lm-sys / lm-sys/FastChat

garbled text generation on Mac

Open
#3,629 2 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

python3 -m fastchat.serve.cli --device mps --model-path lmsys/vicuna-7b-v1.5

The above command produces the following garbled text. What can I do? Thank you for any advcie.

USER: hello
ASSISTANT: We detected that you are passing past_key_values as a tuple of tuples. This is deprecated and will be removed in v4.47. Please convert your cache or use an appropriate Cache class (https://huggingface.co/docs/transformers/kv_cache#legacy-cache-format)
sol assumColor� objects sk requireical During> город emptynotationsel National vb�“.igiниюheckIII пери
viceenen0trz Nan органuallyNotFoundannelaalrivatelag externesstoffográutelyenschapp.emblyUID становникаird Gh, државиW grandessingleusettshoraój printedРСР removnection�l,\ enter host Simple3archivigit Biography фран thing Ottoхівовано Wolfgangeor Invaliddbc Une�� Lynбуnego FrederEpandasindowrachestre hijdep BranVALUE`arium

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 reproducing the reported python3 -m fastchat.serve.cli command with --device mps and lmsys/vicuna-7b-v1.5 on macOS, recording the cache warning and generated output. No files or tests are named; done would require identifying the cause of the garbled generation and documenting a verified fix or limitation for this setup.

Written by the indexing model from the issue text.

Assessment

Tech stack
huggingface, python
Domain
ai, cli
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
28/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.