straightline777 commited on
Commit
a7ca42b
·
verified ·
1 Parent(s): c16758a

Upload app.py

Browse files
Files changed (1) hide show
  1. app.py +13 -0
app.py CHANGED
@@ -6,6 +6,8 @@ from transformers import pipeline, AutoTokenizer, AutoModelForSequenceClassifica
6
  model_NER = AutoModelForTokenClassification.from_pretrained("./NER_model")
7
  tokenizer_NER = AutoTokenizer.from_pretrained("./tokenizer_NER")
8
 
 
 
9
  model_path = './model'
10
  tokenizer_path = './tokenizer'
11
 
@@ -18,14 +20,25 @@ clf_ner = pipeline("ner", model=model_NER, tokenizer=tokenizer_NER)
18
  config = AutoConfig.from_pretrained("./NER_model")
19
  label_list = config.id2label
20
 
 
 
 
 
 
 
 
 
21
  def analyze_sentiment(text):
 
22
  predictions = clf_sentiment(text)
23
  label_mapping = {'LABEL_0': 'Negative', 'LABEL_1': 'Neutral', 'LABEL_2': 'Positive'}
24
  predicted_label = label_mapping[predictions[0]['label']]
25
  confidence = f"{predictions[0]['score'] * 100:.2f}%"
26
  return f"Predicted Label: {predicted_label}\nConfidence: {confidence}"
27
 
 
28
  def analyze_ner(text):
 
29
  predictions = clf_ner(text)
30
  print(predictions) # Debugging statement to see the structure of the output
31
 
 
6
  model_NER = AutoModelForTokenClassification.from_pretrained("./NER_model")
7
  tokenizer_NER = AutoTokenizer.from_pretrained("./tokenizer_NER")
8
 
9
+
10
+
11
  model_path = './model'
12
  tokenizer_path = './tokenizer'
13
 
 
20
  config = AutoConfig.from_pretrained("./NER_model")
21
  label_list = config.id2label
22
 
23
+ def tokenize_NER(input_text):
24
+ tokenized_text = tokenizer_NER(input_text, padding='max_length', truncation=True, max_length=512)
25
+ return tokenized_text
26
+
27
+ def tokenize_sentiment(input_text):
28
+ tokenized_text = tokenizer_sentiment(input_text, padding='max_length', truncation=True, max_length=512)
29
+ return tokenized_text
30
+
31
  def analyze_sentiment(text):
32
+ text = tokenizer_sentiment(text)
33
  predictions = clf_sentiment(text)
34
  label_mapping = {'LABEL_0': 'Negative', 'LABEL_1': 'Neutral', 'LABEL_2': 'Positive'}
35
  predicted_label = label_mapping[predictions[0]['label']]
36
  confidence = f"{predictions[0]['score'] * 100:.2f}%"
37
  return f"Predicted Label: {predicted_label}\nConfidence: {confidence}"
38
 
39
+
40
  def analyze_ner(text):
41
+ text = tokenize_NER(text)
42
  predictions = clf_ner(text)
43
  print(predictions) # Debugging statement to see the structure of the output
44