Update app.py
Browse files
app.py
CHANGED
@@ -1,164 +1,164 @@
|
|
1 |
|
2 |
-
import streamlit as st
|
3 |
-
from openai import OpenAI
|
4 |
-
import os
|
5 |
-
import sys
|
6 |
-
from dotenv import load_dotenv, dotenv_values
|
7 |
-
load_dotenv()
|
8 |
|
9 |
|
10 |
|
11 |
-
# initialize the client
|
12 |
-
client = OpenAI(
|
13 |
-
|
14 |
-
|
15 |
-
)
|
16 |
|
17 |
|
18 |
|
19 |
-
#Create supported models
|
20 |
-
model_links ={
|
21 |
-
|
22 |
-
|
23 |
-
|
24 |
-
|
25 |
-
|
26 |
|
27 |
-
}
|
28 |
|
29 |
-
#Pull info about the model to display
|
30 |
-
model_info ={
|
31 |
-
|
32 |
-
|
33 |
-
|
34 |
-
|
35 |
|
36 |
-
|
37 |
-
|
38 |
-
|
39 |
-
|
40 |
|
41 |
|
42 |
-
|
43 |
-
|
44 |
-
|
45 |
-
|
46 |
-
|
47 |
-
|
48 |
-
|
49 |
-
|
50 |
-
|
51 |
-
|
52 |
-
|
53 |
-
|
54 |
-
|
55 |
-
|
56 |
-
|
57 |
-
|
58 |
-
|
59 |
-
|
60 |
-
|
61 |
-
|
62 |
-
|
63 |
-
|
64 |
-
|
65 |
-
|
66 |
-
|
67 |
-
}
|
68 |
-
|
69 |
-
def reset_conversation():
|
70 |
-
|
71 |
-
|
72 |
-
|
73 |
-
|
74 |
-
|
75 |
-
|
76 |
|
77 |
|
78 |
|
79 |
-
# Define the available models
|
80 |
-
models =[key for key in model_links.keys()]
|
81 |
|
82 |
-
# Create the sidebar with the dropdown for model selection
|
83 |
-
selected_model = st.sidebar.selectbox("Select Model", models)
|
84 |
|
85 |
-
#Create a temperature slider
|
86 |
-
temp_values = st.sidebar.slider('Select a temperature value', 0.0, 1.0, (0.5))
|
87 |
|
88 |
|
89 |
-
#Add reset button to clear conversation
|
90 |
-
st.sidebar.button('Reset Chat', on_click=reset_conversation) #Reset button
|
91 |
|
92 |
|
93 |
-
# Create model description
|
94 |
-
st.sidebar.write(f"You're now chatting with **{selected_model}**")
|
95 |
-
st.sidebar.markdown(model_info[selected_model]['description'])
|
96 |
-
st.sidebar.image(model_info[selected_model]['logo'])
|
97 |
-
# st.sidebar.markdown("*Generated content may be inaccurate or false.*")
|
98 |
-
# st.sidebar.markdown("\nLearn how to build this chatbot [here](https://ngebodh.github.io/projects/2024-03-05/).")
|
99 |
-
# st.sidebar.markdown("\nRun into issues? Try the [back-up](https://huggingface.co/spaces/ngebodh/SimpleChatbot-Backup).")
|
100 |
|
101 |
|
102 |
|
103 |
-
if "prev_option" not in st.session_state:
|
104 |
-
|
105 |
|
106 |
-
if st.session_state.prev_option != selected_model:
|
107 |
-
|
108 |
-
|
109 |
-
|
110 |
-
|
111 |
|
112 |
|
113 |
|
114 |
-
#Pull in the model we want to use
|
115 |
-
repo_id = model_links[selected_model]
|
116 |
|
117 |
|
118 |
-
st.subheader(f'AI - {selected_model}')
|
119 |
-
# st.title(f'ChatBot Using {selected_model}')
|
120 |
|
121 |
-
# Set a default model
|
122 |
-
if selected_model not in st.session_state:
|
123 |
-
|
124 |
|
125 |
-
# Initialize chat history
|
126 |
-
if "messages" not in st.session_state:
|
127 |
-
|
128 |
|
129 |
|
130 |
-
# Display chat messages from history on app rerun
|
131 |
-
for message in st.session_state.messages:
|
132 |
-
|
133 |
-
|
134 |
|
135 |
|
136 |
|
137 |
-
# Accept user input
|
138 |
-
if prompt := st.chat_input(f"Hi I'm {selected_model}, ask me a question"):
|
139 |
|
140 |
-
|
141 |
-
|
142 |
-
|
143 |
-
|
144 |
-
|
145 |
|
146 |
|
147 |
-
|
