单测case覆盖

paconvert/api_mapping.json

@@ -6029,7 +6029,11 @@
     "args_list": [
       "graceful",
       "timeout"
-    ]
+    ],
+    "kwargs_change": {
+      "graceful": "",
+      "timeout": ""
+    }
   },
   "torch.distributed.scatter": {
     "Matcher": "ScatterMatcher",
@@ -11811,8 +11815,8 @@
       "input",
       "grid",
       "mode",
-      "align_corners",
-      "padding_mode"
+      "padding_mode",
+      "align_corners"
     ],
     "kwargs_change": {
       "input": "x"

tests/test_cuda_amp_autocast.py

@@ -90,9 +90,8 @@ def test_case_4():
 
 
 @pytest.mark.skipif(
-    condition=not paddle.device.is_compiled_with_cuda()
-    or not paddle.device.cuda.get_device_properties(0).major >= 8,
-    reason="computational capabilities less 8",
+    condition=not paddle.device.is_compiled_with_cuda(),
+    reason="can only run on paddle with CUDA",
 )
 def test_case_5():
     pytorch_code = textwrap.dedent(
@@ -110,11 +109,9 @@ def test_case_5():
     obj.run(pytorch_code, ["result"])
 
 
-# generated by validate_unittest autofix, based on test_case_5
 @pytest.mark.skipif(
-    condition=not paddle.device.is_compiled_with_cuda()
-    or not paddle.device.cuda.get_device_properties(0).major >= 8,
-    reason="computational capabilities less 8",
+    condition=not paddle.device.is_compiled_with_cuda(),
+    reason="can only run on paddle with CUDA",
 )
 def test_case_6():
     pytorch_code = textwrap.dedent(
@@ -134,9 +131,8 @@ def test_case_6():
 
 # generated by validate_unittest autofix, based on test_case_5
 @pytest.mark.skipif(
-    condition=not paddle.device.is_compiled_with_cuda()
-    or not paddle.device.cuda.get_device_properties(0).major >= 8,
-    reason="computational capabilities less 8",
+    condition=not paddle.device.is_compiled_with_cuda(),
+    reason="can only run on paddle with CUDA",
 )
 def test_case_7():
     pytorch_code = textwrap.dedent(

tests/test_diff.py

@@ -130,3 +130,15 @@ def test_case_9():
         """
     )
     obj.run(pytorch_code, ["result"])
+
+
+def test_case_10():
+    pytorch_code = textwrap.dedent(
+        """
+        import torch
+        x = torch.tensor([1, 3, 2])
+        b = torch.tensor([4, 5])
+        result = torch.diff(x, 1, 0, b, b)
+        """
+    )
+    obj.run(pytorch_code, ["result"])

tests/test_distributed_rpc_shutdown.py

@@ -14,17 +14,11 @@
 
 import textwrap
 
-import paddle
-import pytest
 from apibase import APIBase
 
 obj = APIBase("torch.distributed.rpc.shutdown")
 
