Hugging Face's logo

lingkai
/
open-clip

Zero-Shot Image Classification
OpenCLIP
clip
vision
image-text-retrieval
laion400m
Model card Files
xet
 Community
open-clip / out.log
lingkai's picture
lingkai
Upload folder using huggingface_hub
78276df verified
raw
history blame contribute delete
202 kB
2026-05-28,06:29:25 | INFO | open_clip_train.main | Running in distributed mode with multiple processes. Device: cuda:0.Process (global: 0, local 0), total 8.
2026-05-28,06:29:25 | INFO | open_clip.factory | Parsing model identifier. Schema: None, Identifier: ViT-B-32
2026-05-28,06:29:25 | INFO | open_clip.factory | Loaded built-in ViT-B-32 model config.
2026-05-28,06:29:25 | INFO | open_clip.factory | No potential checkpoint path found from config source or pretrained arg.
2026-05-28,06:29:25 | INFO | open_clip.factory | Instantiating model architecture: CLIP
2026-05-28,06:29:25 | WARNING | open_clip.factory | No pretrained weights loaded for model 'ViT-B-32'. Model initialized randomly.
2026-05-28,06:29:25 | INFO | open_clip.factory | Final image preprocessing configuration set: {'size': (224, 224), 'mode': 'RGB', 'mean': (0.48145466, 0.4578275, 0.40821073), 'std': (0.26862954, 0.26130258, 0.27577711), 'interpolation': 'bicubic', 'resize_mode': 'shortest', 'fill_color': 0}
2026-05-28,06:29:25 | INFO | open_clip.factory | Model ViT-B-32 creation process complete.
2026-05-28,06:29:25 | INFO | open_clip_train.main | Model:
2026-05-28,06:29:25 | INFO | open_clip_train.main | CLIP(
 (visual): VisionTransformer(
 (conv1): Conv2d(3, 768, kernel_size=(32, 32), stride=(32, 32), bias=False)
 (patch_dropout): Identity()
 (ln_pre): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
 (transformer): Transformer(
 (resblocks): ModuleList(
 (0-11): 12 x ResidualAttentionBlock(
 (ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
 (attn): Attention(
 (ln_q): Identity()
 (ln_k): Identity()
 (attn_drop): Dropout(p=0.0, inplace=False)
 (ln_inner): Identity()
 (out_proj): Linear(in_features=768, out_features=768, bias=True)
 (out_drop): Dropout(p=0.0, inplace=False)
 )
 (ls_1): Identity()
 (ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
 (mlp): Sequential(
 (c_fc): Linear(in_features=768, out_features=3072, bias=True)
 (gelu): GELU(approximate='none')
 (c_proj): Linear(in_features=3072, out_features=768, bias=True)
 )
 (ls_2): Identity()
 )
 )
 )
 (ln_post): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
 )
 (transformer): Transformer(
 (resblocks): ModuleList(
 (0-11): 12 x ResidualAttentionBlock(
 (ln_1): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
 (attn): Attention(
 (ln_q): Identity()
 (ln_k): Identity()
 (attn_drop): Dropout(p=0.0, inplace=False)
 (ln_inner): Identity()
 (out_proj): Linear(in_features=512, out_features=512, bias=True)
 (out_drop): Dropout(p=0.0, inplace=False)
 )
 (ls_1): Identity()
 (ln_2): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
 (mlp): Sequential(
 (c_fc): Linear(in_features=512, out_features=2048, bias=True)
 (gelu): GELU(approximate='none')
 (c_proj): Linear(in_features=2048, out_features=512, bias=True)
 )
 (ls_2): Identity()
 )
 )
 )
 (token_embedding): Embedding(49408, 512)
 (ln_final): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
)
2026-05-28,06:29:25 | INFO | open_clip_train.main | Params:
2026-05-28,06:29:25 | INFO | open_clip_train.main | accum_freq: 1
2026-05-28,06:29:25 | INFO | open_clip_train.main | audio_ext: flac
2026-05-28,06:29:25 | INFO | open_clip_train.main | audio_fill: repeatpad
2026-05-28,06:29:25 | INFO | open_clip_train.main | audio_fusion: False
2026-05-28,06:29:25 | INFO | open_clip_train.main | audio_int16_normalize: False
2026-05-28,06:29:25 | INFO | open_clip_train.main | audio_trunc: rand_trunc
2026-05-28,06:29:25 | INFO | open_clip_train.main | audio_zeroshot_audio_key: audio
2026-05-28,06:29:25 | INFO | open_clip_train.main | audio_zeroshot_class_key: category
2026-05-28,06:29:25 | INFO | open_clip_train.main | audio_zeroshot_dataset: None
2026-05-28,06:29:25 | INFO | open_clip_train.main | audio_zeroshot_multiprocessing_context: forkserver
2026-05-28,06:29:25 | INFO | open_clip_train.main | audio_zeroshot_split: train
2026-05-28,06:29:25 | INFO | open_clip_train.main | audio_zeroshot_target_key: target
2026-05-28,06:29:25 | INFO | open_clip_train.main | audio_zeroshot_templates: None
2026-05-28,06:29:25 | INFO | open_clip_train.main | audio_zeroshot_workers: 0
2026-05-28,06:29:25 | INFO | open_clip_train.main | aug_cfg: {}
2026-05-28,06:29:25 | INFO | open_clip_train.main | batch_size: 4096
2026-05-28,06:29:25 | INFO | open_clip_train.main | beta1: 0.9
2026-05-28,06:29:25 | INFO | open_clip_train.main | beta2: 0.98
2026-05-28,06:29:25 | INFO | open_clip_train.main | cache_dir: None
2026-05-28,06:29:25 | INFO | open_clip_train.main | checkpoint_path: ./logs/ViT-B-32-Vanilla-resume/checkpoints
2026-05-28,06:29:25 | INFO | open_clip_train.main | coca_caption_loss_weight: 2.0
2026-05-28,06:29:25 | INFO | open_clip_train.main | coca_contrastive_loss_weight: 1.0
2026-05-28,06:29:25 | INFO | open_clip_train.main | copy_codebase: False
2026-05-28,06:29:25 | INFO | open_clip_train.main | csv_caption_key: title
2026-05-28,06:29:25 | INFO | open_clip_train.main | csv_img_key: filepath
2026-05-28,06:29:25 | INFO | open_clip_train.main | csv_separator:
2026-05-28,06:29:25 | INFO | open_clip_train.main | dataset_resampled: False
2026-05-28,06:29:25 | INFO | open_clip_train.main | dataset_type: webdataset
2026-05-28,06:29:25 | INFO | open_clip_train.main | ddp_static_graph: False
2026-05-28,06:29:25 | INFO | open_clip_train.main | debug: False
2026-05-28,06:29:25 | INFO | open_clip_train.main | delete_previous_checkpoint: False
2026-05-28,06:29:25 | INFO | open_clip_train.main | device: cuda:0
2026-05-28,06:29:25 | INFO | open_clip_train.main | dist_backend: None
2026-05-28,06:29:25 | INFO | open_clip_train.main | dist_url: None
2026-05-28,06:29:25 | INFO | open_clip_train.main | distill: False
2026-05-28,06:29:25 | INFO | open_clip_train.main | distill_model: None
2026-05-28,06:29:25 | INFO | open_clip_train.main | distill_pretrained: None
2026-05-28,06:29:25 | INFO | open_clip_train.main | distributed: True
2026-05-28,06:29:25 | INFO | open_clip_train.main | epochs: 32
2026-05-28,06:29:25 | INFO | open_clip_train.main | epochs_cooldown: None
2026-05-28,06:29:25 | INFO | open_clip_train.main | eps: 1e-06
2026-05-28,06:29:25 | INFO | open_clip_train.main | force_context_length: None
2026-05-28,06:29:25 | INFO | open_clip_train.main | force_custom_text: False
2026-05-28,06:29:25 | INFO | open_clip_train.main | force_image_size: None
2026-05-28,06:29:25 | INFO | open_clip_train.main | force_naflex_vision: False
2026-05-28,06:29:25 | INFO | open_clip_train.main | force_patch_dropout: None
2026-05-28,06:29:25 | INFO | open_clip_train.main | force_quick_gelu: False
2026-05-28,06:29:25 | INFO | open_clip_train.main | fsdp: False
2026-05-28,06:29:25 | INFO | open_clip_train.main | fsdp_checkpoint: full
2026-05-28,06:29:25 | INFO | open_clip_train.main | fsdp_no_reshard_after_forward: False
2026-05-28,06:29:25 | INFO | open_clip_train.main | fsdp_offload_cpu: False
2026-05-28,06:29:25 | INFO | open_clip_train.main | gather_with_grad: True
2026-05-28,06:29:25 | INFO | open_clip_train.main | grad_checkpointing: False
2026-05-28,06:29:25 | INFO | open_clip_train.main | grad_clip_norm: None
2026-05-28,06:29:25 | INFO | open_clip_train.main | image_interpolation: None
2026-05-28,06:29:25 | INFO | open_clip_train.main | image_mean: None
2026-05-28,06:29:25 | INFO | open_clip_train.main | image_resize_mode: None
2026-05-28,06:29:25 | INFO | open_clip_train.main | image_std: None
2026-05-28,06:29:25 | INFO | open_clip_train.main | imagenet_v2: None
2026-05-28,06:29:25 | INFO | open_clip_train.main | imagenet_val: None
2026-05-28,06:29:25 | INFO | open_clip_train.main | local_loss: True
2026-05-28,06:29:25 | INFO | open_clip_train.main | local_rank: 0
2026-05-28,06:29:25 | INFO | open_clip_train.main | lock_image: False
2026-05-28,06:29:25 | INFO | open_clip_train.main | lock_image_freeze_bn_stats: False
2026-05-28,06:29:25 | INFO | open_clip_train.main | lock_image_unlocked_groups: 0
2026-05-28,06:29:25 | INFO | open_clip_train.main | lock_text: False
2026-05-28,06:29:25 | INFO | open_clip_train.main | lock_text_freeze_layer_norm: False
2026-05-28,06:29:25 | INFO | open_clip_train.main | lock_text_unlocked_layers: 0
2026-05-28,06:29:25 | INFO | open_clip_train.main | log_every_n_steps: 100
2026-05-28,06:29:25 | INFO | open_clip_train.main | log_level: 20
2026-05-28,06:29:25 | INFO | open_clip_train.main | log_local: False
2026-05-28,06:29:25 | INFO | open_clip_train.main | log_path: ./logs/ViT-B-32-Vanilla-resume/out.log
2026-05-28,06:29:25 | INFO | open_clip_train.main | logs: ./logs/
2026-05-28,06:29:25 | INFO | open_clip_train.main | loss_dist_impl: None
2026-05-28,06:29:25 | INFO | open_clip_train.main | lr: 0.0005
2026-05-28,06:29:25 | INFO | open_clip_train.main | lr_cooldown_end: 0.0
2026-05-28,06:29:25 | INFO | open_clip_train.main | lr_cooldown_power: 1.0
2026-05-28,06:29:25 | INFO | open_clip_train.main | lr_scheduler: cosine
2026-05-28,06:29:25 | INFO | open_clip_train.main | model: ViT-B-32
2026-05-28,06:29:25 | INFO | open_clip_train.main | momentum: None
2026-05-28,06:29:25 | INFO | open_clip_train.main | naflex_batch_divisor: 8
2026-05-28,06:29:25 | INFO | open_clip_train.main | naflex_loss_scale: none
2026-05-28,06:29:25 | INFO | open_clip_train.main | naflex_max_image_tokens_per_batch: 16384
2026-05-28,06:29:25 | INFO | open_clip_train.main | naflex_num_train_image_tokens: None
2026-05-28,06:29:25 | INFO | open_clip_train.main | naflex_patch_size_probs: None
2026-05-28,06:29:25 | INFO | open_clip_train.main | naflex_patch_sizes: None
2026-05-28,06:29:25 | INFO | open_clip_train.main | naflex_seq_lens: None
2026-05-28,06:29:25 | INFO | open_clip_train.main | name: ViT-B-32-Vanilla-resume
2026-05-28,06:29:25 | INFO | open_clip_train.main | no_set_device_rank: False
2026-05-28,06:29:25 | INFO | open_clip_train.main | opt: adamw
2026-05-28,06:29:25 | INFO | open_clip_train.main | precision: amp_bfloat16
2026-05-28,06:29:25 | INFO | open_clip_train.main | pretrained:
2026-05-28,06:29:25 | INFO | open_clip_train.main | pretrained_audio: None
2026-05-28,06:29:25 | INFO | open_clip_train.main | pretrained_image: False
2026-05-28,06:29:25 | INFO | open_clip_train.main | rank: 0
2026-05-28,06:29:25 | INFO | open_clip_train.main | remote_sync: None
2026-05-28,06:29:25 | INFO | open_clip_train.main | remote_sync_frequency: 300
2026-05-28,06:29:25 | INFO | open_clip_train.main | remote_sync_protocol: s3
2026-05-28,06:29:25 | INFO | open_clip_train.main | report_to: wandb
2026-05-28,06:29:25 | INFO | open_clip_train.main | resume: /scratch/work/zhul2/code/open_clip/logs/ViT-B-32-Vanilla/checkpoints/epoch_24.pt
2026-05-28,06:29:25 | INFO | open_clip_train.main | save_frequency: 1
2026-05-28,06:29:25 | INFO | open_clip_train.main | save_most_recent: False
2026-05-28,06:29:25 | INFO | open_clip_train.main | seed: 0
2026-05-28,06:29:25 | INFO | open_clip_train.main | siglip: False
2026-05-28,06:29:25 | INFO | open_clip_train.main | skip_scheduler: False
2026-05-28,06:29:25 | INFO | open_clip_train.main | tensorboard: False
2026-05-28,06:29:25 | INFO | open_clip_train.main | tensorboard_path:
2026-05-28,06:29:25 | INFO | open_clip_train.main | torchcompile: True
2026-05-28,06:29:25 | INFO | open_clip_train.main | torchcompile_backend: None
2026-05-28,06:29:25 | INFO | open_clip_train.main | torchcompile_mode: None
2026-05-28,06:29:25 | INFO | open_clip_train.main | torchcompile_strategy: task
2026-05-28,06:29:25 | INFO | open_clip_train.main | train_data: /scratch/shareddata/dldata/laion400M/img2dataset/laion400m-data/{00000..41407}.tar
2026-05-28,06:29:25 | INFO | open_clip_train.main | train_data_upsampling_factors: None
2026-05-28,06:29:25 | INFO | open_clip_train.main | train_num_samples: 268836185
2026-05-28,06:29:25 | INFO | open_clip_train.main | use_bn_sync: False
2026-05-28,06:29:25 | INFO | open_clip_train.main | use_bnb_linear: None
2026-05-28,06:29:25 | INFO | open_clip_train.main | use_naflex: False
2026-05-28,06:29:25 | INFO | open_clip_train.main | val_data: None
2026-05-28,06:29:25 | INFO | open_clip_train.main | val_frequency: 1
2026-05-28,06:29:25 | INFO | open_clip_train.main | val_num_samples: None
2026-05-28,06:29:25 | INFO | open_clip_train.main | wandb: True
2026-05-28,06:29:25 | INFO | open_clip_train.main | wandb_notes:
2026-05-28,06:29:25 | INFO | open_clip_train.main | wandb_project_name: open-clip
2026-05-28,06:29:25 | INFO | open_clip_train.main | warmup: 2000
2026-05-28,06:29:25 | INFO | open_clip_train.main | wd: 0.2
2026-05-28,06:29:25 | INFO | open_clip_train.main | workers: 12
2026-05-28,06:29:25 | INFO | open_clip_train.main | world_size: 8
2026-05-28,06:29:25 | INFO | open_clip_train.main | zeroshot_frequency: 2
2026-05-28,06:29:26 | INFO | open_clip_train.main | Created AdamW (adamw) optimizer: lr: 0.0005, betas: (0.9, 0.98), eps: 1e-06, weight_decay: 0.2, amsgrad: False, maximize: False, foreach: None, capturable: False, differentiable: False, fused: None, decoupled_weight_decay: True
2026-05-28,06:29:31 | INFO | open_clip.task.checkpoint | => resuming checkpoint '/scratch/work/zhul2/code/open_clip/logs/ViT-B-32-Vanilla/checkpoints/epoch_24.pt' (epoch 24)
2026-05-28,06:29:31 | INFO | open_clip.factory | Parsing tokenizer identifier. Schema: None, Identifier: ViT-B-32
2026-05-28,06:29:31 | INFO | open_clip.factory | Attempting to load config from built-in: ViT-B-32
2026-05-28,06:29:31 | INFO | open_clip.factory | Using default SimpleTokenizer.
