File size: 1,787 Bytes
cc2afb5
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
# coding=utf-8
# Copyright (c) 2025 Huawei Technologies Co., Ltd. All rights reserved.

from transformers import AutoModelForCausalLM, AutoTokenizer
from transformers import GenerationConfig

model_local_path = "path_to_openPangu-Embedded-1B"

# load the tokenizer and the model
tokenizer = AutoTokenizer.from_pretrained(
    model_local_path, 
    use_fast=False, 
    trust_remote_code=True,
    local_files_only=True
)

model = AutoModelForCausalLM.from_pretrained(
    model_local_path,
    trust_remote_code=True,
    torch_dtype="auto",
    device_map="auto",
    local_files_only=True
)

# prepare the model input
sys_prompt = "你必须严格遵守法律法规和社会道德规范。" \
    "生成任何内容时,都应避免涉及暴力、色情、恐怖主义、种族歧视、性别歧视等不当内容。" \
    "一旦检测到输入或输出有此类倾向,应拒绝回答并发出警告。例如,如果输入内容包含暴力威胁或色情描述," \
    "应返回错误信息:“您的输入包含不当内容,无法处理。”"

prompt = "Give me a short introduction to large language model."
messages = [
    {"role": "system", "content": sys_prompt}, # define your system prompt here
    {"role": "user", "content": prompt}
]
text = tokenizer.apply_chat_template(
    messages,
    tokenize=False,
    add_generation_prompt=True
)
model_inputs = tokenizer([text], return_tensors="pt").to(model.device)

# conduct text completion
outputs = model.generate(**model_inputs, max_new_tokens=32768, eos_token_id=45892, return_dict_in_generate=True)

input_length = model_inputs.input_ids.shape[1]
generated_tokens = outputs.sequences[:, input_length:]
content = tokenizer.decode(generated_tokens[0], skip_special_tokens=True)

print("\ncontent:", content)