 
-@pytest.mark.skipif(
-    condition=paddle.is_compiled_with_cinn(),
-    reason="WITH_RPC = OFF, if WITH_CINN = ON.",
-)
 def test_case_1():
     pytorch_code = textwrap.dedent(
         """
@@ -56,3 +50,102 @@ def test_case_1():
         """
     )
     obj.run(pytorch_code, ["result"])
+
+
+def test_case_2():
+    pytorch_code = textwrap.dedent(
+        """
+        import os
+        import torch
+        import socket
+        s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
+        start = 25000
+        end = 30000
+        for port in range(start, end):
+            try:
+                s.bind(('localhost', port))
+                s.close()
+                break
+            except socket.error:
+                continue
+        print("port: " + str(port))
+
+        from torch.distributed import rpc
+        os.environ['MASTER_ADDR'] = 'localhost'
+        os.environ['MASTER_PORT'] = str(port)
+        os.environ['PADDLE_MASTER_ENDPOINT'] = 'localhost:' + str(port)
+        rpc.init_rpc(
+            "worker1",
+            rank=0,
+            world_size=1
+        )
+        result = rpc.shutdown(graceful=False, timeout=2)
+        """
+    )
+    obj.run(pytorch_code, ["result"])
+
+
+def test_case_3():
+    pytorch_code = textwrap.dedent(
+        """
+        import os
+        import torch
+        import socket
+        s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
+        start = 25000
+        end = 30000
+        for port in range(start, end):
+            try:
+                s.bind(('localhost', port))
+                s.close()
+                break
+            except socket.error:
+                continue
+        print("port: " + str(port))
+
+        from torch.distributed import rpc
+        os.environ['MASTER_ADDR'] = 'localhost'
+        os.environ['MASTER_PORT'] = str(port)
+        os.environ['PADDLE_MASTER_ENDPOINT'] = 'localhost:' + str(port)
+        rpc.init_rpc(
+            "worker1",
+            rank=0,
+            world_size=1
+        )
+        result = rpc.shutdown(timeout=2, graceful=False)
+        """
+    )
+    obj.run(pytorch_code, ["result"])
+
+
+def test_case_4():
+    pytorch_code = textwrap.dedent(
+        """
+        import os
+        import torch
+        import socket
+        s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
+        start = 25000
+        end = 30000
+        for port in range(start, end):
+            try:
+                s.bind(('localhost', port))
+                s.close()
+                break
+            except socket.error:
+                continue
+        print("port: " + str(port))
+
+        from torch.distributed import rpc
+        os.environ['MASTER_ADDR'] = 'localhost'
+        os.environ['MASTER_PORT'] = str(port)
+        os.environ['PADDLE_MASTER_ENDPOINT'] = 'localhost:' + str(port)
+        rpc.init_rpc(
+            "worker1",
+            rank=0,
+            world_size=1
+        )
+        result = rpc.shutdown(True, 1)
+        """
+    )
+    obj.run(pytorch_code, ["result"])

tests/test_distributions_AffineTransform.py

@@ -45,3 +45,45 @@ def test_case_2():
         """
     )
     obj.run(pytorch_code, ["result"])
+
+
+def test_case_3():
+    pytorch_code = textwrap.dedent(
+        """
+        import torch
+
+        x = torch.tensor(1.)
+        y = torch.tensor(2.)
+        affine = torch.distributions.AffineTransform(x, y, 1, 1)
+        result = affine.forward_shape([1, 2])
+        """
+    )
+    obj.run(pytorch_code, ["result"])
+
+
+def test_case_4():
+    pytorch_code = textwrap.dedent(
+        """
+        import torch
+
+        x = torch.tensor(1.)
+        y = torch.tensor(2.)
+        affine = torch.distributions.AffineTransform(loc=x, scale=y, event_dim=1, cache_size=1)
+        result = affine.forward_shape([1, 2])
+        """
+    )
+    obj.run(pytorch_code, ["result"])
+
+
+def test_case_5():
+    pytorch_code = textwrap.dedent(
+        """
+        import torch
+
+        x = torch.tensor(1.)
+        y = torch.tensor(2.)
+        affine = torch.distributions.AffineTransform(event_dim=1, cache_size=1, loc=x, scale=y)
+        result = affine.forward_shape([1, 2])
+        """
+    )
+    obj.run(pytorch_code, ["result"])

tests/test_distributions_ExpTransform.py

@@ -29,3 +29,27 @@ def test_case_1():
         """
     )
     obj.run(pytorch_code, ["result"])
+
+
+def test_case_2():
+    pytorch_code = textwrap.dedent(
+        """
+        import torch
+
+        exp = torch.distributions.ExpTransform(cache_size=1)
+        result = exp.forward_shape([1, 2])
+        """
+    )
+    obj.run(pytorch_code, ["result"])
+
+
+def test_case_3():
+    pytorch_code = textwrap.dedent(
+        """
+        import torch
+
+        exp = torch.distributions.ExpTransform(1)
+        result = exp.forward_shape([1, 2])
+        """
+    )
+    obj.run(pytorch_code, ["result"])

