alpaca-lora-30b / README.md
chansung's picture
Update README.md
597694d
|
raw
history blame
911 Bytes
metadata
license: gpl-3.0
datasets:
  - yahma/alpaca-cleaned
language:
  - en
pipeline_tag: text2text-generation
tags:
  - alpaca
  - llama
  - chat

This repository comes with LoRA checkpoint to make LLaMA into a chatbot like language model. The checkpoint is the output of instruction following fine-tuning process with the following settings on 8xA100(40G) DGX system.

python finetune.py \
    --base_model='decapoda-research/llama-30b-hf' \
    --num_epochs=10 \
    --cutoff_len=512 \
    --group_by_length \
    --output_dir='./lora-alpaca' \
    --lora_target_modules='[q_proj,k_proj,v_proj,o_proj]' \
    --lora_r=16 \
    --batch_size=... \
    --micro_batch_size=...