Add FlashInfer support #2354
Merged
Add FlashInfer support #2354
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
This change adds support for FlashInfer. FlashInfer can be enabled using `FLASH_INFER=1` and is currently only implemented in `FlashCausalLM`. Since this functionality is currently only for testing, FlashInfer is not installed anywhere yet. The FlashInfer API is quite different from FlashAttention/vLLM in that it requires more global bookkeeping: * A wrapper class needs to be contstructed (which we just call *state*). Since this is fairly expensive (due to pinned host memory allocation), we only do this once in a FlashCausalLM instance or for each CUDA Graph size. * Each model forward call needs to be wrapped in `begin_forward` and `end_forward`. This sets up data structures that can be reused for all calls to attention for that forward call. When calling attention, we need access to the state object. To avoid passing an argument down the call chain (which would require changes to all models), we use a context variable. Each model forward call is wrapped using a context manager that does all the bookkeeping for such a call: * Set the context variable to the forward call's state. * Call `begin_forward` on the state. * Yield. * Call `end_forward` on the state. * Reset the context variable. We cannot use a single shared global variable for this, since e.g. CUDA Graphs of different sizes each have their own state.
yuanwu2017
pushed a commit
to yuanwu2017/tgi-gaudi
that referenced
this pull request
Sep 26, 2024
This change adds support for FlashInfer. FlashInfer can be enabled using `FLASH_INFER=1` and is currently only implemented in `FlashCausalLM`. Since this functionality is currently only for testing, FlashInfer is not installed anywhere yet. The FlashInfer API is quite different from FlashAttention/vLLM in that it requires more global bookkeeping: * A wrapper class needs to be contstructed (which we just call *state*). Since this is fairly expensive (due to pinned host memory allocation), we only do this once in a FlashCausalLM instance or for each CUDA Graph size. * Each model forward call needs to be wrapped in `begin_forward` and `end_forward`. This sets up data structures that can be reused for all calls to attention for that forward call. When calling attention, we need access to the state object. To avoid passing an argument down the call chain (which would require changes to all models), we use a context variable. Each model forward call is wrapped using a context manager that does all the bookkeeping for such a call: * Set the context variable to the forward call's state. * Call `begin_forward` on the state. * Yield. * Call `end_forward` on the state. * Reset the context variable. We cannot use a single shared global variable for this, since e.g. CUDA Graphs of different sizes each have their own state.
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.
What does this PR do?
This change adds support for FlashInfer. FlashInfer can be enabled using
FLASH_INFER=1and is currently only implemented inFlashCausalLM. Since this functionality is currently only for testing, FlashInfer is not installed anywhere yet.The FlashInfer API is quite different from FlashAttention/vLLM in that it requires more global bookkeeping:
begin_forwardandend_forward. This sets up data structures that can be reused for all calls to attention for that forward call.When calling attention, we need access to the state object. To avoid passing an argument down the call chain (which would require changes to all models), we use a context variable.
Each model forward call is wrapped using a context manager that does all the bookkeeping for such a call:
begin_forwardon the state.end_forwardon the state.We cannot use a single shared global variable for this, since e.g. CUDA Graphs of different sizes each have their own state.
Before submitting
Pull Request section?
to it if that's the case.
documentation guidelines, and
here are tips on formatting docstrings.
Who can review?
Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.