Skip to content

Adding cuda kernel (optimized for sm80) for block-wise 4b quantized float 16 GEMM. #25760

Adding cuda kernel (optimized for sm80) for block-wise 4b quantized float 16 GEMM.

Adding cuda kernel (optimized for sm80) for block-wise 4b quantized float 16 GEMM. #25760

Triggered via pull request January 30, 2024 18:20
Status Failure
Total duration 23m 5s
Artifacts

lint.yml

on: pull_request
Fit to window
Zoom out
Zoom in

Annotations

1 error and 22 warnings
Lint C++
reviewdog: Too many results (annotations) in diff. You may miss some annotations due to GitHub limitation for annotation created by logging command. Please check GitHub Actions log console to see all results. Limitation: - 10 warning annotations and 10 error annotations per step - 50 annotations per job (sum of annotations from all the steps) - 50 annotations per run (separate from the job annotations, these annotations aren't created by users) Source: https://github.community/t5/GitHub-Actions/Maximum-number-of-annotations-that-can-be-created-using-GitHub/m-p/39085
Optional Lint: onnxruntime/core/mickey/cutlass_ext/q4gemm/warp/quantb_meta_mma_tensor_op_tile_iterator.h#L150
[misspell] reported by reviewdog 🐶 "fragement" is a misspelling of "fragment" Raw Output: ./onnxruntime/core/mickey/cutlass_ext/q4gemm/warp/quantb_meta_mma_tensor_op_tile_iterator.h:150:30: "fragement" is a misspelling of "fragment"
Optional Lint: onnxruntime/core/mickey/cutlass_ext/q4gemm/warp/quantb_meta_mma_tensor_op_tile_iterator.h#L178
[misspell] reported by reviewdog 🐶 "dimention" is a misspelling of "dimension" Raw Output: ./onnxruntime/core/mickey/cutlass_ext/q4gemm/warp/quantb_meta_mma_tensor_op_tile_iterator.h:178:18: "dimention" is a misspelling of "dimension"
Optional Lint: onnxruntime/core/mickey/cutlass_ext/q4gemm/warp/quantb_meta_mma_tensor_op_tile_iterator.h#L184
[misspell] reported by reviewdog 🐶 "fragement" is a misspelling of "fragment" Raw Output: ./onnxruntime/core/mickey/cutlass_ext/q4gemm/warp/quantb_meta_mma_tensor_op_tile_iterator.h:184:11: "fragement" is a misspelling of "fragment"
Optional Lint: onnxruntime/test/cuda_host/blkq4_fp16_quant_sm80.h#L102
[misspell] reported by reviewdog 🐶 "seperate" is a misspelling of "separate" Raw Output: ./onnxruntime/test/cuda_host/blkq4_fp16_quant_sm80.h:102:17: "seperate" is a misspelling of "separate"
Optional Lint: onnxruntime/test/cuda_host/blkq4_fp16_quant_sm80.h#L171
[misspell] reported by reviewdog 🐶 "seperate" is a misspelling of "separate" Raw Output: ./onnxruntime/test/cuda_host/blkq4_fp16_quant_sm80.h:171:15: "seperate" is a misspelling of "separate"
Optional Lint: onnxruntime/test/cuda_host/blkq4_fp16_quant_sm80.h#L237
[misspell] reported by reviewdog 🐶 "seperate" is a misspelling of "separate" Raw Output: ./onnxruntime/test/cuda_host/blkq4_fp16_quant_sm80.h:237:15: "seperate" is a misspelling of "separate"
Python format
Node.js 16 actions are deprecated. Please update the following actions to use Node.js 20: actions-rs/toolchain@v1. For more information see: https://github.blog/changelog/2023-09-22-github-actions-transitioning-from-node-16-to-node-20/.
Python format
The following actions uses node12 which is deprecated and will be forced to run on node16: actions-rs/toolchain@v1. For more info: https://github.blog/changelog/2023-06-13-github-actions-all-actions-will-run-on-node16-instead-of-node12-by-default/
Python format
The `set-output` command is deprecated and will be disabled soon. Please upgrade to using Environment Files. For more information see: https://github.blog/changelog/2022-10-11-github-actions-deprecating-save-state-and-set-output-commands/
Python format
The `set-output` command is deprecated and will be disabled soon. Please upgrade to using Environment Files. For more information see: https://github.blog/changelog/2022-10-11-github-actions-deprecating-save-state-and-set-output-commands/
Python format
