lm-sys / lm-sys/FastChat

[Model Request] Add support for Cartik/Sonexa-1-Chat

Open
#3,901 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Python
Stars
39.5k
Forks
4.8k
PR merge metrics
No merged PRs in 30d

Description

Hello LMSYS and Arena team,

I would like to request adding our new open-source coding model to the Chatbot Arena leaderboard for evaluation. It is fully open and hosted on Hugging Face.

### Model Details:
* **Model Name:** Sonexa-1-Chat
* **Hugging Face Link:** https://huggingface.co/Cartik/Sonexa-1-Chat
* **Base Architecture:** Qwen3
* **Model Size:** 4B parameters
* **Primary Focus:** Code generation and technical instruction following

### Configuration & Templates:
* **Chat Template:** Standard Qwen3 / ChatML format (the model correctly utilizes `<|im_start|>` and `<|im_end|>` tokens).
* **Tokenizer:** Uses the default Qwen3Tokenizer setup.

The model has been tested and shows stable performance on standard programming tasks, outputting clean Markdown and syntactically correct code blocks. We would love to see how it performs in the blind testing environment against other models in its weight class.

Please let me know if you need any additional configurations, API endpoints, or if I should prepare a Pull Request for `fastchat/conversation.py` to ensure the template renders perfectly.

Thank you for your amazing work on the Arena!

Best regards,
Cartik

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 by reviewing fastchat/conversation.py and the Sonexa-1-Chat model details at its Hugging Face link, paying attention to the stated Qwen3/ChatML template and tokenizer configuration. Done means the model is added for Chatbot Arena evaluation and its template renders correctly in the blind testing environment.

Written by the indexing model from the issue text.

Assessment

Tech stack
huggingface, python
Domain
ai, machine-learning
Issue type
Feature
Difficulty
3/5
Estimated time
1-2 days
Activity status
Quiet
Clarity
Mostly clear
Newbie friendliness
48/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.