### Current progress: - General - [x] Random color distortion (from SimCLR) - [x] Random Gaussian blur (from SimCLR) - [ ] Random multi-crop (from SwAV) - [ ] InfoNCE loss - [ ] Momentum encoder - [ ] Selective Kernel (SK) convolution layer (used in SimCLR v2) - [x] LARS optimizer - [x] Consine annealing with linear warmup scheduler - [x] Linear scheduler (for torch<=1.9) - [x] CSV logger - [x] TensorBoard logger - [x] Checkpoint saving and restoring - [x] Command line parameter parser - [x] YAML parameter parser - [ ] PyTorch workflow encapsulation - [ ] Dedicated evaluation script - [ ] Detailed readme file - [ ] Data Parallel - [ ] Distributed Data Parallel - [ ] Global batch normalization - [ ] Shuffling batch normalization - Supervised baseline - [x] ResNet - [ ] ViT - [x] CIFAR-10 - [x] CIFAR-100 - [x] ImageNet-1k - Self-supervised baseline - [ ] SimCLR (contrastive, ResNet) - [ ] MoCo v2 (contrastive, ResNet) - [ ] MoCo v3 (contrastive, ViT) - [ ] BYOL (non-contrastive, ResNet) - [ ] DINO (self-distillation, ViT) To be continued ...