This is a Chinese instruction-tuning lora checkpoint based on llama-7B from this repo's work
You can use it like this:
from transformers import LlamaForCausalLM
from peft import PeftModel
model = LlamaForCausalLM.from_pretrained(
"decapoda-research/llama-7b-hf",
load_in_8bit=True,
torch_dtype=torch.float16,
device_map="auto",
)
model = PeftModel.from_pretrained(
model,
LORA_PATH, # specific checkpoint path from "Chinese-Vicuna/Chinese-Vicuna-lora-7b-belle-and-guanaco"
torch_dtype=torch.float16,
device_map={'': 0}
)