattr_l__self___layer1___1___bn1_num_batches_tracked, b_getattr_l__self___layer1___1___bn2_running_mean, b_getattr_l__self___layer1___1___bn2_running_var, b_getattr_l__self___layer1___1___bn2_num_batches_tracked, b_getattr_l__self___layer2___0___bn1_running_mean, b_getattr_l__self___layer2___0___bn1_running_var, b_getattr_l__self___layer2___0___bn1_num_batches_tracked, b_getattr_l__self___layer2___0___bn2_running_mean, b_getattr_l__self___layer2___0___bn2_running_var, b_getattr_l__self___layer2___0___bn2_num_batches_tracked, b_getattr_l__self___layer2___0___downsample_1_running_mean, b_getattr_l__self___layer2___0___downsample_1_running_var, b_getattr_l__self___layer2___0___downsample_1_num_batches_tracked, b_getattr_l__self___layer2___1___bn1_running_mean, b_getattr_l__self___layer2___1___bn1_running_var, b_getattr_l__self___layer2___1___bn1_num_batches_tracked, b_getattr_l__self___layer2___1___bn2_running_mean, b_getattr_l__self___layer2___1___bn2_running_var, b_getattr_l__self___layer2___1___bn2_num_batches_tracked, b_getattr_l__self___layer3___0___bn1_running_mean, b_getattr_l__self___layer3___0___bn1_running_var, b_getattr_l__self___layer3___0___bn1_num_batches_tracked, b_getattr_l__self___layer3___0___bn2_running_mean, b_getattr_l__self___layer3___0___bn2_running_var, b_getattr_l__self___layer3___0___bn2_num_batches_tracked, b_getattr_l__self___layer3___0___downsample_1_running_mean, b_getattr_l__self___layer3___0___downsample_1_running_var, b_getattr_l__self___layer3___0___downsample_1_num_batches_tracked, b_getattr_l__self___layer3___1___bn1_running_mean, b_getattr_l__self___layer3___1___bn1_running_var, b_getattr_l__self___layer3___1___bn1_num_batches_tracked, b_getattr_l__self___layer3___1___bn2_running_mean, b_getattr_l__self___layer3___1___bn2_running_var, b_getattr_l__self___layer3___1___bn2_num_batches_tracked, b_getattr_l__self___layer4___0___bn1_running_mean, b_getattr_l__self___layer4___0___bn1_running_var, b_getattr_l__self___layer4___0___bn1_num_batches_tracked, b_getattr_l__self___layer4___0___bn2_running_mean, b_getattr_l__self___layer4___0___bn2_running_var, b_getattr_l__self___layer4___0___bn2_num_batches_tracked, b_getattr_l__self___layer4___0___downsample_1_running_mean, b_getattr_l__self___layer4___0___downsample_1_running_var, b_getattr_l__self___layer4___0___downsample_1_num_batches_tracked, b_getattr_l__self___layer4___1___bn1_running_mean, b_getattr_l__self___layer4___1___bn1_running_var, b_getattr_l__self___layer4___1___bn1_num_batches_tracked, b_getattr_l__self___layer4___1___bn2_running_mean, b_getattr_l__self___layer4___1___bn2_running_var, b_getattr_l__self___layer4___1___bn2_num_batches_tracked, x): @@ -77,20 +74,4 @@ def forward(self, p_conv1_weight, p_bn1_weight, p_bn1_bias, p_getattr_l__self___ # To see more debug info, please use `graph_module.print_readable()` -model = GraphModule() - -inputs_params = utils.load_converted_from_text(f'./source_tensor_meta.py') -inputs = inputs_params["input_info"] -inputs = [utils.replay_tensor(i) for i in inputs] -params = inputs_params["weight_info"] - -state_dict = {} -for k, v in params.items(): - k = utils.convert_param_name(k) - v = utils.replay_tensor(v) - state_dict[k] = v - -y = model(x=inputs[0], **state_dict)[0] -print(torch.argmin(y), torch.argmax(y)) -print(y.shape) diff --git a/torch_tools/README.md b/torch_tools/README.md new file mode 100644 index 000000000..72251c153 --- /dev/null +++ b/torch_tools/README.md @@ -0,0 +1,13 @@ +## 生成单侧 + + +``` +python vision_model_generator.py --key "model name like resnet18" --model-path "path to be restored at" +``` + + +## 运行单侧 +''' +python runner.py --model-path ../../sample/torch/resnet18 +''' + diff --git a/torch_tools/execution/runner.py b/torch_tools/execution/runner.py new file mode 100644 index 000000000..d63086afa --- /dev/null +++ b/torch_tools/execution/runner.py @@ -0,0 +1,50 @@ +import utils +import argparse +import importlib.util +import torch +from pathlib import Path +from typing import Type, Any +import sys + +def load_class_from_file(file_path: str, class_name: str) -> Type[torch.nn.Module]: + file = Path(file_path).resolve() + module_name = file.stem + + with open(file_path, 'r', encoding='utf-8') as f: + original_code = f.read() + import_stmt= "import torch" + modified_code = f"{import_stmt}\n{original_code}" + spec = importlib.util.spec_from_loader(module_name, loader=None) + module = importlib.util.module_from_spec(spec) + sys.modules[module_name] = module + compiled_code = compile(modified_code, filename=file, mode='exec') + exec(compiled_code, module.__dict__) + + model_class = getattr(module, class_name, None) + return model_class + +def main(model_path: str): + model_class = load_class_from_file(f"{model_path}/model.py", class_name="GraphModule") + model = model_class() + + inputs_params = utils.load_converted_from_text(f'{model_path}/source_tensor_meta.py') + inputs = inputs_params["input_info"] + inputs = [utils.replay_tensor(i) for i in inputs] + params = inputs_params["weight_info"] + + state_dict = {} + for k, v in params.items(): + k = utils.convert_param_name(k) + v = utils.replay_tensor(v) + state_dict[k] = v + + y = model(x=inputs[0], **state_dict)[0] + print(torch.argmin(y), torch.argmax(y)) + print(y.shape) + +if __name__ == '__main__': + parser = argparse.ArgumentParser(description="load and run model") + parser.add_argument("--model-path", type=str, required=True, + help="模型文件夹的路径，如'../../samples/torch/resnet18'") + args = parser.parse_args() + main(model_path=args.model_path) diff --git a/torch_tools/execution/utils.py b/torch_tools/execution/utils.py new file mode 100644 index 000000000..9ddfccbbb --- /dev/null +++ b/torch_tools/execution/utils.py @@ -0,0 +1,138 @@ +import torch +import torch.nn as nn +from collections import OrderedDict +import os +import re + +def convert_param_name(original_name): + if original_name.endswith(('.weight', '.bias')): + prefix = 'p_' + base_name = original_name + + elif any(x in original_name for x in ['running_mean', 'running_var', 'num_batches_tracked']): + prefix = 'b_' + base_name = original_name + else: + raise ValueError(f"Unrecognized parameter type: {original_name}") + + if '.' in base_name: + parts = base_name.split('.') + if len(parts) == 2 and not parts[0].startswith('layer'): + return prefix + parts[0] + '_' + parts[1] + else: + # layer1.0 -> layer1___0___ + pattern = r'(layer\d+)\.(\d+)\.' + replacement = r'\1___\2___' + converted = re.sub(pattern, replacement, base_name) + converted = converted.replace('.', '_') + return f"{prefix}getattr_l__self___{converted}" + else: + return prefix + base_name + +def load_converted_from_text(file_path): + + def parse_value(value_str): + value_str = value_str.strip() + if value_str == "None": + return None + if value_str == "[]": + return [] + elif value_str.startswith('"') or value_str.startswith("'"): + return value_str[1:-1] + elif value_str.startswith('['): + elements = value_str[1:-1].split(',') + result = [] + for e in elements: + e = e.strip() + try: + result.append(eval(e)) + except: + result.append(e) + return result + else: + try: + return eval(value_str) + except: + return value_str + + with open(file_path, 'r') as f: + lines = f.readlines() + + classes = [] + current_class = None + + for line in lines: + line = line.strip() + if line.startswith("class "): + if current_class is not None: + classes.append(current_class) + current_class = {} + elif "=" in line: + key, val = line.split("=", 1) + key = key.strip() + val = val.strip() + current_class[key] = parse_value(val) + + if current_class is not None: + classes.append(current_class) + + input_info = [] + weight_info = {} + + for cls in classes: + if 'input_' in cls["name"]: + item = { + "type": "random_tensor", + "info": { + "shape": cls.get("shape", []), + "dtype": getattr(torch, cls.get("dtype", "torch.float").split('.')[-1]), + "device": cls.get("device", "cpu"), + "mean": cls.get("mean", 0.0), + "std": cls.get("std", 1.0), + } + } + if cls.get("data") is not None: + if isinstance(cls.get("data"), str): + pass + else: + item["data"] = torch.tensor(cls["data"], dtype=item["info"]["dtype"]).reshape(cls.get("shape"), []) + input_info.append(item) + else: + data_value = None + data_type = getattr(torch, cls.get("dtype", "torch.float").split('.')[-1]) + if cls.get("data") is not None: + if isinstance(cls.get("data"), str): + raise ValueError("Unimplemented") + else: + data_value = torch.tensor(cls["data"], dtype=data_type).reshape(cls.get("shape"), []) + weight_info[cls["name"]] = { + "info": { + "shape": cls.get("shape", []), + "dtype": data_type, + "device": cls.get("device", "cpu"), + "mean": cls.get("mean", 0.0), + "std": cls.get("std", 1.0), + }, + "data": data_value, + } + + return { + "input_info": input_info if len(input_info) > 0 else None, + "weight_info": weight_info, + "dynamic_shapes": None + } + +def extract_dynamic_shapes(example_inputs): + pass + +def replay_tensor(info): + + device = info["info"]["device"] + dtype = info["info"]["dtype"] + shape = info["info"]["shape"] + + mean = info["info"]["mean"] + std = info["info"]["std"] + if info["data"] is not None: + return info["data"].to(device) + return torch.randn(size=shape).to(dtype).to(device) * std * 0.2 + mean diff --git a/samples/torch/utils.py b/torch_tools/generators/utils.py similarity index 66% rename from samples/torch/utils.py rename to torch_tools/generators/utils.py index 79817ea79..ea4ec0d0f 100644 --- a/samples/torch/utils.py +++ b/torch_tools/generators/utils.py @@ -5,10 +5,9 @@ import uuid import json import os +import argparse dyn_template = """ -import torch -from .. import utils %MODULE model = GraphModule() @@ -29,6 +28,100 @@ print(y.shape) """ +def convert_param_name(original_name): + if original_name.endswith(('.weight', '.bias')): + prefix = 'p_' + base_name = original_name + + elif any(x in original_name for x in ['running_mean', 'running_var', 'num_batches_tracked']): + prefix = 'b_' + base_name = original_name + else: + raise ValueError(f"Unrecognized parameter type: {original_name}") + + if '.' in base_name: + parts = base_name.split('.') + if len(parts) == 2 and not parts[0].startswith('layer'): + return prefix + parts[0] + '_' + parts[1] + else: + # layer1.0 -> layer1___0___ + pattern = r'(layer\d+)\.(\d+)\.' + replacement = r'\1___\2___' + converted = re.sub(pattern, replacement, base_name) + converted = converted.replace('.', '_') + return f"{prefix}getattr_l__self___{converted}" + else: + return prefix + base_name + +def indent_with_tab(code: str) -> str: + lines = code.splitlines() + indented_lines = [f" {line}" for line in lines] + return "\n".join(indented_lines) + +def apply_templates(code: str) -> str: + code = indent_with_tab(code) + code = code.replace(" GraphModule()", "class GraphModule(torch.nn.Module):") + code = code.replace(" \n" * 3, "\n") + py_code = dyn_template.replace('%MODULE', code) + return py_code + + +def convert_state_and_inputs(state_dict, example_inputs): + def tensor_info(tensor): + is_float = tensor.dtype.is_floating_point + return { + "shape": list(tensor.shape), + "dtype": str(tensor.dtype), + "device": str(tensor.device), + "mean": float(tensor.mean().item()) if is_float else None, + "std": float(tensor.std().item()) if is_float else None, + } + + def process_tensor(tensor): + if not isinstance(tensor, torch.Tensor): + return {"type": "unknown", "value": tensor} + + info = tensor_info(tensor) + if tensor.dtype in [torch.int8, torch.int16, torch.int32, torch.int64]: + if tensor.numel() < 1024: + return {"type": "small_int_tensor", "data": tensor.clone(), "info": info} + else: + return {"type": "big_int_tensor", "data": tensor.clone(), "info": info} + elif tensor.numel() < 1024: + return {"type": "small_tensor", "data": tensor.clone(), "info": info} + else: + return {"type": "random_tensor", "info": info} + + if isinstance(example_inputs, torch.Tensor): + processed_inputs = process_tensor(example_inputs) + elif isinstance(example_inputs, (list, tuple)): + processed_inputs = [process_tensor(t) for t in example_inputs] + else: + processed_inputs = {"type": "unknown", "value": example_inputs} + + processed_weights = {} + for key, tensor in state_dict.items(): + data_value = None + data_type = "random_tensor" + if tensor.dtype in [torch.int8, torch.int16, torch.int32, torch.int64]: + if tensor.numel() < 1024: + data_type = "small_int_tensor" + data_value = tensor.clone() + else: + data_type = "big_int_tensor" + + info = tensor_info(tensor) + processed_weights[key] = {"info": info} + processed_weights[key]["data"] = data_value + processed_weights[key]["type"] = data_type + + # dynamic_shapes = extract_dynamic_shapes(example_inputs) + return { + "input_info": processed_inputs, + "weight_info": processed_weights, + "dynamic_shapes": None + } + def save_constraints_text(converted, file_path): lines = [] if converted["dynamic_shapes"] is not None: @@ -99,204 +192,3 @@ def process_tensor_info(tensor_info, name_prefix="example_input"): with open(file_path, 'w') as f: f.write("\n".join(lines)) - -def load_converted_from_text(file_path): - - def parse_value(value_str): - value_str = value_str.strip() - if value_str == "None": - return None - if value_str == "[]": - return [] - elif value_str.startswith('"') or value_str.startswith("'"): - return value_str[1:-1] - elif value_str.startswith('['): - elements = value_str[1:-1].split(',') - result = [] - for e in elements: - e = e.strip() - try: - result.append(eval(e)) - except: - result.append(e) - return result - else: - try: - return eval(value_str) - except: - return value_str - - with open(file_path, 'r') as f: - lines = f.readlines() - - classes = [] - current_class = None - - for line in lines: - line = line.strip() - if line.startswith("class "): - if current_class is not None: - classes.append(current_class) - current_class = {} - elif "=" in line: - key, val = line.split("=", 1) - key = key.strip() - val = val.strip() - current_class[key] = parse_value(val) - - if current_class is not None: - classes.append(current_class) - - input_info = [] - weight_info = {} - - for cls in classes: - if 'input_' in cls["name"]: - item = { - "type": "random_tensor", - "info": { - "shape": cls.get("shape", []), - "dtype": getattr(torch, cls.get("dtype", "torch.float").split('.')[-1]), - "device": cls.get("device", "cpu"), - "mean": cls.get("mean", 0.0), - "std": cls.get("std", 1.0), - } - } - if cls.get("data") is not None: - if isinstance(cls.get("data"), str): - pass - else: - item["data"] = torch.tensor(cls["data"], dtype=item["info"]["dtype"]).reshape(cls.get("shape"), []) - input_info.append(item) - else: - data_value = None - data_type = getattr(torch, cls.get("dtype", "torch.float").split('.')[-1]) - if cls.get("data") is not None: - if isinstance(cls.get("data"), str): - raise ValueError("Unimplemented") - else: - data_value = torch.tensor(cls["data"], dtype=data_type).reshape(cls.get("shape"), []) - weight_info[cls["name"]] = { - "info": { - "shape": cls.get("shape", []), - "dtype": data_type, - "device": cls.get("device", "cpu"), - "mean": cls.get("mean", 0.0), - "std": cls.get("std", 1.0), - }, - "data": data_value, - } - - return { - "input_info": input_info if len(input_info) > 0 else None, - "weight_info": weight_info, - "dynamic_shapes": None - } - -def extract_dynamic_shapes(example_inputs): - pass - -def replay_tensor(info): - - device = info["info"]["device"] - dtype = info["info"]["dtype"] - shape = info["info"]["shape"] - - mean = info["info"]["mean"] - std = info["info"]["std"] - if info["data"] is not None: - return info["data"].to(device) - return torch.randn(size=shape).to(dtype).to(device) * std * 0.2 + mean - -def convert_state_and_inputs(state_dict, example_inputs): - def tensor_info(tensor): - is_float = tensor.dtype.is_floating_point - return { - "shape": list(tensor.shape), - "dtype": str(tensor.dtype), - "device": str(tensor.device), - "mean": float(tensor.mean().item()) if is_float else None, - "std": float(tensor.std().item()) if is_float else None, - } - - def process_tensor(tensor): - if not isinstance(tensor, torch.Tensor): - return {"type": "unknown", "value": tensor} - - info = tensor_info(tensor) - if tensor.dtype in [torch.int8, torch.int16, torch.int32, torch.int64]: - if tensor.numel() < 1024: - return {"type": "small_int_tensor", "data": tensor.clone(), "info": info} - else: - return {"type": "big_int_tensor", "data": tensor.clone(), "info": info} - elif tensor.numel() < 1024: - return {"type": "small_tensor", "data": tensor.clone(), "info": info} - else: - return {"type": "random_tensor", "info": info} - - if isinstance(example_inputs, torch.Tensor): - processed_inputs = process_tensor(example_inputs) - elif isinstance(example_inputs, (list, tuple)): - processed_inputs = [process_tensor(t) for t in example_inputs] - else: - processed_inputs = {"type": "unknown", "value": example_inputs} - - processed_weights = {} - for key, tensor in state_dict.items(): - data_value = None - data_type = "random_tensor" - if tensor.dtype in [torch.int8, torch.int16, torch.int32, torch.int64]: - if tensor.numel() < 1024: - data_type = "small_int_tensor" - data_value = tensor.clone() - else: - data_type = "big_int_tensor" - - info = tensor_info(tensor) - processed_weights[key] = {"info": info} - processed_weights[key]["data"] = data_value - processed_weights[key]["type"] = data_type - - # dynamic_shapes = extract_dynamic_shapes(example_inputs) - return { - "input_info": processed_inputs, - "weight_info": processed_weights, - "dynamic_shapes": None - } - -def convert_param_name(original_name): - if original_name.endswith(('.weight', '.bias')): - prefix = 'p_' - base_name = original_name - - elif any(x in original_name for x in ['running_mean', 'running_var', 'num_batches_tracked']): - prefix = 'b_' - base_name = original_name - else: - raise ValueError(f"Unrecognized parameter type: {original_name}") - - if '.' in base_name: - parts = base_name.split('.') - if len(parts) == 2 and not parts[0].startswith('layer'): - return prefix + parts[0] + '_' + parts[1] - else: - # layer1.0 -> layer1___0___ - pattern = r'(layer\d+)\.(\d+)\.' - replacement = r'\1___\2___' - converted = re.sub(pattern, replacement, base_name) - converted = converted.replace('.', '_') - return f"{prefix}getattr_l__self___{converted}" - else: - return prefix + base_name - -def indent_with_tab(code: str) -> str: - lines = code.splitlines() - indented_lines = [f" {line}" for line in lines] - return "\n".join(indented_lines) - -def apply_templates(code: str) -> str: - code = indent_with_tab(code) - code = code.replace(" GraphModule()", "class GraphModule(torch.nn.Module):") - code = code.replace(" \n" * 3, "\n") - py_code = dyn_template.replace('%MODULE', code) - return py_code diff --git a/torch_tools/generators/vision_model_generator.py b/torch_tools/generators/vision_model_generator.py new file mode 100644 index 000000000..069108152 --- /dev/null +++ b/torch_tools/generators/vision_model_generator.py @@ -0,0 +1,101 @@ +import argparse +import os +import json +import torch +import torchvision +from torchvision import transforms +from torch.export import export +from torch import nn +import utils +from utils import convert_param_name, indent_with_tab, apply_templates + + +def main(key, model_path): + # Normalization parameters for ImageNet + normalize = transforms.Normalize( + mean=[0.485, 0.456, 0.406], + std=[0.229, 0.224, 0.225] + ) + + # Create dummy input + batch_size = 1 + height, width = 224, 224 # Standard ImageNet size + num_channels = 3 + random_input = torch.rand(batch_size, num_channels, height, width) + normalized_input = normalize(random_input) + + # Get and initialize model + try: + model = torchvision.models.get_model(key, weights="DEFAULT") + except ValueError as e: + print(f"Error loading model {key}: {e}") + return + + model.eval() + device = torch.device("cuda" if torch.cuda.is_available() else "cpu") + model.to(device) + normalized_input = normalized_input.to(device) + + # Export model + try: + exported = export(model, args=(normalized_input,)) + except Exception as e: + print(f"Error exporting model {key}: {e}") + return + + # Process parameters + params = exported.state_dict + new_params = { + convert_param_name(k): v + for k, v in params.items() + } + + # Generate and save model code + base_code = exported.graph_module.__str__() + write_code = apply_templates(base_code) + + os.makedirs(model_path, exist_ok=True) + + with open(f'{model_path}/model.py', 'w') as fp: + fp.write(write_code) + + # Save metadata + metadata = { + "framework": "torch", + "num_devices_required": 1, + "num_nodes_required": 1 + } + + with open(f'{model_path}/attribute.json', 'w') as f: + json.dump(metadata, f, indent=4) + + # Save tensor metadata and constraints + converted = utils.convert_state_and_inputs(params, exported.example_inputs[0]) + utils.save_converted_to_text( + converted, + file_path=f'{model_path}/source_tensor_meta.py' + ) + utils.save_constraints_text( + converted, + file_path=f'{model_path}/input_tensor_constraints.py' + ) + + +if __name__ == '__main__': + parser = argparse.ArgumentParser( + description="Export torchvision models to txt" + ) + parser.add_argument( + "--key", + type=str, + required=True, + help="Model name from torchvision.models" + ) + parser.add_argument( + "--model-path", + type=str, + required=True, + help="Directory to save the exported model" + ) + args = parser.parse_args() + main(key=args.key, model_path=args.model_path) \ No newline at end of file From 4da0184d8f842c452862ede5895dd8d6724f8d50 Mon Sep 17 00:00:00 2001 From: hxzd5568 <3257591325@qq.com> Date: Thu, 17 Jul 2025 21:12:37 +0800 Subject: [PATCH 04/10] update readme --- README.md | 17 +++++------------ torch_tools/README.md | 4 ++-- 2 files changed, 7 insertions(+), 14 deletions(-) diff --git a/README.md b/README.md index c9e8490c2..93feb372c 100644 --- a/README.md +++ b/README.md @@ -1,20 +1,13 @@ # GraphNet -🧠 GraphNet：高性能编译器优化的基准数据集 -近年来，尽管编译器具备强大的通用优化能力，但其在实际工程中的应用却受到结构复杂、学习成本高、优化 Pass 开发门槛高等因素的限制。