Update app.py
Browse files
app.py
CHANGED
@@ -215,33 +215,38 @@ demo = gr.Interface(
|
|
215 |
demo.launch()
|
216 |
'''
|
217 |
import gradio as gr
|
218 |
-
from transformers import TFBertForSequenceClassification, BertTokenizer
|
219 |
import tensorflow as tf
|
220 |
import praw
|
221 |
import os
|
222 |
|
223 |
-
#
|
|
|
|
|
|
|
|
|
|
|
224 |
model = TFBertForSequenceClassification.from_pretrained("shrish191/sentiment-bert")
|
225 |
tokenizer = BertTokenizer.from_pretrained("shrish191/sentiment-bert")
|
226 |
|
227 |
-
# Load fallback sentiment pipeline model
|
228 |
-
fallback_classifier = pipeline("text-classification", model="VinMir/GordonAI-sentiment_analysis")
|
229 |
-
|
230 |
-
# Label mapping for main model
|
231 |
LABELS = {
|
232 |
0: "Neutral",
|
233 |
1: "Positive",
|
234 |
2: "Negative"
|
235 |
}
|
236 |
|
237 |
-
#
|
|
|
|
|
|
|
|
|
|
|
238 |
reddit = praw.Reddit(
|
239 |
client_id=os.getenv("REDDIT_CLIENT_ID"),
|
240 |
client_secret=os.getenv("REDDIT_CLIENT_SECRET"),
|
241 |
user_agent=os.getenv("REDDIT_USER_AGENT", "sentiment-classifier-script")
|
242 |
)
|
243 |
|
244 |
-
# Fetch content from Reddit URL
|
245 |
def fetch_reddit_text(reddit_url):
|
246 |
try:
|
247 |
submission = reddit.submission(url=reddit_url)
|
@@ -249,7 +254,15 @@ def fetch_reddit_text(reddit_url):
|
|
249 |
except Exception as e:
|
250 |
return f"Error fetching Reddit post: {str(e)}"
|
251 |
|
252 |
-
#
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
253 |
def classify_sentiment(text_input, reddit_url):
|
254 |
if reddit_url.strip():
|
255 |
text = fetch_reddit_text(reddit_url)
|
@@ -262,7 +275,6 @@ def classify_sentiment(text_input, reddit_url):
|
|
262 |
return f"[!] {text}"
|
263 |
|
264 |
try:
|
265 |
-
# Main BERT model prediction
|
266 |
inputs = tokenizer(text, return_tensors="tf", truncation=True, padding=True)
|
267 |
outputs = model(inputs)
|
268 |
probs = tf.nn.softmax(outputs.logits, axis=1)
|
@@ -270,9 +282,7 @@ def classify_sentiment(text_input, reddit_url):
|
|
270 |
pred_label = tf.argmax(probs, axis=1).numpy()[0]
|
271 |
|
272 |
if confidence < 0.5:
|
273 |
-
|
274 |
-
fallback = fallback_classifier(text)[0]['label']
|
275 |
-
return f"Prediction: {fallback}"
|
276 |
|
277 |
return f"Prediction: {LABELS[pred_label]}"
|
278 |
except Exception as e:
|
@@ -301,3 +311,6 @@ demo = gr.Interface(
|
|
301 |
demo.launch()
|
302 |
|
303 |
|
|
|
|
|
|
|
|
215 |
demo.launch()
|
216 |
'''
|
217 |
import gradio as gr
|
218 |
+
from transformers import TFBertForSequenceClassification, BertTokenizer
|
219 |
import tensorflow as tf
|
220 |
import praw
|
221 |
import os
|
222 |
|
223 |
+
# Fallback imports
|
224 |
+
from transformers import AutoTokenizer, AutoModelForSequenceClassification
|
225 |
+
import torch
|
226 |
+
from scipy.special import softmax
|
227 |
+
|
228 |
+
# Load main model and tokenizer
|
229 |
model = TFBertForSequenceClassification.from_pretrained("shrish191/sentiment-bert")
|
230 |
tokenizer = BertTokenizer.from_pretrained("shrish191/sentiment-bert")
|
231 |
|
|
|
|
|
|
|
|
|
232 |
LABELS = {
|
233 |
0: "Neutral",
|
234 |
1: "Positive",
|
235 |
2: "Negative"
|
236 |
}
|
237 |
|
238 |
+
# Load fallback model and tokenizer
|
239 |
+
fallback_model_name = "cardiffnlp/twitter-roberta-base-sentiment"
|
240 |
+
fallback_tokenizer = AutoTokenizer.from_pretrained(fallback_model_name)
|
241 |
+
fallback_model = AutoModelForSequenceClassification.from_pretrained(fallback_model_name)
|
242 |
+
|
243 |
+
# Reddit API
|
244 |
reddit = praw.Reddit(
|
245 |
client_id=os.getenv("REDDIT_CLIENT_ID"),
|
246 |
client_secret=os.getenv("REDDIT_CLIENT_SECRET"),
|
247 |
user_agent=os.getenv("REDDIT_USER_AGENT", "sentiment-classifier-script")
|
248 |
)
|
249 |
|
|
|
250 |
def fetch_reddit_text(reddit_url):
|
251 |
try:
|
252 |
submission = reddit.submission(url=reddit_url)
|
|
|
254 |
except Exception as e:
|
255 |
return f"Error fetching Reddit post: {str(e)}"
|
256 |
|
257 |
+
# Fallback classifier using RoBERTa
|
258 |
+
def fallback_classifier(text):
|
259 |
+
encoded_input = fallback_tokenizer(text, return_tensors='pt', truncation=True, padding=True)
|
260 |
+
with torch.no_grad():
|
261 |
+
output = fallback_model(**encoded_input)
|
262 |
+
scores = softmax(output.logits.numpy()[0])
|
263 |
+
labels = ['Negative', 'Neutral', 'Positive']
|
264 |
+
return f"Prediction: {labels[scores.argmax()]}"
|
265 |
+
|
266 |
def classify_sentiment(text_input, reddit_url):
|
267 |
if reddit_url.strip():
|
268 |
text = fetch_reddit_text(reddit_url)
|
|
|
275 |
return f"[!] {text}"
|
276 |
|
277 |
try:
|
|
|
278 |
inputs = tokenizer(text, return_tensors="tf", truncation=True, padding=True)
|
279 |
outputs = model(inputs)
|
280 |
probs = tf.nn.softmax(outputs.logits, axis=1)
|
|
|
282 |
pred_label = tf.argmax(probs, axis=1).numpy()[0]
|
283 |
|
284 |
if confidence < 0.5:
|
285 |
+
return fallback_classifier(text)
|
|
|
|
|
286 |
|
287 |
return f"Prediction: {LABELS[pred_label]}"
|
288 |
except Exception as e:
|
|
|
311 |
demo.launch()
|
312 |
|
313 |
|
314 |
+
|
315 |
+
|
316 |
+
|