add_wrapped_dataloader: false
add_wrapped_schedulefree: false
amp: false
amp_dtype: float16
amp_impl: native
batch_size: 16
channels_last: false
checkpoint_hist: 10
cl_lambda: 0.1
class_map: ''
clip_grad: null
clip_mode: norm
crop_pct: null
data: train_skeleton.py
data_dir: /gpfs/data/tserre/npant1/ILSVRC/
dataset: torch/imagenet
dataset_download: false
decay_epochs: 30.0
decay_rate: 0.1
device: cuda
device_modules: null
epoch_repeats: 0.0
epochs: 90
experiment: ip_3_hmax_v3_adj_gpu_8_cl_0.1_ip_3_322_322_18432_c1[_6,3,1_]_bypass
fast_norm: false
fuser: ''
gp: null
grad_accum_steps: 1
grad_checkpointing: false
head_init_bias: null
head_init_scale: null
hflip: 0.5
img_size: null
in_chans: null
initial_checkpoint: ''
input_img_mode: null
input_key: null
input_size:
- 3
- 322
- 322
interpolation: ''
layer_decay: null
local_rank: 0
log_interval: 50
lr: 0.01
lr_cycle_decay: 0.1
mean: null
model: hmax_v3_adj
model_kwargs:
  bypass: true
  classifier_input_size: 18432
  ip_scale_bands: 3
momentum: 0.9
no_aug: false
no_ddp_bb: false
no_prefetcher: false
no_resume_opt: false
num_classes: null
opt: sgd
opt_betas: null
opt_eps: null
opt_kwargs: {}
output: /oscar/data/tserre/xyu110/pytorch-output/train/5
pretrained: false
pretrained_path: null
resume: ''
scale:
- 1.0
- 1.0
sched: step
seed: 42
start_epoch: null
std: null
synchronize_step: false
target_key: null
train_crop_mode: rrc
train_num_samples: null
train_split: train
use_multi_epochs_loader: false
val_num_samples: null
val_split: validation
validation_batch_size: null
warmup_epochs: 0
warmup_lr: 1.0e-05
weight_decay: 0.0005
worker_seeding: all
workers: 8