Logs

app.py

@@ -5,7 +5,7 @@ from datasets import load_dataset
 
 import numpy as np
 import os
-from sklearn.metrics import accuracy_score, precision_recall_fscore_support
+from sklearn.metrics import classification_report, accuracy_score, precision_recall_fscore_support
 
 import torch
 import torch.nn as nn
@@ -170,10 +170,11 @@ def evaluate_model(model, dataloader, device):
     print(len(all_labels))
     all_preds = np.asarray(all_preds, dtype=np.float32)
     all_labels = np.asarray(all_labels, dtype=np.float32)
+    report = classification_report(all_labels, all_preds, target_names=id2label.values(), zero_division=0)
     accuracy = accuracy_score(all_labels, all_preds)
     precision, recall, f1, _ = precision_recall_fscore_support(all_labels, all_preds, average='micro')
 
-    return accuracy, precision, recall, f1
+    return report, accuracy, precision, recall, f1
 
 # Function to compute distillation and hard-label loss
 def distillation_loss(student_logits, teacher_logits, true_labels, temperature, alpha):
@@ -212,8 +213,10 @@ dataloader = DataLoader(tokenized_data['train'], batch_size=batch_size, collate_
 # create testing data loader
 test_dataloader = DataLoader(tokenized_data['test'], batch_size=batch_size, collate_fn=data_collator)
 
-untrained_student_accuracy, untrained_student_precision, untrained_student_recall, untrained_student_f1 = evaluate_model(student_model, test_dataloader, device)
-print(f"Untrained Student (test) - Accuracy: {untrained_student_accuracy:.4f}, Precision: {untrained_student_precision:.4f}, Recall: {untrained_student_recall:.4f}, F1 Score: {untrained_student_f1:.4f}")
+untrained_student_report, untrained_student_accuracy, untrained_student_precision, untrained_student_recall, untrained_student_f1 = evaluate_model(student_model, test_dataloader, device)
+print(f"Untrained Student (test) - Report:")
+print(untrained_student_report)
+print(f"Accuracy: {untrained_student_accuracy:.4f}, Precision: {untrained_student_precision:.4f}, Recall: {untrained_student_recall:.4f}, F1 Score: {untrained_student_f1:.4f}")
 
 # put student model in train mode
 student_model.train()
@@ -248,28 +251,22 @@ for epoch in range(num_epochs):
     test_dataloader = DataLoader(tokenized_data['test'], batch_size=batch_size, collate_fn=data_collator, shuffle=True)
 
     # Evaluate the teacher model
-    teacher_accuracy, teacher_precision, teacher_recall, teacher_f1 = evaluate_model(teacher_model, test_dataloader, device)
-    print(f"Teacher (test) - Accuracy: {teacher_accuracy:.4f}, Precision: {teacher_precision:.4f}, Recall: {teacher_recall:.4f}, F1 Score: {teacher_f1:.4f}")
+    teacher_report, teacher_accuracy, teacher_precision, teacher_recall, teacher_f1 = evaluate_model(teacher_model, test_dataloader, device)
+    print(f"Teacher (test) - Report:")
+    print(teacher_report)
+    print(f"Accuracy: {teacher_accuracy:.4f}, Precision: {teacher_precision:.4f}, Recall: {teacher_recall:.4f}, F1 Score: {teacher_f1:.4f}")
+    print("\n")
 
     # Evaluate the student model
-    student_accuracy, student_precision, student_recall, student_f1 = evaluate_model(student_model, test_dataloader, device)
-    print(f"Student (test) - Accuracy: {student_accuracy:.4f}, Precision: {student_precision:.4f}, Recall: {student_recall:.4f}, F1 Score: {student_f1:.4f}")
+    student_report, student_accuracy, student_precision, student_recall, student_f1 = evaluate_model(student_model, test_dataloader, device)
+    print(f"Student (test) - Report:")
+    print(student_report)
+    print(f"Accuracy: {student_accuracy:.4f}, Precision: {student_precision:.4f}, Recall: {student_recall:.4f}, F1 Score: {student_f1:.4f}")
     print("\n")
 
     # put student model back into train mode
     student_model.train()
 
-#Compare the models
-# create testing data loader
-validation_dataloader = DataLoader(tokenized_data['test'], batch_size=8, collate_fn=data_collator)
-# Evaluate the teacher model
-teacher_accuracy, teacher_precision, teacher_recall, teacher_f1 = evaluate_model(teacher_model, validation_dataloader, device)
-print(f"Teacher (validation) - Accuracy: {teacher_accuracy:.4f}, Precision: {teacher_precision:.4f}, Recall: {teacher_recall:.4f}, F1 Score: {teacher_f1:.4f}")
-# Evaluate the student model
-student_accuracy, student_precision, student_recall, student_f1 = evaluate_model(student_model, validation_dataloader, device)
-print(f"Student (validation) - Accuracy: {student_accuracy:.4f}, Precision: {student_precision:.4f}, Recall: {student_recall:.4f}, F1 Score: {student_f1:.4f}")
-     
-
 st.write('Pushing model to huggingface')
 
 # Push model to huggingface