Update README.md

README.md

@@ -1,6 +1,48 @@
 ---
 library_name: peft
 ---
+
+在sharegpt_cn数据集上训练得到的llama2 Chinese chat 13b，为减轻文件大小负担这里只放出了qlora的权重  
+请拉取https://huggingface.co/TheBloke/Llama-2-13B-fp16作为基础权重，使用如下脚步进行合并：
+```python
+from peft import PeftModel
+from transformers import AutoModelForCausalLM, AutoTokenizer
+import torch
+"""
+使用该脚本（引用于firefly项目），将qlora的权重合并大base model中
+"""
+
+
+def merge_lora_to_base_model():
+    model_name_or_path = '/data/TheBloke/Llama-2-13B-fp16'
+    adapter_name_or_path = '/data/llama2-13b-Chinese-chat'
+    save_path = '/data/llama2-13b-Chinese-chat_v1'
+
+    tokenizer = AutoTokenizer.from_pretrained(
+        model_name_or_path,
+        trust_remote_code=True
+    )
+    model = AutoModelForCausalLM.from_pretrained(
+        model_name_or_path,
+        trust_remote_code=True,
+        low_cpu_mem_usage=True,
+        torch_dtype=torch.float16,
+        device_map='auto'
+    )
+    print("load model success")
+    model = PeftModel.from_pretrained(model, adapter_name_or_path)
+    print("load adapter success")
+    model = model.merge_and_unload()
+    print("merge success")
+
+    tokenizer.save_pretrained(save_path)
+    model.save_pretrained(save_path)
+
+
+if __name__ == '__main__':
+    merge_lora_to_base_model()
+
+```
 ## Training procedure