为了释放编译器的潜力，我们提出了 GraphNet —— 一个面向高性能编译器优化的大规模基准数据集，旨在为研究者和开发者提供一个统一、开放、可扩展的实验平台。 +## 🧠 GraphNet：高性能编译器优化的基准数据集 +我们提出了 GraphNet —— 一个面向高性能编译器优化的大规模基准数据集，旨在为研究者和开发者提供一个统一、开放、可扩展的实验平台。 -📌 项目简介 -GraphNet 包含大量来自真实高性能计算任务的图结构表示（如计算图、IR图、融合模式等），可用于评估编译器Pass的优化效果、训练编译器智能体、实现从示例到优化Pass的自动生成。 +## 📌 项目简介 +GraphNet 包含大量来自真实高性能计算任务的图结构表示，可用于评估编译器Pass的优化效果、高性能优化方案。 通过 GraphNet，用户可以： 快速测试不同优化策略的效果； 训练模型以自动生成编译器优化Pass； -推动编译器在统一硬件接口、广泛适配后端方面的能力； -降低高性能算法集成到编译器中的门槛。 - -🔍 核心特性 - -📊 数据集概览 - -🧰 使用工具与接口 +降低高性能算法测评的门槛。 diff --git a/torch_tools/README.md b/torch_tools/README.md index 72251c153..15c5059f9 100644 --- a/torch_tools/README.md +++ b/torch_tools/README.md @@ -7,7 +7,7 @@ python vision_model_generator.py --key "model name like resnet18" --model-path ## 运行单侧 -''' +``` python runner.py --model-path ../../sample/torch/resnet18 -''' +``` From ddb8c1c1adfd8ac1613f3e0e3b5e305503c5fef7 Mon Sep 17 00:00:00 2001 From: hxzd5568 <3257591325@qq.com> Date: Thu, 17 Jul 2025 21:25:40 +0800 Subject: [PATCH 05/10] update README.md --- README.md | 13 +++++++++++++ torch_tools/README.md | 13 ------------- 2 files changed, 13 insertions(+), 13 deletions(-) delete mode 100644 torch_tools/README.md diff --git a/README.md b/README.md index 93feb372c..5d40ae5ed 100644 --- a/README.md +++ b/README.md @@ -11,3 +11,16 @@ GraphNet 包含大量来自真实高性能计算任务的图结构表示，可 快速测试不同优化策略的效果； 训练模型以自动生成编译器优化Pass； 降低高性能算法测评的门槛。 + + +## 生成单侧 + +``` +python vision_model_generator.py --key "model-name like resnet18" --model-path "path to be restored at" +``` + +## 运行单侧 +``` +python runner.py --model-path ../../sample/torch/resnet18 +``` + diff --git a/torch_tools/README.md b/torch_tools/README.md deleted file mode 100644 index 15c5059f9..000000000 --- a/torch_tools/README.md +++ /dev/null @@ -1,13 +0,0 @@ -## 生成单侧 - - -``` -python vision_model_generator.py --key "model name like resnet18" --model-path "path to be restored at" -``` - - -## 运行单侧 -``` -python runner.py --model-path ../../sample/torch/resnet18 -``` - From f155ee85de117702b4b2e2346d7a81f82647fdd7 Mon Sep 17 00:00:00 2001 From: hxzd5568 <3257591325@qq.com> Date: Thu, 17 Jul 2025 21:39:18 +0800 Subject: [PATCH 06/10] update dirname --- README.md | 9 +++++---- {torch_tools => graph_net/torch}/execution/runner.py | 0 {torch_tools => graph_net/torch}/execution/utils.py | 0 .../generators => graph_net/torch/extractor}/utils.py | 0 .../torch/extractor/vision_model_extractor.py | 0 5 files changed, 5 insertions(+), 4 deletions(-) rename {torch_tools => graph_net/torch}/execution/runner.py (100%) rename {torch_tools => graph_net/torch}/execution/utils.py (100%) rename {torch_tools/generators => graph_net/torch/extractor}/utils.py (100%) rename torch_tools/generators/vision_model_generator.py => graph_net/torch/extractor/vision_model_extractor.py (100%) diff --git a/README.md b/README.md index 5d40ae5ed..6388dba33 100644 --- a/README.md +++ b/README.md @@ -13,13 +13,14 @@ GraphNet 包含大量来自真实高性能计算任务的图结构表示，可 降低高性能算法测评的门槛。 -## 生成单侧 - +## 单测生成Demo +### torch ``` -python vision_model_generator.py --key "model-name like resnet18" --model-path "path to be restored at" +python vision_model_generator.py --key resnet18 --model-path "path/to/your/generated/graph_net/" ``` -## 运行单侧 +## 单测运行Demo +### torch ``` python runner.py --model-path ../../sample/torch/resnet18 ``` diff --git a/torch_tools/execution/runner.py b/graph_net/torch/execution/runner.py similarity index 100% rename from torch_tools/execution/runner.py rename to graph_net/torch/execution/runner.py diff --git a/torch_tools/execution/utils.py b/graph_net/torch/execution/utils.py similarity index 100% rename from torch_tools/execution/utils.py rename to graph_net/torch/execution/utils.py diff --git a/torch_tools/generators/utils.py b/graph_net/torch/extractor/utils.py similarity index 100% rename from torch_tools/generators/utils.py rename to graph_net/torch/extractor/utils.py diff --git a/torch_tools/generators/vision_model_generator.py b/graph_net/torch/extractor/vision_model_extractor.py similarity index 100% rename from torch_tools/generators/vision_model_generator.py rename to graph_net/torch/extractor/vision_model_extractor.py From 4a3baa86b77139dd1c3511f908c79f755900fad4 Mon Sep 17 00:00:00 2001 From: hxzd5568 <3257591325@qq.com> Date: Thu, 17 Jul 2025 22:01:53 +0800 Subject: [PATCH 07/10] update README.md --- .gitignore | 1 + README.md | 6 ++++-- .../__pycache__/utils.cpython-310.pyc | Bin 0 -> 6023 bytes .../vision_model_extractor.cpython-310.pyc | Bin 0 -> 2677 bytes graph_net/torch/extractor/utils.py | 17 ----------------- .../torch/extractor/vision_model_extractor.py | 4 ++-- .../single_device_runner.cpython-310.pyc | Bin 0 -> 2147 bytes .../runner/__pycache__/utils.cpython-310.pyc | Bin 0 -> 3269 bytes .../single_device_runner.py} | 2 +- graph_net/torch/{execution => runner}/utils.py | 2 +- 10 files changed, 9 insertions(+), 23 deletions(-) create mode 100644 .gitignore create mode 100644 graph_net/torch/extractor/__pycache__/utils.cpython-310.pyc create mode 100644 graph_net/torch/extractor/__pycache__/vision_model_extractor.cpython-310.pyc create mode 100644 graph_net/torch/runner/__pycache__/single_device_runner.cpython-310.pyc create mode 100644 graph_net/torch/runner/__pycache__/utils.cpython-310.pyc rename graph_net/torch/{execution/runner.py => runner/single_device_runner.py} (97%) rename graph_net/torch/{execution => runner}/utils.py (98%) diff --git a/.gitignore b/.gitignore new file mode 100644 index 000000000..c18dd8d83 --- /dev/null +++ b/.gitignore @@ -0,0 +1 @@ +__pycache__/ diff --git a/README.md b/README.md index 6388dba33..4d469031a 100644 --- a/README.md +++ b/README.md @@ -16,12 +16,14 @@ GraphNet 包含大量来自真实高性能计算任务的图结构表示，可 ## 单测生成Demo ### torch ``` -python vision_model_generator.py --key resnet18 --model-path "path/to/your/generated/graph_net/" +export PYTHONPATH=$PYTHONPATH:/path/to/your/GraphNet/repo +python3 -m graph_net.torch.extractor.vision_model_extractor --key resnet18 --model-path /path/to/your/extracted/graph_net/sample ``` ## 单测运行Demo ### torch ``` -python runner.py --model-path ../../sample/torch/resnet18 +export PYTHONPATH=$PYTHONPATH:/path/to/your/GraphNet/repo +python3 -m graph_net.torch.runner.single_device_runner --model-path /path/to/your/extracted/graph_net/sample ``` diff --git a/graph_net/torch/extractor/__pycache__/utils.cpython-310.pyc b/graph_net/torch/extractor/__pycache__/utils.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..921d38e4476fb07b2abf944fdc5ffe314d2700fe GIT binary patch literal 6023 zcmb7ITW=f372cV>@*?VH%a=Ab>%d7E+fr<&ZtA*Dia1TuCQMq_xiAjc6=x}}O)lx# zl`U1bWdks(Lg+@V`jiRuGeb;D|gcFp78LR)tK4WQWPb$R76?KVC<}@h*{)w zVouB>pBL<^QCoO~FEi>jrb-`35y(JX30p~8D4u-(gDcnGd#l)DG+^{|k=sZnQv6qt z9b<@7?J`*ue86^$0gnzP?4HqP!W7nBPuN$CZMMziQQ?gG-GmEw$N19tif^;uGvQ_Y zwjn>u_|O#IE*qGMm6_X26mD|)W@d`w5o2hJ(k_eN%vfGOWQg)11Lc_`#v$XXAu0n) z%nt0t8M>L3*%{w82TtY;J#eMG%TZ=u8e%SUu&a61z&f1f)#mFpp1R9-f^cm;N$us; z(2o`GH>VEGM+ZJnhdvDQJ0mkbcLyxPfo~fh8O*rF1(tir>JN>FT=C^s6x)2r)J#u? zYhmPfnvvfPBysNXyDRIxIBBl;y1~k3=%-<8-S4gV{ieSbM9IqA((kX=gJflm3e5;`_MO5@sXU#i7uquX5dlh%3=H)-daflz$4nLc+d zl0mDt7KLdbs(ObY31l_d>IWCAiY-4Peo2xVWq16oKq}UJMAdZhC0)NRPSvh2zm#Wb z@%q_jvw8g-NMm6wNc<#`%}%o!2c0(BY6ay5k%;ev$+|NAXiF78@;e*BTT=F`BdLBHd-f|@HyQ~v;vb!_x{XpaZF>WRmD%AuEr%PxUxE7luK)VXno$|rA`A*#Lw7|>$I80V3!|DTU!?79-PWk&D+Qvd>-pBC zGC99zpV)||#tw)Fh8CGzLGc?opjr3|dJ}K}Oe0omHo`9CWmqaICmii|W}pn!25+2t zI#0~eI{AK2Y;=N`YU%T>jvvQUjaoyOqi8uFRJ)kY(!An?$W!=$)EyaBoSZ&a9tTln zu@yCwpbI%o0$c$dtl{>znj{=gal!BRJ6n6hlxx2gRFZP#t#-1Wu+7}>(jV8b8q z0S}FV3EtYf#uel9i`V(U8rZ@ZIHE9cGy8^v)miaNGV^XlC1~H?^D?dAuRF&icl&i38EN=&>JuGq`;K&0nIRK_h@>&0OG*{A+aU4 zz?`qS?R=e{8Jja$(Ija4k-(JxMiNgHdf!SQH?;4y^Drt1`R6Z`7>#=Zd&oazIl0DPc5 zYsI>9H=>(S?@k2AZ0YN;Z<7OgMCMV=#a+MC!BG?F6mb%&`4b#LS;9|z@EUAWmDO-< z-w;*KCyiJ?Bf+d*cRZi#`SPb&q*heLFb*TIG{W`U+Ds_tS9+sd0I)&wnzBL9UeMY3 zb2_{5s{Oe+6-VC<%t*aslV^fQ*D(Er^a(+mX0&rRLR*Qx9)`!21 zmt;<2@`i92#qw+s&hr4I!s2sq9Tsp3xe(wNpcZl`6^1Uxne%Vi#H1;=LSo<{X$SiW z3TuG>&~^sXY-})DPN4Y&@e|lmVQ%w3&BH;MLoTeLNvLI)G2>&5wea6IDe$>!oHgQ; zchFA%Pm!ptX4mD%SU)}e4pa^3V6BSd!8rQRo46Xr)u@+LLt-$2Ho(%YAX!%Ij`q4W zOD)!W$-9&896MUQK2W25N8(Q987yD3?zX9E|wi!I`{K(@LGK=tCW{V>HBjqJg{v$__@;ojF z5!2Ar9=d|@4*ac?xucOTBE)>;kKhcMnAKP2WggdF@CEgWs?NAf~q zs!dY@4>9EART1_9)9#YBHtUbs8ATWILyS|_#zrWVP4rcbmG<4k#Q2pd2$G~~7?;rc zgnbUM^z`_aC>DaIB|Uvj6Nk2c;*^o2Xk`y82vop0W5bXRh0GK0CYQ;V-LT|w!NDSP z=;UEJyb&-1A#{>)_^xrULqYqKy@^MD5vO1^9(=AFvJk{c zxE$>Qw5&`7rnK)ocW4)i+dP;((%NI}7))P!>i^)0nwhfhQ_4=zoFr+H_Gv?|0El^x ziH_jyKvp4HChR3a;e#mKP)~~Uc=WwdkAr8BW5Pwt;ert@jaoglp58xx-{=bo&=1*N zNE#Q#kOG?A?!pkDjOP`!&#Y0qt;={ALK(0c?RXqQ8L%4V1+j=S&=ej|4(zaPdbBYI zg&lgTIhYmDy7ZMvR$R9n())37c$cH+$iAMVnN22XZ{&B#L`{wOEjzF1Ax@kUnlkw$L18b(#%!)gw_s=YQE68#!d%^fsJYFf%e z#2m&GxX*FD+3Ig-RKJEM{nNmHRQu=ajhPXo!wU@%KxlW+9Rw1Xu_N&Xb#xpVGip zBIF`-=uyznLkemj%cn{cyb2^3ip~7X=%e)ZGN6(+Y6B{1E-roZiD^b}uj>&A@0aP^r>(x~o&K zwcUBNsu5CF;(#KMkhm->lI4WNpMba{?Vf>>&>}dXoH%gfG2yd2okZe7S^jK)<@)?R zf3CdQ^btHixjY*E17?5Ki`$O_i?2bKe*nS|!wI7E5mRi|gc37a)H0(@Z5T;nr!`u` zdaso@X`R;1+)k|2rLLLd#7ixLFiL$x8?<5gHQ)y{NSm~owrC4@4y%8FXqzuFcb|wc zN8@GYG5-|P4)2UtSc9*!fHhg`)W*ng+J>{V;GC_UBf8Gpe0_We`UYD*B_J0+H@(-n z0PE1R9PqD=1PB9*Xm_)(Dy1v8RmJ*(;r^6Y>M6&8_*(lCosUbW+l@l1t>u^NRc z+O67AQ6$Hq;%SjYipv4f!mpk+7`4aWhAsnmqtF-+L5q(yB_CsEost=0_RM0n?~<84 zt1)NhC~{_%R_T;>Sv$99^^%nJ6FhULzOu))u~Q;eXD)o{y+qjq8>#y7I~dJ;jHb)V zWnPIsM9eS!579^V3C8FpdJiHq^S|#i1U@jcLwtxz@60=gh`@QiBICwcaE`Yz)pIfuem6SVIQkG^b>hAbn*(q1p2D^I--~W%S zzucCkHqY*tP0;lLyLT6zli8};9KTks8r?QQ>V2c@=J??i@4E_V$mGzy0fvJMi)WWncO1o1g9sk1t+-)BW`)JHt;uG0#qq zP&?&OMlGq>1%~|Uk&8dWAUq%!77$v0^7NgzzWwzVJxuF|JRXjeoUT3J$-*XNt(WpT58ppRQT`Xn3rbLjFOkQkNddmbb+ zo)Lvge1c2zQ}RA}j(&xDSkzz()E3!(qr&?YJ`hb{81hfrALUYoqdevPqd1zzgHe?C zqauukP;&iuMO2KQan&E{NtnS*H=|)L)=T}a=MX5akl*^$p=v)A8+4m1=m$BKZ-%Ce1GcojbJSfA67MUc+aBV z9-sq1%*8%k%_eEc_(442G8Fv%Nh~;{9nHJ_nyB{ef;@yej+hdt$7utsqC})l@ zLc4gn-k14A3^<@mCUX&{Tt#p?Z1eWwgaal&Wedz>;cS@;Ep&`Qg7`FjYrgb z1{fD5@staVNz?}1V^zbLFn1btjB%2BL$2l?to-jp$%LCEsA`1(nU&2CqNH->Hj0Lp z6~r=(4x%`Tb`wqs$kaJec~6f9yigDi>PrRC80rXAwHTmM(qJA$ICGNlsvd@*bQp@d z{*vUfa?@y^vslQ=&I_JZwL<|PR<`z4WwS|IRJDnU6G=A~C_O&}>TZ1UvufI2~AZY#UatmTc?M ze24V6^TTl3eQNCN-b%qa8cASv)e&67v;ieE5V1+3X>Dr@!trAIxo;TE(zkJU zPvq%s3ERe(67a=V#|5YEq6RB_#FJtgY;BnxTc&cp@C3?J@YZNQp9}(PJ*(0tU*XV0^%YUf~Z?y+6E+uYq{_0AUl_~r9*GR|8Lv5 LnY+#^w*UTL(GvL5 literal 0 HcmV?d00001 diff --git a/graph_net/torch/extractor/utils.py b/graph_net/torch/extractor/utils.py index ea4ec0d0f..22db3db25 100644 --- a/graph_net/torch/extractor/utils.py +++ b/graph_net/torch/extractor/utils.py @@ -9,23 +9,6 @@ dyn_template = """ %MODULE - -model = GraphModule() - -inputs_params = utils.load_converted_from_text(f'./source_tensor_meta.py') -inputs = inputs_params["input_info"] -inputs = [utils.replay_tensor(i) for i in inputs] -params = inputs_params["weight_info"] - -state_dict = {} -for k, v in params.items(): - k = utils.convert_param_name(k) - v = utils.replay_tensor(v) - state_dict[k] = v - -y = model(x=inputs[0], **state_dict)[0] -print(torch.argmin(y), torch.argmax(y)) -print(y.shape) """ def convert_param_name(original_name): diff --git a/graph_net/torch/extractor/vision_model_extractor.py b/graph_net/torch/extractor/vision_model_extractor.py index 069108152..7c28881c7 100644 --- a/graph_net/torch/extractor/vision_model_extractor.py +++ b/graph_net/torch/extractor/vision_model_extractor.py @@ -6,8 +6,8 @@ from torchvision import transforms from torch.export import export from torch import nn -import utils -from utils import convert_param_name, indent_with_tab, apply_templates +import graph_net.torch.extractor.utils as utils +from graph_net.torch.extractor.utils import convert_param_name, indent_with_tab, apply_templates def main(key, model_path): diff --git a/graph_net/torch/runner/__pycache__/single_device_runner.cpython-310.pyc b/graph_net/torch/runner/__pycache__/single_device_runner.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..6742c2d10a338d486bbb9ed4c2003ebc9dd6cbfc GIT binary patch literal 2147 zcmZ8i&2Jnv6t`!*GqbZF-K1%nR8<>3q-uqdrMDiUii%2{h$^KBVNs2SwRd;6^RczP zX%fv|NF)vr5El+qMTt});KHp4fjIDI_zEo`A#sO8c=jX}bTDj?c5^T}MQm zjL)(bYad{`z!%2n*cmwAVRLN$fMCQI4i}FQJl+%2-P%+QR<8jN0nhsdOY-4vGsz}}P%Eks{gK8Y zD800Z85f%HZQcj80t1mng=7We-7~_VLVdwyk#2EHq~e(-#f0aY2p%(CuZ*V2dzwsD zlIo6}@P0H9MHZbH?|`W(Aa7+xt6jMxwO@&0(l4?J_!S0q9YhgJ`YMXFKjbP_N_3l| z1{whbMw*x#(#DBNdsRH3twJQjB#+Yw-U{mKQjt`dqAl2)3=+<)P7`|ruQEVeC%#XV z#x4gVh8f;-VK%qIk%2)q6w4lhFoOy`h;zIm|4; zJu%I5E_xE6hsQI%mGpU33D+h&dcmMmd8&{x537Sic@erM_Hf4u@x^1}v~dSFafsW{ zHywh-3kHa=x~idbAA{iZGYN0tvebPH6GL4f9De9nvN|(d;ys|T!iSF4kcSSC%>@e9 zmeGtbNY8y#lAAzyptZ{akk%3e#H}5=N9ZQHGykqL^MR`Pkd!{0^Q=y%4%gE}D#NDhW-X1g zMYOw=#Yw)lo)-N$mDgc%x`{UK^JF-xuBq|1Ffp|Dx(yp(@e&B=QpJ)Y-{L|6g>4Qg zzO8gEfmD=qzF+B}u8BCa6-~ShCv|Nik{q^KGEy%V!z{_GG2YfSIf^H|8;Tc=+e^k( z&4Nu^7Cl+nlhmu0+Ozg04JG6x&>2!xLti_rHC<|ND2}KKW_?uOB~p^5y$i*4BFPC*y3Aa(SAZTtcQ^ zy>_MRYNFs@1PuhgH%$a*nv8fl0g@SLy36$9v&3GrSgxJqu%(B1#XP{PL{-QFfEvJ( zT(_Uq#x=89%)wsjn|njKw3StR;2R2ra6u)A>CPF0u2h0LImD}?e8q@UuuRp9BHF$o zhSLmc&@JOlP!l3EHe<#jxP;Zw&`Ql(YGV2-Y?Wx+Cj(d7$ zt83P~qaMj*gYYGSE5t}M;^YHYBrb5`z`xK(IQfP+p+yjWuXI~e_wLDL~(6e%W z=u1bsAF{9}J?Y~c$eIlBt;@P>;MK{MOx@Ur$Lu%akbl5s zP>4OI?-pX_$Y9Q=j&Td;kjwf5p>GzBY^<@FC!2Gg-z<35UuCkj%D}hR*ebit#jc6!%nSHh-YwQGZLB%^opP2F?t1 zx4)^!S*G;n*d%GbF*+#g*CDU=NBTyK{)=>rM`zgNtKDI?r?iPiv5p7U@W%CY*p1Wt z#>VqC9Vcv&#w6=S0~Kds^V8m5ti#jUco5x>O}DS|$mqEHK*>jMD4oV5F|Cmwe+HIZIQ>gkBmSC|&aC8FeS8FWLUK{6=Sc<5KmRG`6oqQM7#( zBs|qqCN@S#X%yuu-N9I=UIr?Y`NPEYODE0_%Eo(fI#zFKJ=CR_kJ7}HPO7r9k(*eX zN>FH3I{Em1S<`Bi#$DB^=>We_`XjA&lKrrLKhBl)OzB|{CNFCvSXP?cP%RnYEW-N2 zl}g&9(-i%>PMS`A-pMqXzXsCaF8BEvL3Bw3qQTGcv)pxDzQkK-fzan6b}YnB&^ol6 zX!-ZyL&>H1kWINks7dEg{Gv4#$PhkfQ|As+-Nk$8aFzvE5R(r%M-X!MW3(PaF0iRr zIJelnD>e%RR#7$p1dyrM3c$jc>+clq47oDL+63~u7o}JD$i#Ob=@C%O#kQE%ra@7g zk%n3k*l&RB2@W~3tahF4UR-GV$qQ}xgE}-e3KxIyyJ=I_ilz+KkSh95vR*W2zHIc^ zT$q{(Y_CZ<3Y+aYmrHBxP}tEEa(P~HC*`RHcUrE@g?#0SC(k?))0UN6#VXp*A55hi zebYD;`nTB`6TIJq?bS6_wB*?}CYxw$ABkDDvN7lMADCOc&Wbv`(IRi`O8Bq+#eX+l zI%fADpRxN_)0XE5*FC5EP@gDa3vcQUJPftQ24zP>Y2CM${1S%RS4_0OEk% zbCBdVbk;GC!1$S1`*gmBfZBR7IujPc;Z}qz$MLUCox-tpoJpM^z5;i0^BuN(wF+Ar zFJOrRc=u`{ZnKF8ITVgNu(Ob1K4!74`c~;~4Kp=Qgk{O%FI z-0=VnMkgb^f(d;Jr1ay_2uM=8YA;SZZdk`>tRjd3VNyluo|nE>`8YKtub{e0dYkV} zzyu(}E}|UIzYY@c0I5!G1&Q6@r))3a0vu8wJ%Qx+Au)M99mXjJ6vlCBD!?r2=P1o}M;+XrxI<+=_?oce7+LsvN3hN7zYCmD3i5j#DQFBdXHV8a*(hs^+EIFT?=X@l8A?LdrBYyY=o>|z3Q-cQ zdW|a<)2BgTr@oV<3iZuiHx_=HMGAOb`h zHHxccv;eUu9#NTbQF0(lQ76tpf?oGu{(qJl%ZSO#m~q7g#3ij-7Y#=|ml3t)f3CCy zr2K#Dw_x%Za!~oEW{VU3I3b0m8ctKZi zs$b#yhZyWMD@#EI0$ElHv}SjhrmAa_VU}xhc`1gueg%xZi`X-_XJB=_bbc>>BDg*n Z%5kc0P$=Y-dI2N*6oH7!wcuXM`7a4=BKrUU literal 0 HcmV?d00001 diff --git a/graph_net/torch/execution/runner.py b/graph_net/torch/runner/single_device_runner.py similarity index 97% rename from graph_net/torch/execution/runner.py rename to graph_net/torch/runner/single_device_runner.py index d63086afa..12fc9372d 100644 --- a/graph_net/torch/execution/runner.py +++ b/graph_net/torch/runner/single_device_runner.py @@ -1,4 +1,4 @@ -import utils +import graph_net.torch.runner.utils as utils import argparse import importlib.util import torch diff --git a/graph_net/torch/execution/utils.py b/graph_net/torch/runner/utils.py similarity index 98% rename from graph_net/torch/execution/utils.py rename to graph_net/torch/runner/utils.py index 9ddfccbbb..3eb24e1a5 100644 --- a/graph_net/torch/execution/utils.py +++ b/graph_net/torch/runner/utils.py @@ -133,6 +133,6 @@ def replay_tensor(info): mean = info["info"]["mean"] std = info["info"]["std"] - if info["data"] is not None: + if "data" in info and info["data"] is not None: return info["data"].to(device) return torch.randn(size=shape).to(dtype).to(device) * std * 0.2 + mean From 62c25ef89c255ad3081c02289255ee94cb1bb9c6 Mon Sep 17 00:00:00 2001 From: hxzd5568 <3257591325@qq.com> Date: Thu, 17 Jul 2025 22:14:03 +0800 Subject: [PATCH 08/10] Update README.md --- README.md | 14 +++++--------- 1 file changed, 5 insertions(+), 9 deletions(-) diff --git a/README.md b/README.md index 4d469031a..509605194 100644 --- a/README.md +++ b/README.md @@ -1,26 +1,22 @@ # GraphNet -## 🧠 GraphNet：高性能编译器优化的基准数据集 -我们提出了 GraphNet —— 一个面向高性能编译器优化的大规模基准数据集，旨在为研究者和开发者提供一个统一、开放、可扩展的实验平台。 - ## 📌 项目简介 -GraphNet 包含大量来自真实高性能计算任务的图结构表示，可用于评估编译器Pass的优化效果、高性能优化方案。 +GraphNet —— 一个面向编译器开发的大规模数据集，旨在为研究者提供一个统一、开放的实验平台。其中包含大量来自真实模型的计算图，方便评估不同编译器Pass的优化效果。 