OpenNMT / OpenNMT/CTranslate2

nucleus sampler problem?

Open
#1,424 3 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
C++
Stars
4.7k
Forks
536
Avg merge
12h 12m
Merged PRs (30d)
4

Description

In the Github repository: https://github.com/the-crypt-keeper/can-ai-code/tree/main from @the-crpyt-keeper the WizardCoder was investigated with ctranslate2.

8/08 Added cformers2 support and evaluated michaelfeil/ct2fast-WizardCoder-15B-V1.0 it seems this runtime may have a problem with it's nucleus sampler, precise settings hurt the results far more then they should.

Can this be investigated in more detail?

Contributor guide

Open the contributing guide

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 by reviewing the linked can-ai-code WizardCoder evaluation and the referenced ctranslate2 nucleus-sampling settings. Reproduce the comparison between precise and less precise settings, then determine whether the sampler behavior is a CTranslate2 defect. Done means the behavior is explained and, if confirmed, covered by a regression test and fix.

Written by the indexing model from the issue text.

Assessment

Tech stack
cpp
Domain
backend, machine-learning
Issue type
Bug
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.