app.py support local model

[AllentDan]

No description provided.

[irexyc]

Does it depend on a specific gradio version? I can't run it wht latest gradio version.
from gradio.components import * doesn't import @ducoment in latest gradio

After install gradio 3.24.1 and gradio_client 0.0.8, I get this log:

(py38) [chenxin@SH-IDC1-10-140-24-141 chenxin]$ python3 -m lmdeploy.serve.gradio.app workspace/
/home/chenxin/miniconda3/envs/py38/lib/python3.8/site-packages/gradio/utils.py:874: UserWarning: Expected 5 arguments for function functools.partial(<function chat_stream_local at 0x7fec0348baf0>, model=<lmdeploy.model.Vicuna object at 0x7fec0348d5b0>, tm_model=<lmdeploy.turbomind.turbomind.TurboMind object at 0x7fec00a00190>, tokenizer=<lmdeploy.turbomind.tokenizer.Tokenizer object at 0x7fec0348d400>, request2instance={}), received 4.
  warnings.warn(
/home/chenxin/miniconda3/envs/py38/lib/python3.8/site-packages/gradio/utils.py:878: UserWarning: Expected at least 5 arguments for function functools.partial(<function chat_stream_local at 0x7fec0348baf0>, model=<lmdeploy.model.Vicuna object at 0x7fec0348d5b0>, tm_model=<lmdeploy.turbomind.turbomind.TurboMind object at 0x7fec00a00190>, tokenizer=<lmdeploy.turbomind.tokenizer.Tokenizer object at 0x7fec0348d400>, request2instance={}), received 4.
  warnings.warn(
/home/chenxin/miniconda3/envs/py38/lib/python3.8/site-packages/gradio/utils.py:874: UserWarning: Expected 5 arguments for function functools.partial(<function reset_local_func at 0x7fec0348bb80>, tm_model=<lmdeploy.turbomind.turbomind.TurboMind object at 0x7fec00a00190>, request2instance={}), received 4.
  warnings.warn(
/home/chenxin/miniconda3/envs/py38/lib/python3.8/site-packages/gradio/utils.py:878: UserWarning: Expected at least 5 arguments for function functools.partial(<function reset_local_func at 0x7fec0348bb80>, tm_model=<lmdeploy.turbomind.turbomind.TurboMind object at 0x7fec00a00190>, request2instance={}), received 4.
  warnings.warn(
server is gonna mount on: http://localhost:6006
Running on local URL:  http://localhost:6006

And when I run the model, I get

Task exception was never retrieved
future: <Task finished name='bskzc65nxa_0' coro=<Queue.process_events() done, defined at /home/chenxin/miniconda3/envs/py38/lib/python3.8/site-packages/gradio/queueing.py:343> exception=1 validation error for PredictBody
event_id
  Field required [type=missing, input_value={'fn_index': 0, 'data': [...ion_hash': 'bskzc65nxa'}, input_type=dict]
    For further information visit https://errors.pydantic.dev/2.1/v/missing>
Traceback (most recent call last):
  File "/home/chenxin/miniconda3/envs/py38/lib/python3.8/site-packages/gradio/queueing.py", line 347, in process_events
    client_awake = await self.gather_event_data(event)
  File "/home/chenxin/miniconda3/envs/py38/lib/python3.8/site-packages/gradio/queueing.py", line 220, in gather_event_data
    data, client_awake = await self.get_message(event, timeout=receive_timeout)
  File "/home/chenxin/miniconda3/envs/py38/lib/python3.8/site-packages/gradio/queueing.py", line 456, in get_message
    return PredictBody(**data), True
  File "/home/chenxin/miniconda3/envs/py38/lib/python3.8/site-packages/pydantic/main.py", line 159, in __init__
    __pydantic_self__.__pydantic_validator__.validate_python(data, self_instance=__pydantic_self__)
pydantic_core._pydantic_core.ValidationError: 1 validation error for PredictBody
event_id
  Field required [type=missing, input_value={'fn_index': 0, 'data': [...ion_hash': 'bskzc65nxa'}, input_type=dict]
    For further information visit https://errors.pydantic.dev/2.1/v/missing

What dependents requires nh3 ? Shoud we add it to requirements?

[AllentDan]

Tested OK with upgraded gradio-3.39.0 and gradio-client-0.3.0. 🤔