BERT fine-tuning for classification in NLP - Model Metrics & Evaluation

Metrics & Evaluation - BERT fine-tuning for classification

Which metric matters for BERT fine-tuning for classification and WHY

When we fine-tune BERT for classification, the main goal is to correctly label text into categories. The key metrics to check are accuracy, precision, recall, and F1 score. Accuracy tells us overall how many texts were labeled right. Precision shows how many predicted labels were actually correct. Recall tells us how many true labels we found out of all real ones. F1 score balances precision and recall, which is important when classes are uneven or mistakes have different costs.

Confusion matrix example

      | Predicted Positive | Predicted Negative |
      |--------------------|--------------------|
      | True Positive (TP): 80  | False Negative (FN): 20 |
      | False Positive (FP): 10 | True Negative (TN): 90  |

      Total samples = TP + FP + TN + FN = 80 + 10 + 90 + 20 = 200

      Precision = TP / (TP + FP) = 80 / (80 + 10) = 0.89
      Recall = TP / (TP + FN) = 80 / (80 + 20) = 0.80
      F1 Score = 2 * (Precision * Recall) / (Precision + Recall) = 2 * (0.89 * 0.80) / (0.89 + 0.80) ≈ 0.84
    

Precision vs Recall tradeoff with examples

Imagine BERT is classifying emails as spam or not spam.

• High Precision: Few good emails are wrongly marked as spam. This means users don't miss important emails. But some spam might get through.
• High Recall: Most spam emails are caught. But some good emails might be wrongly marked as spam, annoying users.

Depending on what matters more, we adjust the model or threshold. For spam, usually high precision is preferred to avoid losing good emails.

Good vs Bad metric values for BERT classification

Good: Accuracy above 85%, Precision and Recall above 80%, and F1 score balanced near 0.8 or higher. This means the model predicts well and finds most true labels without many mistakes.

Bad: Accuracy near 50% (like random guessing), Precision or Recall below 50%, or very unbalanced F1 score (e.g., high precision but very low recall). This means the model is not reliable or misses many true cases.

Common pitfalls in metrics for BERT fine-tuning

• Accuracy paradox: High accuracy can be misleading if classes are imbalanced. For example, if 90% of texts are class A, predicting all as A gives 90% accuracy but no real learning.
• Data leakage: If test data leaks into training, metrics look too good but model fails in real use.
• Overfitting: Very high training accuracy but low test accuracy means model memorized training data, not learned general patterns.

Self-check question

Your BERT model has 98% accuracy but only 12% recall on the positive class (e.g., fraud detection). Is this good for production? Why or why not?

Answer: No, it is not good. The model misses 88% of actual positive cases, which is very risky in fraud detection. High accuracy is misleading because most data is negative. You need to improve recall to catch more fraud cases.

Key Result

For BERT fine-tuning classification, balanced precision, recall, and F1 score matter most to ensure reliable and fair predictions.

Practice

1. What is the main purpose of fine-tuning BERT for a classification task?

easy

A. To adapt BERT's knowledge to classify specific categories in your data

B. To train BERT from scratch on a large dataset

C. To reduce the size of the BERT model for faster inference

D. To convert text into images for classification

Solution

1. Step 1: Understand BERT's pretraining

   BERT is pretrained on general language tasks and needs adjustment for specific tasks like classification.

2. Step 2: Purpose of fine-tuning

   Fine-tuning adapts BERT's learned language understanding to classify categories in your dataset.

3. Final Answer:

   To adapt BERT's knowledge to classify specific categories in your data -> Option A

4. Quick Check:

   Fine-tuning = adapt BERT for classification [OK]

Hint: Fine-tuning means adjusting BERT for your task, not training from scratch [OK]

Common Mistakes:

• Thinking fine-tuning trains BERT from zero
• Confusing fine-tuning with model compression
• Assuming BERT outputs images

2. Which of the following is the correct way to tokenize text before feeding it to BERT in Python?

easy

A. tokens = text.split(' ')

B. tokens = tokenizer.encode_plus(text, return_tensors='pt')

C. tokens = tokenizer.tokenize(text)

D. tokens = text.lower()

Solution

1. Step 1: Identify proper BERT tokenization method

   BERT uses tokenizer.encode_plus to convert text into token IDs and attention masks.