通过 GraphNet，用户可以： -快速测试不同优化策略的效果； -训练模型以自动生成编译器优化Pass； -降低高性能算法测评的门槛。 +1. 快速测试不同编译器策略的通用优化效果 +2. 训练AI-for-system模型以自动生成编译器优化Pass -## 单测生成Demo +## 计算图抽取Demo ### torch ``` export PYTHONPATH=$PYTHONPATH:/path/to/your/GraphNet/repo python3 -m graph_net.torch.extractor.vision_model_extractor --key resnet18 --model-path /path/to/your/extracted/graph_net/sample ``` -## 单测运行Demo +## 计算图运行Demo ### torch ``` export PYTHONPATH=$PYTHONPATH:/path/to/your/GraphNet/repo From 50a3de944196b38253d838bd61d2db47a15ecdee Mon Sep 17 00:00:00 2001 From: hxzd5568 <3257591325@qq.com> Date: Fri, 18 Jul 2025 08:44:06 +0800 Subject: [PATCH 09/10] Update storage method --- graph_net/torch/extractor/utils.py | 31 ++++++++++--------- .../torch/extractor/vision_model_extractor.py | 2 +- .../torch/runner/single_device_runner.py | 2 +- graph_net/torch/runner/utils.py | 5 +-- 4 files changed, 22 insertions(+), 18 deletions(-) diff --git a/graph_net/torch/extractor/utils.py b/graph_net/torch/extractor/utils.py index 22db3db25..4cb562330 100644 --- a/graph_net/torch/extractor/utils.py +++ b/graph_net/torch/extractor/utils.py @@ -126,12 +126,14 @@ def format_data(data): return "[{}]".format(", ".join(f'{x}' for x in data.tolist())) else: return repr(data) - - lines = [] + + lines = [[],[]] def process_tensor_info(tensor_info, name_prefix="example_input"): data_list = None + file_index = 1 if "input_" in tensor_info["name"]: + file_index = 0 if tensor_info["type"] in ["small_tensor", "small_int_tensor"]: data_list = tensor_info["data"].flatten() elif tensor_info["type"] == "big_int_tensor": @@ -149,17 +151,16 @@ def process_tensor_info(tensor_info, name_prefix="example_input"): device = info.get("device", "cpu") mean = info.get("mean", 0.0) std = info.get("std", 1.0) - uid = f"{name_prefix}_tensor_meta_{generate_uid()}" - lines.append(f"class {uid}:") - lines.append(f"\tname = \"{tensor_info.get('name', '')}\"") - lines.append(f"\tshape = {shape}") - lines.append(f"\tdtype = \"{dtype}\"") - lines.append(f"\tdevice = \"{device}\"") - lines.append(f"\tmean = {mean}") - lines.append(f"\tstd = {std}") - lines.append(f"\tdata = {format_data(data_list)}") - lines.append("") + lines[file_index].append(f"class {uid}:") + lines[file_index].append(f"\tname = \"{tensor_info.get('name', '')}\"") + lines[file_index].append(f"\tshape = {shape}") + lines[file_index].append(f"\tdtype = \"{dtype}\"") + lines[file_index].append(f"\tdevice = \"{device}\"") + lines[file_index].append(f"\tmean = {mean}") + lines[file_index].append(f"\tstd = {std}") + lines[file_index].append(f"\tdata = {format_data(data_list)}") + lines[file_index].append("") input_infos = converted["input_info"] if isinstance(input_infos, dict): @@ -173,5 +174,7 @@ def process_tensor_info(tensor_info, name_prefix="example_input"): weight_info["name"] = name process_tensor_info(weight_info, name_prefix="Program_weight") - with open(file_path, 'w') as f: - f.write("\n".join(lines)) + with open(f"{file_path}/input_meta.py", 'w') as f: + f.write("\n".join(lines[0])) + with open(f"{file_path}/weight_meta.py", 'w') as f: + f.write("\n".join(lines[1])) diff --git a/graph_net/torch/extractor/vision_model_extractor.py b/graph_net/torch/extractor/vision_model_extractor.py index 7c28881c7..3348ce41c 100644 --- a/graph_net/torch/extractor/vision_model_extractor.py +++ b/graph_net/torch/extractor/vision_model_extractor.py @@ -73,7 +73,7 @@ def main(key, model_path): converted = utils.convert_state_and_inputs(params, exported.example_inputs[0]) utils.save_converted_to_text( converted, - file_path=f'{model_path}/source_tensor_meta.py' + file_path=f'{model_path}' ) utils.save_constraints_text( converted, diff --git a/graph_net/torch/runner/single_device_runner.py b/graph_net/torch/runner/single_device_runner.py index 12fc9372d..5982cbff6 100644 --- a/graph_net/torch/runner/single_device_runner.py +++ b/graph_net/torch/runner/single_device_runner.py @@ -27,7 +27,7 @@ def main(model_path: str): model_class = load_class_from_file(f"{model_path}/model.py", class_name="GraphModule") model = model_class() - inputs_params = utils.load_converted_from_text(f'{model_path}/source_tensor_meta.py') + inputs_params = utils.load_converted_from_text(f'{model_path}') inputs = inputs_params["input_info"] inputs = [utils.replay_tensor(i) for i in inputs] params = inputs_params["weight_info"] diff --git a/graph_net/torch/runner/utils.py b/graph_net/torch/runner/utils.py index 3eb24e1a5..d3162dadc 100644 --- a/graph_net/torch/runner/utils.py +++ b/graph_net/torch/runner/utils.py @@ -55,9 +55,10 @@ def parse_value(value_str): except: return value_str - with open(file_path, 'r') as f: + with open(f"{file_path}/input_meta.py", 'r') as f: lines = f.readlines() - + with open(f"{file_path}/weight_meta.py", 'r') as f: + lines += f.readlines() classes = [] current_class = None From d0dd364506967b6fec1f80965a13b2dc89d0af85 Mon Sep 17 00:00:00 2001 From: hxzd5568 <3257591325@qq.com> Date: Fri, 18 Jul 2025 08:46:03 +0800 Subject: [PATCH 10/10] Update utils --- graph_net/torch/extractor/utils.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/graph_net/torch/extractor/utils.py b/graph_net/torch/extractor/utils.py index 4cb562330..594f5419f 100644 --- a/graph_net/torch/extractor/utils.py +++ b/graph_net/torch/extractor/utils.py @@ -126,7 +126,7 @@ def format_data(data): return "[{}]".format(", ".join(f'{x}' for x in data.tolist())) else: return repr(data) - + lines = [[],[]] def process_tensor_info(tensor_info, name_prefix="example_input"):