update

fine-tune/slurm_script/infer_test.py

 import torch
 from transformers import AutoModelForCausalLM, AutoTokenizer
 from transformers.generation.utils import GenerationConfig
-tokenizer = AutoTokenizer.from_pretrained("/public/home/zhaoying1/work/Baichuan2-main/fine-tune/slurm_script/output/checkpoint-420", use_fast=False, trust_remote_code=True)
-model = AutoModelForCausalLM.from_pretrained("/public/home/zhaoying1/work/Baichuan2-main/fine-tune/slurm_script/output/checkpoint-420", device_map="auto", torch_dtype=torch.bfloat16, trust_remote_code=True)
-model.generation_config = GenerationConfig.from_pretrained("/public/home/zhaoying1/work/Baichuan2-main/fine-tune/slurm_script/output/checkpoint-420")
+tokenizer = AutoTokenizer.from_pretrained("/baichuan/baichuan-13b-chat", use_fast=False, trust_remote_code=True)
+
+# lora
+from peft import AutoPeftModelForCausalLM
+model = AutoPeftModelForCausalLM.from_pretrained("/baichuan/baichuan2_pytorch/fine-tune/output/checkpoint-4", trust_remote_code=True)
+
+# full parameter
+# model = AutoModelForCausalLM.from_pretrained("/public/home/zhaoying1/work/Baichuan2-main/fine-tune/slurm_script/output/checkpoint-420", device_map="auto", torch_dtype=torch.bfloat16, trust_remote_code=True)
+model.generation_config = GenerationConfig.from_pretrained("/baichuan/baichuan-13b-chat")
 messages = []
 messages.append({"role": "user", "content": "解释一下“温故而知新”"})
 response = model.chat(tokenizer, messages)

model.properties

@@ -5,6 +5,6 @@ modelName=baichuan2_pytorch
 # 模型描述
 modelDescription=基于pytorch的baichuan2
 # 应用场景
-appScenario=训练,推理,train,inference,nlp,智能聊天助手
+appScenario=训练,推理,对话问答,医疗,教育,科研,金融
 # 框架类型
 frameType=Pytorch,Transformers,Deepspeed