2026-05-28,06:29:33 | INFO | open_clip_train.main | Using torch.compile strategy=task.
2026-05-28,06:29:33 | INFO | open_clip_train.main | Compiling task train/eval forward callables.
2026-05-28,06:29:34 | INFO | open_clip_train.main | Start epoch 24
2026-05-28,06:30:39 | INFO | open_clip_train.train | Train Epoch: 24 [ 32768/268959744 (0%)] Data (t): 20.699 Batch (t): 65.026, 503.924/s, 62.9905/s/gpu LR: 0.000074 Logit Scale: 100.000 Contrastive_loss: 0.46160 (0.46160) Loss: 0.46160 (0.46160) Logit_scale: 100.00 (100.00)
2026-05-28,06:32:59 | INFO | open_clip_train.train | Train Epoch: 24 [ 3309568/268959744 (1%)] Data (t): 0.679 Batch (t): 1.402, 23142.2/s, 2892.78/s/gpu LR: 0.000074 Logit Scale: 100.000 Contrastive_loss: 0.46411 (0.46285) Loss: 0.46411 (0.46285) Logit_scale: 100.00 (100.00)
2026-05-28,06:35:39 | INFO | open_clip_train.train | Train Epoch: 24 [ 6586368/268959744 (2%)] Data (t): 0.666 Batch (t): 1.598, 23092.5/s, 2886.56/s/gpu LR: 0.000074 Logit Scale: 100.000 Contrastive_loss: 0.45716 (0.46096) Loss: 0.45716 (0.46096) Logit_scale: 100.00 (100.00)
2026-05-28,06:38:09 | INFO | open_clip_train.train | Train Epoch: 24 [ 9863168/268959744 (4%)] Data (t): 0.654 Batch (t): 1.506, 23477.2/s, 2934.65/s/gpu LR: 0.000074 Logit Scale: 100.000 Contrastive_loss: 0.42256 (0.45136) Loss: 0.42256 (0.45136) Logit_scale: 100.00 (100.00)
2026-05-28,06:40:48 | INFO | open_clip_train.train | Train Epoch: 24 [ 13139968/268959744 (5%)] Data (t): 0.620 Batch (t): 1.582, 22619.7/s, 2827.47/s/gpu LR: 0.000073 Logit Scale: 100.000 Contrastive_loss: 0.45424 (0.45193) Loss: 0.45424 (0.45193) Logit_scale: 100.00 (100.00)
2026-05-28,06:43:23 | INFO | open_clip_train.train | Train Epoch: 24 [ 16416768/268959744 (6%)] Data (t): 0.688 Batch (t): 1.551, 22733.4/s, 2841.67/s/gpu LR: 0.000073 Logit Scale: 100.000 Contrastive_loss: 0.43631 (0.44933) Loss: 0.43631 (0.44933) Logit_scale: 100.00 (100.00)
2026-05-28,06:45:56 | INFO | open_clip_train.train | Train Epoch: 24 [ 19693568/268959744 (7%)] Data (t): 0.705 Batch (t): 1.534, 22049.6/s, 2756.20/s/gpu LR: 0.000073 Logit Scale: 100.000 Contrastive_loss: 0.49138 (0.45534) Loss: 0.49138 (0.45534) Logit_scale: 100.00 (100.00)
2026-05-28,06:48:30 | INFO | open_clip_train.train | Train Epoch: 24 [ 22970368/268959744 (9%)] Data (t): 0.705 Batch (t): 1.534, 22035.7/s, 2754.46/s/gpu LR: 0.000073 Logit Scale: 100.000 Contrastive_loss: 0.46737 (0.45684) Loss: 0.46737 (0.45684) Logit_scale: 100.00 (100.00)
2026-05-28,06:51:04 | INFO | open_clip_train.train | Train Epoch: 24 [ 26247168/268959744 (10%)] Data (t): 0.669 Batch (t): 1.545, 22540.5/s, 2817.56/s/gpu LR: 0.000073 Logit Scale: 100.000 Contrastive_loss: 0.45181 (0.45628) Loss: 0.45181 (0.45628) Logit_scale: 100.00 (100.00)
2026-05-28,06:53:42 | INFO | open_clip_train.train | Train Epoch: 24 [ 29523968/268959744 (11%)] Data (t): 0.656 Batch (t): 1.578, 21686.5/s, 2710.82/s/gpu LR: 0.000072 Logit Scale: 100.000 Contrastive_loss: 0.48239 (0.45889) Loss: 0.48239 (0.45889) Logit_scale: 100.00 (100.00)
2026-05-28,06:56:14 | INFO | open_clip_train.train | Train Epoch: 24 [ 32800768/268959744 (12%)] Data (t): 0.722 Batch (t): 1.522, 22214.2/s, 2776.78/s/gpu LR: 0.000072 Logit Scale: 100.000 Contrastive_loss: 0.43631 (0.45684) Loss: 0.43631 (0.45684) Logit_scale: 100.00 (100.00)
2026-05-28,06:58:48 | INFO | open_clip_train.train | Train Epoch: 24 [ 36077568/268959744 (13%)] Data (t): 0.677 Batch (t): 1.533, 21274.9/s, 2659.36/s/gpu LR: 0.000072 Logit Scale: 100.000 Contrastive_loss: 0.43587 (0.45509) Loss: 0.43587 (0.45509) Logit_scale: 100.00 (100.00)
2026-05-28,07:01:22 | INFO | open_clip_train.train | Train Epoch: 24 [ 39354368/268959744 (15%)] Data (t): 0.726 Batch (t): 1.548, 24419.7/s, 3052.47/s/gpu LR: 0.000072 Logit Scale: 99.998 Contrastive_loss: 0.45330 (0.45495) Loss: 0.45330 (0.45495) Logit_scale: 99.998 (100.00)
2026-05-28,07:03:58 | INFO | open_clip_train.train | Train Epoch: 24 [ 42631168/268959744 (16%)] Data (t): 0.763 Batch (t): 1.557, 21446.6/s, 2680.82/s/gpu LR: 0.000072 Logit Scale: 100.000 Contrastive_loss: 0.45321 (0.45483) Loss: 0.45321 (0.45483) Logit_scale: 100.00 (100.00)
2026-05-28,07:06:38 | INFO | open_clip_train.train | Train Epoch: 24 [ 45907968/268959744 (17%)] Data (t): 0.701 Batch (t): 1.596, 20512.0/s, 2564.00/s/gpu LR: 0.000071 Logit Scale: 100.000 Contrastive_loss: 0.46041 (0.45520) Loss: 0.46041 (0.45520) Logit_scale: 100.00 (100.00)
2026-05-28,07:09:06 | INFO | open_clip_train.train | Train Epoch: 24 [ 49184768/268959744 (18%)] Data (t): 0.834 Batch (t): 1.482, 22800.7/s, 2850.09/s/gpu LR: 0.000071 Logit Scale: 100.000 Contrastive_loss: 0.45347 (0.45509) Loss: 0.45347 (0.45509) Logit_scale: 100.00 (100.00)
2026-05-28,07:11:46 | INFO | open_clip_train.train | Train Epoch: 24 [ 52461568/268959744 (20%)] Data (t): 0.741 Batch (t): 1.598, 20628.6/s, 2578.58/s/gpu LR: 0.000071 Logit Scale: 100.000 Contrastive_loss: 0.46585 (0.45573) Loss: 0.46585 (0.45573) Logit_scale: 100.00 (100.00)
2026-05-28,07:14:14 | INFO | open_clip_train.train | Train Epoch: 24 [ 55738368/268959744 (21%)] Data (t): 0.789 Batch (t): 1.485, 21620.3/s, 2702.54/s/gpu LR: 0.000071 Logit Scale: 100.000 Contrastive_loss: 0.48650 (0.45744) Loss: 0.48650 (0.45744) Logit_scale: 100.00 (100.00)
2026-05-28,07:16:52 | INFO | open_clip_train.train | Train Epoch: 24 [ 59015168/268959744 (22%)] Data (t): 0.762 Batch (t): 1.576, 22084.5/s, 2760.56/s/gpu LR: 0.000070 Logit Scale: 100.000 Contrastive_loss: 0.49197 (0.45925) Loss: 0.49197 (0.45925) Logit_scale: 100.00 (100.00)
2026-05-28,07:19:31 | INFO | open_clip_train.train | Train Epoch: 24 [ 62291968/268959744 (23%)] Data (t): 0.745 Batch (t): 1.594, 21702.4/s, 2712.80/s/gpu LR: 0.000070 Logit Scale: 100.000 Contrastive_loss: 0.48334 (0.46046) Loss: 0.48334 (0.46046) Logit_scale: 100.00 (100.00)
2026-05-28,07:22:05 | INFO | open_clip_train.train | Train Epoch: 24 [ 65568768/268959744 (24%)] Data (t): 0.756 Batch (t): 1.538, 21600.6/s, 2700.07/s/gpu LR: 0.000070 Logit Scale: 100.000 Contrastive_loss: 0.46647 (0.46074) Loss: 0.46647 (0.46074) Logit_scale: 100.00 (100.00)
2026-05-28,07:24:43 | INFO | open_clip_train.train | Train Epoch: 24 [ 68845568/268959744 (26%)] Data (t): 0.763 Batch (t): 1.583, 20262.4/s, 2532.80/s/gpu LR: 0.000070 Logit Scale: 100.000 Contrastive_loss: 0.45814 (0.46063) Loss: 0.45814 (0.46063) Logit_scale: 100.00 (100.00)
2026-05-28,07:27:21 | INFO | open_clip_train.train | Train Epoch: 24 [ 72122368/268959744 (27%)] Data (t): 0.726 Batch (t): 1.582, 24001.6/s, 3000.20/s/gpu LR: 0.000070 Logit Scale: 100.000 Contrastive_loss: 0.45405 (0.46034) Loss: 0.45405 (0.46034) Logit_scale: 100.00 (100.00)
2026-05-28,07:29:49 | INFO | open_clip_train.train | Train Epoch: 24 [ 75399168/268959744 (28%)] Data (t): 0.825 Batch (t): 1.474, 24378.9/s, 3047.37/s/gpu LR: 0.000069 Logit Scale: 100.000 Contrastive_loss: 0.48451 (0.46135) Loss: 0.48451 (0.46135) Logit_scale: 100.00 (100.00)
2026-05-28,07:32:31 | INFO | open_clip_train.train | Train Epoch: 24 [ 78675968/268959744 (29%)] Data (t): 0.755 Batch (t): 1.620, 21286.5/s, 2660.81/s/gpu LR: 0.000069 Logit Scale: 100.000 Contrastive_loss: 0.42946 (0.46007) Loss: 0.42946 (0.46007) Logit_scale: 100.00 (100.00)
2026-05-28,07:35:09 | INFO | open_clip_train.train | Train Epoch: 24 [ 81952768/268959744 (30%)] Data (t): 0.731 Batch (t): 1.583, 20461.6/s, 2557.69/s/gpu LR: 0.000069 Logit Scale: 100.000 Contrastive_loss: 0.46643 (0.46032) Loss: 0.46643 (0.46032) Logit_scale: 100.00 (100.00)
2026-05-28,07:37:40 | INFO | open_clip_train.train | Train Epoch: 24 [ 85229568/268959744 (32%)] Data (t): 0.781 Batch (t): 1.507, 21722.2/s, 2715.27/s/gpu LR: 0.000069 Logit Scale: 100.000 Contrastive_loss: 0.43997 (0.45956) Loss: 0.43997 (0.45956) Logit_scale: 100.00 (100.00)
2026-05-28,07:40:19 | INFO | open_clip_train.train | Train Epoch: 24 [ 88506368/268959744 (33%)] Data (t): 0.754 Batch (t): 1.589, 21773.5/s, 2721.69/s/gpu LR: 0.000069 Logit Scale: 100.000 Contrastive_loss: 0.44846 (0.45917) Loss: 0.44846 (0.45917) Logit_scale: 100.00 (100.00)
2026-05-28,07:42:56 | INFO | open_clip_train.train | Train Epoch: 24 [ 91783168/268959744 (34%)] Data (t): 0.706 Batch (t): 1.579, 20241.0/s, 2530.13/s/gpu LR: 0.000068 Logit Scale: 100.000 Contrastive_loss: 0.47274 (0.45963) Loss: 0.47274 (0.45963) Logit_scale: 100.00 (100.00)
2026-05-28,07:45:29 | INFO | open_clip_train.train | Train Epoch: 24 [ 95059968/268959744 (35%)] Data (t): 0.831 Batch (t): 1.526, 21371.1/s, 2671.39/s/gpu LR: 0.000068 Logit Scale: 100.000 Contrastive_loss: 0.46034 (0.45966) Loss: 0.46034 (0.45966) Logit_scale: 100.00 (100.00)
2026-05-28,07:48:02 | INFO | open_clip_train.train | Train Epoch: 24 [ 98336768/268959744 (37%)] Data (t): 0.801 Batch (t): 1.533, 24091.3/s, 3011.41/s/gpu LR: 0.000068 Logit Scale: 100.000 Contrastive_loss: 0.49241 (0.46071) Loss: 0.49241 (0.46071) Logit_scale: 100.00 (100.00)
2026-05-28,07:50:42 | INFO | open_clip_train.train | Train Epoch: 24 [101613568/268959744 (38%)] Data (t): 0.817 Batch (t): 1.592, 20943.7/s, 2617.96/s/gpu LR: 0.000068 Logit Scale: 100.000 Contrastive_loss: 0.43627 (0.45995) Loss: 0.43627 (0.45995) Logit_scale: 100.00 (100.00)