tests/test_distributions_PowerTransform.py

@@ -29,3 +29,39 @@ def test_case_1():
         """
     )
     obj.run(pytorch_code, ["result"])
+
+
+def test_case_2():
+    pytorch_code = textwrap.dedent(
+        """
+        import torch
+
+        power = torch.distributions.PowerTransform(exponent=torch.tensor(2.),cache_size=1)
+        result = power.forward_shape([1, 2])
+        """
+    )
+    obj.run(pytorch_code, ["result"])
+
+
+def test_case_3():
+    pytorch_code = textwrap.dedent(
+        """
+        import torch
+
+        power = torch.distributions.PowerTransform(cache_size=1, exponent=torch.tensor(2.))
+        result = power.forward_shape([1, 2])
+        """
+    )
+    obj.run(pytorch_code, ["result"])
+
+
+def test_case_4():
+    pytorch_code = textwrap.dedent(
+        """
+        import torch
+
+        power = torch.distributions.PowerTransform(torch.tensor(2.), 0)
+        result = power.forward_shape([1, 2])
+        """
+    )
+    obj.run(pytorch_code, ["result"])

tests/test_histogramdd.py

@@ -87,3 +87,16 @@ def test_case_5():
         """
     )
     obj.run(pytorch_code, ["result"])
+
+
+def test_case_6():
+    pytorch_code = textwrap.dedent(
+        """
+        import torch
+        x = torch.tensor([[0., 1.], [1., 0.], [2.,0.], [2., 2.]])
+        bins = [3,3]
+        weights = torch.tensor([1., 2., 4., 8.])
+        result = torch.histogramdd(x, bins)
+        """
+    )
+    obj.run(pytorch_code, ["result"])

tests/test_hub_download_url_to_file.py

@@ -65,3 +65,36 @@ def test_case_3():
         """
     )
     obj.run(pytorch_code, ["result"])
+
+
+def test_case_4():
+    pytorch_code = textwrap.dedent(
+        """
+        import torch
+        result = torch.hub.download_url_to_file(url='https://paddle-paconvert.bj.bcebos.com/model.params', dst='/tmp/temporary_file',
+                hash_prefix="e1bf0a03102811bb2168e9952fe4edfa09cceb3343278bd4e5876b33b6889e9b", progress=False)
+        """
+    )
+    obj.run(pytorch_code, ["result"])
+
+
+def test_case_5():
+    pytorch_code = textwrap.dedent(
+        """
+        import torch
+        result = torch.hub.download_url_to_file(url='https://paddle-paconvert.bj.bcebos.com/model.params', dst='/tmp/temporary_file',
+                hash_prefix="e1bf0a03102811bb2168e9952fe4edfa09cceb3343278bd4e5876b33b6889e9b", progress=False)
+        """
+    )
+    obj.run(pytorch_code, ["result"])
+
+
+def test_case_6():
+    pytorch_code = textwrap.dedent(
+        """
+        import torch
+        result = torch.hub.download_url_to_file(dst='/tmp/temporary_file',
+                hash_prefix="e1bf0a03102811bb2168e9952fe4edfa09cceb3343278bd4e5876b33b6889e9b", url='https://paddle-paconvert.bj.bcebos.com/model.params', progress=False)
+        """
+    )
+    obj.run(pytorch_code, ["result"])

tests/test_inference_mode.py

@@ -73,3 +73,17 @@ def test_case_4():
         """
     )
     obj.run(pytorch_code, ["result"])
+
+
+def test_case_5():
+    pytorch_code = textwrap.dedent(
+        """
+        import torch
+        x = torch.ones(1, 2, 3, requires_grad=True)
+        @torch.inference_mode(mode= False)
+        def doubler(x):
+            return x * 2
+        result = (doubler(x).requires_grad, doubler(x))
+        """
+    )
+    obj.run(pytorch_code, ["result"])