chavinlo

/

gpt4-x-alpaca

like 484

When trying to call torch.load on training_args.bin I get an AttributeError: Can't get attribute 'TrainingArguments' on <module '__main__' (built-in)> error.

This prevents it being used with https://github.com/huggingface/text-generation-inference/ even though Llama models are supported.

Any help getting this working would be greatly appreciated