The `set-output` command is deprecated and will be disabled soon. Please upgrade to using Environment Files. For more information see: https://github.blog/changelog/2022-10-11-github-actions-deprecating-save-state-and-set-output-commands/
Python format
The `set-output` command is deprecated and will be disabled soon. Please upgrade to using Environment Files. For more information see: https://github.blog/changelog/2022-10-11-github-actions-deprecating-save-state-and-set-output-commands/
Lint C++: onnxruntime/core/mickey/cutlass_ext/q4gemm/device/quantb_gemm.h#L148
[cpplint] reported by reviewdog 🐶 Do not leave a blank line after "public:" [whitespace/blank_line] [3] Raw Output: onnxruntime/core/mickey/cutlass_ext/q4gemm/device/quantb_gemm.h:148: Do not leave a blank line after "public:" [whitespace/blank_line] [3]
Lint C++: onnxruntime/core/mickey/cutlass_ext/q4gemm/device/quantb_gemm.h#L223
[cpplint] reported by reviewdog 🐶 Redundant blank line at the start of a code block should be deleted. [whitespace/blank_line] [2] Raw Output: onnxruntime/core/mickey/cutlass_ext/q4gemm/device/quantb_gemm.h:223: Redundant blank line at the start of a code block should be deleted. [whitespace/blank_line] [2]
Lint C++: onnxruntime/core/mickey/cutlass_ext/q4gemm/device/quantb_gemm.h#L252
[cpplint] reported by reviewdog 🐶 Redundant blank line at the start of a code block should be deleted. [whitespace/blank_line] [2] Raw Output: onnxruntime/core/mickey/cutlass_ext/q4gemm/device/quantb_gemm.h:252: Redundant blank line at the start of a code block should be deleted. [whitespace/blank_line] [2]
Lint C++: onnxruntime/core/mickey/cutlass_ext/q4gemm/device/quantb_gemm.h#L252
[cpplint] reported by reviewdog 🐶 Redundant blank line at the end of a code block should be deleted. [whitespace/blank_line] [3] Raw Output: onnxruntime/core/mickey/cutlass_ext/q4gemm/device/quantb_gemm.h:252: Redundant blank line at the end of a code block should be deleted. [whitespace/blank_line] [3]
Lint C++: onnxruntime/core/mickey/cutlass_ext/q4gemm/device/quantb_gemm.h#L266
[cpplint] reported by reviewdog 🐶 Closing ) should be moved to the previous line [whitespace/parens] [2] Raw Output: onnxruntime/core/mickey/cutlass_ext/q4gemm/device/quantb_gemm.h:266: Closing ) should be moved to the previous line [whitespace/parens] [2]
Lint C++: onnxruntime/core/mickey/cutlass_ext/q4gemm/device/quantb_gemm.h#L288
[cpplint] reported by reviewdog 🐶 Closing ) should be moved to the previous line [whitespace/parens] [2] Raw Output: onnxruntime/core/mickey/cutlass_ext/q4gemm/device/quantb_gemm.h:288: Closing ) should be moved to the previous line [whitespace/parens] [2]
Lint C++: onnxruntime/core/mickey/cutlass_ext/q4gemm/device/quantb_gemm.h#L301
[cpplint] reported by reviewdog 🐶 private: should be indented +1 space inside class QuantBGemm [whitespace/indent] [3] Raw Output: onnxruntime/core/mickey/cutlass_ext/q4gemm/device/quantb_gemm.h:301: private: should be indented +1 space inside class QuantBGemm [whitespace/indent] [3]
Lint C++: onnxruntime/core/mickey/cutlass_ext/q4gemm/device/quantb_gemm.h#L302
[cpplint] reported by reviewdog 🐶 Do not leave a blank line after "private:" [whitespace/blank_line] [3] Raw Output: onnxruntime/core/mickey/cutlass_ext/q4gemm/device/quantb_gemm.h:302: Do not leave a blank line after "private:" [whitespace/blank_line] [3]
Lint C++: onnxruntime/core/mickey/cutlass_ext/q4gemm/device/quantb_gemm.h#L306
[cpplint] reported by reviewdog 🐶 public: should be indented +1 space inside class QuantBGemm [whitespace/indent] [3] Raw Output: onnxruntime/core/mickey/cutlass_ext/q4gemm/device/quantb_gemm.h:306: public: should be indented +1 space inside class QuantBGemm [whitespace/indent] [3]
Lint C++: onnxruntime/core/mickey/cutlass_ext/q4gemm/device/quantb_gemm.h#L307
[cpplint] reported by reviewdog 🐶 Do not leave a blank line after "public:" [whitespace/blank_line] [3] Raw Output: onnxruntime/core/mickey/cutlass_ext/q4gemm/device/quantb_gemm.h:307: Do not leave a blank line after "public:" [whitespace/blank_line] [3]