148 |
-
|
149 |
-
|
150 |
-
|
151 |
-
|
152 |
-
|
153 |
-
|
154 |
-
|
155 |
-
|
156 |
-
|
157 |
-
|
158 |
-
|
159 |
|
160 |
-
|
161 |
-
|
162 |
|
163 |
|
164 |
|
@@ -168,54 +168,54 @@ if prompt := st.chat_input(f"Hi I'm {selected_model}, ask me a question"):
|
|
168 |
|
169 |
|
170 |
|
171 |
-
|
172 |
-
|
173 |
-
|
174 |
|
175 |
-
|
176 |
-
|
177 |
-
|
178 |
-
|
179 |
-
|
180 |
-
|
181 |
-
|
182 |
|
183 |
-
|
184 |
-
|
185 |
-
|
186 |
-
|
187 |
-
|
188 |
-
|
189 |
-
|
190 |
|
191 |
-
|
192 |
-
|
193 |
-
|
194 |
-
|
195 |
-
|
196 |
-
|
197 |
-
|
198 |
-
|
199 |
|
200 |
-
|
201 |
|
202 |
-
|
203 |
-
|
204 |
|
205 |
-
|
206 |
-
|
207 |
-
|
208 |
-
|
209 |
|
210 |
|
211 |
-
|
212 |
-
|
213 |
|
214 |
-
|
215 |
-
|
216 |
-
|
217 |
-
|
218 |
-
|
219 |
-
|
220 |
|
221 |
-
|
|
|
1 |
|
2 |
+
# import streamlit as st
|
3 |
+
# from openai import OpenAI
|
4 |
+
# import os
|
5 |
+
# import sys
|
6 |
+
# from dotenv import load_dotenv, dotenv_values
|
7 |
+
# load_dotenv()
|
8 |
|
9 |
|
10 |
|
11 |
+
# # initialize the client
|
12 |
+
# client = OpenAI(
|
13 |
+
# base_url="https://api-inference.huggingface.co/v1",
|
14 |
+
# api_key=os.environ.get('HUGGINGFACEHUB_API_TOKEN')#"hf_xxx" # Replace with your token
|
15 |
+
# )
|
16 |
|
17 |
|
18 |
|
19 |
+
# #Create supported models
|
20 |
+
# model_links ={
|
21 |
+
# "Mistral-7b":"mistralai/Mistral-7B-Instruct-v0.2",
|
22 |
+
# "Mistral-8x7b":"mistralai/Mixtral-8x7B-Instruct-v0.1"
|
23 |
+
# # "Gemma-7B":"google/gemma-7b-it",
|
24 |
+
# # "Gemma-2B":"google/gemma-2b-it",
|
25 |
+
# # "Zephyr-7B-β":"HuggingFaceH4/zephyr-7b-beta",
|
26 |
|
27 |
+
# }
|
28 |
|
29 |
+
# #Pull info about the model to display
|
30 |
+
# model_info ={
|
31 |
+
# "Mistral-7b":
|
32 |
+
# {'description':"""The Mistral 7B model is a **Large Language Model (LLM)** that's able to have question and answer interactions.\n \
|
33 |
+
# \nIt was created by the [**Mistral AI**](https://mistral.ai/news/announcing-mistral-7b/) team as has over **7 billion parameters.** \n""",
|
34 |
+
# 'logo':'https://mistral.ai/images/logo_hubc88c4ece131b91c7cb753f40e9e1cc5_2589_256x0_resize_q97_h2_lanczos_3.webp'},
|
35 |
|
36 |
+
# "Mistral-8x7b":
|
37 |
+
# {'description':"""The Mistral 8x7B model is a **Large Language Model (LLM)** that's able to have question and answer interactions.\n \
|
38 |
+
# \nIt was created by the [**Mistral AI**](https://mistral.ai/news/announcing-mistral-8x7b/) team as has based on MOE arch.** \n""",
|
39 |
+
# 'logo':'https://mistral.ai/images/logo_hubc88c4ece131b91c7cb753f40e9e1cc5_2589_256x0_resize_q97_h2_lanczos_3.webp'},
|
40 |
|
41 |
|
42 |
+
# # "Gemma-7B":
|
43 |
+
# # {'description':"""The Gemma model is a **Large Language Model (LLM)** that's able to have question and answer interactions.\n \
|
44 |
+
# # \nIt was created by the [**Google's AI Team**](https://blog.google/technology/developers/gemma-open-models/) team as has over **7 billion parameters.** \n""",
|
45 |
+
# # 'logo':'https://pbs.twimg.com/media/GG3sJg7X0AEaNIq.jpg'},
|
46 |
+
# # "Gemma-2B":
|
47 |
+
# # {'description':"""The Gemma model is a **Large Language Model (LLM)** that's able to have question and answer interactions.\n \
|
