added baichuan2

fine-tune/slurm_script/hostfile/hostfile-dl-46547085

+f14r1n19 slots=4
+f14r2n00 slots=4
+f14r2n01 slots=4
+f14r2n02 slots=4
+f14r2n03 slots=4
+f14r2n04 slots=4
+f14r2n05 slots=4
+f14r2n06 slots=4
+f14r2n07 slots=4
+f14r2n08 slots=4
+f14r2n09 slots=4
+f14r2n10 slots=4
+f14r2n11 slots=4
+f14r2n12 slots=4
+f14r2n13 slots=4
+f14r2n14 slots=4
+f14r2n15 slots=4
+f14r2n16 slots=4
+f14r2n17 slots=4
+f14r2n18 slots=4
+f14r2n19 slots=4
+f14r3n00 slots=4
+f14r3n01 slots=4
+f14r3n02 slots=4

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")
+messages = []
+messages.append({"role": "user", "content": "解释一下“温故而知新”"})
+response = model.chat(tokenizer, messages)
+print(response)
\ No newline at end of file