2026-05-28,07:53:18 | INFO | open_clip_train.train | Train Epoch: 24 [104890368/268959744 (39%)] Data (t): 0.707 Batch (t): 1.560, 9454.76/s, 1181.85/s/gpu LR: 0.000068 Logit Scale: 100.000 Contrastive_loss: 0.44784 (0.45958) Loss: 0.44784 (0.45958) Logit_scale: 100.00 (100.00)
2026-05-28,07:55:57 | INFO | open_clip_train.train | Train Epoch: 24 [108167168/268959744 (40%)] Data (t): 0.727 Batch (t): 1.590, 21109.4/s, 2638.68/s/gpu LR: 0.000067 Logit Scale: 100.000 Contrastive_loss: 0.43875 (0.45897) Loss: 0.43875 (0.45897) Logit_scale: 100.00 (100.00)
2026-05-28,07:58:29 | INFO | open_clip_train.train | Train Epoch: 24 [111443968/268959744 (41%)] Data (t): 0.819 Batch (t): 1.525, 22965.5/s, 2870.69/s/gpu LR: 0.000067 Logit Scale: 100.000 Contrastive_loss: 0.46502 (0.45914) Loss: 0.46502 (0.45914) Logit_scale: 100.00 (100.00)
2026-05-28,08:01:09 | INFO | open_clip_train.train | Train Epoch: 24 [114720768/268959744 (43%)] Data (t): 0.763 Batch (t): 1.598, 20545.4/s, 2568.18/s/gpu LR: 0.000067 Logit Scale: 100.000 Contrastive_loss: 0.41565 (0.45793) Loss: 0.41565 (0.45793) Logit_scale: 100.00 (100.00)
2026-05-28,08:03:48 | INFO | open_clip_train.train | Train Epoch: 24 [117997568/268959744 (44%)] Data (t): 0.708 Batch (t): 1.588, 20231.3/s, 2528.91/s/gpu LR: 0.000067 Logit Scale: 100.000 Contrastive_loss: 0.40935 (0.45662) Loss: 0.40935 (0.45662) Logit_scale: 100.00 (100.00)
2026-05-28,08:06:25 | INFO | open_clip_train.train | Train Epoch: 24 [121274368/268959744 (45%)] Data (t): 0.738 Batch (t): 1.576, 21369.1/s, 2671.14/s/gpu LR: 0.000067 Logit Scale: 100.000 Contrastive_loss: 0.43994 (0.45618) Loss: 0.43994 (0.45618) Logit_scale: 100.00 (100.00)
2026-05-28,08:09:00 | INFO | open_clip_train.train | Train Epoch: 24 [124551168/268959744 (46%)] Data (t): 0.748 Batch (t): 1.547, 21398.6/s, 2674.82/s/gpu LR: 0.000066 Logit Scale: 100.000 Contrastive_loss: 0.45253 (0.45609) Loss: 0.45253 (0.45609) Logit_scale: 100.00 (100.00)
2026-05-28,08:11:32 | INFO | open_clip_train.train | Train Epoch: 24 [127827968/268959744 (48%)] Data (t): 0.746 Batch (t): 1.518, 21506.7/s, 2688.34/s/gpu LR: 0.000066 Logit Scale: 100.000 Contrastive_loss: 0.43730 (0.45562) Loss: 0.43730 (0.45562) Logit_scale: 100.00 (100.00)
2026-05-28,08:14:13 | INFO | open_clip_train.train | Train Epoch: 24 [131104768/268959744 (49%)] Data (t): 0.809 Batch (t): 1.613, 20107.1/s, 2513.38/s/gpu LR: 0.000066 Logit Scale: 100.000 Contrastive_loss: 0.45211 (0.45553) Loss: 0.45211 (0.45553) Logit_scale: 100.00 (100.00)
2026-05-28,08:16:52 | INFO | open_clip_train.train | Train Epoch: 24 [134381568/268959744 (50%)] Data (t): 0.751 Batch (t): 1.592, 20594.1/s, 2574.26/s/gpu LR: 0.000066 Logit Scale: 100.000 Contrastive_loss: 0.43590 (0.45507) Loss: 0.43590 (0.45507) Logit_scale: 100.00 (100.00)
2026-05-28,08:19:22 | INFO | open_clip_train.train | Train Epoch: 24 [137658368/268959744 (51%)] Data (t): 0.750 Batch (t): 1.501, 20639.1/s, 2579.89/s/gpu LR: 0.000066 Logit Scale: 100.000 Contrastive_loss: 0.40255 (0.45384) Loss: 0.40255 (0.45384) Logit_scale: 100.00 (100.00)
2026-05-28,08:22:02 | INFO | open_clip_train.train | Train Epoch: 24 [140935168/268959744 (52%)] Data (t): 0.718 Batch (t): 1.593, 21067.5/s, 2633.43/s/gpu LR: 0.000065 Logit Scale: 100.000 Contrastive_loss: 0.40915 (0.45283) Loss: 0.40915 (0.45283) Logit_scale: 100.00 (100.00)
2026-05-28,08:24:44 | INFO | open_clip_train.train | Train Epoch: 24 [144211968/268959744 (54%)] Data (t): 0.689 Batch (t): 1.623, 20789.9/s, 2598.74/s/gpu LR: 0.000065 Logit Scale: 100.000 Contrastive_loss: 0.49230 (0.45371) Loss: 0.49230 (0.45371) Logit_scale: 100.00 (100.00)
2026-05-28,08:27:17 | INFO | open_clip_train.train | Train Epoch: 24 [147488768/268959744 (55%)] Data (t): 0.744 Batch (t): 1.527, 21556.3/s, 2694.54/s/gpu LR: 0.000065 Logit Scale: 100.000 Contrastive_loss: 0.43753 (0.45335) Loss: 0.43753 (0.45335) Logit_scale: 100.00 (100.00)
2026-05-28,08:29:54 | INFO | open_clip_train.train | Train Epoch: 24 [150765568/268959744 (56%)] Data (t): 0.770 Batch (t): 1.576, 20262.0/s, 2532.75/s/gpu LR: 0.000065 Logit Scale: 100.000 Contrastive_loss: 0.41887 (0.45262) Loss: 0.41887 (0.45262) Logit_scale: 100.00 (100.00)
2026-05-28,08:32:33 | INFO | open_clip_train.train | Train Epoch: 24 [154042368/268959744 (57%)] Data (t): 0.755 Batch (t): 1.587, 20372.7/s, 2546.59/s/gpu LR: 0.000065 Logit Scale: 99.998 Contrastive_loss: 0.46771 (0.45294) Loss: 0.46771 (0.45294) Logit_scale: 99.998 (100.00)
2026-05-28,08:35:03 | INFO | open_clip_train.train | Train Epoch: 24 [157319168/268959744 (58%)] Data (t): 0.815 Batch (t): 1.499, 21240.2/s, 2655.03/s/gpu LR: 0.000064 Logit Scale: 100.000 Contrastive_loss: 0.42112 (0.45229) Loss: 0.42112 (0.45229) Logit_scale: 100.00 (100.00)
2026-05-28,08:37:39 | INFO | open_clip_train.train | Train Epoch: 24 [160595968/268959744 (60%)] Data (t): 0.732 Batch (t): 1.562, 15675.4/s, 1959.42/s/gpu LR: 0.000064 Logit Scale: 100.000 Contrastive_loss: 0.47447 (0.45273) Loss: 0.47447 (0.45273) Logit_scale: 100.00 (100.00)
2026-05-28,08:40:21 | INFO | open_clip_train.train | Train Epoch: 24 [163872768/268959744 (61%)] Data (t): 0.735 Batch (t): 1.616, 20005.7/s, 2500.71/s/gpu LR: 0.000064 Logit Scale: 100.000 Contrastive_loss: 0.42817 (0.45225) Loss: 0.42817 (0.45225) Logit_scale: 100.00 (100.00)
2026-05-28,08:43:00 | INFO | open_clip_train.train | Train Epoch: 24 [167149568/268959744 (62%)] Data (t): 0.774 Batch (t): 1.593, 20368.2/s, 2546.03/s/gpu LR: 0.000064 Logit Scale: 100.000 Contrastive_loss: 0.43634 (0.45194) Loss: 0.43634 (0.45194) Logit_scale: 100.00 (100.00)
2026-05-28,08:45:32 | INFO | open_clip_train.train | Train Epoch: 24 [170426368/268959744 (63%)] Data (t): 0.745 Batch (t): 1.522, 24247.5/s, 3030.93/s/gpu LR: 0.000063 Logit Scale: 100.000 Contrastive_loss: 0.44985 (0.45190) Loss: 0.44985 (0.45190) Logit_scale: 100.00 (100.00)
2026-05-28,08:48:07 | INFO | open_clip_train.train | Train Epoch: 24 [173703168/268959744 (65%)] Data (t): 0.763 Batch (t): 1.545, 12937.2/s, 1617.14/s/gpu LR: 0.000063 Logit Scale: 100.000 Contrastive_loss: 0.41325 (0.45119) Loss: 0.41325 (0.45119) Logit_scale: 100.00 (100.00)
2026-05-28,08:50:48 | INFO | open_clip_train.train | Train Epoch: 24 [176979968/268959744 (66%)] Data (t): 0.755 Batch (t): 1.609, 8172.18/s, 1021.52/s/gpu LR: 0.000063 Logit Scale: 100.000 Contrastive_loss: 0.45013 (0.45117) Loss: 0.45013 (0.45117) Logit_scale: 100.00 (100.00)
2026-05-28,08:53:26 | INFO | open_clip_train.train | Train Epoch: 24 [180256768/268959744 (67%)] Data (t): 0.736 Batch (t): 1.581, 20437.7/s, 2554.71/s/gpu LR: 0.000063 Logit Scale: 100.000 Contrastive_loss: 0.45506 (0.45124) Loss: 0.45506 (0.45124) Logit_scale: 100.00 (100.00)
2026-05-28,08:56:02 | INFO | open_clip_train.train | Train Epoch: 24 [183533568/268959744 (68%)] Data (t): 0.759 Batch (t): 1.560, 19933.1/s, 2491.64/s/gpu LR: 0.000063 Logit Scale: 99.999 Contrastive_loss: 0.41740 (0.45064) Loss: 0.41740 (0.45064) Logit_scale: 99.999 (100.00)
2026-05-28,08:58:40 | INFO | open_clip_train.train | Train Epoch: 24 [186810368/268959744 (69%)] Data (t): 0.739 Batch (t): 1.580, 21435.6/s, 2679.45/s/gpu LR: 0.000062 Logit Scale: 99.999 Contrastive_loss: 0.45089 (0.45065) Loss: 0.45089 (0.45065) Logit_scale: 99.999 (100.00)
2026-05-28,09:01:16 | INFO | open_clip_train.train | Train Epoch: 24 [190087168/268959744 (71%)] Data (t): 0.789 Batch (t): 1.560, 23672.2/s, 2959.02/s/gpu LR: 0.000062 Logit Scale: 100.000 Contrastive_loss: 0.44772 (0.45060) Loss: 0.44772 (0.45060) Logit_scale: 100.00 (100.00)
2026-05-28,09:03:43 | INFO | open_clip_train.train | Train Epoch: 24 [193363968/268959744 (72%)] Data (t): 0.794 Batch (t): 1.476, 26743.3/s, 3342.91/s/gpu LR: 0.000062 Logit Scale: 100.000 Contrastive_loss: 0.42787 (0.45022) Loss: 0.42787 (0.45022) Logit_scale: 100.00 (100.00)
2026-05-28,09:06:24 | INFO | open_clip_train.train | Train Epoch: 24 [196640768/268959744 (73%)] Data (t): 0.792 Batch (t): 1.611, 21779.9/s, 2722.49/s/gpu LR: 0.000062 Logit Scale: 100.000 Contrastive_loss: 0.40929 (0.44955) Loss: 0.40929 (0.44955) Logit_scale: 100.00 (100.00)
2026-05-28,09:09:00 | INFO | open_clip_train.train | Train Epoch: 24 [199917568/268959744 (74%)] Data (t): 0.773 Batch (t): 1.554, 21449.1/s, 2681.14/s/gpu LR: 0.000062 Logit Scale: 100.000 Contrastive_loss: 0.43123 (0.44925) Loss: 0.43123 (0.44925) Logit_scale: 100.00 (100.00)
2026-05-28,09:11:41 | INFO | open_clip_train.train | Train Epoch: 24 [203194368/268959744 (76%)] Data (t): 0.723 Batch (t): 1.609, 21048.8/s, 2631.10/s/gpu LR: 0.000062 Logit Scale: 100.000 Contrastive_loss: 0.47119 (0.44960) Loss: 0.47119 (0.44960) Logit_scale: 100.00 (100.00)
2026-05-28,09:14:12 | INFO | open_clip_train.train | Train Epoch: 24 [206471168/268959744 (77%)] Data (t): 0.802 Batch (t): 1.514, 22097.8/s, 2762.22/s/gpu LR: 0.000061 Logit Scale: 99.999 Contrastive_loss: 0.44569 (0.44954) Loss: 0.44569 (0.44954) Logit_scale: 99.999 (100.00)
2026-05-28,09:16:50 | INFO | open_clip_train.train | Train Epoch: 24 [209747968/268959744 (78%)] Data (t): 0.775 Batch (t): 1.580, 20469.5/s, 2558.69/s/gpu LR: 0.000061 Logit Scale: 100.000 Contrastive_loss: 0.43849 (0.44937) Loss: 0.43849 (0.44937) Logit_scale: 100.00 (100.00)
