Spaces:
Sleeping
Sleeping
File size: 8,285 Bytes
c551c64 b06572a c551c64 b06572a c551c64 b06572a c551c64 b06572a c551c64 b06572a c551c64 b06572a c551c64 b06572a c551c64 b06572a c551c64 b06572a c551c64 b06572a c551c64 b06572a c551c64 b06572a c551c64 b06572a 91956a4 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 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 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 | # import tensorflow as tf
# import numpy as np
# import os
# import pandas as pd
# from flask import Flask, request, jsonify
# import joblib
# import sklearn
# import re
# from helpers.helper_functions import remove_html_tags, remove_url, remove_digits, remove_punc, lower
# from models import Model
# from werkzeug.exceptions import RequestEntityTooLarge
# # Configure TensorFlow logging
# os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'
# os.environ['TF_ENABLE_ONEDNN_OPTS'] = '0'
# app = Flask(__name__)
# app.config['MAX_CONTENT_LENGTH'] = 16 * 1024 * 1024 * 1024 # 16 GB max-limit
# # Initialize model
# model = Model()
# # Define label mappings
# SENTIMENT_LABELS = ['negative', 'neutral', 'positive']
# EMOTION_LABELS = ['sadness', 'joy', 'love', 'anger', 'fear', 'surprise']
# @app.route('/')
# def index():
# return 'Model deployed successfully'
# @app.route('/predict', methods=['POST'])
# def predict():
# try:
# # Validate request
# data = request.get_json()
# if not data:
# return jsonify({'error': 'No input provided'}), 400
# # Handle array of objects
# if isinstance(data, list):
# texts = []
# tweet_ids = []
# for item in data:
# if not isinstance(item, dict) or 'text' not in item or 'id' not in item or not isinstance(item['text'], str):
# return jsonify({'error': f'Invalid object format: {item}. Each object must have "id" and "text" keys with a string value for "text"'}), 400
# texts.append(item['text'])
# tweet_ids.append(item['id'])
# else:
# return jsonify({'error': 'Input must be an array of objects'}), 400
# results = []
# for idx, text in enumerate(texts):
# # Preprocess text for sentiment
# text_sentiment = text
# text_sentiment = remove_html_tags(text_sentiment)
# text_sentiment = lower(text_sentiment)
# # Preprocess text for emotion
# text_emotion = text
# text_emotion = remove_html_tags(text_emotion)
# text_emotion = remove_url(text_emotion)
# text_emotion = remove_digits(text_emotion)
# text_emotion = remove_punc(text_emotion)
# # Convert to tensors and get predictions
# sentiment_tensor = tf.convert_to_tensor([text_sentiment])
# emotion_tensor = tf.convert_to_tensor([text_emotion])
# print(f"Sentiment tensor shape: {sentiment_tensor.shape}")
# print(f"Emotion tensor shape: {emotion_tensor.shape}")
# sentiment_pred, emotion_pred = model.predict(
# text_sentiment=sentiment_tensor,
# text_emotion=emotion_tensor
# )
# print(f"Sentiment prediction shape: {sentiment_pred.shape}")
# print(f"Emotion prediction shape: {emotion_pred.shape}")
# print(f"Sentiment prediction: {sentiment_pred}")
# print(f"Emotion prediction: {emotion_pred}")
# # Process predictions
# sentiment_idx = int(np.argmax(sentiment_pred, axis=1)[0])
# emotion_idx = int(np.argmax(emotion_pred, axis=1)[0])
# print(f"Sentiment index: {sentiment_idx}")
# print(f"Emotion index: {emotion_idx}")
# # Create result object including tweet_id
# result = {
# 'id': tweet_ids[idx], # Add the tweet_id
# 'text': text,
# 'sentiment': {
# 'label': SENTIMENT_LABELS[sentiment_idx],
# 'score': float(sentiment_pred[0][sentiment_idx]),
# 'raw_scores': [float(score) for score in sentiment_pred[0]]
# },
# 'emotion': {
# 'label': EMOTION_LABELS[emotion_idx],
# 