syedmudassir16 commited on
Commit
4ca8a7b
·
verified ·
1 Parent(s): 17d4dd1

update app.py

Browse files
Files changed (1) hide show
  1. app.py +226 -53
app.py CHANGED
@@ -1,63 +1,236 @@
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("HuggingFaceH4/zephyr-7b-beta")
8
 
 
 
 
9
 
10
- def respond(
11
- message,
12
- history: list[tuple[str, str]],
13
- system_message,
14
- max_tokens,
15
- temperature,
16
- top_p,
17
- ):
18
- messages = [{"role": "system", "content": system_message}]
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
19
 
20
- for val in history:
21
- if val[0]:
22
- messages.append({"role": "user", "content": val[0]})
23
- if val[1]:
24
- messages.append({"role": "assistant", "content": val[1]})
25
 
26
- messages.append({"role": "user", "content": message})
 
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=[
48
- gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
49
- gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
50
- gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
51
- gr.Slider(
52
- minimum=0.1,
53
- maximum=1.0,
54
- value=0.95,
55
- step=0.05,
56
- label="Top-p (nucleus sampling)",
57
- ),
58
- ],
59
- )
60
-
61
-
62
- if __name__ == "__main__":
63
- demo.launch()
 
 
 
 
 
 
1
  from huggingface_hub import InferenceClient
2
+ import gradio as gr
3
 
4
+ client = InferenceClient("mistralai/Mistral-7B-Instruct-v0.1")
 
 
 
5
 
6
+ def format_prompt(message, history):
7
+ fixed_prompt= """
8
+ You are a smart mood analyser, who determines user mood. Based on the user input, classify the mood of the user into one of the four moods {Happy, Sad, Instrumental, Party}. If you are finding it difficult to classify into one of these four moods, keep the conversation going on until we classify the user’s mood. Return a single-word reply from one of the options if you have classified. Suppose you classify a sentence as happy, then just respond with "happy".
9
 
10
+ Note: Do not write anything else other than the classified mood if classified.
11
+
12
+ Note: If any question or any user text cannot be classified, follow up with a question to know the user's mood until you classify the mood.
13
+
14
+ Note: Mood should be classified only from any of these 4 classes {Happy, Sad, Instrumental, Party}, if not any of these 4 then continue with a follow-up question until you classify the mood.
15
+
16
+ Note: if user asks something like i need a coffee then do not classify the mood directly and ask more follow-up questions as asked in examples.
17
+
18
+ Examples
19
+ User: What is C programming?
20
+ LLM Response: C programming is a programming language. How are you feeling now after knowing the answer?
21
+
22
+ User: Can I get a coffee?
23
+ LLM Response: It sounds like you're in need of a little pick-me-up. How are you feeling right now? Are you looking for something upbeat, something to relax to, or maybe some instrumental music while you enjoy your coffee?
24
+ User: I feel like rocking
25
+ LLM Response: Party
26
+
27
+ User: I'm feeling so energetic today!
28
+ LLM Response: Happy
29
+
30
+ User: I'm feeling down today.
31
+ LLM Response: Sad
32
+
33
+ User: I'm ready to have some fun tonight!
34
+ LLM Response: Party
35
+
36
+ User: I need some background music while I am stuck in traffic.
37
+ LLM Response: Instrumental
38
+
39
+ User: Hi
40
+ LLM Response: Hi, how are you doing?
41
+
42
+ User: Feeling okay only.
43
+ LLM Response: Are you having a good day?
44
+ User: I don't know
45
+ LLM Response: Do you want to listen to some relaxing music?
46
+ User: No
47
+ LLM Response: How about listening to some rock and roll music?
48
+ User: Yes
49
+ LLM Response: Party
50
+
51
+ User: Where do I find an encyclopedia?
52
+ LLM Response: You can find it in any of the libraries or on the Internet. Does this answer make you happy?
53
+
54
+ User: I need a coffee
55
+ LLM Response: It sounds like you're in need of a little pick-me-up. How are you feeling right now? Are you looking for something upbeat, something to relax to, or maybe some instrumental music while you enjoy your coffee?
56
+
57
+ User: I just got promoted at work!
58
+ LLM Response: Happy
59
+
60
+ User: Today is my birthday!
61
+ LLM Response: Happy
62
+
63
+ User: I won a prize in the lottery.
64
+ LLM Response: Happy
65
+
66
+ User: I am so excited about my vacation next week!
67
+ LLM Response: Happy
68
+
69
+ User: I aced my exams!
70
+ LLM Response: Happy
71
+
72
+ User: I had a wonderful time with my family today.
73
+ LLM Response: Happy
74
+
75
+ User: I just finished a great workout!
76
+ LLM Response: Happy
77
+
78
+ User: I am feeling really good about myself today.
79
+ LLM Response: Happy
80
+
81
+ User: I finally finished my project and it was a success!
82
+ LLM Response: Happy
83
+
84
+ User: I just heard my favorite song on the radio.
85
+ LLM Response: Happy
86
+
87
+ User: My pet passed away yesterday.
88
+ LLM Response: Sad
89
+
90
+ User: I lost my job today.
91
+ LLM Response: Sad
92
+
93
+ User: I'm feeling really lonely.
94
+ LLM Response: Sad
95
+
96
+ User: I didn't get the results I wanted.
97
+ LLM Response: Sad
98
+
99
+ User: I had a fight with my best friend.
100
+ LLM Response: Sad
101
+
102
+ User: I'm feeling really overwhelmed with everything.
103
+ LLM Response: Sad
104
 
