Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
|
@@ -1,85 +1,44 @@
|
|
| 1 |
-
import os
|
| 2 |
-
# This automatically picks up the secret you added in Settings
|
| 3 |
-
token = os.getenv("admin_token")
|
| 4 |
-
|
| 5 |
import gradio as gr
|
| 6 |
from huggingface_hub import InferenceClient
|
| 7 |
-
import os
|
| 8 |
-
|
| 9 |
-
# 1. Initialize Client (Uses HF servers, saves your Space's memory)
|
| 10 |
-
# This will automatically use the HF_TOKEN secret you added to Settings
|
| 11 |
-
client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
|
| 12 |
-
|
| 13 |
-
def respond(message, history):
|
| 14 |
-
messages = []
|
| 15 |
-
|
| 16 |
-
# Build history for the API
|
| 17 |
-
for user_msg, bot_msg in history:
|
| 18 |
-
messages.append({"role": "user", "content": user_msg})
|
| 19 |
-
messages.append({"role": "assistant", "content": bot_msg})
|
| 20 |
-
|
| 21 |
-
messages.append({"role": "user", "content": message})
|
| 22 |
-
|
| 23 |
-
response = ""
|
| 24 |
|
| 25 |
-
|
| 26 |
-
|
| 27 |
-
messages,
|
| 28 |
-
max_tokens=512,
|
| 29 |
-
stream=True,
|
| 30 |
-
temperature=0.7,
|
| 31 |
-
top_p=0.95,
|
| 32 |
-
):
|
| 33 |
-
token = message.choices[0].delta.content
|
| 34 |
-
response += token
|
| 35 |
-
yield response
|
| 36 |
|
| 37 |
-
# 2. Simplified UI
|
| 38 |
def chat_function(message, history):
|
| 39 |
-
|
| 40 |
-
|
| 41 |
-
|
| 42 |
-
"""
|
| 43 |
-
messages = [{"role": "system", "content": "You are a helpful and professional AI assistant."}]
|
| 44 |
|
| 45 |
-
# Add previous conversation turns to the memory
|
| 46 |
for user_msg, bot_msg in history:
|
| 47 |
messages.append({"role": "user", "content": user_msg})
|
| 48 |
messages.append({"role": "assistant", "content": bot_msg})
|
| 49 |
|
| 50 |
-
# Add the current user message
|
| 51 |
messages.append({"role": "user", "content": message})
|
| 52 |
|
| 53 |
response = ""
|
| 54 |
|
| 55 |
-
# Call the model and stream the result
|
| 56 |
for message_obj in client.chat_completion(
|
| 57 |
-
messages,
|
| 58 |
max_tokens=512,
|
| 59 |
stream=True,
|
| 60 |
temperature=0.7,
|
| 61 |
top_p=0.95,
|
| 62 |
):
|
| 63 |
-
token = message_obj.choices[0].delta.content
|
| 64 |
-
|
| 65 |
-
|
| 66 |
-
yield response
|
| 67 |
|
| 68 |
-
# 3. BUILD THE UI (Linking to the function defined above)
|
| 69 |
demo = gr.ChatInterface(
|
| 70 |
-
fn=chat_function,
|
| 71 |
title="Premium AI Chatbot",
|
| 72 |
description="A high-performance assistant hosted on Hugging Face.",
|
| 73 |
-
theme="soft",
|
| 74 |
-
examples=[
|
|
|
|
|
|
|
|
|
|
|
|
|
| 75 |
cache_examples=False,
|
| 76 |
-
chatbot=gr.Chatbot(
|
| 77 |
-
show_label=False,
|
| 78 |
-
show_share_button=False,
|
| 79 |
-
show_copy_button=True,
|
| 80 |
-
likeable=True,
|
| 81 |
-
layout="bubble"
|
| 82 |
-
)
|
| 83 |
)
|
| 84 |
|
| 85 |
if __name__ == "__main__":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
import gradio as gr
|
| 2 |
from huggingface_hub import InferenceClient
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3 |
|
| 4 |
+
# Initialize client
|
| 5 |
+
client = InferenceClient(model="HuggingFaceH4/zephyr-7b-beta")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 6 |
|
|
|
|
| 7 |
def chat_function(message, history):
|
| 8 |
+
messages = [
|
| 9 |
+
{"role": "system", "content": "You are a helpful and professional AI assistant."}
|
| 10 |
+
]
|
|
|
|
|
|
|
| 11 |
|
|
|
|
| 12 |
for user_msg, bot_msg in history:
|
| 13 |
messages.append({"role": "user", "content": user_msg})
|
| 14 |
messages.append({"role": "assistant", "content": bot_msg})
|
| 15 |
|
|
|
|
| 16 |
messages.append({"role": "user", "content": message})
|
| 17 |
|
| 18 |
response = ""
|
| 19 |
|
|
|
|
| 20 |
for message_obj in client.chat_completion(
|
| 21 |
+
messages=messages,
|
| 22 |
max_tokens=512,
|
| 23 |
stream=True,
|
| 24 |
temperature=0.7,
|
| 25 |
top_p=0.95,
|
| 26 |
):
|
| 27 |
+
token = message_obj.choices[0].delta.content or ""
|
| 28 |
+
response += token
|
| 29 |
+
yield response
|
|
|
|
| 30 |
|
|
|
|
| 31 |
demo = gr.ChatInterface(
|
| 32 |
+
fn=chat_function,
|
| 33 |
title="Premium AI Chatbot",
|
| 34 |
description="A high-performance assistant hosted on Hugging Face.",
|
| 35 |
+
theme="soft",
|
| 36 |
+
examples=[
|
| 37 |
+
"What are the benefits of RPA?",
|
| 38 |
+
"How to optimize SIP investments?",
|
| 39 |
+
"Explain financial planning."
|
| 40 |
+
],
|
| 41 |
cache_examples=False,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 42 |
)
|
| 43 |
|
| 44 |
if __name__ == "__main__":
|