Yeyu/hf eagle medusa #664
Merged
Yeyu/hf eagle medusa #664
Conversation
This file contains hidden or 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
Contributor
Author
|
/ok to test cdea9ed |
Contributor
Author
|
/ok to test f8a1088 |
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@
## main #664 +/- ##
=======================================
Coverage 74.78% 74.79%
=======================================
Files 192 192
Lines 18814 18810 -4
=======================================
- Hits 14070 14068 -2
+ Misses 4744 4742 -2 ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
h-guo18
reviewed
Dec 11, 2025
h-guo18
reviewed
Dec 11, 2025
h-guo18
reviewed
Dec 11, 2025
| draft_logits_list = [eagle_logits] | ||
| if self.eagle_config.parallel_draft_step > 1: | ||
| # Get additional draft logits from parallel draft heads | ||
| for draft_head in self.eagle_module.parallel_draft_heads: |
Contributor
There was a problem hiding this comment.
I think we can optimize this for-loop and do it parallelly. Can be done in following PRs perhaps.
h-guo18
reviewed
Dec 11, 2025
| ( | ||
| torch.zeros(b, ttt_step, dtype=loss_mask.dtype, device=loss_mask.device), | ||
| loss_mask[:, 1 + ttt_step :], | ||
| for i in range(self.eagle_config.parallel_draft_step): |
Contributor
There was a problem hiding this comment.
Similar as above, this for-loop seems parallelizable
h-guo18
reviewed
Dec 11, 2025
| eagle_input_hidden_states = base_model_hidden_states | ||
|
|
||
| draft_tokens = [] | ||
| for _ in range(steps): |
Contributor
There was a problem hiding this comment.
The semantic of this steps argument seems ambiguous to me. Shall we rename it to eagle_steps? Then it means we do eagle_steps sequential drafting + num_medusa_heads parallel drafrting.
h-guo18
reviewed
Dec 11, 2025
h-guo18
approved these changes
Dec 11, 2025
Contributor
h-guo18
left a comment
h-guo18
left a comment
There was a problem hiding this comment.
Left some comments and questions. Other changes in HF LGTM.
I think it would be great to test PTQ and AR regression test before merging. Thanks
sugunav14
reviewed
Dec 11, 2025
sugunav14
approved these changes
Dec 12, 2025
Signed-off-by: Ye Yu <[email protected]>
Signed-off-by: Ye Yu <[email protected]>
Signed-off-by: Ye Yu <[email protected]>
Signed-off-by: Ye Yu <[email protected]>
Signed-off-by: Ye Yu <[email protected]>
Signed-off-by: Ye Yu <[email protected]>
Signed-off-by: Ye Yu <[email protected]>
Signed-off-by: Ye Yu <[email protected]>
Signed-off-by: Ye Yu <[email protected]>
Signed-off-by: Ye Yu <[email protected]>
Signed-off-by: Ye Yu <[email protected]>
Signed-off-by: Ye Yu <[email protected]>
Signed-off-by: Ye Yu <[email protected]>
yeyu-nvidia
force-pushed
the
yeyu/hf_eagle_medusa
branch
from
0e40d90 to
10f46d7
Compare
Signed-off-by: Ye Yu <[email protected]>
h-guo18
reviewed
Dec 12, 2025
Signed-off-by: Ye Yu <[email protected]>
Signed-off-by: Ye Yu <[email protected]>
Signed-off-by: Ye Yu <[email protected]>
Signed-off-by: Ye Yu <[email protected]>
Signed-off-by: Ye Yu <[email protected]>
b7r6
pushed a commit
to weyl-ai/Model-Optimizer
that referenced
this pull request
Dec 18, 2025
## What does this PR do? new feature, **Overview:** This PR implements HF parallel draft by combining eagle and medusa. In training, multiple medusa heads are added and trained together with eagle. In inference, medusa heads are used to generate draft tokens after all eagle tokens. ## Usage Set parallel_draft_step>1 in eagle_config to enable parallel draft. ```python # Add a code snippet demonstrating how to use this ``` ## Testing <!-- Mention how have you tested your change if applicable. --> ## Before your PR is "*Ready for review*" <!-- If you haven't finished some of the above items you can still open `Draft` PR. --> - **Make sure you read and follow [Contributor guidelines](https://github.com/NVIDIA/Model-Optimizer/blob/main/CONTRIBUTING.md)** and your commits are signed. - **Is this change backward compatible?**: Yes/No <!--- If No, explain why. --> - **Did you write any new necessary tests?**: Yes/No - **Did you add or update any necessary documentation?**: Yes/No - **Did you update [Changelog](https://github.com/NVIDIA/Model-Optimizer/blob/main/CHANGELOG.rst)?**: Yes/No <!--- Only for new features, API changes, critical bug fixes or bw breaking changes. --> ## Additional Information <!-- E.g. related issue. --> --------- Signed-off-by: Ye Yu <[email protected]>
soodoshll
pushed a commit
to soodoshll/TensorRT-Model-Optimizer
that referenced
this pull request
Dec 18, 2025
## What does this PR do? new feature, **Overview:** This PR implements HF parallel draft by combining eagle and medusa. In training, multiple medusa heads are added and trained together with eagle. In inference, medusa heads are used to generate draft tokens after all eagle tokens. ## Usage Set parallel_draft_step>1 in eagle_config to enable parallel draft. ```python # Add a code snippet demonstrating how to use this ``` ## Testing <!-- Mention how have you tested your change if applicable. --> ## Before your PR is "*Ready for review*" <!-- If you haven't finished some of the above items you can still open `Draft` PR. --> - **Make sure you read and follow [Contributor guidelines](https://github.com/NVIDIA/Model-Optimizer/blob/main/CONTRIBUTING.md)** and your commits are signed. - **Is this change backward compatible?**: Yes/No <!--- If No, explain why. --> - **Did you write any new necessary tests?**: Yes/No - **Did you add or update any necessary documentation?**: Yes/No - **Did you update [Changelog](https://github.com/NVIDIA/Model-Optimizer/blob/main/CHANGELOG.rst)?**: Yes/No <!--- Only for new features, API changes, critical bug fixes or bw breaking changes. --> ## Additional Information <!-- E.g. related issue. --> --------- Signed-off-by: Ye Yu <[email protected]>
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?
new feature,
Overview:
This PR implements HF parallel draft by combining eagle and medusa. In training, multiple medusa heads are added and trained together with eagle. In inference, medusa heads are used to generate draft tokens after all eagle tokens.
Usage
Set parallel_draft_step>1 in eagle_config to enable parallel draft.
# Add a code snippet demonstrating how to use thisTesting
Before your PR is "Ready for review"
Additional Information