'score': float(emotion_pred[0][emotion_idx]),
# 'raw_scores': [float(score) for score in emotion_pred[0]]
# }
# }
# results.append(result)
# return jsonify({'results': results})
# except RequestEntityTooLarge:
# return jsonify({'error': 'Request Entity Too Large'}), 413
# except IndexError as e:
# return jsonify({'error': f'IndexError: {str(e)}'}), 500
# except Exception as e:
# return jsonify({'error': f'Unexpected error: {str(e)}'}), 400
# if __name__ == '__main__':
# app.run(host='0.0.0.0' ,port=7860, debug=False)
import tensorflow as tf
import numpy as np
import os
import pandas as pd
from flask import Flask, request, jsonify
from helpers.helper_functions import remove_html_tags, remove_url, remove_digits, remove_punc, lower
from models import Model
from werkzeug.exceptions import RequestEntityTooLarge
# Configure TensorFlow logging and GPU
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2' # Suppress warnings
os.environ['TF_ENABLE_ONEDNN_OPTS'] = '0'
physical_devices = tf.config.list_physical_devices('GPU')
if physical_devices:
tf.config.experimental.set_memory_growth(physical_devices[0], True)
app = Flask(__name__)
app.config['MAX_CONTENT_LENGTH'] = 16 * 1024 * 1024 * 1024 # 16 GB max-limit
# Initialize model globally
model = Model()
# Define label mappings
SENTIMENT_LABELS = ['negative', 'neutral', 'positive']
EMOTION_LABELS = ['sadness', 'joy', 'love', 'anger', 'fear', 'surprise']
# Vectorized preprocessing function
def preprocess_batch(texts, for_emotion=False):
df = pd.Series(texts)
df = df.apply(remove_html_tags).str.lower()
if for_emotion:
df = df.apply(remove_url).apply(remove_digits).apply(remove_punc)
return df.tolist()
@app.route('/')
def index():
return 'Model deployed successfully'
@app.route('/predict', methods=['POST'])
def predict():
try:
# Validate request
data = request.get_json()
if not data or not isinstance(data, list):
return jsonify({'error': 'Input must be an array of objects'}), 400
# Extract texts and IDs
texts = []
tweet_ids = []
for item in data:
if not isinstance(item, dict) or 'text' not in item or 'id' not in item or not isinstance(item['text'], str):
return jsonify({'error': f'Invalid object format: {item}'}), 400
texts.append(item['text'])
tweet_ids.append(item['id'])
# Batch preprocessing
texts_sentiment = preprocess_batch(texts, for_emotion=False)
texts_emotion = preprocess_batch(texts, for_emotion=True)
# Convert to tensors and ensure tf.string dtype
sentiment_tensor = tf.convert_to_tensor(texts_sentiment, dtype=tf.string)
emotion_tensor = tf.convert_to_tensor(texts_emotion, dtype=tf.string)
# Batch inference
sentiment_preds, emotion_preds = model.predict(
text_sentiment=sentiment_tensor,
text_emotion=emotion_tensor
)
# Process predictions
results = []
for idx, (sentiment_pred, emotion_pred) in enumerate(zip(sentiment_preds, emotion_preds)):
sentiment_idx = int(np.argmax(sentiment_pred))
emotion_idx = int(np.argmax(emotion_pred))
result = {
'id': tweet_ids[idx],
'text': texts[idx],
'sentiment': {
'label': SENTIMENT_LABELS[sentiment_idx],
'score': float(sentiment_pred[sentiment_idx]),
'raw_scores': [float(score) for score in sentiment_pred]
},
'emotion': {
'label': EMOTION_LABELS[emotion_idx],
'score': float(emotion_pred[emotion_idx]),
'raw_scores': [float(score) for score in emotion_pred]
}
}
results.append(result)
return jsonify({'results': results})
except RequestEntityTooLarge:
return jsonify({'error': 'Request Entity Too Large'}), 413
except Exception as e:
return jsonify({'error': f'Unexpected error: {str(e)}'}), 500
if __name__ == '__main__':
app.run(host='0.0.0.0' ,port=7860, debug=False) |