File size: 13,740 Bytes
7eaaff0
 
7b9d8b2
7eaaff0
 
 
 
7b9d8b2
 
 
 
7eaaff0
 
 
 
 
7b9d8b2
7eaaff0
 
 
 
 
7b9d8b2
7eaaff0
7b9d8b2
7eaaff0
 
 
 
 
7b9d8b2
7eaaff0
 
 
 
 
 
 
 
 
 
 
 
7b9d8b2
 
 
 
 
 
 
 
 
 
 
7eaaff0
 
 
 
 
 
7b9d8b2
7eaaff0
 
 
 
7b9d8b2
7eaaff0
 
 
 
 
 
 
 
 
 
 
7b9d8b2
7eaaff0
 
7b9d8b2
7eaaff0
 
7b9d8b2
 
 
 
 
3fe982e
7b9d8b2
3fe982e
 
 
7b9d8b2
 
3fe982e
 
7b9d8b2
 
 
 
 
 
 
 
 
 
 
 
3fe982e
7b9d8b2
 
 
 
 
 
 
 
7eaaff0
7b9d8b2
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
7eaaff0
7b9d8b2
 
 
 
 
7eaaff0
 
7b9d8b2
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
7eaaff0
 
 
 
 
 
 
 
 
 
 
 
7b9d8b2
 
 
7eaaff0
 
 
 
 
 
 
 
 
 
 
 
 
7b9d8b2
7eaaff0
 
 
 
 
 
7b9d8b2
 
 
 
 
7eaaff0
 
 
 
 
 
 
 
 
 
 
 
 
 
7b9d8b2
 
 
7eaaff0
 
7b9d8b2
7eaaff0
7b9d8b2
7eaaff0
 
7b9d8b2
 
 
 
7eaaff0
 
 
7b9d8b2
7eaaff0
 
7b9d8b2
7eaaff0
7b9d8b2
7eaaff0
7b9d8b2
7eaaff0
7b9d8b2
7eaaff0
7b9d8b2
 
 
7eaaff0
7b9d8b2
 
 
 
7eaaff0
 
7b9d8b2
 
7eaaff0
 
 
 
 
 
 
 
 
 
 
 
 
7b9d8b2
 
7eaaff0
 
 
 
 
7b9d8b2
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
7eaaff0
7b9d8b2
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
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
import torch
import torch.nn.functional as F
from transformers import AutoTokenizer, AutoModelForSequenceClassification
import spacy
from typing import List, Dict
import logging
import os
from colorama import init, Fore, Back, Style

# Initialize colorama for colored terminal output
init()

# Configure logging
logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)

# Constants - matching original implementations
MAX_LENGTH = 512
MODEL_NAME = "microsoft/deberta-v3-small"
WINDOW_SIZE = 17
WINDOW_OVERLAP = 2
CONFIDENCE_THRESHOLD = 0.65
BATCH_SIZE = 16  # Matching original batch size

class TextProcessor:
    def __init__(self):
        try:
            self.nlp = spacy.load("en_core_web_sm")
        except OSError:
            logger.info("Downloading spacy model...")
            os.system("python -m spacy download en_core_web_sm")
            self.nlp = spacy.load("en_core_web_sm")

        if 'sentencizer' not in self.nlp.pipe_names:
            self.nlp.add_pipe('sentencizer')

        disabled_pipes = [pipe for pipe in self.nlp.pipe_names if pipe != 'sentencizer']
        self.nlp.disable_pipes(*disabled_pipes)

    def split_into_sentences(self, text: str) -> List[str]:
        doc = self.nlp(text)
        return [str(sent).strip() for sent in doc.sents]

    def create_windows(self, sentences: List[str], window_size: int, overlap: int) -> List[str]:
        if len(sentences) < window_size:
            return [" ".join(sentences)]

        windows = []
        stride = window_size - overlap
        for i in range(0, len(sentences) - window_size + 1, stride):
            window = sentences[i:i + window_size]
            windows.append(" ".join(window))
        return windows

    def create_centered_windows(self, sentences: List[str], window_size: int) -> tuple[List[str], List[List[int]]]:
        """Create windows centered around each sentence for detailed analysis."""
        windows = []
        window_sentence_indices = []

        for i in range(len(sentences)):
            # Calculate window boundaries centered on current sentence
            half_window = window_size // 2
            start_idx = max(0, i - half_window)
            end_idx = min(len(sentences), i + half_window + 1)

