Examples

Examples to help you get started with Mixtrain.

• Simple Workflow - Basic workflow structure with inputs, lifecycle methods, and CLI usage.

Training

Train models on Mixtrain infrastructure with GPU support and checkpointing.

• PyTorch Training - Train a CNN on MNIST with a GPU
• Distributed Training - Multi-node PyTorch DDP with 4x A100 GPUs, ResNet on CIFAR-10

Evaluations

Eval models on a dataset and visualize outputs side-by-side from different models.

• Evaluations - Compare text-to-image, video, and other generative models on the same inputs