How to loading local model

[mingyang-wang26]

Hi,

Thank you for your excellent work!

I'm facing some issues while loading my fine-tuned GPT2-XL model locally. I referred to #28, but I'm still unclear on how to configure the local.json file properly.

For the GPT2-XL model, I understand there should be an entry like "gpt2-xl": null in the JSON file. How can I modify this to point to my local model?

In #28, it mentions passing two strings—model_name and supported_model_name. Could you clarify where I should pass these strings?

Thank you for your help!

[himanshubeniwal]

+1, I am also unable to figure out which strings to update. I completely changed my code, but it seems I am missing something.