1 Star 0 Fork 26

wuzeen/HuatuoGPT

加入 Gitee
与超过 1200万 开发者一起发现、参与优秀开源项目,私有仓库也完全免费 :)
免费加入
克隆/下载
apply_delta.py 1.53 KB
一键复制 编辑 原始数据 按行查看 历史
HenSt 提交于 2023-07-08 02:13 . delta
"""
Apply the delta weights on top of a base model.
Usage:
"""
import argparse
import torch
from tqdm import tqdm
from transformers import AutoTokenizer, AutoModelForCausalLM
def apply_delta(base_model_path, target_model_path, delta_path):
print(f"Loading the base model from {base_model_path}")
base = AutoModelForCausalLM.from_pretrained(
base_model_path, torch_dtype=torch.float16, low_cpu_mem_usage=True)
print(f"Loading the delta from {delta_path}")
delta = AutoModelForCausalLM.from_pretrained(delta_path, torch_dtype=torch.float16, low_cpu_mem_usage=True)
delta_tokenizer = AutoTokenizer.from_pretrained(delta_path, use_fast=False)
print("Applying the delta")
for name, param in tqdm(delta.state_dict().items(), desc="Applying delta"):
assert name in base.state_dict()
if "embed_tokens" in name or "lm_head.weight" in name or "self_attn.rotary_emb.inv_freq" in name:
continue
else:
param.data += base.state_dict()[name]
print(f"Saving the target model to {target_model_path}")
delta.save_pretrained(target_model_path)
delta_tokenizer.save_pretrained(target_model_path)
if __name__ == "__main__":
parser = argparse.ArgumentParser()
parser.add_argument("--base-model-path", type=str, required=True)
parser.add_argument("--target-model-path", type=str, required=True)
parser.add_argument("--delta-path", type=str, required=True)
args = parser.parse_args()
apply_delta(args.base_model_path, args.target_model_path, args.delta_path)
马建仓 AI 助手
尝试更多
代码解读
代码找茬
代码优化
Python
1
https://gitee.com/zeenwu/HuatuoGPT.git
[email protected]:zeenwu/HuatuoGPT.git
zeenwu
HuatuoGPT
HuatuoGPT
main

搜索帮助