NSGAD4K357E46W8.yaml

model_name: mobilenetv3_1d
num_classes: 6

kernel_size_list: [3, 5, 7]
expand_ratio_list: [0.5, 1, 3, 4, 6]
depth_list: [1, 2, 3, 4]
width_mult: 0.25
input_seq_len: 23
patch_size: 1
feature_cross: false
arch: "Patchemb"

act_func_list: ["h_swish","h_swish","relu6"]

Task: "expand_ratio:6->46"
NSGA: True
checkpoint_path: "./nsga_output/NSGA_D4K357E6W8_1x/checkpoint/model_best.pth.tar"
work_path: "./nsga_output/NSGA_D4K357E46W8_1x"
load_train_settings: "./nsga_output/NSGA_D4K357E6W8_1x/record/16-train.pkl"
load_val_record: "./nsga_output/NSGA_D4K357E6W8_1x/record/16-val.pkl"
finetune: True
distilation:
  kd_ratio: 1.0
  teacher_model_path: "./nsga_output/OFA_D4K7E6W64_10x/checkpoint/model_best.pth.tar"
  kd_type: "ce"
max_no_improved_epochs: 10
runtime:
  n_epochs: 64
  batch_size: 2048
  num_workers: 3

  opt_type: "sgd"
  init_lr: 0.0005
  momentum: 0.99
  no_nesterov: False
  lr_schedule_type: "cosine"
  weight_decay: 3e-5
  no_decay_keys: "bn#bias"
  warmup_epochs: 5
  warmup_lr: 0.000001
  
  model_init: "he_fin"
  label_smoothing: 0.1
  validation_frequency: 1
  dynamic_batch_size: 2


num_sample_train_per_epoch: 1000
num_sample_val_per_epoch: 32
object_type: "params"
input_shape: [23]
deleted_keys: []

search_space:
  active_stem0_depth: [4]
  active_stem1_depth: [4]
  active_embed_dim: [8]
  active_kernel_size: [3, 5, 7]
  active_expand_ratio: [4, 6]


xu: [3, 2, 2, 2, 2, 2, 2, 2, 2, 3, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 0]
xl: [3, 0, 0, 0, 0, 1, 1, 1, 1, 3, 0, 0, 0, 0, 1, 1, 1, 1, 0, 1, 0]