I finetuned llama 3.1 8b instruct bnb 4bit, and uploaded it to HF
I used this colab notebook ChatML + chat templates + Mistral 7b full example.ipynb
And when I try to use the finetuned model, I get errors that it’s missing config.json, which I later created manually, but model.safetensors, I don’t understand, where and how it can be created?