Can any one provide a script for sft?

[zlh1992]

The script of sft_alpaca is just the same as inferenece.py

[Nianhui Guo]

Both the evaluations for the alpaca finetuned LLM and the pre-trained LLM are accessible. The primary distinction lies in the loading of the Lora adapter parameters. Kindly refer to the following links to understand the differences:

llama_2b_inference.py L20

llama_2b_sft_alpaca.py L31