frozenleaves

frozenleaves synced new reference refs/tags/ciflow/inductor/143544 to frozenleaves/pytorch from mirror 2025-10-20 12:54:14 +08:00

frozenleaves synced commits to refs/tags/ciflow/inductor/143544 at frozenleaves/pytorch from mirror 2025-10-20 12:54:14 +08:00

frozenleaves synced new reference refs/tags/ciflow/h100/164790 to frozenleaves/pytorch from mirror 2025-10-20 12:54:14 +08:00

frozenleaves synced commits to refs/tags/ciflow/h100/164790 at frozenleaves/pytorch from mirror 2025-10-20 12:54:14 +08:00

frozenleaves synced commits to viable/strict at frozenleaves/pytorch from mirror 2025-10-20 12:54:14 +08:00

61d9a5180e [Fix XPU CI] [Inductor UT] Fix test cases broken by community. (#165714)

8a8329b51f [ATen] Switch order of blocked reduce when vectorize loads (#165178)

6b80c94901 [FlexAttention] Fix dynamic shaped heads flex_flash check (#165866)

8951df03de test_scaled_matmul_cuda: fix infer_scale_swizzle (#165788)

8139f33fa5 [dynamo] Add recompile reason for set_stance fail_on_recompile (#165445)

Compare 16 commits »

frozenleaves synced commits to update-vllm-commit-hash/18451675449-1879-1 at frozenleaves/pytorch from mirror 2025-10-20 12:54:14 +08:00

449b5f9222 update vllm commit hash

61d9a5180e [Fix XPU CI] [Inductor UT] Fix test cases broken by community. (#165714)

8a8329b51f [ATen] Switch order of blocked reduce when vectorize loads (#165178)

6b80c94901 [FlexAttention] Fix dynamic shaped heads flex_flash check (#165866)

8951df03de test_scaled_matmul_cuda: fix infer_scale_swizzle (#165788)

Compare 23 commits »

frozenleaves synced commits to update-triton-commit-hash/13663274526-1487-2 at frozenleaves/pytorch from mirror 2025-10-20 12:54:14 +08:00

68ef8e35af update triton commit hash

61d9a5180e [Fix XPU CI] [Inductor UT] Fix test cases broken by community. (#165714)

8a8329b51f [ATen] Switch order of blocked reduce when vectorize loads (#165178)

6b80c94901 [FlexAttention] Fix dynamic shaped heads flex_flash check (#165866)

8951df03de test_scaled_matmul_cuda: fix infer_scale_swizzle (#165788)

Compare 23 commits »

frozenleaves synced commits to tianren/customOp_fusion at frozenleaves/pytorch from mirror 2025-10-20 12:54:14 +08:00

23cba2d68c fix typo and lint

3c28278dff add prologue fusion

2bd0ab5db6 initial code

06890019d2 resolve comments:fix input_gen_fns API; tensor shape infer with siz_hints instead of ir_node_to_tensor; seperated non_tensor_args for Functools.partial for kwargs ; added Single output - layout assertion; lint

8daee127db simplify API and improve code readability

Compare 9 commits »

frozenleaves synced commits to tianren/customOp_enable_max_autotune at frozenleaves/pytorch from mirror 2025-10-20 12:54:14 +08:00

f194b740f7 resolve comments:fix input_gen_fns API; tensor shape infer with siz_hints instead of ir_node_to_tensor; seperated non_tensor_args for Functools.partial for kwargs ; added Single output - layout assertion; lint

frozenleaves synced commits to main at frozenleaves/pytorch from mirror 2025-10-20 12:54:14 +08:00

e1e8491b31 [1/N] Change C-style casts to static_cast or reinterpret_cast (#165750)

767199fd9b [flex_attention] replace sliced BlockMask noop with helpful error (#164702)

602ace5eb4 Revert "[ATen] Fix CUDA reduction warp shuffle order (#164790)"

47804ce467 Revert "12/n : Remove fbandroid_compiler_flags (#165558)"

e8cb34dd52 [Inductor] support masked vectorization for the tail_loop for fp8 datatype (#163324)

Compare 12 commits »

frozenleaves synced commits to gh/laithsakka/311/orig at frozenleaves/pytorch from mirror 2025-10-20 12:54:14 +08:00

a2c2c3295c WIP Support python slicing with data depedennt inptu tensors maybe

frozenleaves synced commits to gh/laithsakka/311/head at frozenleaves/pytorch from mirror 2025-10-20 12:54:14 +08:00

6e61ce45fe Update on "[WIP] Support python slicing with tensor inputs."

b87d3ddd0d Update base for Update on "[WIP] Support python slicing with tensor inputs."

Compare 2 commits »

frozenleaves synced commits to gh/laithsakka/311/base at frozenleaves/pytorch from mirror 2025-10-20 12:54:14 +08:00

b87d3ddd0d Update base for Update on "[WIP] Support python slicing with tensor inputs."

frozenleaves synced new reference gh/jiayisunx/86/orig to frozenleaves/pytorch from mirror 2025-10-20 12:54:14 +08:00

frozenleaves synced commits to gh/jiayisunx/86/orig at frozenleaves/pytorch from mirror 2025-10-20 12:54:14 +08:00

frozenleaves synced new reference gh/jiayisunx/86/head to frozenleaves/pytorch from mirror 2025-10-20 12:54:14 +08:00

frozenleaves synced commits to gh/jiayisunx/86/head at frozenleaves/pytorch from mirror 2025-10-20 12:54:14 +08:00

frozenleaves synced new reference gh/jiayisunx/86/base to frozenleaves/pytorch from mirror 2025-10-20 12:54:14 +08:00

frozenleaves synced commits to gh/jiayisunx/86/base at frozenleaves/pytorch from mirror 2025-10-20 12:54:14 +08:00

frozenleaves synced commits to gh/jiayisunx/84/orig at frozenleaves/pytorch from mirror 2025-10-20 12:54:14 +08:00

7bc36099ec [Inductor] support masked vectorization for the tail_loop for bool datatype

eedf6e950b [Inductor] support masked vectorization for the tail_loop for integer datatypes

e8cb34dd52 [Inductor] support masked vectorization for the tail_loop for fp8 datatype (#163324)

e9d8973427 [Inductor] support masked vectorization for the tail_loop for float64 datatype (#163316)

61d9a5180e [Fix XPU CI] [Inductor UT] Fix test cases broken by community. (#165714)

Compare 260 commits »