from transformers import AutoTokenizer, AutoModelForCausalLM model_id = "Qwen/Qwen3-0.6B" # 加载 tokenizer 和模型 tokenizer = AutoTokenizer.from_pretrained(model_id, trust_remote_code=True) model = AutoModelForCausalLM.from_pretrained(model_id, trust_remote_code=True).eval() # 支持 GPU(如可用) import torch device = "cuda" if torch.cuda.is_available() else "cpu" model = model.to(device) # 推理 prompt = "你好,请介绍一下你自己。" inputs = tokenizer(prompt, return_tensors="pt").to(device) outputs = model.generate(**inputs, max_new_tokens=100) response = tokenizer.decode(outputs[0], skip_special_tokens=True) print(response)