You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi ! I want just try litgpt and start to just do simple inference with doc found here.
So i start with simple chat :
litgpt chat microsoft/phi-2
And at first char not english language crash of many line with :
/litgtp/lib/python3.11/site-packages/litgpt/chat/base.py", line 120, in interact
prompt = input(">> Prompt: ")
^^^^^^^^^^^^^^^^^^^^
UnicodeDecodeError: 'utf-8' codec can't decode byte 0xc3 in position 53: invalid continuation byte
So this is the utf8 witch is bad handle in your code I think. Because i have never this error with all the python
app i work on ..pytorch..and so and so..
I have add the header for utf8 encodind to the base.py and now this work but not for all the case.;sometime
the unicodedecodeerrror come again. Perhaps you will find a better system to handle well the utf8 encodind ?
Or perhaps add the # -- coding: utf-8 -- to all your header python files ?
So to have this work about good i have add this to line 2:
# -*- coding: utf-8 -*-
Many thanks for litgpt look like very promising.
What operating system are you using?
Linux
LitGPT Version
Using Python 3.11.11 environment at:
Version: 0.5.7
The text was updated successfully, but these errors were encountered:
Bug description
Hi ! I want just try litgpt and start to just do simple inference with doc found here.
So i start with simple chat :
litgpt chat microsoft/phi-2
And at first char not english language crash of many line with :
/litgtp/lib/python3.11/site-packages/litgpt/chat/base.py", line 120, in interact
prompt = input(">> Prompt: ")
^^^^^^^^^^^^^^^^^^^^
UnicodeDecodeError: 'utf-8' codec can't decode byte 0xc3 in position 53: invalid continuation byte
So this is the utf8 witch is bad handle in your code I think. Because i have never this error with all the python
app i work on ..pytorch..and so and so..
I have add the header for utf8 encodind to the base.py and now this work but not for all the case.;sometime
the unicodedecodeerrror come again. Perhaps you will find a better system to handle well the utf8 encodind ?
Or perhaps add the # -- coding: utf-8 -- to all your header python files ?
So to have this work about good i have add this to line 2:
# -*- coding: utf-8 -*-
Many thanks for litgpt look like very promising.
What operating system are you using?
Linux
LitGPT Version
The text was updated successfully, but these errors were encountered: