frozenleaves/pytorch
1
0
Fork 0
mirror of https://github.com/pytorch/pytorch.git synced 2025-10-20 21:14:14 +08:00
Code Issues Packages Projects Releases Wiki Activity

[fbgemm_gpu] Break up fbgemm_cuda_utils.cuh, pt 10 (#130468)

Browse Source 
Summary:
X-link: https://github.com/pytorch/FBGEMM/pull/2814

X-link: https://github.com/facebookresearch/FBGEMM/pull/19

- Break up `fbgemm_cuda_utils.cuh`, pt 10

Test Plan:
```
buck2 targets //deeplearning/fbgemm/fbgemm_gpu/test/jagged/... | grep -v '-' | xargs -I % sh -c 'buck2 run @//mode/opt -c fbcode.nvcc_arch=v100 -c fbcode.platform=platform010 % || exit 255'

buck2 targets //deeplearning/fbgemm/fbgemm_gpu/test/tbe/... | grep -v '-' | xargs -I % sh -c 'buck2 run @//mode/opt -c fbcode.nvcc_arch=v100 -c fbcode.platform=platform010 % || exit 255'

buck2 targets //deeplearning/fbgemm/fbgemm_gpu/test/sparse/... | grep -v '-' | xargs -I % sh -c 'buck2 run @//mode/opt -c fbcode.nvcc_arch=v100 -c fbcode.platform=platform010 % || exit 255'

buck2 build --config fbcode.enable_gpu_sections=true --flagfile fbcode//mode/dev-nosan-amd-gpu fbcode//smart/inference_platform_sp/llm_predictor_amd:service

buck2 build --flagfile fbcode//mode/amd-gpu fbcode//hpc/ops:sparse_ops

buck2 build --flagfile fbcode//mode/dev-nosan-amd-gpu fbcode//caffe2/benchmarks/operator_benchmark/pt:add_test
```

Reviewed By: spcyppt

Differential Revision: D59545097

Pull Request resolved: https://github.com/pytorch/pytorch/pull/130468
Approved by: https://github.com/ezyang
This commit is contained in:
Benson Ma
2024-07-11 07:10:27 +00:00
committed by PyTorch MergeBot
parent 79c41bb58a
commit b1942a1af4
1 changed files with 4 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
tools/amd_build/build_amd.py
View File

@ -159,7 +159,10 @@ hip_platform_files = [
"third_party/fbgemm/fbgemm_gpu/codegen/embedding_backward_split_host_template.cpp",
"third_party/fbgemm/fbgemm_gpu/codegen/embedding_backward_split_template.cu",
"third_party/fbgemm/fbgemm_gpu/codegen/embedding_forward_quantized_split_lookup.cu",
"third_party/fbgemm/fbgemm_gpu/include/fbgemm_gpu/fbgemm_cuda_utils.cuh",
"third_party/fbgemm/fbgemm_gpu/include/fbgemm_gpu/utils/cuda_prelude.cuh",
"third_party/fbgemm/fbgemm_gpu/include/fbgemm_gpu/utils/stochastic_rounding.cuh",
"third_party/fbgemm/fbgemm_gpu/include/fbgemm_gpu/utils/vec4.cuh",
"third_party/fbgemm/fbgemm_gpu/include/fbgemm_gpu/utils/weight_row.cuh",
"third_party/fbgemm/fbgemm_gpu/include/fbgemm_gpu/sparse_ops.cuh",
"third_party/fbgemm/fbgemm_gpu/src/jagged_tensor_ops.cu",
"third_party/fbgemm/fbgemm_gpu/src/quantize_ops.cu",