參考 Fine Turning
https://colab.research.google.com/drive/12dVqXZMIVxGI0uutU6HG9RWbWPXL3vts?usp=sharing#scrollTo=mNnkgBq7Q3EU

Model：meta-llama/Llama-2-7b-chat-hf
dataset：COT_Answer.txt

Training procedure

The following bitsandbytes quantization config was used during training:

• load_in_8bit: False
• load_in_4bit: True
• llm_int8_threshold: 6.0
• llm_int8_skip_modules: None
• llm_int8_enable_fp32_cpu_offload: False
• llm_int8_has_fp16_weight: False
• bnb_4bit_quant_type: nf4
• bnb_4bit_use_double_quant: False
• bnb_4bit_compute_dtype: float16

Framework versions

• PEFT 0.5.0.dev0