48 |
+
# # \nIt was created by the [**Google's AI Team**](https://blog.google/technology/developers/gemma-open-models/) team as has over **2 billion parameters.** \n""",
|
49 |
+
# # 'logo':'https://pbs.twimg.com/media/GG3sJg7X0AEaNIq.jpg'},
|
50 |
+
# # "Zephyr-7B":
|
51 |
+
# # {'description':"""The Zephyr model is a **Large Language Model (LLM)** that's able to have question and answer interactions.\n \
|
52 |
+
# # \nFrom Huggingface: \n\
|
53 |
+
# # Zephyr is a series of language models that are trained to act as helpful assistants. \
|
54 |
+
# # [Zephyr 7B Gemma](https://huggingface.co/HuggingFaceH4/zephyr-7b-gemma-v0.1)\
|
55 |
+
# # is the third model in the series, and is a fine-tuned version of google/gemma-7b \
|
56 |
+
# # that was trained on on a mix of publicly available, synthetic datasets using Direct Preference Optimization (DPO)\n""",
|
57 |
+
# # 'logo':'https://huggingface.co/HuggingFaceH4/zephyr-7b-gemma-v0.1/resolve/main/thumbnail.png'},
|
58 |
+
# # "Zephyr-7B-β":
|
59 |
+
# # {'description':"""The Zephyr model is a **Large Language Model (LLM)** that's able to have question and answer interactions.\n \
|
60 |
+
# # \nFrom Huggingface: \n\
|
61 |
+
# # Zephyr is a series of language models that are trained to act as helpful assistants. \
|
62 |
+
# # [Zephyr-7B-β](https://huggingface.co/HuggingFaceH4/zephyr-7b-beta)\
|
63 |
+
# # is the second model in the series, and is a fine-tuned version of mistralai/Mistral-7B-v0.1 \
|
64 |
+
# # that was trained on on a mix of publicly available, synthetic datasets using Direct Preference Optimization (DPO)\n""",
|
65 |
+
# # 'logo':'https://huggingface.co/HuggingFaceH4/zephyr-7b-alpha/resolve/main/thumbnail.png'},
|
66 |
+
|
67 |
+
# }
|
68 |
+
|
69 |
+
# def reset_conversation():
|
70 |
+
# '''
|
71 |
+
# Resets Conversation
|
72 |
+
# '''
|
73 |
+
# st.session_state.conversation = []
|
74 |
+
# st.session_state.messages = []
|
75 |
+
# return None
|
76 |
|
77 |
|
78 |
|
79 |
+
# # Define the available models
|
80 |
+
# models =[key for key in model_links.keys()]
|
81 |
|
82 |
+
# # Create the sidebar with the dropdown for model selection
|
83 |
+
# selected_model = st.sidebar.selectbox("Select Model", models)
|
84 |
|
85 |
+
# #Create a temperature slider
|
86 |
+
# temp_values = st.sidebar.slider('Select a temperature value', 0.0, 1.0, (0.5))
|
87 |
|
88 |
|
89 |
+
# #Add reset button to clear conversation
|
90 |
+
# st.sidebar.button('Reset Chat', on_click=reset_conversation) #Reset button
|
91 |
|
92 |
|
93 |
+
# # Create model description
|
94 |
+
# st.sidebar.write(f"You're now chatting with **{selected_model}**")
|
95 |
+
# st.sidebar.markdown(model_info[selected_model]['description'])
|
96 |
+
# st.sidebar.image(model_info[selected_model]['logo'])
|
97 |
+
# # st.sidebar.markdown("*Generated content may be inaccurate or false.*")
|
98 |
+
# # st.sidebar.markdown("\nLearn how to build this chatbot [here](https://ngebodh.github.io/projects/2024-03-05/).")
|
99 |
+
# # st.sidebar.markdown("\nRun into issues? Try the [back-up](https://huggingface.co/spaces/ngebodh/SimpleChatbot-Backup).")
