skzz commited on
Commit
5a307f6
1 Parent(s): 99159d2

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +17 -11
app.py CHANGED
@@ -50,9 +50,13 @@ class ChatClient:
50
  output_sequences = self.model.generate(**inputs, **gen_kwargs)
51
 
52
  # 解码生成的文本
53
- result_text = self.tokenizer.decode(output_sequences[0], skip_special_tokens=True)
 
54
 
55
- yield result_text
 
 
 
56
 
57
  # 创建客户端实例,指定模型路径
58
  model_path = 'model/v3/'
@@ -63,7 +67,7 @@ client = ChatClient(model_path)
63
 
64
 
65
 
66
- def respond(
67
  message,
68
  history: list[tuple[str, str]],
69
  system_message,
@@ -86,19 +90,21 @@ def respond(
86
 
87
  response = ""
88
 
89
- for message in client.chat_completion(
90
  messages,
91
  max_tokens=max_tokens,
92
  stream=True,
93
  temperature=temperature,
94
  top_p=top_p,
95
  ):
96
- print(message)
97
- token = message
98
- #token = message.choices[0].delta.content
 
 
 
99
 
100
- response += token
101
- yield response
102
 
103
  """
104
  For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
@@ -107,8 +113,8 @@ demo = gr.ChatInterface(
107
  respond,
108
  additional_inputs=[
109
  gr.Textbox(value="Yahoo!ショッピングについての質問を回答してください。", label="System message"),
110
- gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
111
- gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
112
  gr.Slider(
113
  minimum=0.1,
114
  maximum=1.0,
 
50
  output_sequences = self.model.generate(**inputs, **gen_kwargs)
51
 
52
  # 解码生成的文本
53
+ # result_text = self.tokenizer.decode(output_sequences[0], skip_special_tokens=True)
54
+ # yield result_text
55
 
56
+ # 解码生成的文本
57
+ for sequence in output_sequences:
58
+ result_text = self.tokenizer.decode(sequence, skip_special_tokens=True)
59
+ yield result_text
60
 
61
  # 创建客户端实例,指定模型路径
62
  model_path = 'model/v3/'
 
67
 
68
 
69
 
70
+ async def respond(
71
  message,
72
  history: list[tuple[str, str]],
73
  system_message,
 
90
 
91
  response = ""
92
 
93
+ async for message in client.chat_completion(
94
  messages,
95
  max_tokens=max_tokens,
96
  stream=True,
97
  temperature=temperature,
98
  top_p=top_p,
99
  ):
100
+ # print(message)
101
+ # token = message
102
+ # #token = message.choices[0].delta.content
103
+
104
+ # response += token
105
+ # yield response
106
 
107
+ yield message
 
108
 
109
  """
110
  For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
 
113
  respond,
114
  additional_inputs=[
115
  gr.Textbox(value="Yahoo!ショッピングについての質問を回答してください。", label="System message"),
116
+ gr.Slider(minimum=1, maximum=2048, value=1024, step=1, label="Max new tokens"),
117
+ gr.Slider(minimum=0.1, maximum=4.0, value=0.1, step=0.1, label="Temperature"),
118
  gr.Slider(
119
  minimum=0.1,
120
  maximum=1.0,