DL & GenAI · PyTorch training workflow

Indian Institute of Technology Madras

BS Degree Programme · Examination & Study Paper

Introduction to Deep Learning and Generative AI

PyTorch training workflow

Official Paper

Examination Instructions & Guidelines

Solve on notebook/paper first before checking solutions.
MCQ: Exactly 1 option is correct. Negative marking applies.
Scientific calculator allowed for all calculations.
Official answer key is sealed at the end of this sheet.

Useful Reference Data for this Paper
Note: In all the questions, assume necessary imports wherever applicable.
Q1MCQ2 Marks·Choose one option
Key ↓
1. Amodel.cuda()
2. Binput_batch.cuda()
3. Cinput_batch = input_batch.requires_grad_(True)
4. Dinput_batch = input_batch.to(device)
5. Etorch.set_default_tensor_type('torch.cuda.FloatTensor')
Q2MCQ2 Marks·Choose one option
Key ↓
1. AIt prevents parameters from being changed by the optimizer during the forward pass.
2. BIt deallocates the model's parameters from the GPU to save memory.
3. CIt disables the creation of the computational graph, saving memory and speeding up execution.
4. DIt ensures that the model operates entirely on the CPU, preventing unnecessary GPU transfers.
Q3MCQ3 Marks·Choose one option
Key ↓
1. A
2. B
3. C
4. D
5. E

Official Key

Master Answer Key & Explanations

Official answer keys from IIT Madras BS past papers. Click any question to review.