|
100 |
|
101 |
|
102 |
|
103 |
+
# if "prev_option" not in st.session_state:
|
104 |
+
# st.session_state.prev_option = selected_model
|
105 |
|
106 |
+
# if st.session_state.prev_option != selected_model:
|
107 |
+
# st.session_state.messages = []
|
108 |
+
# # st.write(f"Changed to {selected_model}")
|
109 |
+
# st.session_state.prev_option = selected_model
|
110 |
+
# reset_conversation()
|
111 |
|
112 |
|
113 |
|
114 |
+
# #Pull in the model we want to use
|
115 |
+
# repo_id = model_links[selected_model]
|
116 |
|
117 |
|
118 |
+
# st.subheader(f'AI - {selected_model}')
|
119 |
+
# # st.title(f'ChatBot Using {selected_model}')
|
120 |
|
121 |
+
# # Set a default model
|
122 |
+
# if selected_model not in st.session_state:
|
123 |
+
# st.session_state[selected_model] = model_links[selected_model]
|
124 |
|
125 |
+
# # Initialize chat history
|
126 |
+
# if "messages" not in st.session_state:
|
127 |
+
# st.session_state.messages = []
|
128 |
|
129 |
|
130 |
+
# # Display chat messages from history on app rerun
|
131 |
+
# for message in st.session_state.messages:
|
132 |
+
# with st.chat_message(message["role"]):
|
133 |
+
# st.markdown(message["content"])
|
134 |
|
135 |
|
136 |
|
137 |
+
# # Accept user input
|
138 |
+
# if prompt := st.chat_input(f"Hi I'm {selected_model}, ask me a question"):
|
139 |
|
140 |
+
# # Display user message in chat message container
|
141 |
+
# with st.chat_message("user"):
|
142 |
+
# st.markdown(prompt)
|
143 |
+
# # Add user message to chat history
|
144 |
+
# st.session_state.messages.append({"role": "user", "content": prompt})
|
145 |
|
146 |
|
147 |
+
# # Display assistant response in chat message container
|
148 |
+
# with st.chat_message("assistant"):
|
149 |
+
# stream = client.chat.completions.create(
|
150 |
+
# model=model_links[selected_model],
|
151 |
+
# messages=[
|
152 |
+
# {"role": m["role"], "content": m["content"]}
|
153 |
+
# for m in st.session_state.messages
|
154 |
+
# ],
|
155 |
+
# temperature=temp_values,#0.5,
|
156 |
+
# stream=True,
|
157 |
+
# max_tokens=3000,
|
158 |
+
# )
|
159 |
|
160 |
+
# response = st.write_stream(stream)
|
161 |
+
# st.session_state.messages.append({"role": "assistant", "content": response})
|
162 |
|
163 |
|
164 |
|
|
|
168 |
|
169 |
|
170 |
|
171 |
+
from huggingface_hub import InferenceClient
|
172 |
+
import gradio as gr
|
173 |
+
client = InferenceClient("mistralai/Mixtral-8x7B-Instruct-v0.1")
|
174 |
|
175 |
+
def format_prompt(message, history):
|
176 |
+
prompt = "<s>"
|
177 |
+
for user_prompt, bot_response in history:
|
178 |
+
prompt += f"[INST] {user_prompt} [/INST]"
|
179 |
+
prompt += f" {bot_response}</s> "
|
180 |
+
prompt += f"[INST] {message} [/INST]"
|
181 |
+
return prompt
|
182 |
|
183 |
+
def generate(
|
184 |
+
prompt, history, temperature=0.2, max_new_tokens=3000, top_p=0.95, repetition_penalty=1.0,
|
185 |
+
):
|
186 |
+
temperature = float(temperature)
|
187 |
+
if temperature < 1e-2:
|
188 |
+
temperature = 1e-2
|
189 |
+
top_p = float(top_p)
|
190 |
|
191 |
+
generate_kwargs = dict(
|
192 |
+
temperature=temperature,
|
193 |
+
max_new_tokens=max_new_tokens,
|
194 |
+
top_p=top_p,
|
195 |
+
repetition_penalty=repetition_penalty,
|
196 |
+
do_sample=True,
|
197 |
+
seed=42,
|
198 |
+
)
|
199 |
|
200 |
+
formatted_prompt = format_prompt(prompt, history)
|
201 |
|
202 |
+
stream = client.text_generation(formatted_prompt, **generate_kwargs, stream=True, details=True, return_full_text=False)
|
203 |
+
output = ""
|
204 |
|
205 |
+
for response in stream:
|
206 |
+
output += response.token.text
|
207 |
+
yield output
|
208 |
+
return output
|
209 |
|
210 |
|
211 |
+
mychatbot = gr.Chatbot(
|
212 |
+
avatar_images=["./user.png", "./bot.png"], bubble_full_width=False, show_label=False, show_copy_button=True, likeable=True,)
|
213 |
|
214 |
+
demo = gr.ChatInterface(fn=generate,
|
215 |
+
chatbot=mychatbot,
|
216 |
+
title="Mistral-Chat",
|
217 |
+
retry_btn=None,
|
218 |
+
undo_btn=None
|
219 |
+
)
|
220 |
|
221 |
+
demo.queue().launch(show_api=False)
|