fix detach tensor

src/main.py

@@ -2,6 +2,7 @@
 import argparse
 import logging
 
+logging.basicConfig(level=logging.INFO)
 logger = logging.getLogger(__name__)
 logger.setLevel(logging.INFO)
 
@@ -33,7 +34,7 @@ def main(path_to_data: str,
     optimal_length = get_length(df, texts_col)
     X, vocab_size = encode_texts(df, texts_col, max_seq_length=optimal_length, return_vocab_size=True)
 
-    y = get_labels(df, labels_col)
+    y = get_labels(df, labels_col, n_classes)
 
     train_loader, test_loader = create_TorchLoaders(X, y, test_size=0.10, batch_size=batch_size, batch_size_eval=batch_size_eval)
 
@@ -86,15 +87,12 @@ def main(path_to_data: str,
             labels = labels.to(device, dtype=torch.float)
 
             preds = Model(inputs)
-
             loss = criterion(preds, labels)
 
             ## Metrics computation
-
-
             metrics["training_loss"].append(loss.item())
 
-            tmp_f1 = f1_score(labels.to("cpu").numpy(), preds.to("cpu").numpy())
+            tmp_f1 = f1_score(labels.to("cpu").detach().numpy(), preds.to("cpu").detach().numpy())
 
             metrics["training_f1"].append(tmp_f1)
 
@@ -119,13 +117,12 @@ def main(path_to_data: str,
                 labels = labels.to(device, dtype=torch.float)
 
                 preds = Model(inputs)
-
                 eval_loss = criterion(preds, labels)
 
                 ## Eval metrics
                 metrics["eval_loss"].append(eval_loss.item())
 
-                tmp_f1 = f1_score(labels.to("cpu").numpy(), preds.to("cpu").numpy()) ## detach 
+                tmp_f1 = f1_score(labels.to("cpu").detach().numpy(), preds.to("cpu").detach().numpy()) ## detach 
 
                 metrics["eval_f1"].append(tmp_f1)