text
stringclasses 206
values |
|---|
debug: False |
delete_previous_checkpoint: False |
denormalize_features: False |
device: cuda:0 |
dist_backend: None |
dist_url: None |
distill: False |
distill_model: None |
distill_pretrained: None |
distributed: True |
ema_z_mode: ema_n_model |
epochs: 33 |
epochs_cooldown: None |
eps: 1e-06 |
fit_neg_only: True |
fit_w_prev: False |
force_custom_text: False |
force_image_size: None |
force_patch_dropout: None |
force_quick_gelu: False |
freeze_lambda_after_num_epochs: 100 |
gather_with_grad: True |
grad_checkpointing: False |
grad_clip_norm: 1.0 |
h_dim_factor: 4 |
horovod: False |
image_interpolation: None |
image_mean: None |
image_resize_mode: None |
image_std: None |
imagenet_v2: None |
imagenet_val: /scratch/imagenet/val/ |
init_lambda: 1.0 |
init_logit_bias: -10.0 |
init_logit_scale: None |
lambda_eps: 1e-12 |
lambda_lr: 0.001 |
lambda_tolerance: 0.0005 |
lambda_update_frequency: [20, 10, 1] |
learn_logit_bias: False |
learn_logit_scale: True |
local_loss: True |
local_rank: 0 |
lock_image: False |
lock_image_freeze_bn_stats: False |
lock_image_unlocked_groups: 0 |
lock_text: False |
lock_text_freeze_layer_norm: False |
lock_text_unlocked_layers: 0 |
log_every_n_steps: 100 |
log_level: 20 |
log_local: False |
log_path: ./logs/emn_fix-bias-10.0sample-negFalse-scaleFalse-pos1.0-eps1e-06-stocFalselr0.0004-ViT-B-32-1990_0124122059/out.log |
logit_scale_clamp: 100 |
logs: ./logs/ |
loss_type: COND_EXP |
lr: 0.0004 |
lr_cooldown_end: 0.0 |
lr_cooldown_power: 1.0 |
lr_scheduler: cosine |
lr_tau: 0.0001 |
model: ViT-B-32 |
model_beta_init: 0.0 |
model_ema_beta: 0.99 |
model_update_type: ONE_STEP |
n_class_tokens: -1 |
name: emn_fix-bias-10.0sample-negFalse-scaleFalse-pos1.0-eps1e-06-stocFalselr0.0004-ViT-B-32-1990_0124122059 |
no_set_device_rank: False |
norm_cap: 1.0 |
normalize_type: L2 |
note: emn_fix |
pos_coef: 1.0 |
precision: amp |
pretrained: |
pretrained_image: False |
rank: 0 |
remote_sync: None |
remote_sync_frequency: 300 |
remote_sync_protocol: s3 |
report_to: wandb |
resume: None |
save_frequency: 1 |
save_most_recent: False |
scale_loss: False |
seed: 42 |
siglip: False |
skip_scheduler: False |
stoc_fit_lambda: False |
tensorboard: False |
tensorboard_path: |
torchcompile: False |
torchscript: False |
trace: False |
train_data: /scratch/cc12m/{00000..01240}.tar |
train_data_upsampling_factors: None |
train_num_samples: 9187328 |
update_ema_mlps_every_n_steps: 2000 |
use_bn_sync: False |
use_bnb_linear: None |
use_feature_diff: False |
Subsets and Splits