jiaxiang zhang

icon company jiaxiang zhang icon location I am a student,love learning and code, hope to learn more and make more friends in Github.

Results 1 issues of jiaxiang zhang

Reranker微调好后，做推理时加载原始minicpm_layerwise报错，是否可以调用本地已下好的模型？

6 comment

#微调模型加载 model_path = '/home/fintuned_model/remini_hardmined_fintuned10_rawmodel/checkpoint-1083' tokenizer = AutoTokenizer.from_pretrained(model_path, trust_remote_code=True) model = AutoModelForCausalLM.from_pretrained(model_path, trust_remote_code=True, torch_dtype=torch.bfloat16) # 将模型转移到 GPU device = torch.device("cuda" if torch.cuda.is_available() else "cpu") model = model.to(device) model.eval() file_path = '/home/test_model/result/全部底库_500测试集召回结果.xlsx'...