includesL:
  - headers/data.yaml
  - headers/device.yaml
  - headers/model.yaml

data_root: /home/xtx/datasets/cifar100
image_size: &image_size 224
num_workers: &num_workers 16

save_path: ./

class_order: &class_order [87, 0, 52, 58, 44, 91, 68, 97, 51, 15, 94, 92, 10, 72, 49, 78, 61, 14, 8, 86, 84, 96, 18, 24, 32, 45, 88, 11, 4, 67, 69, 66, 77, 47, 79, 93, 29, 50, 57, 83, 17, 81, 41, 12, 37, 59, 25, 20, 80, 73, 1, 28, 6, 46, 62, 82, 53, 9, 31, 75, 38, 63, 33, 74, 27, 22, 36, 3, 16, 21, 60, 19, 70, 90, 89, 43, 5, 42, 65, 76, 40, 30, 23, 85, 2, 95, 56, 48, 71, 64, 98, 13, 99, 7, 34, 55, 54, 26, 35, 39]
seed: &seed 1919810

is_rapf: True


# Control B and 
init_cls_num: &init_cls_num 50
inc_cls_num: &inc_cls_num 10
task_num: &task_num 6

epoch: &epoch 15
batch_size: &batch_size 128
train_batch_size: &train_batch_size 100
n_gpu: 1
beta: &beta 2
shrinkage: &shrinkage False
threshold: &threshold 0.55
val_per_epoch: &val_per_epoch 10

optimizer:
  name: Adam
  kwargs:
    lr: 0.001
    weight_decay: 0.0000

lr_scheduler:
  name: MultiStepLR
  kwargs:
    gamma: 0.1
    milestones: [4, 10]
    last_epoch: -1

backbone:
  name: clip
  kwargs:
    model_name: ViT-B/16
    device: cuda
    experts_num: 1
    block_layer: ResidualAttentionBlock_MoE_MLP
    top_k : 1
    step: 1
    act_layer: QuickGELU
    norm_layer: LayerNorm
 
classifier:
  name: RAPF
  kwargs:
    init_cls_num: *init_cls_num
    inc_cls_num: *inc_cls_num
    threshold: *threshold
    beta: *beta
    shrinkage: *shrinkage
    train_batch_size: *train_batch_size
    batch_size: *batch_size
    num_workers: *num_workers
    prompt_template: "a good photo of a {}"
    seed: *seed
    fp16: False
    class_order: *class_order
    mix_bias: 0.6