Replies: 6 comments 8 replies
-
It will be released in v0.0.6. |
Beta Was this translation helpful? Give feedback.
-
Hi, @realhaik Sorry for the inconvenience. Besides, when you install lmdeploy via |
Beta Was this translation helpful? Give feedback.
-
@realhaik You may try to |
Beta Was this translation helpful? Give feedback.
-
@AllentDan could you help investigate it? |
Beta Was this translation helpful? Give feedback.
-
@realhaik Can you run |
Beta Was this translation helpful? Give feedback.
-
I see. Seems the architectures with compute capability is lower than 7.0. Is it right? We primarily did not support this yet. Check this out. |
Beta Was this translation helpful? Give feedback.
-
I have followed the documentation and did: pip install lmdeploy
Unfortunately I am unable to use lmdeploy.serve.openai, as I get ModuleNotFoundError: No module named 'lmdeploy.serve.openai'
This is running in Ubuntu 22.04 docker container image.
root@3b1c8b744e04:/hostfs# pip show lmdeploy
Name: lmdeploy
Version: 0.0.5
Summary: A toolset for compressing, deploying and serving LLM
Home-page: None
Author: OpenMMLab
Author-email: openmmlab@gmail.com
License: None
Location: /usr/local/lib/python3.8/dist-packages
Requires: gradio, mmengine, pybind11, numpy, nvidia-nccl-cu11, torch, datasets, tritonclient, setuptools, transformers, sentencepiece, fire, safetensors
Required-by:
root@3b1c8b744e04:/hostfs# pip show lmdeploy.serve.openai
WARNING: Package(s) not found: lmdeploy.serve.openai
Beta Was this translation helpful? Give feedback.
All reactions