2026-05-28,09:19:30 | INFO | open_clip_train.train | Train Epoch: 24 [213024768/268959744 (79%)] Data (t): 0.761 Batch (t): 1.595, 20914.2/s, 2614.28/s/gpu LR: 0.000061 Logit Scale: 100.000 Contrastive_loss: 0.43572 (0.44916) Loss: 0.43572 (0.44916) Logit_scale: 100.00 (100.00)
2026-05-28,09:21:58 | INFO | open_clip_train.train | Train Epoch: 24 [216301568/268959744 (80%)] Data (t): 0.751 Batch (t): 1.482, 22466.7/s, 2808.33/s/gpu LR: 0.000061 Logit Scale: 100.000 Contrastive_loss: 0.41743 (0.44869) Loss: 0.41743 (0.44869) Logit_scale: 100.00 (100.00)
2026-05-28,09:24:33 | INFO | open_clip_train.train | Train Epoch: 24 [219578368/268959744 (82%)] Data (t): 0.753 Batch (t): 1.556, 20832.9/s, 2604.11/s/gpu LR: 0.000061 Logit Scale: 100.000 Contrastive_loss: 0.44844 (0.44869) Loss: 0.44844 (0.44869) Logit_scale: 100.00 (100.00)
2026-05-28,09:27:07 | INFO | open_clip_train.train | Train Epoch: 24 [222855168/268959744 (83%)] Data (t): 0.744 Batch (t): 1.532, 9579.27/s, 1197.41/s/gpu LR: 0.000060 Logit Scale: 100.000 Contrastive_loss: 0.39893 (0.44796) Loss: 0.39893 (0.44796) Logit_scale: 100.00 (100.00)
2026-05-28,09:29:47 | INFO | open_clip_train.train | Train Epoch: 24 [226131968/268959744 (84%)] Data (t): 0.750 Batch (t): 1.604, 20641.9/s, 2580.24/s/gpu LR: 0.000060 Logit Scale: 100.000 Contrastive_loss: 0.43153 (0.44773) Loss: 0.43153 (0.44773) Logit_scale: 100.00 (100.00)
2026-05-28,09:32:23 | INFO | open_clip_train.train | Train Epoch: 24 [229408768/268959744 (85%)] Data (t): 0.752 Batch (t): 1.557, 20914.8/s, 2614.35/s/gpu LR: 0.000060 Logit Scale: 99.999 Contrastive_loss: 0.43799 (0.44759) Loss: 0.43799 (0.44759) Logit_scale: 99.999 (100.00)
2026-05-28,09:34:55 | INFO | open_clip_train.train | Train Epoch: 24 [232685568/268959744 (87%)] Data (t): 0.747 Batch (t): 1.520, 19917.5/s, 2489.69/s/gpu LR: 0.000060 Logit Scale: 100.000 Contrastive_loss: 0.38734 (0.44676) Loss: 0.38734 (0.44676) Logit_scale: 100.00 (100.00)
2026-05-28,09:37:31 | INFO | open_clip_train.train | Train Epoch: 24 [235962368/268959744 (88%)] Data (t): 0.770 Batch (t): 1.561, 21594.5/s, 2699.31/s/gpu LR: 0.000060 Logit Scale: 100.000 Contrastive_loss: 0.40956 (0.44625) Loss: 0.40956 (0.44625) Logit_scale: 100.00 (100.00)
2026-05-28,09:40:06 | INFO | open_clip_train.train | Train Epoch: 24 [239239168/268959744 (89%)] Data (t): 0.805 Batch (t): 1.554, 21008.5/s, 2626.06/s/gpu LR: 0.000059 Logit Scale: 100.000 Contrastive_loss: 0.44231 (0.44619) Loss: 0.44231 (0.44619) Logit_scale: 100.00 (100.00)
2026-05-28,09:42:46 | INFO | open_clip_train.train | Train Epoch: 24 [242515968/268959744 (90%)] Data (t): 0.852 Batch (t): 1.598, 21033.3/s, 2629.16/s/gpu LR: 0.000059 Logit Scale: 100.000 Contrastive_loss: 0.44000 (0.44611) Loss: 0.44000 (0.44611) Logit_scale: 100.00 (100.00)
2026-05-28,09:45:22 | INFO | open_clip_train.train | Train Epoch: 24 [245792768/268959744 (91%)] Data (t): 0.832 Batch (t): 1.554, 21010.2/s, 2626.27/s/gpu LR: 0.000059 Logit Scale: 100.000 Contrastive_loss: 0.41917 (0.44576) Loss: 0.41917 (0.44576) Logit_scale: 100.00 (100.00)
2026-05-28,09:48:00 | INFO | open_clip_train.train | Train Epoch: 24 [249069568/268959744 (93%)] Data (t): 0.781 Batch (t): 1.586, 21965.9/s, 2745.74/s/gpu LR: 0.000059 Logit Scale: 100.000 Contrastive_loss: 0.42621 (0.44550) Loss: 0.42621 (0.44550) Logit_scale: 100.00 (100.00)
2026-05-28,09:50:37 | INFO | open_clip_train.train | Train Epoch: 24 [252346368/268959744 (94%)] Data (t): 0.732 Batch (t): 1.572, 24355.8/s, 3044.47/s/gpu LR: 0.000059 Logit Scale: 100.000 Contrastive_loss: 0.40840 (0.44503) Loss: 0.40840 (0.44503) Logit_scale: 100.00 (100.00)
2026-05-28,09:53:10 | INFO | open_clip_train.train | Train Epoch: 24 [255623168/268959744 (95%)] Data (t): 0.808 Batch (t): 1.531, 21565.4/s, 2695.68/s/gpu LR: 0.000058 Logit Scale: 100.000 Contrastive_loss: 0.40298 (0.44449) Loss: 0.40298 (0.44449) Logit_scale: 100.00 (100.00)
2026-05-28,09:55:45 | INFO | open_clip_train.train | Train Epoch: 24 [258899968/268959744 (96%)] Data (t): 0.809 Batch (t): 1.546, 21141.9/s, 2642.74/s/gpu LR: 0.000058 Logit Scale: 100.000 Contrastive_loss: 0.43832 (0.44442) Loss: 0.43832 (0.44442) Logit_scale: 100.00 (100.00)
2026-05-28,09:58:20 | INFO | open_clip_train.train | Train Epoch: 24 [262176768/268959744 (97%)] Data (t): 0.745 Batch (t): 1.552, 20903.7/s, 2612.96/s/gpu LR: 0.000058 Logit Scale: 100.000 Contrastive_loss: 0.48457 (0.44491) Loss: 0.48457 (0.44491) Logit_scale: 100.00 (100.00)
2026-05-28,10:00:59 | INFO | open_clip_train.train | Train Epoch: 24 [265453568/268959744 (99%)] Data (t): 0.819 Batch (t): 1.587, 20888.6/s, 2611.07/s/gpu LR: 0.000058 Logit Scale: 100.000 Contrastive_loss: 0.49469 (0.44552) Loss: 0.49469 (0.44552) Logit_scale: 100.00 (100.00)
2026-05-28,10:03:28 | INFO | open_clip_train.train | Train Epoch: 24 [268730368/268959744 (100%)] Data (t): 0.815 Batch (t): 1.486, 29048.1/s, 3631.01/s/gpu LR: 0.000058 Logit Scale: 100.000 Contrastive_loss: 0.42852 (0.44532) Loss: 0.42852 (0.44532) Logit_scale: 100.00 (100.00)
2026-05-28,10:03:35 | INFO | open_clip_train.train | Train Epoch: 24 [268959744/268959744 (100%)] Data (t): 0.634 Batch (t): 1.122, 29884.7/s, 3735.59/s/gpu LR: 0.000058 Logit Scale: 99.995 Contrastive_loss: 0.46738 (0.44558) Loss: 0.46738 (0.44558) Logit_scale: 99.995 (100.00)
2026-05-28,10:03:40 | INFO | open_clip_train.main | Start epoch 25
2026-05-28,10:04:02 | INFO | open_clip_train.train | Train Epoch: 25 [ 32768/268959744 (0%)] Data (t): 15.053 Batch (t): 21.341, 1535.47/s, 191.934/s/gpu LR: 0.000058 Logit Scale: 99.994 Contrastive_loss: 0.42066 (0.42066) Loss: 0.42066 (0.42066) Logit_scale: 99.994 (99.994)
2026-05-28,10:06:43 | INFO | open_clip_train.train | Train Epoch: 25 [ 3309568/268959744 (1%)] Data (t): 0.703 Batch (t): 1.614, 23396.3/s, 2924.54/s/gpu LR: 0.000057 Logit Scale: 100.000 Contrastive_loss: 0.42281 (0.42174) Loss: 0.42281 (0.42174) Logit_scale: 100.00 (99.997)
2026-05-28,10:09:22 | INFO | open_clip_train.train | Train Epoch: 25 [ 6586368/268959744 (2%)] Data (t): 0.850 Batch (t): 1.592, 20892.7/s, 2611.58/s/gpu LR: 0.000057 Logit Scale: 99.999 Contrastive_loss: 0.44483 (0.42943) Loss: 0.44483 (0.42943) Logit_scale: 99.999 (99.998)
2026-05-28,10:11:59 | INFO | open_clip_train.train | Train Epoch: 25 [ 9863168/268959744 (4%)] Data (t): 0.793 Batch (t): 1.564, 20860.7/s, 2607.58/s/gpu LR: 0.000057 Logit Scale: 100.000 Contrastive_loss: 0.42111 (0.42735) Loss: 0.42111 (0.42735) Logit_scale: 100.00 (99.998)
2026-05-28,10:14:32 | INFO | open_clip_train.train | Train Epoch: 25 [ 13139968/268959744 (5%)] Data (t): 0.823 Batch (t): 1.535, 20970.6/s, 2621.32/s/gpu LR: 0.000057 Logit Scale: 100.000 Contrastive_loss: 0.43133 (0.42815) Loss: 0.43133 (0.42815) Logit_scale: 100.00 (99.999)
2026-05-28,10:17:09 | INFO | open_clip_train.train | Train Epoch: 25 [ 16416768/268959744 (6%)] Data (t): 0.746 Batch (t): 1.565, 20228.4/s, 2528.55/s/gpu LR: 0.000057 Logit Scale: 100.000 Contrastive_loss: 0.40710 (0.42464) Loss: 0.40710 (0.42464) Logit_scale: 100.00 (99.999)
2026-05-28,10:19:47 | INFO | open_clip_train.train | Train Epoch: 25 [ 19693568/268959744 (7%)] Data (t): 0.733 Batch (t): 1.579, 21119.2/s, 2639.90/s/gpu LR: 0.000056 Logit Scale: 100.000 Contrastive_loss: 0.43909 (0.42670) Loss: 0.43909 (0.42670) Logit_scale: 100.00 (99.999)
2026-05-28,10:22:23 | INFO | open_clip_train.train | Train Epoch: 25 [ 22970368/268959744 (9%)] Data (t): 0.758 Batch (t): 1.561, 21282.0/s, 2660.24/s/gpu LR: 0.000056 Logit Scale: 100.000 Contrastive_loss: 0.42566 (0.42657) Loss: 0.42566 (0.42657) Logit_scale: 100.00 (99.999)
2026-05-28,10:24:59 | INFO | open_clip_train.train | Train Epoch: 25 [ 26247168/268959744 (10%)] Data (t): 0.752 Batch (t): 1.567, 22221.7/s, 2777.71/s/gpu LR: 0.000056 Logit Scale: 100.000 Contrastive_loss: 0.42087 (0.42594) Loss: 0.42087 (0.42594) Logit_scale: 100.00 (99.999)
2026-05-28,10:27:31 | INFO | open_clip_train.train | Train Epoch: 25 [ 29523968/268959744 (11%)] Data (t): 0.770 Batch (t): 1.517, 22021.4/s, 2752.68/s/gpu LR: 0.000056 Logit Scale: 100.000 Contrastive_loss: 0.42173 (0.42552) Loss: 0.42173 (0.42552) Logit_scale: 100.00 (99.999)
2026-05-28,10:30:12 | INFO | open_clip_train.train | Train Epoch: 25 [ 32800768/268959744 (12%)] Data (t): 0.694 Batch (t): 1.612, 21225.5/s, 2653.19/s/gpu LR: 0.000056 Logit Scale: 100.000 Contrastive_loss: 0.44655 (0.42743) Loss: 0.44655 (0.42743) Logit_scale: 100.00 (99.999)
2026-05-28,10:32:48 | INFO | open_clip_train.train | Train Epoch: 25 [ 36077568/268959744 (13%)] Data (t): 0.719 Batch (t): 1.559, 20878.0/s, 2609.75/s/gpu LR: 0.000055 Logit Scale: 100.000 Contrastive_loss: 0.40997 (0.42598) Loss: 0.40997 (0.42598) Logit_scale: 100.00 (99.999)
2026-05-28,10:35:22 | INFO | open_clip_train.train | Train Epoch: 25 [ 39354368/268959744 (15%)] Data (t): 0.779 Batch (t): 1.536, 21831.8/s, 2728.97/s/gpu LR: 0.000055 Logit Scale: 100.000 Contrastive_loss: 0.40582 (0.42443) Loss: 0.40582 (0.42443) Logit_scale: 100.00 (99.999)
2026-05-28,10:37:59 | INFO | open_clip_train.train | Train Epoch: 25 [ 42631168/268959744 (16%)] Data (t): 0.673 Batch (t): 1.570, 21112.0/s, 2639.00/s/gpu LR: 0.000055 Logit Scale: 100.000 Contrastive_loss: 0.41049 (0.42343) Loss: 0.41049 (0.42343) Logit_scale: 100.00 (100.00)
