Skip to content

feat(vendor): add LLM2Vec embedding model #637

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.

Sign up for GitHub

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

Open
davidberenstein1957 wants to merge 2 commits into
base: main
Choose a base branch
from
Open

feat(vendor): add LLM2Vec embedding model #637

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
8a0faab
feat(vendor): add LLM2Vec embedding model
davidberenstein1957 Apr 25, 2026
fb6d967
fix(vendor): honor llm2vec length and numpy flags
May 5, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
17 changes: 17 additions & 0 deletions src/pruna/evaluation/metrics/vendor/NOTICE.oneig_llm2vec
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,17 @@
LLM2Vec (llm2vec package) vendored from OneIG-Benchmark.

Source: https://github.com/OneIG-Bench/OneIG-Benchmark
Commit: 41b49831e79e6dde5323618c164da1c4cf0f699d
Path: scripts/utils/llm2clip/llm2vec/

OneIG-Benchmark is licensed under the Apache License 2.0.
See the project repository for full license text.

``oneig_llm2vec/modeling_llama_encoder.py`` is derived from
McGill-NLP/LLM2Vec-Meta-Llama-3-8B-Instruct-mntp (Hugging Face Hub);
Pruna relaxes the upstream flash-attention-only constraint for CPU use.

Pruna also includes two minimal compatibility fixes in
``oneig_llm2vec/llm2vec.py``:
- Preserve constructor-provided ``doc_max_length`` instead of hardcoding 512.
- Honor ``convert_to_numpy=True`` in ``encode()`` by returning ``numpy.ndarray``.
Loading
Loading