| accum_freq: 1 | |
| aug_cfg: {} | |
| batch_size: 1024 | |
| beta1: 0.9 | |
| beta2: 0.98 | |
| checkpoint_path: $HOME/clip-synthetic-captions/output/ye-pop-llava_caption/checkpoints | |
| coca_caption_loss_weight: 2.0 | |
| coca_contrastive_loss_weight: 1.0 | |
| copy_codebase: False | |
| csv_caption_key: title | |
| csv_img_key: filepath | |
| csv_separator: | |
| dataset_resampled: False | |
| dataset_type: webdataset | |
| ddp_static_graph: True | |
| debug: False | |
| delete_previous_checkpoint: False | |
| device: cuda:0 | |
| dist_backend: nccl | |
| dist_url: env:// | |
| distill: False | |
| distill_model: None | |
| distill_pretrained: None | |
| distributed: True | |
| epochs: 4 | |
| epochs_cooldown: None | |
| eps: 1e-06 | |
| force_custom_text: False | |
| force_image_size: None | |
| force_patch_dropout: None | |
| force_quick_gelu: False | |
| gather_with_grad: True | |
| grad_checkpointing: True | |
| grad_clip_norm: None | |
| horovod: False | |
| image_mean: None | |
| image_std: None | |
| imagenet_v2: None | |
| imagenet_val: None | |
| 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: $HOME/clip-synthetic-captions/output/ye-pop-llava_caption/out.log | |
| logs: $HOME/clip-synthetic-captions/output | |
| lr: 0.0005 | |
| lr_cooldown_end: 0.0 | |
| lr_cooldown_power: 1.0 | |
| lr_scheduler: cosine | |
| model: ViT-B-32 | |
| name: ye-pop-llava_caption | |
| no_set_device_rank: False | |
| precision: amp | |
| pretrained: | |
| pretrained_image: False | |
| rank: 0 | |
| remote_sync: None | |
| remote_sync_frequency: 300 | |
| remote_sync_protocol: s3 | |
| report_to: | |
| resume: None | |
| save_frequency: 0 | |
| save_most_recent: True | |
| seed: 0 | |
| skip_scheduler: False | |
| tensorboard: False | |
| tensorboard_path: | |
| torchscript: False | |
| trace: False | |
| train_data: $HOME/clip-synthetic-captions/data/postprocessed/ye-pop-img2dataset-llava_caption/shards/{00000000..00000049}.tar | |
| train_data_upsampling_factors: None | |
| train_num_samples: 122880 | |
| use_bn_sync: False | |
| val_data: None | |
| val_frequency: 1 | |
| val_num_samples: None | |
| wandb: False | |
| wandb_notes: | |
| wandb_project_name: open-clip | |
| warmup: 500 | |
| wd: 0.2 | |
| workers: 2 | |
| world_size: 4 | |
| zeroshot_frequency: 2 | |