Multi-GPU training in PyTorch - Model Pipeline Trace

Model Pipeline - Multi-GPU training

This pipeline shows how training a neural network can be sped up by using multiple GPUs at the same time. The data is split and sent to each GPU, the model trains in parallel, and results are combined to improve learning speed.

Data Flow - 4 Stages

1Data Loading

10000 rows x 20 features→Load dataset and batch into groups of 100→100 batches x 100 rows x 20 features

Batch 1: [[0.5, 1.2, ..., 0.3], ..., [0.7, 0.8, ..., 1.1]]

↓

2Data Distribution to GPUs

100 batches x 100 rows x 20 features→Split each batch evenly across 2 GPUs→100 batches x 2 GPUs x 50 rows x 20 features

GPU 0 batch slice: 50 rows, GPU 1 batch slice: 50 rows

↓

3Model Training on GPUs

50 rows x 20 features per GPU→Each GPU trains model on its data slice in parallel→50 rows x 10 output classes per GPU

GPU 0 output: [0.1, 0.7, ..., 0.05], GPU 1 output: [0.2, 0.6, ..., 0.1]

↓

4Gradient Aggregation

Gradients from 2 GPUs→Combine gradients from both GPUs to update model weights→Single updated model weights

Weights updated using average gradients from GPU 0 and GPU 1

Training Trace - Epoch by Epoch

Loss
1.2 |****
0.9 |***
0.7 |**
0.55|*
0.45| 
    +------------
    Epochs 1 to 5

Epoch	Loss ↓	Accuracy ↑	Observation
1	1.2	0.45	Initial training with high loss and low accuracy
2	0.9	0.60	Loss decreased, accuracy improved as model learns
3	0.7	0.72	Continued improvement, model converging
4	0.55	0.80	Loss dropping steadily, accuracy rising
5	0.45	0.85	Training nearing good performance

Prediction Trace - 4 Layers

Layer 1: Input Split

Layer 2: Forward Pass on GPU 0

Layer 3: Forward Pass on GPU 1

Layer 4: Combine Outputs

Model Quiz - 3 Questions

Why do we split data batches across multiple GPUs?

A To reduce the size of the dataset

B To increase the number of model layers

C To train the model faster by parallel processing

D To avoid using GPUs

Key Insight

Using multiple GPUs allows the model to train faster by splitting data and computations. This parallelism helps the model learn more quickly while maintaining accuracy improvements.

Practice

1. What is the main purpose of using torch.nn.DataParallel in PyTorch?

easy

A. To reduce the model size by pruning layers

B. To convert the model to a CPU-only version

C. To split the model and data across multiple GPUs for faster training

D. To visualize the model architecture

Solution

1. Step 1: Understand the role of DataParallel

   torch.nn.DataParallel is designed to distribute the model and input data across multiple GPUs to speed up training.

2. Step 2: Compare options

   Reducing the model size by pruning layers, converting the model to a CPU-only version, and visualizing the model architecture do not relate to multi-GPU training. Only splitting the model and data across multiple GPUs for faster training correctly describes the purpose.

3. Final Answer:

   To split the model and data across multiple GPUs for faster training -> Option C

4. Quick Check:

   Multi-GPU training = DataParallel usage [OK]

Hint: DataParallel splits model and data across GPUs for speed [OK]

Common Mistakes:

• Thinking DataParallel reduces model size
• Confusing DataParallel with CPU conversion
• Assuming DataParallel is for visualization

2. Which of the following is the correct way to wrap a PyTorch model for multi-GPU training?

easy

A. model = torch.nn.DataParallel(model)

B. model = torch.nn.Sequential(model)

C. model = torch.nn.ModuleList(model)

D. model = torch.nn.Parameter(model)

Solution

1. Step 1: Identify the correct wrapper for multi-GPU

   To enable multi-GPU training, PyTorch provides torch.nn.DataParallel which wraps the model.

2. Step 2: Check other options

   Sequential chains layers, ModuleList stores layers, and Parameter is for tensors, none enable multi-GPU.

3. Final Answer:

   model = torch.nn.DataParallel(model) -> Option A

4. Quick Check:

   Wrap model with DataParallel for multi-GPU [OK]

Hint: Use DataParallel(model) to enable multi-GPU training [OK]

Common Mistakes:

• Using Sequential or ModuleList instead of DataParallel
• Trying to wrap model with Parameter
• Forgetting to wrap the model at all

3. Consider this code snippet for multi-GPU training:

import torch
model = MyModel()
model = torch.nn.DataParallel(model)
model.cuda()
inputs = inputs.cuda()
outputs = model(inputs)
print(outputs.device)

What will be printed if multiple GPUs are available?

medium

A. cuda:0

B. cpu

C. cuda:1

D. cuda:all

Solution

1. Step 1: Understand DataParallel device behavior

   When using DataParallel, the model replicates on all GPUs but the output is gathered on the default device, usually cuda:0.

2. Step 2: Analyze the print statement

   The outputs.device will show the device of the gathered output tensor, which is cuda:0, not all GPUs or CPU.

3. Final Answer:

   cuda:0 -> Option A

4. Quick Check:

   DataParallel output device = cuda:0 [OK]

Hint: DataParallel gathers output on cuda:0 device [OK]

Common Mistakes:

• Expecting output on all GPUs simultaneously
• Thinking output is on CPU
• Assuming output device is cuda:1 or others

4. You wrote this code for multi-GPU training but get an error:

model = MyModel()
model = torch.nn.DataParallel(model)
inputs = inputs.cuda()
outputs = model(inputs)

The error says: 'Expected all tensors to be on the same device'. What is the likely cause?

medium

A. Model should be wrapped with Sequential instead

B. Model is not moved to GPU before wrapping with DataParallel

C. DataParallel requires inputs on CPU

D. Inputs are not on CPU

Solution

1. Step 1: Check model and input device placement

   The model must be moved to GPU (e.g., model.cuda()) before wrapping with DataParallel to ensure all parts are on the same device.

2. Step 2: Analyze error cause

   If the model is on CPU but inputs are on GPU, the error about device mismatch occurs.

3. Final Answer:

   Model is not moved to GPU before wrapping with DataParallel -> Option B

4. Quick Check:

   Model and inputs must be on same device [OK]

Hint: Move model to GPU before DataParallel wrap [OK]

Common Mistakes:

• Wrapping model before moving to GPU
• Assuming inputs must be on CPU
• Using wrong wrapper like Sequential

5. You want to train a model on 4 GPUs using PyTorch. Which of these steps is NOT necessary for correct multi-GPU training?

hard

A. Move input data to GPU before passing to the model

B. Move the model to GPU before wrapping

C. Wrap the model with torch.nn.DataParallel

D. Manually split input batches and send each to a different GPU

Solution

1. Step 1: Review multi-GPU training steps

   Using DataParallel automatically splits input batches and sends them to multiple GPUs, so manual splitting is not needed.

2. Step 2: Confirm necessary steps

   Wrapping the model, moving it to GPU, and moving inputs to GPU are all required for proper multi-GPU training.

3. Final Answer:

   Manually split input batches and send each to a different GPU -> Option D

4. Quick Check:

   DataParallel handles input splitting automatically [OK]

Hint: DataParallel auto-splits inputs; manual split not needed [OK]

Common Mistakes:

• Trying to manually split inputs across GPUs
• Forgetting to move model to GPU
• Not wrapping model with DataParallel