lucidrains / lucidrains/vector-quantize-pytorch
What is the difference between Grouped VQ and Multi-codebook VQ when used for Autoregressive LLM training?
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 4k
- Forks
- 338
- PR merge metrics
- No merged PRs in 30d
Description
If I use a Grouped-VQ with two groups and codebook size =100, then the count of final speech tokens is 100 * 100 = 10000;
If I use a 2-codebook VQ with codebook size= 100, then the count of the final speech tokens is also 100 * 100 = 10000?
Contributor guide
No contributing guide indexed for this repository
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
No files, tests, or entry points are named. Start by locating the Grouped VQ and multi-codebook VQ implementations, then compare how groups, codebooks, and autoregressive tokens are represented; done means documenting whether the two configurations are equivalent and explaining any difference in training or tokenization.
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
- 25/100