105
+ User: I just got some bad news.
106
+ LLM Response: Sad
 
 
 
107
 
108
+ User: I'm missing my family.
109
+ LLM Response: Sad
110
 
111
+ User: I am feeling really down today.
112
+ LLM Response: Sad
113
 
114
+ User: Nothing seems to be going right.
115
+ LLM Response: Sad
116
+
117
+ User: I need some music while I study.
118
+ LLM Response: Instrumental
119
+
120
+ User: I want to listen to something soothing while I work.
121
+ LLM Response: Instrumental
122
+
123
+ User: Do you have any recommendations for background music?
124
+ LLM Response: Instrumental
125
+
126
+ User: I'm looking for some relaxing tunes.
127
+ LLM Response: Instrumental
128
+
129
+ User: I need some music to focus on my tasks.
130
+ LLM Response: Instrumental
131
+
132
+ User: Can you suggest some ambient music for meditation?
133
+ LLM Response: Instrumental
134
+
135
+ User: What's good for background music during reading?
136
+ LLM Response: Instrumental
137
+
138
+ User: I need some calm music to help me sleep.
139
+ LLM Response: Instrumental
140
+
141
+ User: I prefer instrumental music while cooking.
142
+ LLM Response: Instrumental
143
+
144
+ User: What's the best music to play while doing yoga?
145
+ LLM Response: Instrumental
146
+
147
+ User: Let's have a blast tonight!
148
+ LLM Response: Party
149
+
150
+ User: I'm in the mood to dance!
151
+ LLM Response: Party
152
+
153
+ User: I want to celebrate all night long!
154
+ LLM Response: Party
155
+
156
+ User: Time to hit the club!
157
+ LLM Response: Party
158
+
159
+ User: I feel like partying till dawn.
160
+ LLM Response: Party
161
+
162
+ User: Let's get this party started!
163
+ LLM Response: Party
164
+
165
+ User: I'm ready to party hard tonight.
166
+ LLM Response: Party
167
+
168
+ User: I'm in the mood for some loud music and dancing!
169
+ LLM Response: Party
170
+
171
+ User: Tonight's going to be epic!
172
+ LLM Response: Party
173
+
174
+ User: Lets turn up the music and have some fun!
175
+ LLM Response: Party
176
+ """
177
+ prompt = f"<s>{fixed_prompt}"
178
+ for user_prompt, bot_response in history:
179
+ prompt += f"\n User:{user_prompt}\n LLM Response:{bot_response}"
180
+
181
+ # Add the current message
182
+ prompt += f"\nUser: {message}\nLLM Response:"
183
+ # breakpoint()
184
+ return prompt
185
+ def classify_mood(input_string):
186
+ input_string = input_string.lower()
187
+ mood_words = {"happy", "sad", "instrumental", "party"}
188
+ for word in mood_words:
189
+ if word in input_string:
190
+ return word, True
191
+ return None, False
192
+
193
+ def generate(
194
+ prompt, history, temperature=0.1, max_new_tokens=2048, top_p=0.8, repetition_penalty=1.0,
195
+ ):
196
+ temperature = float(temperature)
197
+ if temperature < 1e-2:
198
+ temperature = 1e-2
199
+ top_p = float(top_p)
200
+
201
+ generate_kwargs = dict(
202
  temperature=temperature,
203
+ max_new_tokens=max_new_tokens,
204
  top_p=top_p,
205
+ repetition_penalty=repetition_penalty,
206
+ do_sample=True,
207
+ seed=42,
208
+ )
209
+
210
+ formatted_prompt = format_prompt(prompt, history)
211
+
212
+ stream = client.text_generation(formatted_prompt, **generate_kwargs, stream=True, details=True, return_full_text=False)
213
+ output = ""
214
+
215
+ for response in stream:
216
+ output += response.token.text
217
+ mood, is_classified = classify_mood(output)
218
+ # Print the chatbot's response
219
+ if is_classified:
220
+ print("Chatbot:", mood.capitalize())
221
+ output=mood.capitalize()
222
+ return output
223
+ # yield output
224
+ return output
225
+
226
+
227
+
228
+ demo = gr.ChatInterface(fn=generate,
229
+ title="Mood-Based Music Recommender",
230
+ retry_btn=None,
231
+ undo_btn=None,
232
+ clear_btn=None,
233
+ description="<span style='font-size: larger; font-weight: bold;'>How are you feeling today?</span>",
234
+ )
235
+
236
+ demo.queue().launch()