vericudebuget commited on
Commit
1854dfd
1 Parent(s): 93f834f

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +18 -70
app.py CHANGED
@@ -1,25 +1,19 @@
1
  from huggingface_hub import InferenceClient
2
  import gradio as gr
3
 
4
- client = InferenceClient(
5
- "mistralai/Mixtral-8x7B-Instruct-v0.1"
6
- )
7
-
8
 
9
  def format_prompt(message, history):
10
- prompt = "<s>"
11
- for user_prompt, bot_response in history:
12
- prompt += f"[INST] {user_prompt} [/INST]"
13
- prompt += f" {bot_response}</s> "
14
- prompt += f"[INST] {message} [/INST]"
15
- return prompt
16
-
17
- def generate(
18
- prompt, history, system_prompt, temperature=0.9, max_new_tokens=256, top_p=0.95, repetition_penalty=1.0,
19
- ):
20
- temperature = float(temperature)
21
- if temperature < 1e-2:
22
- temperature = 1e-2
23
  top_p = float(top_p)
24
 
25
  generate_kwargs = dict(
@@ -40,64 +34,18 @@ def generate(
40
  yield output
41
  return output
42
 
43
-
44
- additional_inputs=[
45
- gr.Textbox(
46
- label="System Prompt",
47
- max_lines=1,
48
- interactive=True,
49
- ),
50
- gr.Slider(
51
- label="Temperature",
52
- value=0.9,
53
- minimum=0.0,
54
- maximum=1.0,
55
- step=0.05,
56
- interactive=True,
57
- info="Higher values produce more diverse outputs",
58
- ),
59
- gr.Slider(
60
- label="Max new tokens",
61
- value=256,
62
- minimum=0,
63
- maximum=9048,
64
- step=64,
65
- interactive=True,
66
- info="The maximum numbers of new tokens",
67
- ),
68
- gr.Slider(
69
- label="Top-p (nucleus sampling)",
70
- value=0.90,
71
- minimum=0.0,
72
- maximum=1,
73
- step=0.05,
74
- interactive=True,
75
- info="Higher values sample more low-probability tokens",
76
- ),
77
- gr.Slider(
78
- label="Repetition penalty",
79
- value=1.2,
80
- minimum=1.0,
81
- maximum=2.0,
82
- step=0.05,
83
- interactive=True,
84
- info="Penalize repeated tokens",
85
- )
86
  ]
87
 
88
- examples=[["I'm planning a vacation to Japan. Can you suggest a one-week itinerary including must-visit places and local cuisines to try?", None, None, None, None, None, ],
89
- ["Can you write a short story about a time-traveling detective who solves historical mysteries?", None, None, None, None, None,],
90
- ["I'm trying to learn French. Can you provide some common phrases that would be useful for a beginner, along with their pronunciations?", None, None, None, None, None,],
91
- ["I have chicken, rice, and bell peppers in my kitchen. Can you suggest an easy recipe I can make with these ingredients?", None, None, None, None, None,],
92
- ["Can you explain how the QuickSort algorithm works and provide a Python implementation?", None, None, None, None, None,],
93
- ["What are some unique features of Rust that make it stand out compared to other systems programming languages like C++?", None, None, None, None, None,],
94
- ]
95
-
96
  gr.ChatInterface(
97
  fn=generate,
98
  chatbot=gr.Chatbot(show_label=False, show_share_button=False, show_copy_button=True, likeable=True, layout="panel"),
99
  additional_inputs=additional_inputs,
100
  title="Mixtral 46.7B",
101
- examples=examples,
102
  concurrency_limit=20,
103
- ).launch(show_api=False)
 
1
  from huggingface_hub import InferenceClient
2
  import gradio as gr
3
 
4
+ # Initialize the InferenceClient
5
+ client = InferenceClient("mistralai/Mixtral-8x7B-Instruct-v0.1")
 
 
6
 
7
  def format_prompt(message, history):
8
+ prompt = "<s>"
9
+ for user_prompt, bot_response in history:
10
+ prompt += f"[INST] {user_prompt} [/INST]"
11
+ prompt += f" {bot_response}</s> "
12
+ prompt += f"[INST] {message} [/INST]"
13
+ return prompt
14
+
15
+ def generate(prompt, history, system_prompt, temperature=0.9, max_new_tokens=9048, top_p=0.95, repetition_penalty=1.0):
16
+ temperature = max(float(temperature), 1e-2)
 
 
 
 
17
  top_p = float(top_p)
18
 
19
  generate_kwargs = dict(
 
34
  yield output
35
  return output
36
 
37
+ additional_inputs = [
38
+ gr.Textbox(label="System Prompt", max_lines=1, interactive=True),
39
+ gr.Slider(label="Temperature", value=0.9, minimum=0.0, maximum=1.0, step=0.05, interactive=True, info="Higher values produce more diverse outputs"),
40
+ gr.Slider(label="Max new tokens", value=256, minimum=9000, maximum=9048, step=64, interactive=True, info="The maximum numbers of new tokens"),
41
+ gr.Slider(label="Top-p (nucleus sampling)", value=0.90, minimum=0.0, maximum=1, step=0.05, interactive=True, info="Higher values sample more low-probability tokens"),
42
+ gr.Slider(label="Repetition penalty", value=1.2, minimum=1.0, maximum=2.0, step=0.05, interactive=True, info="Penalize repeated tokens")
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
43
  ]
44
 
 
 
 
 
 
 
 
 
45
  gr.ChatInterface(
46
  fn=generate,
47
  chatbot=gr.Chatbot(show_label=False, show_share_button=False, show_copy_button=True, likeable=True, layout="panel"),
48
  additional_inputs=additional_inputs,
49
  title="Mixtral 46.7B",
 
50
  concurrency_limit=20,
51
+ ).launch(show_api=False)