-
-
Notifications
You must be signed in to change notification settings - Fork 4.4k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[Core] Subclass ModelRunner to support cross-attention & encoder sequences (towards eventual encoder/decoder model support) #4942
Merged
robertgshaw2-neuralmagic
merged 624 commits into
vllm-project:main
from
neuralmagic:afeldman-nm/infra_enc_dec_model_runner
Aug 6, 2024
Merged
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
afeldman-nm
changed the title
[Core] Correctly invoke prefill & decode kernels for cross-attention (towards eventual encoder/decoder model support)
[Core] Subclass ModelRunner to support cross-attention & encoder sequences (towards eventual encoder/decoder model support)
May 21, 2024
Open
afeldman-nm
force-pushed
the
afeldman-nm/infra_enc_dec_model_runner
branch
from
June 11, 2024 03:49
62f7903
to
0285548
Compare
This was referenced Jun 25, 2024
laishzh
reviewed
Jul 10, 2024
afeldman-nm
force-pushed
the
afeldman-nm/infra_enc_dec_model_runner
branch
from
July 15, 2024 09:06
a5c28fc
to
db5539a
Compare
/ready |
github-actions
bot
added
the
ready
ONLY add when PR is ready to merge/full CI is needed
label
Aug 6, 2024
njhill
approved these changes
Aug 6, 2024
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks again @afeldman-nm!
/ready |
robertgshaw2-neuralmagic
deleted the
afeldman-nm/infra_enc_dec_model_runner
branch
August 6, 2024 20:51
Once again thanks @sroy745 @njhill @maxdebayser @DarkLight1337 @WoosukKwon @khluu @robertgshaw2-neuralmagic for all of your excellent help in landing this PR!
|
This was referenced Aug 7, 2024
This was referenced Aug 9, 2024
sfc-gh-mkeralapura
pushed a commit
to sfc-gh-mkeralapura/vllm
that referenced
this pull request
Aug 12, 2024
…ences (towards eventual encoder/decoder model support) (vllm-project#4942) Co-authored-by: Andrew Feldman <afeld2012@gmail.com> Co-authored-by: Nick Hill <nickhill@us.ibm.com>
kylesayrs
pushed a commit
to neuralmagic/vllm
that referenced
this pull request
Aug 17, 2024
…ences (towards eventual encoder/decoder model support) (vllm-project#4942) Co-authored-by: Andrew Feldman <afeld2012@gmail.com> Co-authored-by: Nick Hill <nickhill@us.ibm.com>
fialhocoelho
pushed a commit
to opendatahub-io/vllm
that referenced
this pull request
Aug 22, 2024
…ences (towards eventual encoder/decoder model support) (vllm-project#4942) Co-authored-by: Andrew Feldman <afeld2012@gmail.com> Co-authored-by: Nick Hill <nickhill@us.ibm.com>
Alvant
pushed a commit
to compressa-ai/vllm
that referenced
this pull request
Oct 26, 2024
…ences (towards eventual encoder/decoder model support) (vllm-project#4942) Co-authored-by: Andrew Feldman <afeld2012@gmail.com> Co-authored-by: Nick Hill <nickhill@us.ibm.com> Signed-off-by: Alvant <alvasian@yandex.ru>
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.
This PR is a step towards encoder/decoder model support. This PR creates a specialized ModelRunner subclass for encoder/decoder models; it differs from the base ModelRunner class primarily in that it (1) expects each SequenceGroup to have an encoder sequence, and (2) it properly constructs the AttentionMetadata structure to support both self- and cross-attention.
A quick overview of the plan for supporting encoder/decoder models in vLLM:
Prefill phase: (1) Non-autoregressive encoder inference yields encoder hidden states in a single pass; no KV caching occurs. (2) decoder prefill yields first-token-prediction & cached KVs. Within the decoder, cross-attention layers cache the KVs derived from encoder hidden states:
Key_{cross-attn, layer-n} = W_{K, cross-attn, layer-n} x (Encoder hidden states)
Value_{cross-attn, layer-n} = W_{V, cross-attn, layer-n} x (Encoder hidden states)
Note that all cross-attention layers consume the same encoder hidden states; however each cross-attention layers' keys and values differ because each layer has unique W_{K, cross-attn, layer-n} and W_{V, cross-attn, layer-n}. Therefore, the cross-attention KV cache must store KVs for each decoder layer, even though these KVs are all derived from a single set of encoder hidden states.
Note that self-attention layer behavior is unchanged compared to what it would be in a decoder-only model (cache KVs computed from the previous decoder layer outputs.)
Decode phase: during each iteration of the autoregressive decode process,
To implement the above encoder/decoder inference process, the following functionality will be added to vLLM over the course of multiple PRs:
Note 1: because this PR makes an incremental contribution (cross-attention KV-caching and memory management), this PR will not enable end-to-end encoder/decoder support (this will rely on later PRs.)
Note 2: the best effort is being made to ensure that encoder/decoder models are compatible with existing vLLM features. At this time, encoder/decoder models are unlikely to be compatible with the following vLLM features:
INCREMENTAL FIX TOWARDS #187
BEFORE SUBMITTING, PLEASE READ THE CHECKLIST BELOW AND FILL IN THE DESCRIPTION ABOVE
PR Checklist (Click to Expand)
Thank you for your contribution to vLLM! Before submitting the pull request, please ensure the PR meets the following criteria. This helps vLLM maintain the code quality and improve the efficiency of the review process.
PR Title and Classification
Only specific types of PRs will be reviewed. The PR title is prefixed appropriately to indicate the type of change. Please use one of the following:
[Bugfix]
for bug fixes.[CI/Build]
for build or continuous integration improvements.[Doc]
for documentation fixes and improvements.[Model]
for adding a new model or improving an existing model. Model name should appear in the title.[Frontend]
For changes on the vLLM frontend (e.g., OpenAI API server,LLM
class, etc.)[Kernel]
for changes affecting CUDA kernels or other compute kernels.[Core]
for changes in the core vLLM logic (e.g.,LLMEngine
,AsyncLLMEngine
,Scheduler
, etc.)[Hardware][Vendor]
for hardware-specific changes. Vendor name should appear in the prefix (e.g.,[Hardware][AMD]
).[Misc]
for PRs that do not fit the above categories. Please use this sparingly.Note: If the PR spans more than one category, please include all relevant prefixes.
Code Quality
The PR need to meet the following code quality standards:
format.sh
to format your code.docs/source/
if the PR modifies the user-facing behaviors of vLLM. It helps vLLM user understand and utilize the new features or changes.Notes for Large Changes
Please keep the changes as concise as possible. For major architectural changes (>500 LOC excluding kernel/data/config/test), we would expect a GitHub issue (RFC) discussing the technical design and justification. Otherwise, we will tag it with
rfc-required
and might not go through the PR.What to Expect for the Reviews
The goal of the vLLM team is to be a transparent reviewing machine. We would like to make the review process transparent and efficient and make sure no contributor feel confused or frustrated. However, the vLLM team is small, so we need to prioritize some PRs over others. Here is what you can expect from the review process:
action-required
label on the PR if there are changes required. The contributor should address the comments and ping the reviewer to re-review the PR.Thank You
Finally, thank you for taking the time to read these guidelines and for your interest in contributing to vLLM. Your contributions make vLLM a great tool for everyone!