deve

app.py

@@ -37,9 +37,11 @@ def fine_tune_model(model, dataset_name, hub_id, api_key, num_epochs, batch_size
             if isinstance(preds, tuple):
                 preds = preds[0]
             # Replace -100s used for padding as we can't decode them
-            #preds = np.where(preds != -100, preds, tokenizer.pad_token_id)
+           # preds = np.where(preds != -100, preds, tokenizer.pad_token_id)
             #preds = np.array(preds)
             decoded_preds = tokenizer.batch_decode(preds, skip_special_tokens=True)
+            if isinstance(labels, tuple):
+                labels = labels[0]            
             #labels = np.where(labels != -100, labels, tokenizer.pad_token_id)
             #labels = np.array(labels)
             decoded_labels = tokenizer.batch_decode(labels, skip_special_tokens=True)
@@ -51,7 +53,7 @@ def fine_tune_model(model, dataset_name, hub_id, api_key, num_epochs, batch_size
             accuracy = accuracy_score(decoded_labels, decoded_preds)
             result["eval_accuracy"] = round(accuracy * 100, 4)            
             return result
-        
+        tokenizer.
         login(api_key.strip())
    
     
@@ -232,7 +234,7 @@ def fine_tune_model(model, dataset_name, hub_id, api_key, num_epochs, batch_size
         )    
     
         # Fine-tune the model
-        trainer.train()
+        trainer.evaluate()
         #if os.path.exists(training_args.output_dir) and os.listdir(training_args.output_dir):
             #train_result = trainer.train(resume_from_checkpoint=True)
         #else: