OpenNMT / OpenNMT/CTranslate2

Support Baichuan2?

Open
#1,479 7 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

Baichuan2 is a generative model similar to llama, I find the following two differences:

    1. qkv merge as W_pack, so i change the file /ctranslate2/converters/transformers.py
      image
    1. rotary changes to alibi, so i change the file /ctranslate2/converters/transformers.py
      image

But I found that the conversion only worked on simple sentences, complex sentences would not work.

Could you help me or is there a plan to support Baichuan2?

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 with /ctranslate2/converters/transformers.py and the linked Baichuan2 modeling_baichuan.py to compare W_pack and ALiBI handling. Reproduce the conversion on simple and complex sentences, then verify that Baichuan2 conversion and inference work beyond simple sentences.

Written by the indexing model from the issue text.

Assessment

Tech stack
cpp, python
Domain
machine-learning
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.