from transformers import AutoModelForCausalLM, AutoTokenizer
import torch
from config import Config,Default,Dir


def get_model_train_dataset_json_file():
    model_train_json_file = Dir.MODEL_DATASET_DIR + "/" + Default.TRAIN_JSONL_NEW_FILE
    model_test_json_file = Dir.MODEL_DATASET_DIR + "/" + Default.TEST_JSONL_NEW_FILE
    return model_train_json_file, model_test_json_file


def predict(messages, model, tokenizer):
    device = "cuda"
    text = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
    model_inputs = tokenizer([text], return_tensors="pt").to(device)
    generated_ids = model.generate(model_inputs.input_ids, max_new_tokens=2048)
    generated_ids = [output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)]
    response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
    return response
# 加载原下载路径的tokenizer和model
tokenizer = AutoTokenizer.from_pretrained("./Qwen3-0.6B/checkpoint-1084", use_fast=False, trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("./Qwen3-0.6B/checkpoint-1084", device_map="auto", torch_dtype=torch.bfloat16)
test_texts = {
    'instruction': "你是一个医学专家，你需要根据用户的问题，给出带有思考的回答。",
    'input': "医生，我在研究内耳的前庭部分时，发现了一些特殊的结构，比如前庭嵴。请问前庭内还有哪些特殊的结构，它们的作用是什么？"
}
instruction = test_texts['instruction']
input_value = test_texts['input']
messages = [
    {"role": "system", "content": f"{instruction}"},
    {"role": "user", "content": f"{input_value}"}
]
response = predict(messages, model, tokenizer)
print(response)