mahin1234 commited on
Commit
cab3e58
·
verified ·
1 Parent(s): 9fb6e6c

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +24 -26
app.py CHANGED
@@ -1,6 +1,11 @@
1
  import gradio as gr
2
  from huggingface_hub import InferenceClient
 
3
 
 
 
 
 
4
 
5
  def respond(
6
  message,
@@ -12,18 +17,15 @@ def respond(
12
  hf_token: gr.OAuthToken,
13
  ):
14
  """
15
- 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
16
  """
17
- client = InferenceClient(token=hf_token.token, model="openai/gpt-oss-20b")
18
 
19
  messages = [{"role": "system", "content": system_message}]
20
-
21
  messages.extend(history)
22
-
23
  messages.append({"role": "user", "content": message})
24
 
25
  response = ""
26
-
27
  for message in client.chat_completion(
28
  messages,
29
  max_tokens=max_tokens,
@@ -31,39 +33,35 @@ def respond(
31
  temperature=temperature,
32
  top_p=top_p,
33
  ):
34
- choices = message.choices
35
- token = ""
36
- if len(choices) and choices[0].delta.content:
37
- token = choices[0].delta.content
38
-
39
  response += token
40
  yield response
41
 
42
-
43
- """
44
- For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
45
- """
46
  chatbot = gr.ChatInterface(
47
  respond,
48
  additional_inputs=[
49
- gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
50
- gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
51
- gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
52
- gr.Slider(
53
- minimum=0.1,
54
- maximum=1.0,
55
- value=0.95,
56
- step=0.05,
57
- label="Top-p (nucleus sampling)",
58
  ),
 
 
 
59
  ],
 
 
60
  )
61
 
62
- with gr.Blocks() as demo:
 
 
 
63
  with gr.Sidebar():
64
  gr.LoginButton()
65
  chatbot.render()
66
 
67
-
68
  if __name__ == "__main__":
69
- demo.launch()
 
1
  import gradio as gr
2
  from huggingface_hub import InferenceClient
3
+ import os
4
 
5
+ # ============================================================
6
+ # ১. মডেল আইডি (সিক্রেট থেকে অথবা ডিফল্ট)
7
+ # ============================================================
8
+ model_id = os.getenv("MODEL_ID", "mx-llms/BLM") # আপনার মডেল আইডি
9
 
10
  def respond(
11
  message,
 
17
  hf_token: gr.OAuthToken,
18
  ):
19
  """
20
+ Hugging Face Inference API ব্যবহার করে স্ট্রিমিং রেসপন্স।
21
  """
22
+ client = InferenceClient(token=hf_token.token, model=model_id)
23
 
24
  messages = [{"role": "system", "content": system_message}]
 
25
  messages.extend(history)
 
26
  messages.append({"role": "user", "content": message})
27
 
28
  response = ""
 
29
  for message in client.chat_completion(
30
  messages,
31
  max_tokens=max_tokens,
 
33
  temperature=temperature,
34
  top_p=top_p,
35
  ):
36
+ token = message.choices[0].delta.content or ""
 
 
 
 
37
  response += token
38
  yield response
39
 
40
+ # ============================================================
41
+ # ২. ChatInterface UI
42
+ # ============================================================
 
43
  chatbot = gr.ChatInterface(
44
  respond,
45
  additional_inputs=[
46
+ gr.Textbox(
47
+ value="You are BLM, a helpful AI assistant created by MD Mushfiqur Rahim.",
48
+ label="System message"
 
 
 
 
 
 
49
  ),
50
+ gr.Slider(minimum=1, maximum=4096, value=512, step=1, label="Max new tokens"),
51
+ gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
52
+ gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p"),
53
  ],
54
+ title="BLM AI Assistant",
55
+ description="Created by MD Mushfiqur Rahim. Powered by Hugging Face Inference API.",
56
  )
57
 
58
+ # ============================================================
59
+ # ৩. Login Button সহ ডেমো
60
+ # ============================================================
61
+ with gr.Blocks(title="BLM AI Assistant") as demo:
62
  with gr.Sidebar():
63
  gr.LoginButton()
64
  chatbot.render()
65
 
 
66
  if __name__ == "__main__":
67
+ demo.launch()