2026-05-28,10:40:32 | INFO | open_clip_train.train | Train Epoch: 25 [ 45907968/268959744 (17%)] Data (t): 0.691 Batch (t): 1.531, 21648.8/s, 2706.10/s/gpu LR: 0.000055 Logit Scale: 100.000 Contrastive_loss: 0.39976 (0.42185) Loss: 0.39976 (0.42185) Logit_scale: 100.00 (100.00)
2026-05-28,10:43:09 | INFO | open_clip_train.train | Train Epoch: 25 [ 49184768/268959744 (18%)] Data (t): 0.724 Batch (t): 1.568, 21339.6/s, 2667.45/s/gpu LR: 0.000055 Logit Scale: 100.000 Contrastive_loss: 0.43398 (0.42261) Loss: 0.43398 (0.42261) Logit_scale: 100.00 (100.00)
2026-05-28,10:45:44 | INFO | open_clip_train.train | Train Epoch: 25 [ 52461568/268959744 (20%)] Data (t): 0.743 Batch (t): 1.555, 24418.4/s, 3052.30/s/gpu LR: 0.000055 Logit Scale: 100.000 Contrastive_loss: 0.40437 (0.42154) Loss: 0.40437 (0.42154) Logit_scale: 100.00 (100.00)
2026-05-28,10:48:18 | INFO | open_clip_train.train | Train Epoch: 25 [ 55738368/268959744 (21%)] Data (t): 0.765 Batch (t): 1.541, 20427.1/s, 2553.39/s/gpu LR: 0.000054 Logit Scale: 100.000 Contrastive_loss: 0.40659 (0.42071) Loss: 0.40659 (0.42071) Logit_scale: 100.00 (100.00)
2026-05-28,10:50:58 | INFO | open_clip_train.train | Train Epoch: 25 [ 59015168/268959744 (22%)] Data (t): 0.685 Batch (t): 1.597, 21293.1/s, 2661.63/s/gpu LR: 0.000054 Logit Scale: 100.000 Contrastive_loss: 0.40350 (0.41980) Loss: 0.40350 (0.41980) Logit_scale: 100.00 (100.00)
2026-05-28,10:53:34 | INFO | open_clip_train.train | Train Epoch: 25 [ 62291968/268959744 (23%)] Data (t): 0.748 Batch (t): 1.558, 20758.1/s, 2594.76/s/gpu LR: 0.000054 Logit Scale: 100.000 Contrastive_loss: 0.41513 (0.41957) Loss: 0.41513 (0.41957) Logit_scale: 100.00 (100.00)
2026-05-28,10:56:10 | INFO | open_clip_train.train | Train Epoch: 25 [ 65568768/268959744 (24%)] Data (t): 0.729 Batch (t): 1.559, 23222.2/s, 2902.78/s/gpu LR: 0.000054 Logit Scale: 100.000 Contrastive_loss: 0.37479 (0.41744) Loss: 0.37479 (0.41744) Logit_scale: 100.00 (100.00)
2026-05-28,10:58:39 | INFO | open_clip_train.train | Train Epoch: 25 [ 68845568/268959744 (26%)] Data (t): 0.691 Batch (t): 1.489, 21080.6/s, 2635.08/s/gpu LR: 0.000054 Logit Scale: 100.000 Contrastive_loss: 0.41830 (0.41747) Loss: 0.41830 (0.41747) Logit_scale: 100.00 (100.00)
2026-05-28,11:01:19 | INFO | open_clip_train.train | Train Epoch: 25 [ 72122368/268959744 (27%)] Data (t): 0.711 Batch (t): 1.608, 20000.9/s, 2500.11/s/gpu LR: 0.000053 Logit Scale: 100.000 Contrastive_loss: 0.42203 (0.41767) Loss: 0.42203 (0.41767) Logit_scale: 100.00 (100.00)
2026-05-28,11:03:56 | INFO | open_clip_train.train | Train Epoch: 25 [ 75399168/268959744 (28%)] Data (t): 0.745 Batch (t): 1.561, 20864.9/s, 2608.11/s/gpu LR: 0.000053 Logit Scale: 100.000 Contrastive_loss: 0.39095 (0.41656) Loss: 0.39095 (0.41656) Logit_scale: 100.00 (100.00)
2026-05-28,11:06:29 | INFO | open_clip_train.train | Train Epoch: 25 [ 78675968/268959744 (29%)] Data (t): 0.729 Batch (t): 1.533, 20782.0/s, 2597.75/s/gpu LR: 0.000053 Logit Scale: 100.000 Contrastive_loss: 0.41185 (0.41637) Loss: 0.41185 (0.41637) Logit_scale: 100.00 (100.00)
2026-05-28,11:09:02 | INFO | open_clip_train.train | Train Epoch: 25 [ 81952768/268959744 (30%)] Data (t): 0.733 Batch (t): 1.531, 21760.4/s, 2720.05/s/gpu LR: 0.000053 Logit Scale: 100.000 Contrastive_loss: 0.40763 (0.41604) Loss: 0.40763 (0.41604) Logit_scale: 100.00 (100.00)
2026-05-28,11:11:32 | INFO | open_clip_train.train | Train Epoch: 25 [ 85229568/268959744 (32%)] Data (t): 0.769 Batch (t): 1.497, 8533.92/s, 1066.74/s/gpu LR: 0.000053 Logit Scale: 100.000 Contrastive_loss: 0.40512 (0.41563) Loss: 0.40512 (0.41563) Logit_scale: 100.00 (100.00)
2026-05-28,11:14:13 | INFO | open_clip_train.train | Train Epoch: 25 [ 88506368/268959744 (33%)] Data (t): 0.658 Batch (t): 1.611, 20702.8/s, 2587.85/s/gpu LR: 0.000052 Logit Scale: 100.000 Contrastive_loss: 0.41062 (0.41545) Loss: 0.41062 (0.41545) Logit_scale: 100.00 (100.00)
2026-05-28,11:16:50 | INFO | open_clip_train.train | Train Epoch: 25 [ 91783168/268959744 (34%)] Data (t): 0.694 Batch (t): 1.576, 19798.8/s, 2474.86/s/gpu LR: 0.000052 Logit Scale: 100.000 Contrastive_loss: 0.39898 (0.41488) Loss: 0.39898 (0.41488) Logit_scale: 100.00 (100.00)
2026-05-28,11:19:27 | INFO | open_clip_train.train | Train Epoch: 25 [ 95059968/268959744 (35%)] Data (t): 0.684 Batch (t): 1.569, 21018.4/s, 2627.30/s/gpu LR: 0.000052 Logit Scale: 100.000 Contrastive_loss: 0.40426 (0.41453) Loss: 0.40426 (0.41453) Logit_scale: 100.00 (100.00)
2026-05-28,11:22:01 | INFO | open_clip_train.train | Train Epoch: 25 [ 98336768/268959744 (37%)] Data (t): 0.704 Batch (t): 1.542, 21759.0/s, 2719.88/s/gpu LR: 0.000052 Logit Scale: 100.000 Contrastive_loss: 0.38654 (0.41363) Loss: 0.38654 (0.41363) Logit_scale: 100.00 (100.00)
2026-05-28,11:24:34 | INFO | open_clip_train.train | Train Epoch: 25 [101613568/268959744 (38%)] Data (t): 0.727 Batch (t): 1.524, 22039.0/s, 2754.87/s/gpu LR: 0.000052 Logit Scale: 100.000 Contrastive_loss: 0.40030 (0.41321) Loss: 0.40030 (0.41321) Logit_scale: 100.00 (100.00)
2026-05-28,11:27:15 | INFO | open_clip_train.train | Train Epoch: 25 [104890368/268959744 (39%)] Data (t): 0.661 Batch (t): 1.612, 20102.2/s, 2512.77/s/gpu LR: 0.000052 Logit Scale: 100.000 Contrastive_loss: 0.41241 (0.41319) Loss: 0.41241 (0.41319) Logit_scale: 100.00 (100.00)
2026-05-28,11:29:43 | INFO | open_clip_train.train | Train Epoch: 25 [108167168/268959744 (40%)] Data (t): 0.784 Batch (t): 1.484, 22193.4/s, 2774.18/s/gpu LR: 0.000051 Logit Scale: 100.000 Contrastive_loss: 0.43298 (0.41377) Loss: 0.43298 (0.41377) Logit_scale: 100.00 (100.00)
2026-05-28,11:32:22 | INFO | open_clip_train.train | Train Epoch: 25 [111443968/268959744 (41%)] Data (t): 0.706 Batch (t): 1.584, 21122.7/s, 2640.34/s/gpu LR: 0.000051 Logit Scale: 100.000 Contrastive_loss: 0.38773 (0.41302) Loss: 0.38773 (0.41302) Logit_scale: 100.00 (100.00)
2026-05-28,11:34:58 | INFO | open_clip_train.train | Train Epoch: 25 [114720768/268959744 (43%)] Data (t): 0.739 Batch (t): 1.557, 23129.6/s, 2891.19/s/gpu LR: 0.000051 Logit Scale: 100.000 Contrastive_loss: 0.40466 (0.41279) Loss: 0.40466 (0.41279) Logit_scale: 100.00 (100.00)
2026-05-28,11:37:30 | INFO | open_clip_train.train | Train Epoch: 25 [117997568/268959744 (44%)] Data (t): 0.713 Batch (t): 1.520, 26087.6/s, 3260.95/s/gpu LR: 0.000051 Logit Scale: 100.000 Contrastive_loss: 0.38182 (0.41195) Loss: 0.38182 (0.41195) Logit_scale: 100.00 (100.00)
2026-05-28,11:40:07 | INFO | open_clip_train.train | Train Epoch: 25 [121274368/268959744 (45%)] Data (t): 0.770 Batch (t): 1.571, 24605.5/s, 3075.69/s/gpu LR: 0.000051 Logit Scale: 100.000 Contrastive_loss: 0.38023 (0.41112) Loss: 0.38023 (0.41112) Logit_scale: 100.00 (100.00)
2026-05-28,11:42:44 | INFO | open_clip_train.train | Train Epoch: 25 [124551168/268959744 (46%)] Data (t): 0.682 Batch (t): 1.578, 22215.9/s, 2776.99/s/gpu LR: 0.000050 Logit Scale: 100.000 Contrastive_loss: 0.39103 (0.41060) Loss: 0.39103 (0.41060) Logit_scale: 100.00 (100.00)
2026-05-28,11:45:21 | INFO | open_clip_train.train | Train Epoch: 25 [127827968/268959744 (48%)] Data (t): 0.688 Batch (t): 1.562, 20955.8/s, 2619.48/s/gpu LR: 0.000050 Logit Scale: 100.000 Contrastive_loss: 0.39469 (0.41021) Loss: 0.39469 (0.41021) Logit_scale: 100.00 (100.00)
2026-05-28,11:47:50 | INFO | open_clip_train.train | Train Epoch: 25 [131104768/268959744 (49%)] Data (t): 0.787 Batch (t): 1.496, 21518.2/s, 2689.78/s/gpu LR: 0.000050 Logit Scale: 100.000 Contrastive_loss: 0.43462 (0.41080) Loss: 0.43462 (0.41080) Logit_scale: 100.00 (100.00)
2026-05-28,11:50:33 | INFO | open_clip_train.train | Train Epoch: 25 [134381568/268959744 (50%)] Data (t): 0.687 Batch (t): 1.627, 21310.4/s, 2663.81/s/gpu LR: 0.000050 Logit Scale: 100.000 Contrastive_loss: 0.41067 (0.41080) Loss: 0.41067 (0.41080) Logit_scale: 100.00 (100.00)
2026-05-28,11:53:08 | INFO | open_clip_train.train | Train Epoch: 25 [137658368/268959744 (51%)] Data (t): 0.722 Batch (t): 1.555, 21144.5/s, 2643.07/s/gpu LR: 0.000050 Logit Scale: 100.000 Contrastive_loss: 0.42018 (0.41102) Loss: 0.42018 (0.41102) Logit_scale: 100.00 (100.00)
2026-05-28,11:55:43 | INFO | open_clip_train.train | Train Epoch: 25 [140935168/268959744 (52%)] Data (t): 0.728 Batch (t): 1.546, 21888.0/s, 2736.00/s/gpu LR: 0.000050 Logit Scale: 100.000 Contrastive_loss: 0.41304 (0.41106) Loss: 0.41304 (0.41106) Logit_scale: 100.00 (100.00)
2026-05-28,11:58:15 | INFO | open_clip_train.train | Train Epoch: 25 [144211968/268959744 (54%)] Data (t): 0.832 Batch (t): 1.521, 21766.2/s, 2720.77/s/gpu LR: 0.000049 Logit Scale: 100.000 Contrastive_loss: 0.41891 (0.41124) Loss: 0.41891 (0.41124) Logit_scale: 100.00 (100.00)
2026-05-28,12:00:52 | INFO | open_clip_train.train | Train Epoch: 25 [147488768/268959744 (55%)] Data (t): 0.677 Batch (t): 1.572, 22508.4/s, 2813.55/s/gpu LR: 0.000049 Logit Scale: 100.000 Contrastive_loss: 0.40912 (0.41119) Loss: 0.40912 (0.41119) Logit_scale: 100.00 (100.00)
2026-05-28,12:03:29 | INFO | open_clip_train.train | Train Epoch: 25 [150765568/268959744 (56%)] Data (t): 0.626 Batch (t): 1.563, 21011.4/s, 2626.42/s/gpu LR: 0.000049 Logit Scale: 100.000 Contrastive_loss: 0.39037 (0.41075) Loss: 0.39037 (0.41075) Logit_scale: 100.00 (100.00)
