File size: 2,926 Bytes
506c93a
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
52
53
54
55
56
57
58
59
model_name = "Qwen_Local"
cmd_to_install = "`pip install -r request_llms/requirements_qwen_local.txt`"

from toolbox import ProxyNetworkActivate, get_conf
from .local_llm_class import LocalLLMHandle, get_local_llm_predict_fns



# ------------------------------------------------------------------------------------------------------------------------
# ๐Ÿ”Œ๐Ÿ’ป Local Model
# ------------------------------------------------------------------------------------------------------------------------
class GetQwenLMHandle(LocalLLMHandle):

    def load_model_info(self):
        # ๐Ÿƒโ€โ™‚๏ธ๐Ÿƒโ€โ™‚๏ธ๐Ÿƒโ€โ™‚๏ธ ๅญ่ฟ›็จ‹ๆ‰ง่กŒ
        self.model_name = model_name
        self.cmd_to_install = cmd_to_install

    def load_model_and_tokenizer(self):
        # ๐Ÿƒโ€โ™‚๏ธ๐Ÿƒโ€โ™‚๏ธ๐Ÿƒโ€โ™‚๏ธ ๅญ่ฟ›็จ‹ๆ‰ง่กŒ
        # from modelscope import AutoModelForCausalLM, AutoTokenizer, GenerationConfig
        from transformers import AutoModelForCausalLM, AutoTokenizer
        from transformers.generation import GenerationConfig
        with ProxyNetworkActivate('Download_LLM'):
            model_id = get_conf('QWEN_LOCAL_MODEL_SELECTION')
            self._tokenizer = AutoTokenizer.from_pretrained(model_id, trust_remote_code=True, resume_download=True)
            # use fp16
            model = AutoModelForCausalLM.from_pretrained(model_id, device_map="auto", trust_remote_code=True).eval()
            model.generation_config = GenerationConfig.from_pretrained(model_id, trust_remote_code=True)  # ๅฏๆŒ‡ๅฎšไธๅŒ็š„็”Ÿๆˆ้•ฟๅบฆใ€top_p็ญ‰็›ธๅ…ณ่ถ…ๅ‚
            self._model = model

        return self._model, self._tokenizer

    def llm_stream_generator(self, **kwargs):
        # ๐Ÿƒโ€โ™‚๏ธ๐Ÿƒโ€โ™‚๏ธ๐Ÿƒโ€โ™‚๏ธ ๅญ่ฟ›็จ‹ๆ‰ง่กŒ
        def adaptor(kwargs):
            query = kwargs['query']
            max_length = kwargs['max_length']
            top_p = kwargs['top_p']
            temperature = kwargs['temperature']
            history = kwargs['history']
            return query, max_length, top_p, temperature, history

        query, max_length, top_p, temperature, history = adaptor(kwargs)

        for response in self._model.chat_stream(self._tokenizer, query, history=history):
            yield response
        
    def try_to_import_special_deps(self, **kwargs):
        # import something that will raise error if the user does not install requirement_*.txt
        # ๐Ÿƒโ€โ™‚๏ธ๐Ÿƒโ€โ™‚๏ธ๐Ÿƒโ€โ™‚๏ธ ไธป่ฟ›็จ‹ๆ‰ง่กŒ
        import importlib
        importlib.import_module('modelscope')


# ------------------------------------------------------------------------------------------------------------------------
# ๐Ÿ”Œ๐Ÿ’ป GPT-Academic Interface
# ------------------------------------------------------------------------------------------------------------------------
predict_no_ui_long_connection, predict = get_local_llm_predict_fns(GetQwenLMHandle, model_name)