Spaces:
Running
Running
Update app.py
Browse files
app.py
CHANGED
@@ -1,9 +1,23 @@
|
|
|
|
|
|
1 |
from huggingface_hub import InferenceClient
|
2 |
import gradio as gr
|
|
|
3 |
|
4 |
-
|
5 |
-
|
6 |
-
)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
7 |
|
8 |
def format_prompt(message, history):
|
9 |
prompt = "<s>"
|
@@ -16,32 +30,47 @@ def format_prompt(message, history):
|
|
16 |
def generate(
|
17 |
prompt, history, temperature=0.9, max_new_tokens=256, top_p=0.95, repetition_penalty=1.0,
|
18 |
):
|
19 |
-
|
20 |
-
|
21 |
-
temperature
|
22 |
-
|
|
|
23 |
|
24 |
-
|
25 |
-
|
26 |
-
|
27 |
-
|
28 |
-
|
29 |
-
|
30 |
-
|
31 |
-
|
32 |
|
33 |
-
|
|
|
34 |
|
35 |
-
|
36 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
37 |
|
38 |
-
|
39 |
-
|
40 |
-
|
41 |
-
|
42 |
|
|
|
|
|
|
|
|
|
|
|
|
|
43 |
|
44 |
-
|
|
|
45 |
gr.Slider(
|
46 |
label="Temperature",
|
47 |
value=0.9,
|
@@ -58,7 +87,7 @@ additional_inputs=[
|
|
58 |
maximum=1048,
|
59 |
step=64,
|
60 |
interactive=True,
|
61 |
-
info="The maximum
|
62 |
),
|
63 |
gr.Slider(
|
64 |
label="Top-p (nucleus sampling)",
|
@@ -77,21 +106,26 @@ additional_inputs=[
|
|
77 |
step=0.05,
|
78 |
interactive=True,
|
79 |
info="Penalize repeated tokens",
|
80 |
-
)
|
81 |
]
|
82 |
|
83 |
-
# Create a Chatbot object
|
84 |
-
chatbot = gr.Chatbot(height=450,
|
85 |
-
layout="bubble")
|
86 |
|
|
|
87 |
with gr.Blocks() as demo:
|
88 |
-
gr.HTML("<h1><center>🤖 Mistral-7B-Chat
|
89 |
gr.ChatInterface(
|
90 |
-
generate,
|
91 |
-
chatbot=chatbot,
|
92 |
additional_inputs=additional_inputs,
|
93 |
-
examples=[
|
94 |
-
|
|
|
|
|
|
|
95 |
)
|
96 |
|
97 |
-
|
|
|
|
|
|
1 |
+
import os
|
2 |
+
import logging
|
3 |
from huggingface_hub import InferenceClient
|
4 |
import gradio as gr
|
5 |
+
from requests.exceptions import ConnectionError
|
6 |
|
7 |
+
# Configure logging
|
8 |
+
logging.basicConfig(level=logging.INFO)
|
9 |
+
logger = logging.getLogger(__name__)
|
10 |
+
|
11 |
+
# Initialize the Hugging Face Inference Client
|
12 |
+
try:
|
13 |
+
client = InferenceClient(
|
14 |
+
model="mistralai/Mistral-7B-Instruct-v0.3",
|
15 |
+
token=os.getenv("HF_TOKEN"), # Ensure HF_TOKEN is set in your environment
|
16 |
+
timeout=30,
|
17 |
+
)
|
18 |
+
except Exception as e:
|
19 |
+
logger.error(f"Failed to initialize InferenceClient: {e}")
|
20 |
+
raise
|
21 |
|
22 |
def format_prompt(message, history):
|
23 |
prompt = "<s>"
|
|
|
30 |
def generate(
|
31 |
prompt, history, temperature=0.9, max_new_tokens=256, top_p=0.95, repetition_penalty=1.0,
|
32 |
):
|
33 |
+
try:
|
34 |
+
temperature = float(temperature)
|
35 |
+
if temperature < 1e-2:
|
36 |
+
temperature = 1e-2
|
37 |
+
top_p = float(top_p)
|
38 |
|
39 |
+
generate_kwargs = dict(
|
40 |
+
temperature=temperature,
|
41 |
+
max_new_tokens=max_new_tokens,
|
42 |
+
top_p=top_p,
|
43 |
+
repetition_penalty=repetition_penalty,
|
44 |
+
do_sample=True,
|
45 |
+
seed=42,
|
46 |
+
)
|
47 |
|
48 |
+
formatted_prompt = format_prompt(prompt, history)
|
49 |
+
logger.info("Sending request to Hugging Face API")
|
50 |
|
51 |
+
stream = client.text_generation(
|
52 |
+
formatted_prompt,
|
53 |
+
**generate_kwargs,
|
54 |
+
stream=True,
|
55 |
+
details=True,
|
56 |
+
return_full_text=False,
|
57 |
+
)
|
58 |
+
output = ""
|
59 |
|
60 |
+
for response in stream:
|
61 |
+
output += response.token.text
|
62 |
+
yield output
|
63 |
+
return output
|
64 |
|
65 |
+
except ConnectionError as e:
|
66 |
+
logger.error(f"Network error: {e}")
|
67 |
+
yield "Error: Unable to connect to the Hugging Face API. Please check your internet connection and try again."
|
68 |
+
except Exception as e:
|
69 |
+
logger.error(f"Error during text generation: {e}")
|
70 |
+
yield f"Error: {str(e)}"
|
71 |
|
72 |
+
# Define additional inputs for Gradio interface
|
73 |
+
additional_inputs = [
|
74 |
gr.Slider(
|
75 |
label="Temperature",
|
76 |
value=0.9,
|
|
|
87 |
maximum=1048,
|
88 |
step=64,
|
89 |
interactive=True,
|
90 |
+
info="The maximum number of new tokens",
|
91 |
),
|
92 |
gr.Slider(
|
93 |
label="Top-p (nucleus sampling)",
|
|
|
106 |
step=0.05,
|
107 |
interactive=True,
|
108 |
info="Penalize repeated tokens",
|
109 |
+
),
|
110 |
]
|
111 |
|
112 |
+
# Create a Chatbot object
|
113 |
+
chatbot = gr.Chatbot(height=450, layout="bubble")
|
|
|
114 |
|
115 |
+
# Build the Gradio interface
|
116 |
with gr.Blocks() as demo:
|
117 |
+
gr.HTML("<h1><center>🤖 Mistral-7B-Chat 💬</center></h1>")
|
118 |
gr.ChatInterface(
|
119 |
+
fn=generate,
|
120 |
+
chatbot=chatbot,
|
121 |
additional_inputs=additional_inputs,
|
122 |
+
examples=[
|
123 |
+
["Give me the code for Binary Search in C++"],
|
124 |
+
["Explain the chapter of The Grand Inquisitor from The Brothers Karamazov."],
|
125 |
+
["Explain Newton's second law."],
|
126 |
+
],
|
127 |
)
|
128 |
|
129 |
+
if __name__ == "__main__":
|
130 |
+
logger.info("Starting Gradio application")
|
131 |
+
demo.launch()
|