2026-05-28,12:06:04 | INFO | open_clip_train.train | Train Epoch: 25 [154042368/268959744 (57%)] Data (t): 0.692 Batch (t): 1.555, 21246.5/s, 2655.82/s/gpu LR: 0.000049 Logit Scale: 100.000 Contrastive_loss: 0.42263 (0.41100) Loss: 0.42263 (0.41100) Logit_scale: 100.00 (100.00)
2026-05-28,12:08:35 | INFO | open_clip_train.train | Train Epoch: 25 [157319168/268959744 (58%)] Data (t): 0.734 Batch (t): 1.512, 21803.4/s, 2725.43/s/gpu LR: 0.000049 Logit Scale: 100.000 Contrastive_loss: 0.37556 (0.41027) Loss: 0.37556 (0.41027) Logit_scale: 100.00 (100.00)
2026-05-28,12:11:13 | INFO | open_clip_train.train | Train Epoch: 25 [160595968/268959744 (60%)] Data (t): 0.670 Batch (t): 1.573, 21794.5/s, 2724.32/s/gpu LR: 0.000049 Logit Scale: 100.000 Contrastive_loss: 0.41621 (0.41039) Loss: 0.41621 (0.41039) Logit_scale: 100.00 (100.00)
2026-05-28,12:13:49 | INFO | open_clip_train.train | Train Epoch: 25 [163872768/268959744 (61%)] Data (t): 0.669 Batch (t): 1.563, 20077.3/s, 2509.66/s/gpu LR: 0.000048 Logit Scale: 100.000 Contrastive_loss: 0.43902 (0.41095) Loss: 0.43902 (0.41095) Logit_scale: 100.00 (100.00)
2026-05-28,12:16:21 | INFO | open_clip_train.train | Train Epoch: 25 [167149568/268959744 (62%)] Data (t): 0.686 Batch (t): 1.524, 21572.8/s, 2696.59/s/gpu LR: 0.000048 Logit Scale: 100.000 Contrastive_loss: 0.42833 (0.41129) Loss: 0.42833 (0.41129) Logit_scale: 100.00 (100.00)
2026-05-28,12:19:00 | INFO | open_clip_train.train | Train Epoch: 25 [170426368/268959744 (63%)] Data (t): 0.743 Batch (t): 1.589, 21392.9/s, 2674.12/s/gpu LR: 0.000048 Logit Scale: 100.000 Contrastive_loss: 0.38886 (0.41086) Loss: 0.38886 (0.41086) Logit_scale: 100.00 (100.00)
2026-05-28,12:21:39 | INFO | open_clip_train.train | Train Epoch: 25 [173703168/268959744 (65%)] Data (t): 0.711 Batch (t): 1.590, 20052.3/s, 2506.54/s/gpu LR: 0.000048 Logit Scale: 100.000 Contrastive_loss: 0.41267 (0.41090) Loss: 0.41267 (0.41090) Logit_scale: 100.00 (100.00)
2026-05-28,12:24:08 | INFO | open_clip_train.train | Train Epoch: 25 [176979968/268959744 (66%)] Data (t): 0.769 Batch (t): 1.492, 21884.8/s, 2735.60/s/gpu LR: 0.000048 Logit Scale: 100.000 Contrastive_loss: 0.38657 (0.41046) Loss: 0.38657 (0.41046) Logit_scale: 100.00 (100.00)
2026-05-28,12:26:44 | INFO | open_clip_train.train | Train Epoch: 25 [180256768/268959744 (67%)] Data (t): 0.687 Batch (t): 1.552, 20044.8/s, 2505.60/s/gpu LR: 0.000047 Logit Scale: 100.000 Contrastive_loss: 0.39580 (0.41019) Loss: 0.39580 (0.41019) Logit_scale: 100.00 (100.00)
2026-05-28,12:29:19 | INFO | open_clip_train.train | Train Epoch: 25 [183533568/268959744 (68%)] Data (t): 0.706 Batch (t): 1.551, 20345.0/s, 2543.12/s/gpu LR: 0.000047 Logit Scale: 100.000 Contrastive_loss: 0.37128 (0.40951) Loss: 0.37128 (0.40951) Logit_scale: 100.00 (100.00)
2026-05-28,12:31:58 | INFO | open_clip_train.train | Train Epoch: 25 [186810368/268959744 (69%)] Data (t): 0.703 Batch (t): 1.590, 20698.4/s, 2587.30/s/gpu LR: 0.000047 Logit Scale: 100.000 Contrastive_loss: 0.44072 (0.41005) Loss: 0.44072 (0.41005) Logit_scale: 100.00 (100.00)
2026-05-28,12:34:31 | INFO | open_clip_train.train | Train Epoch: 25 [190087168/268959744 (71%)] Data (t): 0.773 Batch (t): 1.534, 21582.5/s, 2697.81/s/gpu LR: 0.000047 Logit Scale: 100.000 Contrastive_loss: 0.41523 (0.41014) Loss: 0.41523 (0.41014) Logit_scale: 100.00 (100.00)
2026-05-28,12:37:06 | INFO | open_clip_train.train | Train Epoch: 25 [193363968/268959744 (72%)] Data (t): 0.758 Batch (t): 1.552, 21237.6/s, 2654.71/s/gpu LR: 0.000047 Logit Scale: 100.000 Contrastive_loss: 0.39741 (0.40993) Loss: 0.39741 (0.40993) Logit_scale: 100.00 (100.00)
2026-05-28,12:39:39 | INFO | open_clip_train.train | Train Epoch: 25 [196640768/268959744 (73%)] Data (t): 0.656 Batch (t): 1.523, 21455.0/s, 2681.87/s/gpu LR: 0.000047 Logit Scale: 100.000 Contrastive_loss: 0.38632 (0.40954) Loss: 0.38632 (0.40954) Logit_scale: 100.00 (100.00)
2026-05-28,12:42:14 | INFO | open_clip_train.train | Train Epoch: 25 [199917568/268959744 (74%)] Data (t): 0.706 Batch (t): 1.556, 21903.8/s, 2737.98/s/gpu LR: 0.000046 Logit Scale: 100.000 Contrastive_loss: 0.41809 (0.40968) Loss: 0.41809 (0.40968) Logit_scale: 100.00 (100.00)
2026-05-28,12:44:49 | INFO | open_clip_train.train | Train Epoch: 25 [203194368/268959744 (76%)] Data (t): 0.769 Batch (t): 1.551, 20963.2/s, 2620.41/s/gpu LR: 0.000046 Logit Scale: 100.000 Contrastive_loss: 0.40816 (0.40965) Loss: 0.40816 (0.40965) Logit_scale: 100.00 (100.00)
2026-05-28,12:47:23 | INFO | open_clip_train.train | Train Epoch: 25 [206471168/268959744 (77%)] Data (t): 0.776 Batch (t): 1.542, 21705.4/s, 2713.18/s/gpu LR: 0.000046 Logit Scale: 100.000 Contrastive_loss: 0.41007 (0.40966) Loss: 0.41007 (0.40966) Logit_scale: 100.00 (100.00)
2026-05-28,12:49:57 | INFO | open_clip_train.train | Train Epoch: 25 [209747968/268959744 (78%)] Data (t): 0.730 Batch (t): 1.541, 13248.6/s, 1656.07/s/gpu LR: 0.000046 Logit Scale: 100.000 Contrastive_loss: 0.39562 (0.40944) Loss: 0.39562 (0.40944) Logit_scale: 100.00 (100.00)
2026-05-28,12:52:31 | INFO | open_clip_train.train | Train Epoch: 25 [213024768/268959744 (79%)] Data (t): 0.700 Batch (t): 1.538, 21859.4/s, 2732.42/s/gpu LR: 0.000046 Logit Scale: 100.000 Contrastive_loss: 0.42187 (0.40963) Loss: 0.42187 (0.40963) Logit_scale: 100.00 (100.00)
2026-05-28,12:55:04 | INFO | open_clip_train.train | Train Epoch: 25 [216301568/268959744 (80%)] Data (t): 0.768 Batch (t): 1.525, 24063.4/s, 3007.93/s/gpu LR: 0.000046 Logit Scale: 100.000 Contrastive_loss: 0.37757 (0.40915) Loss: 0.37757 (0.40915) Logit_scale: 100.00 (100.00)
2026-05-28,12:57:42 | INFO | open_clip_train.train | Train Epoch: 25 [219578368/268959744 (82%)] Data (t): 0.680 Batch (t): 1.588, 21175.5/s, 2646.94/s/gpu LR: 0.000045 Logit Scale: 100.000 Contrastive_loss: 0.37729 (0.40868) Loss: 0.37729 (0.40868) Logit_scale: 100.00 (100.00)
2026-05-28,13:00:20 | INFO | open_clip_train.train | Train Epoch: 25 [222855168/268959744 (83%)] Data (t): 0.726 Batch (t): 1.572, 20004.5/s, 2500.56/s/gpu LR: 0.000045 Logit Scale: 100.000 Contrastive_loss: 0.39995 (0.40856) Loss: 0.39995 (0.40856) Logit_scale: 100.00 (100.00)
2026-05-28,13:02:56 | INFO | open_clip_train.train | Train Epoch: 25 [226131968/268959744 (84%)] Data (t): 0.710 Batch (t): 1.561, 21328.8/s, 2666.10/s/gpu LR: 0.000045 Logit Scale: 100.000 Contrastive_loss: 0.41111 (0.40859) Loss: 0.41111 (0.40859) Logit_scale: 100.00 (100.00)
2026-05-28,13:05:32 | INFO | open_clip_train.train | Train Epoch: 25 [229408768/268959744 (85%)] Data (t): 0.801 Batch (t): 1.560, 21004.8/s, 2625.61/s/gpu LR: 0.000045 Logit Scale: 100.000 Contrastive_loss: 0.40011 (0.40847) Loss: 0.40011 (0.40847) Logit_scale: 100.00 (100.00)
2026-05-28,13:08:08 | INFO | open_clip_train.train | Train Epoch: 25 [232685568/268959744 (87%)] Data (t): 0.727 Batch (t): 1.561, 20841.1/s, 2605.13/s/gpu LR: 0.000045 Logit Scale: 100.000 Contrastive_loss: 0.41902 (0.40862) Loss: 0.41902 (0.40862) Logit_scale: 100.00 (100.00)
2026-05-28,13:10:44 | INFO | open_clip_train.train | Train Epoch: 25 [235962368/268959744 (88%)] Data (t): 0.702 Batch (t): 1.562, 21616.9/s, 2702.11/s/gpu LR: 0.000044 Logit Scale: 99.999 Contrastive_loss: 0.36174 (0.40798) Loss: 0.36174 (0.40798) Logit_scale: 99.999 (100.00)
2026-05-28,13:13:17 | INFO | open_clip_train.train | Train Epoch: 25 [239239168/268959744 (89%)] Data (t): 0.657 Batch (t): 1.534, 20555.7/s, 2569.46/s/gpu LR: 0.000044 Logit Scale: 100.000 Contrastive_loss: 0.39381 (0.40779) Loss: 0.39381 (0.40779) Logit_scale: 100.00 (100.00)
2026-05-28,13:15:50 | INFO | open_clip_train.train | Train Epoch: 25 [242515968/268959744 (90%)] Data (t): 0.670 Batch (t): 1.528, 21381.2/s, 2672.65/s/gpu LR: 0.000044 Logit Scale: 100.000 Contrastive_loss: 0.39217 (0.40758) Loss: 0.39217 (0.40758) Logit_scale: 100.00 (100.00)
2026-05-28,13:18:27 | INFO | open_clip_train.train | Train Epoch: 25 [245792768/268959744 (91%)] Data (t): 0.715 Batch (t): 1.571, 20223.9/s, 2527.99/s/gpu LR: 0.000044 Logit Scale: 100.000 Contrastive_loss: 0.47215 (0.40843) Loss: 0.47215 (0.40843) Logit_scale: 100.00 (100.00)
2026-05-28,13:21:02 | INFO | open_clip_train.train | Train Epoch: 25 [249069568/268959744 (93%)] Data (t): 0.701 Batch (t): 1.545, 20496.4/s, 2562.06/s/gpu LR: 0.000044 Logit Scale: 100.000 Contrastive_loss: 0.41261 (0.40848) Loss: 0.41261 (0.40848) Logit_scale: 100.00 (100.00)
2026-05-28,13:23:38 | INFO | open_clip_train.train | Train Epoch: 25 [252346368/268959744 (94%)] Data (t): 0.708 Batch (t): 1.562, 24579.4/s, 3072.43/s/gpu LR: 0.000044 Logit Scale: 100.000 Contrastive_loss: 0.40728 (0.40847) Loss: 0.40728 (0.40847) Logit_scale: 100.00 (100.00)
2026-05-28,13:26:11 | INFO | open_clip_train.train | Train Epoch: 25 [255623168/268959744 (95%)] Data (t): 0.755 Batch (t): 1.526, 22411.5/s, 2801.44/s/gpu LR: 0.000043 Logit Scale: 100.000 Contrastive_loss: 0.39813 (0.40834) Loss: 0.39813 (0.40834) Logit_scale: 100.00 (100.00)
2026-05-28,13:28:41 | INFO | open_clip_train.train | Train Epoch: 25 [258899968/268959744 (96%)] Data (t): 0.723 Batch (t): 1.505, 21883.9/s, 2735.48/s/gpu LR: 0.000043 Logit Scale: 99.999 Contrastive_loss: 0.40417 (0.40828) Loss: 0.40417 (0.40828) Logit_scale: 99.999 (100.00)
