Aksh1t commited on
Commit
f846aa8
1 Parent(s): a75ac91

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +10 -8
app.py CHANGED
@@ -1,11 +1,15 @@
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("Aksh1t/mistral-7b-oig-unsloth-merged")
8
 
 
 
 
 
 
 
 
9
 
10
  def respond(
11
  message,
@@ -27,21 +31,20 @@ def respond(
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=[
@@ -58,6 +61,5 @@ demo = gr.ChatInterface(
58
  ],
59
  )
60
 
61
-
62
  if __name__ == "__main__":
63
- demo.launch()
 
1
  import gradio as gr
2
  from huggingface_hub import InferenceClient
3
 
 
 
 
4
  client = InferenceClient("Aksh1t/mistral-7b-oig-unsloth-merged")
5
 
6
+ # Define the custom chat template
7
+ custom_template = {
8
+ "chat": {
9
+ "prompt": "The following is a conversation with an AI assistant. The assistant is helpful, creative, clever, and very friendly.\n\nHuman: {input}\nAI:",
10
+ "stop": ["\nHuman:"]
11
+ }
12
+ }
13
 
14
  def respond(
15
  message,
 
31
 
32
  response = ""
33
 
34
+ # Call chat_completion with custom template and other parameters
35
  for message in client.chat_completion(
36
  messages,
37
  max_tokens=max_tokens,
38
  stream=True,
39
  temperature=temperature,
40
  top_p=top_p,
41
+ template=custom_template # Pass custom template here
42
  ):
43
  token = message.choices[0].delta.content
44
 
45
  response += token
46
  yield response
47
 
 
 
 
48
  demo = gr.ChatInterface(
49
  respond,
50
  additional_inputs=[
 
61
  ],
62
  )
63
 
 
64
  if __name__ == "__main__":
65
+ demo.launch()