[Prototype] Add vLLM Apriel2 model with plugin-based registration #447
+2,418
−0
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
- Add README.md documenting the algebraic structure of the conversion system (surgery monoid, action law, plan composition, total vs partial operations) - Add prune_supernet_step1.yaml and prune_supernet_step2.yaml examples demonstrating the two-step workflow for pruning a homogeneous supernet to a heterogeneous network with different mixer types per layer Co-Authored-By: Claude Opus 4.5 <[email protected]>
- Add modeling_apriel2.py with full vLLM-optimized implementation supporting attention, mamba, GDN, and KDA mixer types - Add register() function for runtime model registration via vLLM's ModelRegistry (no patching required) - Based on Nanda's vllm_diff.patch, adapted for external package use Co-Authored-By: Claude Opus 4.5 <[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Outdated suggestions cannot be applied.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.
Summary
fast_llm_external_modelsModelRegistry.register_model()for runtime registration (no vLLM patching required)Attribution
Model implementation based on work by @nandahkrishna from the
apriel2-vllmbranch. This PR adapts that implementation for plugin-based registration as an alternative to patching vLLM directly.Goal
Evaluate whether vLLM's plugin/registration mechanism can work for us as a short-term solution, avoiding the need to maintain a patched vLLM fork.
Usage
Test plan
🤖 Generated with Claude Code