2026-05-28,13:31:20 | INFO | open_clip_train.train | Train Epoch: 25 [262176768/268959744 (97%)] Data (t): 0.692 Batch (t): 1.589, 22261.6/s, 2782.70/s/gpu LR: 0.000043 Logit Scale: 99.999 Contrastive_loss: 0.39272 (0.40809) Loss: 0.39272 (0.40809) Logit_scale: 99.999 (100.00)
2026-05-28,13:33:54 | INFO | open_clip_train.train | Train Epoch: 25 [265453568/268959744 (99%)] Data (t): 0.795 Batch (t): 1.537, 23985.3/s, 2998.17/s/gpu LR: 0.000043 Logit Scale: 100.000 Contrastive_loss: 0.41937 (0.40823) Loss: 0.41937 (0.40823) Logit_scale: 100.00 (100.00)
2026-05-28,13:36:25 | INFO | open_clip_train.train | Train Epoch: 25 [268730368/268959744 (100%)] Data (t): 0.727 Batch (t): 1.513, 29522.7/s, 3690.33/s/gpu LR: 0.000043 Logit Scale: 100.000 Contrastive_loss: 0.48322 (0.40913) Loss: 0.48322 (0.40913) Logit_scale: 100.00 (100.00)
2026-05-28,13:36:33 | INFO | open_clip_train.train | Train Epoch: 25 [268959744/268959744 (100%)] Data (t): 0.617 Batch (t): 1.131, 28712.3/s, 3589.04/s/gpu LR: 0.000043 Logit Scale: 99.998 Contrastive_loss: 0.46090 (0.40975) Loss: 0.46090 (0.40975) Logit_scale: 99.998 (100.00)
2026-05-28,13:36:37 | INFO | open_clip_train.main | Start epoch 26
2026-05-28,13:37:00 | INFO | open_clip_train.train | Train Epoch: 26 [ 32768/268959744 (0%)] Data (t): 19.770 Batch (t): 22.229, 1474.12/s, 184.265/s/gpu LR: 0.000043 Logit Scale: 99.997 Contrastive_loss: 0.38443 (0.38443) Loss: 0.38443 (0.38443) Logit_scale: 99.997 (99.997)
2026-05-28,13:39:41 | INFO | open_clip_train.train | Train Epoch: 26 [ 3309568/268959744 (1%)] Data (t): 0.700 Batch (t): 1.609, 20700.2/s, 2587.52/s/gpu LR: 0.000043 Logit Scale: 100.000 Contrastive_loss: 0.37687 (0.38065) Loss: 0.37687 (0.38065) Logit_scale: 100.00 (99.999)
2026-05-28,13:42:20 | INFO | open_clip_train.train | Train Epoch: 26 [ 6586368/268959744 (2%)] Data (t): 0.739 Batch (t): 1.597, 21379.5/s, 2672.44/s/gpu LR: 0.000042 Logit Scale: 100.000 Contrastive_loss: 0.41019 (0.39049) Loss: 0.41019 (0.39049) Logit_scale: 100.00 (99.999)
2026-05-28,13:44:54 | INFO | open_clip_train.train | Train Epoch: 26 [ 9863168/268959744 (4%)] Data (t): 0.731 Batch (t): 1.539, 22052.8/s, 2756.60/s/gpu LR: 0.000042 Logit Scale: 100.000 Contrastive_loss: 0.41112 (0.39565) Loss: 0.41112 (0.39565) Logit_scale: 100.00 (99.999)
2026-05-28,13:47:26 | INFO | open_clip_train.train | Train Epoch: 26 [ 13139968/268959744 (5%)] Data (t): 0.757 Batch (t): 1.519, 24704.3/s, 3088.03/s/gpu LR: 0.000042 Logit Scale: 100.000 Contrastive_loss: 0.40691 (0.39790) Loss: 0.40691 (0.39790) Logit_scale: 100.00 (99.999)
2026-05-28,13:49:56 | INFO | open_clip_train.train | Train Epoch: 26 [ 16416768/268959744 (6%)] Data (t): 0.679 Batch (t): 1.497, 24602.8/s, 3075.35/s/gpu LR: 0.000042 Logit Scale: 100.000 Contrastive_loss: 0.35775 (0.39121) Loss: 0.35775 (0.39121) Logit_scale: 100.00 (100.00)
2026-05-28,13:52:36 | INFO | open_clip_train.train | Train Epoch: 26 [ 19693568/268959744 (7%)] Data (t): 0.714 Batch (t): 1.602, 20844.4/s, 2605.55/s/gpu LR: 0.000042 Logit Scale: 100.000 Contrastive_loss: 0.38081 (0.38972) Loss: 0.38081 (0.38972) Logit_scale: 100.00 (100.00)
2026-05-28,13:55:10 | INFO | open_clip_train.train | Train Epoch: 26 [ 22970368/268959744 (9%)] Data (t): 0.684 Batch (t): 1.542, 21749.2/s, 2718.65/s/gpu LR: 0.000042 Logit Scale: 100.000 Contrastive_loss: 0.40670 (0.39184) Loss: 0.40670 (0.39184) Logit_scale: 100.00 (100.00)
2026-05-28,13:57:47 | INFO | open_clip_train.train | Train Epoch: 26 [ 26247168/268959744 (10%)] Data (t): 0.681 Batch (t): 1.569, 20737.6/s, 2592.20/s/gpu LR: 0.000041 Logit Scale: 100.000 Contrastive_loss: 0.38349 (0.39092) Loss: 0.38349 (0.39092) Logit_scale: 100.00 (100.00)
2026-05-28,14:00:25 | INFO | open_clip_train.train | Train Epoch: 26 [ 29523968/268959744 (11%)] Data (t): 0.735 Batch (t): 1.579, 19753.9/s, 2469.24/s/gpu LR: 0.000041 Logit Scale: 100.000 Contrastive_loss: 0.38850 (0.39067) Loss: 0.38850 (0.39067) Logit_scale: 100.00 (100.00)
2026-05-28,14:03:02 | INFO | open_clip_train.train | Train Epoch: 26 [ 32800768/268959744 (12%)] Data (t): 0.745 Batch (t): 1.572, 20954.5/s, 2619.32/s/gpu LR: 0.000041 Logit Scale: 100.000 Contrastive_loss: 0.38049 (0.38975) Loss: 0.38049 (0.38975) Logit_scale: 100.00 (100.00)
2026-05-28,14:05:36 | INFO | open_clip_train.train | Train Epoch: 26 [ 36077568/268959744 (13%)] Data (t): 0.730 Batch (t): 1.537, 21556.6/s, 2694.57/s/gpu LR: 0.000041 Logit Scale: 100.000 Contrastive_loss: 0.39633 (0.39030) Loss: 0.39633 (0.39030) Logit_scale: 100.00 (100.00)
2026-05-28,14:08:09 | INFO | open_clip_train.train | Train Epoch: 26 [ 39354368/268959744 (15%)] Data (t): 0.693 Batch (t): 1.533, 20977.8/s, 2622.22/s/gpu LR: 0.000041 Logit Scale: 100.000 Contrastive_loss: 0.37392 (0.38904) Loss: 0.37392 (0.38904) Logit_scale: 100.00 (100.00)
2026-05-28,14:10:45 | INFO | open_clip_train.train | Train Epoch: 26 [ 42631168/268959744 (16%)] Data (t): 0.732 Batch (t): 1.558, 20486.2/s, 2560.77/s/gpu LR: 0.000041 Logit Scale: 100.000 Contrastive_loss: 0.38446 (0.38871) Loss: 0.38446 (0.38871) Logit_scale: 100.00 (100.00)
2026-05-28,14:13:24 | INFO | open_clip_train.train | Train Epoch: 26 [ 45907968/268959744 (17%)] Data (t): 0.730 Batch (t): 1.592, 20624.9/s, 2578.11/s/gpu LR: 0.000040 Logit Scale: 100.000 Contrastive_loss: 0.35630 (0.38655) Loss: 0.35630 (0.38655) Logit_scale: 100.00 (100.00)
2026-05-28,14:16:00 | INFO | open_clip_train.train | Train Epoch: 26 [ 49184768/268959744 (18%)] Data (t): 0.755 Batch (t): 1.554, 25410.1/s, 3176.27/s/gpu LR: 0.000040 Logit Scale: 100.000 Contrastive_loss: 0.36634 (0.38529) Loss: 0.36634 (0.38529) Logit_scale: 100.00 (100.00)
2026-05-28,14:18:29 | INFO | open_clip_train.train | Train Epoch: 26 [ 52461568/268959744 (20%)] Data (t): 0.671 Batch (t): 1.491, 23185.1/s, 2898.14/s/gpu LR: 0.000040 Logit Scale: 100.000 Contrastive_loss: 0.39493 (0.38585) Loss: 0.39493 (0.38585) Logit_scale: 100.00 (100.00)
2026-05-28,14:21:06 | INFO | open_clip_train.train | Train Epoch: 26 [ 55738368/268959744 (21%)] Data (t): 0.705 Batch (t): 1.573, 21009.3/s, 2626.16/s/gpu LR: 0.000040 Logit Scale: 100.000 Contrastive_loss: 0.38042 (0.38555) Loss: 0.38042 (0.38555) Logit_scale: 100.00 (100.00)
2026-05-28,14:23:43 | INFO | open_clip_train.train | Train Epoch: 26 [ 59015168/268959744 (22%)] Data (t): 0.697 Batch (t): 1.572, 21195.7/s, 2649.46/s/gpu LR: 0.000040 Logit Scale: 100.000 Contrastive_loss: 0.39309 (0.38595) Loss: 0.39309 (0.38595) Logit_scale: 100.00 (100.00)
2026-05-28,14:26:16 | INFO | open_clip_train.train | Train Epoch: 26 [ 62291968/268959744 (23%)] Data (t): 0.647 Batch (t): 1.531, 23656.4/s, 2957.05/s/gpu LR: 0.000040 Logit Scale: 100.000 Contrastive_loss: 0.38516 (0.38591) Loss: 0.38516 (0.38591) Logit_scale: 100.00 (100.00)
2026-05-28,14:28:46 | INFO | open_clip_train.train | Train Epoch: 26 [ 65568768/268959744 (24%)] Data (t): 0.741 Batch (t): 1.501, 24122.5/s, 3015.31/s/gpu LR: 0.000039 Logit Scale: 99.999 Contrastive_loss: 0.38577 (0.38590) Loss: 0.38577 (0.38590) Logit_scale: 99.999 (100.00)
2026-05-28,14:31:26 | INFO | open_clip_train.train | Train Epoch: 26 [ 68845568/268959744 (26%)] Data (t): 0.726 Batch (t): 1.597, 21094.8/s, 2636.85/s/gpu LR: 0.000039 Logit Scale: 100.000 Contrastive_loss: 0.38076 (0.38567) Loss: 0.38076 (0.38567) Logit_scale: 100.00 (100.00)
2026-05-28,14:34:00 | INFO | open_clip_train.train | Train Epoch: 26 [ 72122368/268959744 (27%)] Data (t): 0.737 Batch (t): 1.537, 23958.4/s, 2994.79/s/gpu LR: 0.000039 Logit Scale: 100.000 Contrastive_loss: 0.39689 (0.38616) Loss: 0.39689 (0.38616) Logit_scale: 100.00 (100.00)
2026-05-28,14:36:38 | INFO | open_clip_train.train | Train Epoch: 26 [ 75399168/268959744 (28%)] Data (t): 0.798 Batch (t): 1.585, 21466.3/s, 2683.29/s/gpu LR: 0.000039 Logit Scale: 100.000 Contrastive_loss: 0.36348 (0.38521) Loss: 0.36348 (0.38521) Logit_scale: 100.00 (100.00)
2026-05-28,14:39:15 | INFO | open_clip_train.train | Train Epoch: 26 [ 78675968/268959744 (29%)] Data (t): 0.715 Batch (t): 1.563, 21793.5/s, 2724.19/s/gpu LR: 0.000039 Logit Scale: 100.000 Contrastive_loss: 0.37584 (0.38484) Loss: 0.37584 (0.38484) Logit_scale: 100.00 (100.00)
2026-05-28,14:41:52 | INFO | open_clip_train.train | Train Epoch: 26 [ 81952768/268959744 (30%)] Data (t): 0.686 Batch (t): 1.577, 21437.7/s, 2679.71/s/gpu LR: 0.000039 Logit Scale: 100.000 Contrastive_loss: 0.36360 (0.38402) Loss: 0.36360 (0.38402) Logit_scale: 100.00 (100.00)
2026-05-28,14:44:25 | INFO | open_clip_train.train | Train Epoch: 26 [ 85229568/268959744 (32%)] Data (t): 0.738 Batch (t): 1.529, 22036.4/s, 2754.54/s/gpu LR: 0.000038 Logit Scale: 100.000 Contrastive_loss: 0.38806 (0.38417) Loss: 0.38806 (0.38417) Logit_scale: 100.00 (100.00)
2026-05-28,14:46:58 | INFO | open_clip_train.train | Train Epoch: 26 [ 88506368/268959744 (33%)] Data (t): 0.717 Batch (t): 1.525, 21543.7/s, 2692.97/s/gpu LR: 0.000038 Logit Scale: 100.000 Contrastive_loss: 0.39366 (0.38451) Loss: 0.39366 (0.38451) Logit_scale: 100.00 (100.00)
2026-05-28,14:49:29 | INFO | open_clip_train.train | Train Epoch: 26 [ 91783168/268959744 (34%)] Data (t): 0.738 Batch (t): 1.510, 21663.2/s, 2707.90/s/gpu LR: 0.000038 Logit Scale: 99.999 Contrastive_loss: 0.40646 (0.38527) Loss: 0.40646 (0.38527) Logit_scale: 99.999 (100.00)
