feat: support custom attention mask in prefill/append attention kernels #266
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
yzh119
added a commit
that referenced
this pull request
May 28, 2024
Followup of #266, add guard to mask array access.
yzh119
added a commit
that referenced
this pull request
May 28, 2024
Followup of #266 , this pr adds some docstring and diagrams for 2D ragged tensor mask layout.
yzh119
added a commit
that referenced
this pull request
Jun 20, 2024
🤖 I have created a release *beep* *boop* --- ## [0.1.0](v0.0.4...v0.1.0) (2024-06-20) ### Highlights * Support any GQA group size support for tensor-cores kernels. * Support any page size support for tensor-cores kernels. * Support CUDA-Graph for prefill/decode APIs. * Add an option to accelerate decode kernels with Tensor Cores. * Support custom attention mask. (https://docs.flashinfer.ai/tutorials/kv_layout.html#mask-layout-2d-ragged-tensor) * Support logits cap in Grok-1 models. * Fused GPU-sampling kernels: top-p, top-k, speculative verification. (https://docs.flashinfer.ai/api/python/sampling.html) * PyTorch wrapper of group-gemm cutlass kernels. (https://docs.flashinfer.ai/api/python/sampling.html) ### Acknowledgement We thank [@ibsidorenko](https://github.com/ibsidorenko), [@LiuXiaoxuanPKU](https://github.com/LiuXiaoxuanPKU), [@Yard1](https://github.com/Yard1) [@AgrawalAmey](https://github.com/AgrawalAmey), [@xuzhenqi](https://github.com/xuzhenqi), [@mgerstgrasser](https://github.com/mgerstgrasser), [@esmeetu](https://github.com/esmeetu), [@yz-tang](https://github.com/yz-tang), [@HSQ79815](https://github.com/HSQ79815), [@Qubitium](https://github.com/Qubitium), [@shreygupta2809](https://github.com/shreygupta2809), [@sighingnow](https://github.com/sighingnow), [@vinx13](https://github.com/vinx13), [@tqchen](https://github.com/tqchen), [@merrymercy](https://github.com/merrymercy), [@comaniac](https://github.com/comaniac) and many others for their contributions and helpful discussions for 0.0.5 release. ### Refactor * support any GQA group size for tensor-cores kernels ([#301](#301)) ([c111ca](c111ca6)) * support any page size for tensor-cores kernels ([#306](#306)) ([82fd8c](82fd8c7)) ### Features * add `use_tensor_cores` option to decode kernels to accelerate GQA ([#317](#317)) ([3b50dd5](3b50dd5)) * add group gemm operators ([#282](#282)) ([e08ba42](e08ba42)) * initial support of distributed operators ([#289](#289)) ([03553da](03553da)) * initial support of logits hook ([#298](#298)) ([ab1e2ad](ab1e2ad)) * Separate Q and KV dtypes for decode ([#286](#286)) ([5602659](5602659)) * support cuda graph for batched multi-query(prefill/append) attention ([#275](#275)) ([83ceb67](83ceb67)) * support cuda graph for batched multi-query(prefill/append) attention ([#277](#277)) ([24cc583](24cc583)) * support custom attention mask in prefill/append attention kernels ([#266](#266)) ([7304282](7304282)) * fused speculative sampilng kernels ([#259](#259)) ([cea2bb](cea2bb9)) * expose sampling APIs in pytorch ([#238](#238)) ([092902](0929023)) ### Performance Improvements * initial cuda graph support ([#256](#256)) ([7e9cc7f](7e9cc7f)) * split kv-cache for prefill/append kernels ([#310](#310)) ([f0bb0a3](f0bb0a3)) * use packed bit array for attention mask ([#308](#308)) ([3d43dc9](3d43dc9)) --- This PR was generated with [Release Please](https://github.com/googleapis/release-please). See [documentation](https://github.com/googleapis/release-please#release-please). --------- Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com> Co-authored-by: Zihao Ye <expye@outlook.com>
Lunderberg
added a commit
to Lunderberg/flashinfer
that referenced
this pull request
Jun 28, 2024
This commit resolves a build-time error with the following message: ``` CMake Error at 3rdparty/flashinfer/CMakeLists.txt:313 (add_library): No SOURCES given to target: prefill_kernels ``` This occurred after flashinfer-ai#266, which replaces the `FLASHINFER_GEN_CASUALS` option with `FLASHINFER_GEN_MASK_MODES`. However, the definition of `flashinfer_option(FLASHINFER_GEN_CASUALS ... )` was not replaced. As a result, loop over the empty `MASK_MODES` does not produce any kernels that should be compiled. This commit updates the `flashinfer_option(FLASH_GEN_CASUALS ...)` line to instead define `FLASH_GEN_MASK_MODES`, using the same default value as `config.cmake`.
Lunderberg
added a commit
to Lunderberg/flashinfer
that referenced
this pull request
Jun 28, 2024
This commit resolves a build-time error with the following message: ``` CMake Error at 3rdparty/flashinfer/CMakeLists.txt:313 (add_library): No SOURCES given to target: prefill_kernels ``` This occurred after flashinfer-ai#266, which replaces the `FLASHINFER_GEN_CASUALS` option with `FLASHINFER_GEN_MASK_MODES`. However, the definition of `flashinfer_option(FLASHINFER_GEN_CASUALS ... )` was not replaced. As a result, loop over the empty `MASK_MODES` does not produce any kernels that should be compiled. This commit updates the `flashinfer_option(FLASH_GEN_CASUALS ...)` line to instead define `FLASH_GEN_MASK_MODES`, using the same default value as `config.cmake`.
yzh119
pushed a commit
that referenced
this pull request
Jun 28, 2024
This commit resolves a build-time error with the following message: ``` CMake Error at 3rdparty/flashinfer/CMakeLists.txt:313 (add_library): No SOURCES given to target: prefill_kernels ``` This occurred after #266, which replaces the `FLASHINFER_GEN_CASUALS` option with `FLASHINFER_GEN_MASK_MODES`. However, the definition of `flashinfer_option(FLASHINFER_GEN_CASUALS ... )` was not replaced. As a result, loop over the empty `MASK_MODES` does not produce any kernels that should be compiled. This commit updates the `flashinfer_option(FLASH_GEN_CASUALS ...)` line to instead define `FLASH_GEN_MASK_MODES`, using the same default value as `config.cmake`.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Some speculative decoding algorithms requires tree attention, which could be supported via prefill/append attention kernels with custom attention mask.
This PR supports this feature.
Related issues: #152
API Breaking Changes
The
begin_forwardfunction inBatchPrefillWithPagedKVCacheWrappernow has an additional argumentpage_sizeto accomodate this new feature.