Views
No views yet
1import torch
2from transformers import AutoTokenizer, AutoModelForCausalLM
3
4tokenizer = AutoTokenizer.from_pretrained('qhduan/aquilachat-7b')
5model = AutoModelForCausalLM.from_pretrained('qhduan/aquilachat-7b', trust_remote_code=True)
6model = model.eval().half().cuda()
7
8question = '北京为什么是中国的首都？'
9prompt = (
10 '''A chat between a curious human and an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the human's questions.'''
11 f'''###Human: {question}###Assistant:'''
12)
13with torch.no_grad():
14 ret = model.generate(
15 **tokenizer(prompt, return_tensors='pt').to('cuda'),
16 do_sample=False,
17 max_new_tokens=200,
18 use_cache=True
19 )
20 output_ids = ret[0].detach().cpu().numpy().tolist()
21 if 100007 in output_ids:
22 output_ids = output_ids[:output_ids.index(100007)]
23 elif 0 in output_ids:
24 output_ids = output_ids[:output_ids.index(0)]
25 # 北京之所以成为中国的首都，是因为它在中国历史和文化中的重要地位和政治、经济、文化等方面的影响力。
26 print(tokenizer.decode(output_ids))