lucidrains / lucidrains/vector-quantize-pytorch

Extracting learnt embeddings for analysis

Open
#182 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Python
Stars
4k
Forks
338
PR merge metrics
No merged PRs in 30d

Description

Hi @lucidrains,
Amazing work on this package, it is really helpful!
I am using my the QINCo (ResVQ + implicit_neural_codebook = True) for my work. I have trained my model on the data with 8 quantizers. I would now like to assess the quality of my embeddings by linear probing. I am a little confused with what to use as my embeddings for the same. I could either use `raw_embeddings=res_vq_layer.codebook` or `embeddings=residual_vq_layer.get_codes_from_indices(indices)` (as per issue #44). Could you guide me with the same?

Contributor guide

No contributing guide indexed for this repository

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start with issue #44 and inspect the two mentioned entry points: res_vq_layer.codebook and residual_vq_layer.get_codes_from_indices(indices). Determine which representation is intended for linear probing in the QINCo ResVQ configuration, then document that guidance and any relevant usage details.

Written by the indexing model from the issue text.

Assessment

Tech stack
python, pytorch
Domain
machine-learning
Issue type
Documentation
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
20/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.