Skip to content

Activity

format

jiazhihaopushed 12 commits to deepseek_fix • 2c75cb2…1ff225d • 
yesterday

Bf16 (#190)

Pull request merge
jiazhihaopushed 1 commit to main • cae8b85…f1c71d7 • 
yesterday

checkpoint

jiazhihaocreated deepseek_fix • 2c75cb2 • 
3 days ago

[New Operators] Add kernel- and threadblock-level ReLU, CLAMP (#176)

AMKCodecreated program_partitioning • cae8b85 • 
5 days ago

[New Operators] Add kernel- and threadblock-level ReLU, CLAMP (#176)

Pull request merge
jiazhihaopushed 1 commit to main • 45ad5aa…cae8b85 • 
5 days ago

clamp working

AMKCodepushed 1 commit to new_ops • fb1a704…2b889fd • 
5 days ago

relu working

AMKCodepushed 1 commit to new_ops • 7c2f4e3…fb1a704 • 
9 days ago

compiles but dont run

AMKCodepushed 1 commit to new_ops • 00c1700…7c2f4e3 • 
9 days ago

clamp relu need debug

AMKCodepushed 1 commit to new_ops • b439e0e…00c1700 • 
9 days ago

Update group-query-attention.rst (#186)

Pull request merge
jiazhihaopushed 1 commit to main • 5823678…45ad5aa • 
10 days ago

Added partition_graph.py function (#183)

Pull request merge
jiazhihaopushed 1 commit to main • b6c1f60…5823678 • 
11 days ago

added abstract and element unary need debug

AMKCodepushed 1 commit to new_ops • eee2cf3…b439e0e • 
12 days ago
jiazhihaopushed 2 commits to new_ops • cdf8edf…eee2cf3 • 
18 days ago

Triton warnings fix & triton kernel added (#181)

Pull request merge
jiazhihaopushed 1 commit to main • 546551d…b6c1f60 • 
18 days ago

add more parsers in benchmarks

NorthmanPKUpushed 1 commit to triton_ws_fix • 8583c3b…4d9cd03 • 
18 days ago

fix compile errors

jiazhihaopushed 4 commits to new_ops • 629ac8c…cdf8edf • 
19 days ago

Added GeLU (#180)

Pull request merge
jiazhihaopushed 1 commit to main • 96c7bfd…546551d • 
19 days ago

merge main

NorthmanPKUpushed 4 commits to triton_ws_fix • 7761abc…8583c3b • 
19 days ago

restore notes

NorthmanPKUpushed 1 commit to triton_ws_fix • c470416…7761abc • 
19 days ago

delete warning of size mismatch

NorthmanPKUpushed 1 commit to triton_ws_fix • c349d42…c470416 • 
19 days ago

Add triton kernels and use default to handle warnings

NorthmanPKUpushed 1 commit to triton_ws_fix • 6a5f1ab…c349d42 • 
19 days ago

1. fix bugs of middle tensor and copy_() for output in execute_mugrap…

NorthmanPKUcreated triton_ws_fix • 6a5f1ab • 
20 days ago

Set num_warp_groups and pipeline_stages with default value in generat…

Pull request merge
jiazhihaopushed 1 commit to main • 49a176a…96c7bfd • 
24 days ago
jiazhihaopushed 2 commits to new_ops • 6995225…629ac8c • 
25 days ago

Grace Hopper: let users assign tasks to different warp groups (#165)

Pull request merge
jiazhihaopushed 1 commit to main • f06356d…49a176a • 
25 days ago

resolve merge conflicts

jiazhihaopushed 2 commits to new_ops • 24a5631…6995225 • 
26 days ago

[Fingerprint] Unify fingerprint calculation (#171)

Pull request merge
wmdipushed 1 commit to main • 1a0af05…f06356d • 
26 days ago

Add initial implementation

jiazhihaocreated new_ops • 24a5631 • 
27 days ago

type fix

jiazhihaopushed 1 commit to fingerprint • 9ce297e…1c3f0db • 
on Jan 30

unify fingerprint calculation

jiazhihaocreated fingerprint • 9ce297e • 
on Jan 30