            # Adjust window if we're near the edges
            if start_idx == 0:
                end_idx = min(len(sentences), window_size)
            elif end_idx == len(sentences):
                start_idx = max(0, len(sentences) - window_size)

            window = sentences[start_idx:end_idx]
            windows.append(" ".join(window))
            window_sentence_indices.append(list(range(start_idx, end_idx)))

        return windows, window_sentence_indices

class AITextDetector:
    def __init__(self):
        self.device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
        self.processor = TextProcessor()
        self.tokenizer = None
        self.model = None
        self._initialize_model()

    def _initialize_model(self):
        """Initialize model and tokenizer."""
        self.tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
        self.model = AutoModelForSequenceClassification.from_pretrained(
            MODEL_NAME,
            num_labels=2
        ).to(self.device)
        
        try:
            model_path = "model_20250209_184929_acc1.0000.pt"
            checkpoint = torch.load(model_path, map_location=self.device)
            self.model.load_state_dict(checkpoint['model_state_dict'])
            logger.info(f"Loaded model from {model_path}")
        except Exception as e:
            logger.error(f"Failed to load model: {e}")
            raise

    def quick_scan(self, text: str) -> Dict:
        """
        Quick scan implementation matching the second original program's predict method.
        """
        if self.model is None or self.tokenizer is None:
            self._initialize_model()

        self.model.eval()
        sentences = self.processor.split_into_sentences(text)
        windows = self.processor.create_windows(sentences, WINDOW_SIZE, WINDOW_OVERLAP)

        predictions = []

        # Process windows in batches to save memory
        for i in range(0, len(windows), BATCH_SIZE):
            batch_windows = windows[i:i + BATCH_SIZE]

            inputs = self.tokenizer(
                batch_windows,
                truncation=True,
                padding=True,
                max_length=MAX_LENGTH,
                return_tensors="pt"
            ).to(self.device)

            with torch.no_grad():
                outputs = self.model(**inputs)
                probs = F.softmax(outputs.logits, dim=-1)

                for idx, window in enumerate(batch_windows):
                    prediction = {
                        'window': window,
                        'human_prob': probs[idx][1].item(),
                        'ai_prob': probs[idx][0].item(),
                        'prediction': 'human' if probs[idx][1] > probs[idx][0] else 'ai'
                    }
                    predictions.append(prediction)

            # Clear memory
            del inputs, outputs, probs
            if torch.cuda.is_available():
                torch.cuda.empty_cache()

        return self._aggregate_quick_predictions(predictions)

    def _aggregate_quick_predictions(self, predictions: List[Dict]) -> Dict:
        """
        Aggregate predictions matching the second original program.
        """
        if not predictions:
            return {
                'human_prob': 0.0,
                'ai_prob': 0.0,
                'prediction': 'unknown',
                'confidence': 0.0,
                'num_windows': 0
            }

        avg_human_prob = sum(p['human_prob'] for p in predictions) / len(predictions)
        avg_ai_prob = sum(p['ai_prob'] for p in predictions) / len(predictions)

        return {
            'human_prob': avg_human_prob,
            'ai_prob': avg_ai_prob,
            'prediction': 'human' if avg_human_prob > avg_ai_prob else 'ai',
            'confidence': max(avg_human_prob, avg_ai_prob),
            'num_windows': len(predictions)
        }

    def detailed_scan(self, text: str) -> Dict:
        """
        Detailed scan implementation matching the first original program's 
        predict_with_sentence_scores method.
        """
        if self.model is None or self.tokenizer is None:
            self._initialize_model()

        self.model.eval()
        sentences = self.processor.split_into_sentences(text)
        if not sentences:
            return {}

        # Create centered windows for each sentence
        windows, window_sentence_indices = self.processor.create_centered_windows(sentences, WINDOW_SIZE)

        # Track scores for each sentence
        sentence_appearances = {i: 0 for i in range(len(sentences))}
        sentence_scores = {i: {'human_prob': 0.0, 'ai_prob': 0.0} for i in range(len(sentences))}

        # Process windows in batches to save memory
        for i in range(0, len(windows), BATCH_SIZE):
            batch_windows = windows[i:i + BATCH_SIZE]
            batch_indices = window_sentence_indices[i:i + BATCH_SIZE]

            inputs = self.tokenizer(
                batch_windows,
                truncation=True,
                padding=True,
                max_length=MAX_LENGTH,
                return_tensors="pt"
            ).to(self.device)

            with torch.no_grad():
                outputs = self.model(**inputs)
                probs = F.softmax(outputs.logits, dim=-1)

                # Attribute window predictions back to individual sentences
                for window_idx, indices in enumerate(batch_indices):
                    for sent_idx in indices:
                        sentence_appearances[sent_idx] += 1
                        sentence_scores[sent_idx]['human_prob'] += probs[window_idx][1].item()
                        sentence_scores[sent_idx]['ai_prob'] += probs[window_idx][0].item()

            # Clear memory
            del inputs, outputs, probs
            if torch.cuda.is_available():
                torch.cuda.empty_cache()

