Upload app.py
Browse files
app.py
ADDED
@@ -0,0 +1,126 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# Import necessary libraries
|
2 |
+
import streamlit as st
|
3 |
+
import pandas as pd
|
4 |
+
import transformers
|
5 |
+
import torch
|
6 |
+
from huggingface_hub import login
|
7 |
+
from dotenv import load_dotenv
|
8 |
+
import os
|
9 |
+
|
10 |
+
# Import data
|
11 |
+
music_data = pd.read_csv("Spotify_Youtube.csv")
|
12 |
+
|
13 |
+
# Login to HuggingFace Hub
|
14 |
+
load_dotenv()
|
15 |
+
HUGGINGFACE_API_KEY = os.environ.get("HUGGINGFACE_API_KEY")
|
16 |
+
login(HUGGINGFACE_API_KEY)
|
17 |
+
|
18 |
+
# Load Meta LLaMA model
|
19 |
+
model_id = "meta-llama/Meta-Llama-3-8B-Instruct"
|
20 |
+
pipeline = transformers.pipeline(
|
21 |
+
"text-generation",
|
22 |
+
model=model_id,
|
23 |
+
model_kwargs={"torch_dtype": torch.bfloat16},
|
24 |
+
device_map="auto"
|
25 |
+
)
|
26 |
+
|
27 |
+
# Function to parse user input using Meta LLaMA model
|
28 |
+
def parse_user_input(user_input):
|
29 |
+
messages = [
|
30 |
+
{
|
31 |
+
"role": "system",
|
32 |
+
"content": """You will be provided with an input: '{user_input}', and your task is to determine the following:
|
33 |
+
- Valence: a number that is equal to the mood. Positive moods are closer to 1 and negative moods are closer to 0.
|
34 |
+
- Number of songs: the number of songs the user requests.
|
35 |
+
- Tempo: the tempo of the songs.
|
36 |
+
- Danceability: the danceability of the songs.
|
37 |
+
|
38 |
+
Provide this information in the following format with each value separated by a space:
|
39 |
+
'valence number_of_songs tempo danceability'
|
40 |
+
Example: '0.5 20 120 0.8'
|
41 |
+
"""
|
42 |
+
},
|
43 |
+
{
|
44 |
+
"role": "user",
|
45 |
+
"content": user_input
|
46 |
+
},
|
47 |
+
]
|
48 |
+
|
49 |
+
prompt = pipeline.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
|
50 |
+
|
51 |
+
terminators = [
|
52 |
+
pipeline.tokenizer.eos_token_id,
|
53 |
+
pipeline.tokenizer.convert_tokens_to_ids("")
|
54 |
+
]
|
55 |
+
|
56 |
+
outputs = pipeline(
|
57 |
+
prompt,
|
58 |
+
max_new_tokens=256,
|
59 |
+
eos_token_id=terminators,
|
60 |
+
do_sample=True,
|
61 |
+
temperature=0.6,
|
62 |
+
top_p=0.9,
|
63 |
+
)
|
64 |
+
return outputs[0]["generated_text"][len(prompt):]
|
65 |
+
|
66 |
+
# Function to create a new dataframe from the music dataframe based on valence, number of tracks, tempo, and danceability
|
67 |
+
def get_tracks_by_artist_and_danceability(music_data, valence, num_tracks, tempo, danceability):
|
68 |
+
filtered_tracks = music_data[
|
69 |
+
(music_data['Valence'].between(valence - 0.1, valence + 0.1)) &
|
70 |
+
(music_data['Tempo'].between(tempo - 30, tempo + 30)) &
|
71 |
+
(music_data['Danceability'].between(danceability - 0.2, danceability + 0.2))
|
72 |
+
]
|
73 |
+
return filtered_tracks.head(num_tracks)[['Track', 'Artist']]
|
74 |
+
|
75 |
+
# Streamlit Application
|
76 |
+
logo = "music_logo.png"
|
77 |
+
|
78 |
+
# Sidebar
|
79 |
+
with st.sidebar:
|
80 |
+
st.image(logo, width=100)
|
81 |
+
st.header("Navigation")
|
82 |
+
tab_selection = st.sidebar.radio("Go to", ["Music Generator", "Browse Music", "About Us"])
|
83 |
+
|
84 |
+
# Music generator page
|
85 |
+
if tab_selection == "Music Generator":
|
86 |
+
st.header("Mood Playlist Generator")
|
87 |
+
st.write("Enter your music preferences in a detailed format and receive a personalized playlist based on your mood")
|
88 |
+
user_prompt = st.text_input("Example: 'I want 20 happy songs with high tempo that I can dance to!'")
|
89 |
+
|
90 |
+
if st.button("Generate Playlist"):
|
91 |
+
try:
|
92 |
+
with st.spinner("Processing your request..."):
|
93 |
+
parsed_input = parse_user_input(user_prompt)
|
94 |
+
# st.write(f"Parsed input: {parsed_input}")
|
95 |
+
|
96 |
+
# Extract parameters from the parsed input
|
97 |
+
valence, num_tracks, tempo, danceability = parsed_input.split()
|
98 |
+
valence = float(valence)
|
99 |
+
num_tracks = int(num_tracks)
|
100 |
+
tempo = int(tempo)
|
101 |
+
danceability = float(danceability)
|
102 |
+
|
103 |
+
# st.write(f"Number of tracks: {num_tracks}, Valence: {valence}, Tempo: {tempo}, Danceability: {danceability}")
|
104 |
+
|
105 |
+
tracks = get_tracks_by_artist_and_danceability(music_data, valence, num_tracks, tempo, danceability)
|
106 |
+
# st.write(f"Found {len(tracks)} tracks.")
|
107 |
+
|
108 |
+
if tracks.empty:
|
109 |
+
st.write("No tracks found. Please try a different query.")
|
110 |
+
else:
|
111 |
+
st.write("Here are your recommended playlist:")
|
112 |
+
st.table(tracks)
|
113 |
+
st.button("Add playlist to Spotify")
|
114 |
+
except ValueError:
|
115 |
+
st.write("Error: Unable to parse the input. Please make sure the format is correct.")
|
116 |
+
|
117 |
+
# Browse music page
|
118 |
+
elif tab_selection == "Browse Music":
|
119 |
+
st.header("Browse Music")
|
120 |
+
st.write("Explore the music data used for generating your playlists.")
|
121 |
+
df = pd.read_csv("Spotify_Youtube.csv")
|
122 |
+
st.dataframe(df)
|
123 |
+
|
124 |
+
# About us page
|
125 |
+
elif tab_selection == "About Us":
|
126 |
+
st.header("About Us")
|