Autotrain container seems to be broken

Spaces
1

Starting a space, I got a runtime error:
Successfully installed autotrain-advanced-0.7.119
Your installed package nvidia-ml-py is corrupted. Skip patch functions nvmlDeviceGet{Compute,Graphics,MPSCompute}RunningProcesses. You may get incorrect or incomplete results. Please consider reinstall package nvidia-ml-py via pip3 install --force-reinstall nvidia-ml-py nvitop.
Your installed package nvidia-ml-py is corrupted. Skip patch functions nvmlDeviceGetMemoryInfo. You may get incorrect or incomplete results. Please consider reinstall package nvidia-ml-py via pip3 install --force-reinstall nvidia-ml-py nvitop.

2

same here did you manage to fix it?

3

Actually, it is a token error, he said. Canā€™t tell from the error messageā€¦

4

I never did. Just gave up.

5

I was starting it using Spaces, so I donā€™t see how a token could be specified. Nothing in the referenced documentation states how to set an environment variable using Spaces.

6

Set HF_TOKEN into your spaceā€™s Secret.

7

I am facing this issue since last update from @hugginfaces. I have tried factory rebuild, updating nltk which is already 3.9.1. Hoping to get this fixed or help sorted by @huggingface !

1 Like
8

Apparently issues is due to warning in transformers 4.46.2 version with mismatch ā€œTrainer.tokenizer is now deprecated. You should use Trainer.processing_class instead.ā€, reverted to 4.43.2 seems to be working, will try 4.45.2 now.

issue not resolved but getting it work fine with older version, hopefully, 4.46.2 arguments will be updated or resolved soon. @huggingface

Thanks to : Tokenizer deprecating in ORPO - #4 by John6666

1 Like