Update tasks/text.py

tasks/text.py

@@ -71,69 +71,23 @@ async def evaluate_text(request: TextEvaluationRequest):
     MAX_LENGTH = 365
 
     model = AutoModelForSequenceClassification.from_pretrained(MODEL_REPO)
-    #model.eval()  # Set to evaluation mode
+    model.eval()  # Set to evaluation mode
 
-    class QuotesDataset(Dataset):
-        def __init__(self, encodings, labels):
-            self.encodings = encodings
-            self.labels = labels
-    
-        def __getitem__(self, idx):
-            item = {key: torch.tensor(val[idx]) for key, val in self.encodings.items()}
-            item['labels'] = torch.tensor(self.labels[idx], dtype=torch.long)
-            return item
-    
-        def __len__(self):
-            return len(self.labels)
-
-    def encode_data(tokenizer, texts, labels, max_length):
-        try:
-            if isinstance(texts, pd.Series):
-                texts = texts.tolist()
-            if isinstance(labels, pd.Series):
-                labels = labels.tolist()
-                
-            encodings = tokenizer(texts, truncation=True, padding='max_length', max_length=max_length, return_tensors='pt')
-            return QuotesDataset(encodings, labels)
-    
-        except Exception as e:
-            print(f"Error during tokenization: {e}")
-            return None
-            
-    val_dataset = encode_data(tokenizer, test_dataset['quote'], test_dataset['label'], MAX_LENGTH)        
-    val_loader = DataLoader(val_dataset, batch_size= 16, shuffle=False) 
-  
 
-    device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
-    print(f"Using device: {device}")
-    
-    def validate_model(model, val_loader, device):
-        model.eval()
-        predictions = []
-        with torch.no_grad():
-            for batch in val_loader:
-                batch = {k: v.to(device) for k, v in batch.items()}
-                outputs = model(**batch)
-                preds = torch.argmax(outputs.logits, dim=-1)
-                predictions.extend(preds.cpu().numpy())
-        return predictions
-
-    
     # tokenize texts
-    #test_encodings = tokenizer(test_dataset["quote"], padding='max_length', truncation=True, max_length=MAX_LENGTH, return_tensors="pt")
-    #test_labels = torch.tensor(test_dataset["label"])
+    test_encodings = tokenizer(test_dataset["quote"], padding='max_length', truncation=True, max_length=MAX_LENGTH, return_tensors="pt")
 
-    #test_dataset = TensorDataset(test_encodings["input_ids"], test_encodings["attention_mask"], test_labels)
-    #test_loader = DataLoader(test_dataset, batch_size=16)
+    test_dataset = TensorDataset(test_encodings["input_ids"], test_encodings["attention_mask"], test_labels)
+    test_loader = DataLoader(test_dataset, batch_size=16)
     
-    #predictions = []
-    #with torch.no_grad():
-    #for batch in test_loader:
-    #    input_ids, attention_mask, labels = [x.to(device) for x in batch]
-    #    outputs = model(input_ids, attention_mask=attention_mask)
-    #   predictions = torch.argmax(outputs.logits, dim=1)
-
-    predictions = validate_model(model, val_loader, device)
+    predictions = []
+    with torch.no_grad():
+    for batch in test_loader:
+        input_ids, attention_mask, labels = [x.to(device) for x in batch]
+        outputs = model(input_ids, attention_mask=attention_mask)
+        preds = torch.argmax(outputs.logits, dim=1)
+        predictions.extend(preds.cpu().numpy())
+
     true_labels = test_dataset["label"]
 
     #--------------------------------------------------------------------------------------------