0x7o commited on
Commit
d57f720
1 Parent(s): 6d7bf51

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +27 -32
app.py CHANGED
@@ -1,11 +1,8 @@
1
  import gradio as gr
2
  from huggingface_hub import InferenceClient
3
 
4
- """
5
- For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
6
- """
7
- client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
8
-
9
 
10
  def respond(
11
  message,
@@ -15,49 +12,47 @@ def respond(
15
  temperature,
16
  top_p,
17
  ):
 
18
  messages = [{"role": "system", "content": system_message}]
19
-
20
- for val in history:
21
- if val[0]:
22
- messages.append({"role": "user", "content": val[0]})
23
- if val[1]:
24
- messages.append({"role": "assistant", "content": val[1]})
25
-
26
  messages.append({"role": "user", "content": message})
27
 
28
  response = ""
29
 
30
- for message in client.chat_completion(
31
- messages,
32
- max_tokens=max_tokens,
 
33
  stream=True,
34
  temperature=temperature,
35
  top_p=top_p,
36
  ):
37
- token = message.choices[0].delta.content
38
-
39
- response += token
40
  yield response
41
 
42
- """
43
- For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
44
- """
45
  demo = gr.ChatInterface(
46
  respond,
47
  additional_inputs=[
48
- gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
49
- gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
50
- gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
51
- gr.Slider(
52
- minimum=0.1,
53
- maximum=1.0,
54
- value=0.95,
55
- step=0.05,
56
- label="Top-p (nucleus sampling)",
57
- ),
 
58
  ],
59
  )
60
 
61
-
62
  if __name__ == "__main__":
63
  demo.launch()
 
1
  import gradio as gr
2
  from huggingface_hub import InferenceClient
3
 
4
+ # Инициализация клиента для модели Aeonium
5
+ client = InferenceClient("aeonium/Aeonium-v1.1-Chat-4B")
 
 
 
6
 
7
  def respond(
8
  message,
 
12
  temperature,
13
  top_p,
14
  ):
15
+ # Формирование сообщений для чата
16
  messages = [{"role": "system", "content": system_message}]
17
+
18
+ for user_msg, assistant_msg in history:
19
+ if user_msg:
20
+ messages.append({"role": "user", "content": user_msg})
21
+ if assistant_msg:
22
+ messages.append({"role": "assistant", "content": assistant_msg})
23
+
24
  messages.append({"role": "user", "content": message})
25
 
26
  response = ""
27
 
28
+ # Генерация ответа с использованием Inference API
29
+ for chunk in client.text_generation(
30
+ prompt=client.tokenizer.apply_chat_template(messages, tokenize=False),
31
+ max_new_tokens=max_tokens,
32
  stream=True,
33
  temperature=temperature,
34
  top_p=top_p,
35
  ):
36
+ response += chunk.token.text
 
 
37
  yield response
38
 
39
+ # Настройка интерфейса Gradio
 
 
40
  demo = gr.ChatInterface(
41
  respond,
42
  additional_inputs=[
43
+ gr.Textbox(value="Ты дружелюбный ассистент, созданный компанией Aeonium.", label="Системное сообщение"),
44
+ gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Максимальное количество новых токенов"),
45
+ gr.Slider(minimum=0.1, maximum=2.0, value=0.7, step=0.1, label="Температура"),
46
+ gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p (nucleus sampling)"),
47
+ ],
48
+ title="Чат с Aeonium v1.1",
49
+ description="Это чат-интерфейс для модели Aeonium v1.1 Chat 4B. Задавайте вопросы и получайте ответы!",
50
+ examples=[
51
+ "Привет! Как дела?",
52
+ "Расскажи мне о квантовой физике",
53
+ "Какие интересные книги ты можешь порекомендовать?",
54
  ],
55
  )
56
 
 
57
  if __name__ == "__main__":
58
  demo.launch()