mv3_distill.yml

class_dim: 100
total_images: 50000
epochs: 1000
topk: 5
save_model_dir: ./output/
use_gpu: True
model_type: cls_distill

LEARNING_RATE:
    function: 'Cosine'
    params:
        lr: 0.001
        warmup_epoch: 5

OPTIMIZER:
    function: 'Momentum'
    params:
        momentum: 0.9
    regularizer:
        function: 'L2'
        factor: 0.00002

TRAIN:
    batch_size: 1280
    num_workers: 4

VALID:
    batch_size: 64
    num_workers: 4