lm-sys / lm-sys/FastChat

Llama 3.1 - Wrong context length reported in `/token_check` endpoint

Open
#3,452 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

When the Llama 3.1 70B model is loaded in FastChat, the /token_check endpoint reports a context length of 1M instead of the expected 128K.

{
    "prompts": [
        {
            "fits": true,
            "tokenCount": 2,
            "contextLength": 1048576
        }
    ]
}

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 tracing the /token_check endpoint and how it obtains the context length for the Llama 3.1 70B model. Compare its response with the expected 128K limit and verify that the endpoint reports the corrected value while preserving the token count and fit result.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
api
Issue type
Bug
Difficulty
3/5
Estimated time
1-2 days
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
45/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.