        # Average the scores and create final sentence-level predictions
        sentence_predictions = []
        for i in range(len(sentences)):
            if sentence_appearances[i] > 0:
                human_prob = sentence_scores[i]['human_prob'] / sentence_appearances[i]
                ai_prob = sentence_scores[i]['ai_prob'] / sentence_appearances[i]
                sentence_predictions.append({
                    'sentence': sentences[i],
                    'human_prob': human_prob,
                    'ai_prob': ai_prob,
                    'prediction': 'human' if human_prob > ai_prob else 'ai',
                    'confidence': max(human_prob, ai_prob)
                })

        # Generate highlighted text output
        highlighted_text = self._generate_highlighted_text(sentence_predictions)

        return {
            'sentence_predictions': sentence_predictions,
            'highlighted_text': highlighted_text,
            'full_text': text,
            'overall_prediction': self._aggregate_detailed_predictions(sentence_predictions)
        }

    def _generate_highlighted_text(self, sentence_predictions: List[Dict]) -> str:
        """Generate colored text output with highlighting based on predictions."""
        highlighted_parts = []

        for pred in sentence_predictions:
            sentence = pred['sentence']
            confidence = pred['confidence']

            if confidence >= CONFIDENCE_THRESHOLD:
                if pred['prediction'] == 'human':
                    highlighted_parts.append(f"{Back.GREEN}{sentence}{Style.RESET_ALL}")
                else:
                    highlighted_parts.append(f"{Back.RED}{sentence}{Style.RESET_ALL}")
            else:
                # Low confidence predictions get a lighter highlight
                if pred['prediction'] == 'human':
                    highlighted_parts.append(f"{Back.LIGHTGREEN_EX}{sentence}{Style.RESET_ALL}")
                else:
                    highlighted_parts.append(f"{Back.LIGHTRED_EX}{sentence}{Style.RESET_ALL}")

        return " ".join(highlighted_parts)

    def _aggregate_detailed_predictions(self, predictions: List[Dict]) -> Dict:
        """
        Aggregate predictions matching the first original program.
        """
        if not predictions:
            return {
                'human_prob': 0.0,
                'ai_prob': 0.0,
                'prediction': 'unknown',
                'confidence': 0.0,
                'num_sentences': 0
            }

        total_human_prob = sum(p['human_prob'] for p in predictions)
        total_ai_prob = sum(p['ai_prob'] for p in predictions)
        num_sentences = len(predictions)

        avg_human_prob = total_human_prob / num_sentences
        avg_ai_prob = total_ai_prob / num_sentences

        return {
            'human_prob': avg_human_prob,
            'ai_prob': avg_ai_prob,
            'prediction': 'human' if avg_human_prob > avg_ai_prob else 'ai',
            'confidence': max(avg_human_prob, avg_ai_prob),
            'num_sentences': num_sentences
        }

def main():
    try:
        detector = AITextDetector()
        
        while True:
            print("\nAI Text Detector")
            print("===============")
            print("1. Quick Scan")
            print("2. Detailed Scan")
            print("3. Exit")
            
            choice = input("\nSelect an option (1-3): ").strip()
            
            if choice == "3":
                break
                
            if choice not in ["1", "2"]:
                print("Invalid choice. Please select 1, 2, or 3.")
                continue
                
            text = input("\nEnter text to analyze: ").strip()
            
            if choice == "1":
                # Quick scan
                result = detector.quick_scan(text)
                print("\nQuick Scan Results:")
                print("==================")
                print(f"Prediction: {result['prediction'].upper()}")
                print(f"Confidence: {result['confidence']*100:.1f}%")
                print(f"Human Probability: {result['human_prob']*100:.1f}%")
                print(f"AI Probability: {result['ai_prob']*100:.1f}%")
                print(f"Number of windows analyzed: {result['num_windows']}")
                
            else:
                # Detailed scan
                result = detector.detailed_scan(text)
                print("\nDetailed Analysis:")
                print("=================")
                
                # Print sentence-level predictions
                for pred in result['sentence_predictions']:
                    confidence = pred['confidence'] * 100
                    print(f"\nSentence: {pred['sentence']}")
                    print(f"Prediction: {pred['prediction'].upper()}")
                    print(f"Confidence: {confidence:.1f}%")
                
                # Print highlighted text
                print("\nHighlighted Text Analysis:")
                print("=========================")
                print(result['highlighted_text'])
                
                # Print final prediction
                final_pred = result['overall_prediction']
                print(f"\nFINAL PREDICTION: {final_pred['prediction'].upper()}")
                print(f"Overall confidence: {final_pred['confidence']*100:.1f}%")
                print(f"Number of sentences analyzed: {final_pred['num_sentences']}")

    except Exception as e:
        logger.error(f"An error occurred: {e}")
        raise

if __name__ == "__main__":
    main()