LoRA and QLoRA concepts in Prompt Engineering / GenAI - Model Metrics & Evaluation

Metrics & Evaluation - LoRA and QLoRA concepts

Which metric matters for LoRA and QLoRA and WHY

LoRA and QLoRA are methods to make large AI models smaller and faster to train. The key metrics to check are model accuracy or task performance after training, and memory usage or speed improvements. We want to keep accuracy high while reducing memory and training time. So, accuracy and efficiency metrics matter most.

Confusion matrix or equivalent visualization

For classification tasks, a confusion matrix shows how well the model predicts each class. For LoRA and QLoRA, the confusion matrix before and after applying these methods helps us see if accuracy dropped.

    Confusion Matrix Example:

          Predicted
          P     N
    Actual P  90    10
           N  15    85

    Total samples = 200
    

If LoRA or QLoRA keeps similar numbers here, it means they preserved accuracy well.

Precision vs Recall tradeoff with examples

LoRA and QLoRA reduce model size and speed up training but might slightly reduce accuracy. This is a tradeoff:

• Precision: How many predicted positives are correct?
• Recall: How many actual positives did the model find?

Example: In spam detection, if LoRA reduces recall, some spam emails might be missed. But if precision stays high, fewer good emails are wrongly marked spam. Depending on the task, you decide which metric to prioritize.

What "good" vs "bad" metric values look like for LoRA and QLoRA

Good: Accuracy or F1 score close to the original full model (e.g., within 1-2%), with much lower memory use and faster training.

Bad: Large drops in accuracy or recall (e.g., more than 5%), meaning the model misses many correct answers, even if it is smaller or faster.

Common pitfalls in metrics for LoRA and QLoRA

• Ignoring accuracy drop: Focusing only on speed or size but losing too much accuracy.
• Data leakage: Testing on data the model saw during training, making metrics look better than real.
• Overfitting: Model performs well on training data but poorly on new data, hiding true performance.
• Not comparing to baseline: Without the original model's metrics, it's hard to judge if LoRA or QLoRA helped or hurt.

Self-check question

Your model uses QLoRA and has 98% accuracy but only 12% recall on fraud cases. Is it good for production? Why or why not?

Answer: No, it is not good. Even though accuracy is high, the very low recall means the model misses most fraud cases. For fraud detection, recall is critical because missing fraud is costly. So this model would not be reliable in real use.

Key Result

LoRA and QLoRA aim to keep accuracy high while reducing model size and training time, balancing precision and recall based on task needs.

Practice

1. What is the main purpose of LoRA in training large AI models?

easy

A. To increase the size of the model for better accuracy

B. To add small trainable parts that make training easier and cheaper

C. To replace the entire model with a smaller one

D. To remove layers from the model to speed up training

Solution

1. Step 1: Understand LoRA's role in model training

   LoRA adds small trainable parts to a big model instead of retraining the whole model, making training easier and cheaper.

2. Step 2: Compare options with LoRA's purpose

   Options B, C, and D describe changing model size or structure, which is not what LoRA does.

3. Final Answer:

   To add small trainable parts that make training easier and cheaper -> Option B

4. Quick Check:

   LoRA = small trainable parts for easier training [OK]

Hint: LoRA adds small parts to big models for easier training [OK]

Common Mistakes:

• Thinking LoRA replaces the whole model
• Confusing LoRA with model size increase
• Assuming LoRA removes layers

2. Which of the following correctly describes QLoRA?

easy

A. A method that combines LoRA with quantization to save memory

B. A technique that trains models without any compression

C. A way to increase model size by adding layers

D. A method that removes LoRA parts to speed up training

Solution

1. Step 1: Recall QLoRA's definition

   QLoRA combines LoRA with quantization (number compression) to reduce memory use and speed up training.

2. Step 2: Eliminate incorrect options

   Options B, C, and D contradict QLoRA's purpose by ignoring compression or removing LoRA parts.

3. Final Answer:

   A method that combines LoRA with quantization to save memory -> Option A

4. Quick Check:

   QLoRA = LoRA + quantization for memory saving [OK]

Hint: QLoRA = LoRA plus compression to save memory [OK]

Common Mistakes:

• Ignoring quantization in QLoRA
• Thinking QLoRA removes LoRA parts
• Believing QLoRA increases model size

3. Given this Python snippet using LoRA and QLoRA concepts:

model_size = 1000  # in MB
lora_size = 10    # LoRA adds 10 MB
quantization_factor = 0.25  # QLoRA compresses to 25%

lora_model_size = model_size + lora_size
qlora_model_size = int(lora_model_size * quantization_factor)
print(qlora_model_size)

What is the printed output?

medium

A. 252

B. 250

C. 260

D. 275

Solution

1. Step 1: Calculate LoRA model size

   LoRA adds 10 MB to 1000 MB, so lora_model_size = 1000 + 10 = 1010 MB.

2. Step 2: Apply QLoRA compression

   QLoRA compresses to 25%, so qlora_model_size = int(1010 * 0.25) = int(252.5) = 252 MB.

3. Final Answer:

   252 -> Option A

4. Quick Check:

   1010 * 0.25 = 252.5 -> 252 [OK]

Hint: Add LoRA size, then multiply by compression factor [OK]

Common Mistakes:

• Multiplying before adding LoRA size
• Rounding incorrectly
• Using 0.2 instead of 0.25 for compression

4. This code tries to calculate QLoRA model size but has an error:

model_size = 800
lora_size = 20
quantization_factor = 0.3

qlora_model_size = model_size + lora_size * quantization_factor
print(qlora_model_size)

What is the error and how to fix it?

medium

A. Wrong variable name; change quantization_factor to quant_factor

B. No error; code is correct

C. Should use integer division // instead of *

D. Missing parentheses; fix with (model_size + lora_size) * quantization_factor

Solution

1. Step 1: Identify operator precedence issue

   Multiplication (*) happens before addition (+), so only lora_size is multiplied by quantization_factor, not the sum.

2. Step 2: Fix with parentheses

   Use (model_size + lora_size) * quantization_factor to multiply the total size by compression factor.

3. Final Answer:

   Missing parentheses; fix with (model_size + lora_size) * quantization_factor -> Option D

4. Quick Check:

   Parentheses fix operator order [OK]

Hint: Use parentheses to control addition before multiplication [OK]

Common Mistakes:

• Ignoring operator precedence
• Changing variable names incorrectly
• Using wrong operators like //

5. You want to fine-tune a large language model on a small laptop with limited memory. Which approach best balances training speed and memory use?

hard

A. Only add LoRA layers without any compression

B. Train the full large model from scratch without compression

C. Use QLoRA to compress the model and add LoRA layers for efficient training

D. Use full precision training without LoRA or compression

Solution

1. Step 1: Understand resource limits

   Small laptops have limited memory, so full model training or full precision is too heavy.

2. Step 2: Choose best method

   QLoRA combines LoRA's small trainable parts with quantization compression, saving memory and speeding training.

3. Step 3: Compare options

   Options B and D ignore memory limits; A lacks compression benefits.

4. Final Answer:

   Use QLoRA to compress the model and add LoRA layers for efficient training -> Option C

5. Quick Check:

   QLoRA = LoRA + compression for small devices [OK]

Hint: Combine LoRA and compression for small device training [OK]

Common Mistakes:

• Ignoring compression benefits
• Trying full model training on small memory
• Using only LoRA without compression