2. Step 2: Compare options

   tokens = tokenizer.encode_plus(text, return_tensors='pt') uses encode_plus with return_tensors='pt' for PyTorch tensors, which is correct for BERT input.

3. Final Answer:

   tokens = tokenizer.encode_plus(text, return_tensors='pt') -> Option B

4. Quick Check:

   Use encode_plus for BERT tokenization [OK]

Hint: Use tokenizer.encode_plus or tokenizer() for BERT input [OK]

Common Mistakes:

• Using simple split instead of tokenizer
• Only tokenizing without encoding IDs
• Not returning tensors for model input

3. Given this code snippet for fine-tuning BERT, what will be the output of print(predictions.argmax(dim=1)) if the model predicts logits [[2.0, 1.0], [0.5, 1.5]] for two samples?

logits = torch.tensor([[2.0, 1.0], [0.5, 1.5]])
predictions = logits
print(predictions.argmax(dim=1))

medium

A. tensor([2, 1])

B. tensor([1, 0])

C. tensor([1, 1])

D. tensor([0, 1])

Solution

1. Step 1: Understand argmax(dim=1)

   Argmax along dim=1 finds the index of max value in each row (sample).

2. Step 2: Calculate argmax for each sample

   First row: max is 2.0 at index 0; second row: max is 1.5 at index 1.

3. Final Answer:

   tensor([0, 1]) -> Option D

4. Quick Check:

   Argmax per row = [0, 1] [OK]

Hint: Argmax dim=1 picks max index per sample row [OK]

Common Mistakes:

• Confusing dim=0 with dim=1
• Mixing up indices and values
• Expecting values instead of indices

4. You run this training loop snippet but get a runtime error: TypeError: forward() missing 1 required positional argument: 'labels'. What is the likely fix?

outputs = model(input_ids, attention_mask)
loss = outputs.loss
loss.backward()

medium

A. Pass labels to the model call: model(input_ids, attention_mask, labels=labels)

B. Remove loss.backward() call

C. Change input_ids to input_id

D. Call model with only input_ids

Solution

1. Step 1: Understand error cause

   The model expects labels to compute loss but they are missing in the call.

2. Step 2: Fix by passing labels

   Include labels argument in model call to get loss: model(input_ids, attention_mask, labels=labels).

3. Final Answer:

   Pass labels to the model call: model(input_ids, attention_mask, labels=labels) -> Option A

4. Quick Check:

   Missing labels argument causes loss error [OK]

Hint: Always pass labels to get loss during training [OK]

Common Mistakes:

• Ignoring the missing labels argument
• Removing backward call instead of fixing input
• Changing variable names incorrectly

5. You want to fine-tune BERT on a small dataset for sentiment classification. Which strategy helps avoid overfitting during training?

hard

A. Train BERT without tokenization to save time

B. Increase batch size to maximum and train longer

C. Use a small learning rate and add dropout layers

D. Remove the classification head and train only embeddings

Solution

1. Step 1: Identify overfitting risks

   Small datasets can cause the model to memorize instead of generalize.

2. Step 2: Apply regularization techniques

   Using a small learning rate and dropout helps the model learn smoothly and avoid overfitting.

3. Final Answer:

   Use a small learning rate and add dropout layers -> Option C

4. Quick Check:

   Small LR + dropout reduces overfitting [OK]

Hint: Small learning rate + dropout helps generalize on small data [OK]

Common Mistakes:

• Training longer without regularization
• Skipping tokenization
• Removing classification head incorrectly