Spaces:
Runtime error
Runtime error
Update app.py
Browse files
app.py
CHANGED
@@ -4,6 +4,7 @@ from tensorflow.keras.preprocessing.sequence import pad_sequences
|
|
4 |
import joblib
|
5 |
import pandas as pd
|
6 |
import numpy as np
|
|
|
7 |
|
8 |
# Check if scikit-learn is installed
|
9 |
try:
|
@@ -19,15 +20,14 @@ emotion_model = load_model('lstm_model.h5')
|
|
19 |
try:
|
20 |
recommender_model = joblib.load('knn_model.pkl')
|
21 |
except Exception as e:
|
22 |
-
st.error(f"Error loading model: {e}")
|
23 |
|
24 |
# Load the tokenizer (ensure it's the one used during training)
|
25 |
tokenizer = joblib.load('tokenizer.pkl') # Correct the path
|
26 |
|
27 |
# Load the dataset and preprocess
|
28 |
df = pd.read_csv('df1.csv')
|
29 |
-
|
30 |
-
df = df.drop(['Unnamed: 0', 'lyrics_filename', 'analysis_url', 'track_href', "type", "id", "uri"], axis=1)
|
31 |
|
32 |
# Set up the title of the app
|
33 |
st.title('Emotion and Audio Feature-based Song Recommendation System')
|
@@ -39,26 +39,34 @@ lyrics = st.text_area("Input the lyrics of the song here:")
|
|
39 |
# Input fields for audio features
|
40 |
st.header('Enter Audio Features')
|
41 |
audio_features = []
|
42 |
-
|
43 |
-
|
|
|
|
|
|
|
|
|
44 |
feature = st.number_input(f"Enter value for {feature_name}:", step=0.01)
|
45 |
audio_features.append(feature)
|
46 |
|
47 |
# Predict and Recommend button
|
48 |
if st.button('Predict Emotion and Recommend Songs'):
|
49 |
if lyrics and all(audio_features):
|
|
|
50 |
sequence = tokenizer.texts_to_sequences([lyrics])
|
51 |
padded_sequence = pad_sequences(sequence, maxlen=128)
|
52 |
emotion = emotion_model.predict(padded_sequence).flatten()
|
53 |
|
|
|
54 |
combined_features = np.concatenate([emotion, audio_features])
|
55 |
|
56 |
-
|
57 |
-
|
|
|
58 |
|
59 |
st.write("Emotion Detected:", emotion[0])
|
60 |
-
st.header('Recommended Songs')
|
61 |
-
for _, song in
|
62 |
st.write(song)
|
|
|
63 |
else:
|
64 |
st.error("Please fill in all the fields.")
|
|
|
4 |
import joblib
|
5 |
import pandas as pd
|
6 |
import numpy as np
|
7 |
+
from sklearn.metrics.pairwise import cosine_similarity
|
8 |
|
9 |
# Check if scikit-learn is installed
|
10 |
try:
|
|
|
20 |
try:
|
21 |
recommender_model = joblib.load('knn_model.pkl')
|
22 |
except Exception as e:
|
23 |
+
st.error(f"Error loading KNN model: {e}")
|
24 |
|
25 |
# Load the tokenizer (ensure it's the one used during training)
|
26 |
tokenizer = joblib.load('tokenizer.pkl') # Correct the path
|
27 |
|
28 |
# Load the dataset and preprocess
|
29 |
df = pd.read_csv('df1.csv')
|
30 |
+
df = df.drop(['Unnamed: 0', 'lyrics_filename', 'analysis_url', 'track_href', "type", "id", "uri"], axis=1)
|
|
|
31 |
|
32 |
# Set up the title of the app
|
33 |
st.title('Emotion and Audio Feature-based Song Recommendation System')
|
|
|
39 |
# Input fields for audio features
|
40 |
st.header('Enter Audio Features')
|
41 |
audio_features = []
|
42 |
+
|
43 |
+
# Display only relevant columns for audio features
|
44 |
+
audio_feature_columns = ['danceability', 'energy', 'key', 'loudness', 'mode', 'speechiness', 'acousticness',
|
45 |
+
'instrumentalness', 'liveness', 'valence', 'tempo']
|
46 |
+
|
47 |
+
for feature_name in audio_feature_columns:
|
48 |
feature = st.number_input(f"Enter value for {feature_name}:", step=0.01)
|
49 |
audio_features.append(feature)
|
50 |
|
51 |
# Predict and Recommend button
|
52 |
if st.button('Predict Emotion and Recommend Songs'):
|
53 |
if lyrics and all(audio_features):
|
54 |
+
# Process the lyrics
|
55 |
sequence = tokenizer.texts_to_sequences([lyrics])
|
56 |
padded_sequence = pad_sequences(sequence, maxlen=128)
|
57 |
emotion = emotion_model.predict(padded_sequence).flatten()
|
58 |
|
59 |
+
# Combine emotion and audio features for recommendation
|
60 |
combined_features = np.concatenate([emotion, audio_features])
|
61 |
|
62 |
+
# Generate recommendations using the KNN model
|
63 |
+
knn_distances, knn_indices = recommender_model.kneighbors([combined_features], n_neighbors=5)
|
64 |
+
knn_recommended_songs = df.iloc[knn_indices.flatten()]
|
65 |
|
66 |
st.write("Emotion Detected:", emotion[0])
|
67 |
+
st.header('Recommended Songs (KNN)')
|
68 |
+
for _, song in knn_recommended_songs.iterrows():
|
69 |
st.write(song)
|
70 |
+
|
71 |
else:
|
72 |
st.error("Please fill in all the fields.")
|