2026-05-28,14:52:09 | INFO | open_clip_train.train | Train Epoch: 26 [ 95059968/268959744 (35%)] Data (t): 0.754 Batch (t): 1.600, 20586.4/s, 2573.30/s/gpu LR: 0.000038 Logit Scale: 100.000 Contrastive_loss: 0.37085 (0.38479) Loss: 0.37085 (0.38479) Logit_scale: 100.00 (100.00)
2026-05-28,14:54:46 | INFO | open_clip_train.train | Train Epoch: 26 [ 98336768/268959744 (37%)] Data (t): 0.736 Batch (t): 1.577, 20525.8/s, 2565.73/s/gpu LR: 0.000038 Logit Scale: 100.000 Contrastive_loss: 0.40043 (0.38529) Loss: 0.40043 (0.38529) Logit_scale: 100.00 (100.00)
2026-05-28,14:57:23 | INFO | open_clip_train.train | Train Epoch: 26 [101613568/268959744 (38%)] Data (t): 0.709 Batch (t): 1.568, 20871.3/s, 2608.91/s/gpu LR: 0.000038 Logit Scale: 100.000 Contrastive_loss: 0.37402 (0.38494) Loss: 0.37402 (0.38494) Logit_scale: 100.00 (100.00)
2026-05-28,14:59:56 | INFO | open_clip_train.train | Train Epoch: 26 [104890368/268959744 (39%)] Data (t): 0.725 Batch (t): 1.530, 21107.0/s, 2638.37/s/gpu LR: 0.000038 Logit Scale: 100.000 Contrastive_loss: 0.35704 (0.38409) Loss: 0.35704 (0.38409) Logit_scale: 100.00 (100.00)
2026-05-28,15:02:28 | INFO | open_clip_train.train | Train Epoch: 26 [108167168/268959744 (40%)] Data (t): 0.685 Batch (t): 1.520, 21986.5/s, 2748.32/s/gpu LR: 0.000037 Logit Scale: 100.000 Contrastive_loss: 0.38760 (0.38420) Loss: 0.38760 (0.38420) Logit_scale: 100.00 (100.00)
2026-05-28,15:05:06 | INFO | open_clip_train.train | Train Epoch: 26 [111443968/268959744 (41%)] Data (t): 0.632 Batch (t): 1.581, 20264.0/s, 2533.00/s/gpu LR: 0.000037 Logit Scale: 100.000 Contrastive_loss: 0.37043 (0.38380) Loss: 0.37043 (0.38380) Logit_scale: 100.00 (100.00)
2026-05-28,15:07:40 | INFO | open_clip_train.train | Train Epoch: 26 [114720768/268959744 (43%)] Data (t): 0.704 Batch (t): 1.540, 20706.0/s, 2588.25/s/gpu LR: 0.000037 Logit Scale: 100.000 Contrastive_loss: 0.40354 (0.38435) Loss: 0.40354 (0.38435) Logit_scale: 100.00 (100.00)
2026-05-28,15:10:16 | INFO | open_clip_train.train | Train Epoch: 26 [117997568/268959744 (44%)] Data (t): 0.682 Batch (t): 1.556, 21327.9/s, 2665.98/s/gpu LR: 0.000037 Logit Scale: 100.000 Contrastive_loss: 0.37684 (0.38415) Loss: 0.37684 (0.38415) Logit_scale: 100.00 (100.00)
2026-05-28,15:12:52 | INFO | open_clip_train.train | Train Epoch: 26 [121274368/268959744 (45%)] Data (t): 0.750 Batch (t): 1.558, 22192.2/s, 2774.03/s/gpu LR: 0.000037 Logit Scale: 100.000 Contrastive_loss: 0.39356 (0.38440) Loss: 0.39356 (0.38440) Logit_scale: 100.00 (100.00)
2026-05-28,15:15:27 | INFO | open_clip_train.train | Train Epoch: 26 [124551168/268959744 (46%)] Data (t): 0.744 Batch (t): 1.554, 21689.8/s, 2711.22/s/gpu LR: 0.000037 Logit Scale: 100.000 Contrastive_loss: 0.36602 (0.38392) Loss: 0.36602 (0.38392) Logit_scale: 100.00 (100.00)
2026-05-28,15:18:02 | INFO | open_clip_train.train | Train Epoch: 26 [127827968/268959744 (48%)] Data (t): 0.759 Batch (t): 1.553, 20851.3/s, 2606.41/s/gpu LR: 0.000036 Logit Scale: 100.000 Contrastive_loss: 0.37967 (0.38382) Loss: 0.37967 (0.38382) Logit_scale: 100.00 (100.00)
2026-05-28,15:20:39 | INFO | open_clip_train.train | Train Epoch: 26 [131104768/268959744 (49%)] Data (t): 0.680 Batch (t): 1.571, 20557.6/s, 2569.70/s/gpu LR: 0.000036 Logit Scale: 100.000 Contrastive_loss: 0.36677 (0.38340) Loss: 0.36677 (0.38340) Logit_scale: 100.00 (100.00)
2026-05-28,15:23:15 | INFO | open_clip_train.train | Train Epoch: 26 [134381568/268959744 (50%)] Data (t): 0.720 Batch (t): 1.555, 21543.1/s, 2692.89/s/gpu LR: 0.000036 Logit Scale: 100.000 Contrastive_loss: 0.37814 (0.38328) Loss: 0.37814 (0.38328) Logit_scale: 100.00 (100.00)
2026-05-28,15:25:50 | INFO | open_clip_train.train | Train Epoch: 26 [137658368/268959744 (51%)] Data (t): 0.668 Batch (t): 1.547, 24025.4/s, 3003.18/s/gpu LR: 0.000036 Logit Scale: 100.000 Contrastive_loss: 0.34643 (0.38242) Loss: 0.34643 (0.38242) Logit_scale: 100.00 (100.00)
2026-05-28,15:28:19 | INFO | open_clip_train.train | Train Epoch: 26 [140935168/268959744 (52%)] Data (t): 0.712 Batch (t): 1.497, 22092.3/s, 2761.54/s/gpu LR: 0.000036 Logit Scale: 100.000 Contrastive_loss: 0.40743 (0.38299) Loss: 0.40743 (0.38299) Logit_scale: 100.00 (100.00)
2026-05-28,15:30:59 | INFO | open_clip_train.train | Train Epoch: 26 [144211968/268959744 (54%)] Data (t): 0.725 Batch (t): 1.599, 21219.5/s, 2652.44/s/gpu LR: 0.000036 Logit Scale: 100.000 Contrastive_loss: 0.37239 (0.38275) Loss: 0.37239 (0.38275) Logit_scale: 100.00 (100.00)
2026-05-28,15:33:35 | INFO | open_clip_train.train | Train Epoch: 26 [147488768/268959744 (55%)] Data (t): 0.731 Batch (t): 1.556, 21938.2/s, 2742.27/s/gpu LR: 0.000035 Logit Scale: 100.000 Contrastive_loss: 0.37610 (0.38261) Loss: 0.37610 (0.38261) Logit_scale: 100.00 (100.00)
2026-05-28,15:36:02 | INFO | open_clip_train.train | Train Epoch: 26 [150765568/268959744 (56%)] Data (t): 0.753 Batch (t): 1.471, 21356.7/s, 2669.59/s/gpu LR: 0.000035 Logit Scale: 100.000 Contrastive_loss: 0.35267 (0.38197) Loss: 0.35267 (0.38197) Logit_scale: 100.00 (100.00)
2026-05-28,15:38:42 | INFO | open_clip_train.train | Train Epoch: 26 [154042368/268959744 (57%)] Data (t): 0.745 Batch (t): 1.605, 20954.6/s, 2619.32/s/gpu LR: 0.000035 Logit Scale: 100.000 Contrastive_loss: 0.33593 (0.38101) Loss: 0.33593 (0.38101) Logit_scale: 100.00 (100.00)
2026-05-28,15:41:15 | INFO | open_clip_train.train | Train Epoch: 26 [157319168/268959744 (58%)] Data (t): 0.740 Batch (t): 1.523, 24844.0/s, 3105.50/s/gpu LR: 0.000035 Logit Scale: 100.000 Contrastive_loss: 0.36264 (0.38064) Loss: 0.36264 (0.38064) Logit_scale: 100.00 (100.00)
2026-05-28,15:43:56 | INFO | open_clip_train.train | Train Epoch: 26 [160595968/268959744 (60%)] Data (t): 0.695 Batch (t): 1.612, 21547.0/s, 2693.37/s/gpu LR: 0.000035 Logit Scale: 100.000 Contrastive_loss: 0.35687 (0.38016) Loss: 0.35687 (0.38016) Logit_scale: 100.00 (100.00)
2026-05-28,15:46:32 | INFO | open_clip_train.train | Train Epoch: 26 [163872768/268959744 (61%)] Data (t): 0.757 Batch (t): 1.557, 21339.9/s, 2667.48/s/gpu LR: 0.000035 Logit Scale: 100.000 Contrastive_loss: 0.35880 (0.37974) Loss: 0.35880 (0.37974) Logit_scale: 100.00 (100.00)
2026-05-28,15:49:00 | INFO | open_clip_train.train | Train Epoch: 26 [167149568/268959744 (62%)] Data (t): 0.757 Batch (t): 1.485, 22057.5/s, 2757.19/s/gpu LR: 0.000035 Logit Scale: 100.000 Contrastive_loss: 0.40329 (0.38020) Loss: 0.40329 (0.38020) Logit_scale: 100.00 (100.00)
2026-05-28,15:51:35 | INFO | open_clip_train.train | Train Epoch: 26 [170426368/268959744 (63%)] Data (t): 0.721 Batch (t): 1.553, 21562.2/s, 2695.28/s/gpu LR: 0.000034 Logit Scale: 100.000 Contrastive_loss: 0.38948 (0.38037) Loss: 0.38948 (0.38037) Logit_scale: 100.00 (100.00)
2026-05-28,15:54:07 | INFO | open_clip_train.train | Train Epoch: 26 [173703168/268959744 (65%)] Data (t): 0.750 Batch (t): 1.513, 21647.1/s, 2705.88/s/gpu LR: 0.000034 Logit Scale: 100.000 Contrastive_loss: 0.39619 (0.38066) Loss: 0.39619 (0.38066) Logit_scale: 100.00 (100.00)
2026-05-28,15:56:43 | INFO | open_clip_train.train | Train Epoch: 26 [176979968/268959744 (66%)] Data (t): 0.708 Batch (t): 1.566, 21594.2/s, 2699.28/s/gpu LR: 0.000034 Logit Scale: 100.000 Contrastive_loss: 0.38588 (0.38076) Loss: 0.38588 (0.38076) Logit_scale: 100.00 (100.00)
2026-05-28,15:59:18 | INFO | open_clip_train.train | Train Epoch: 26 [180256768/268959744 (67%)] Data (t): 0.732 Batch (t): 1.551, 21414.5/s, 2676.82/s/gpu LR: 0.000034 Logit Scale: 100.000 Contrastive_loss: 0.38440 (0.38082) Loss: 0.38440 (0.38082) Logit_scale: 100.00 (100.00)
2026-05-28,16:01:55 | INFO | open_clip_train.train | Train Epoch: 26 [183533568/268959744 (68%)] Data (t): 0.751 Batch (t): 1.565, 21109.7/s, 2638.72/s/gpu LR: 0.000034 Logit Scale: 100.000 Contrastive_loss: 0.40118 (0.38118) Loss: 0.40118 (0.38118) Logit_scale: 100.00 (100.00)
2026-05-28,16:04:31 | INFO | open_clip_train.train | Train Epoch: 26 [186810368/268959744 (69%)] Data (t): 0.711 Batch (t): 1.556, 20267.2/s, 2533.40/s/gpu LR: 0.000034 Logit Scale: 100.000 Contrastive_loss: 0.36677 (0.38093) Loss: 0.36677 (0.38093) Logit_scale: 100.00 (100.00)
2026-05-28,16:07:05 | INFO | open_clip_train.train | Train Epoch: 26 [190087168/268959744 (71%)] Data (t): 0.770 Batch (t): 1.539, 20543.8/s, 2567.97/s/gpu LR: 0.000034 Logit Scale: 100.000 Contrastive_loss: 0.36567 (0.38067) Loss: 0.36567 (0.38067) Logit_scale: 100.00 (100.00)
2026-05-28,16:09:38 | INFO | open_clip_train.train | Train Epoch: 26 [193363968/268959744 (72%)] Data (t): 0.776 Batch (t): 1.538, 21397.1/s, 2674.64/s/gpu LR: 0.000033 Logit Scale: 100.000 Contrastive_loss: 0.37410 (0.38056) Loss: 0.37410 (0.38056) Logit_scale: 100.00 (100.00)
2026-05-28,16:12:12 | INFO | open_clip_train.train | Train Epoch: 26 [196640768/268959744 (73%)] Data (t): 0.733 Batch (t): 1.542, 25235.8/s, 3154.48/s/gpu LR: 0.000033 Logit Scale: 100.000 Contrastive_loss: 0.38083 (0.38057) Loss: 0.38083 (0.38057) Logit_scale: 100.00 (100.00)
2026-05-28,16:14:49 | INFO | open_clip_train.train | Train Epoch: 26 [199917568/268959744 (74%)] Data (t): 0.683 Batch (t): 1.570, 8900.80/s, 1112.60/s/gpu LR: 0.000033 Logit Scale: 100.000 Contrastive_loss: 0.37724 (0.38051) Loss: 0.37724 (0.38051) Logit_scale: 100.00 (100.00)
2026-05-28,16:17:25 | INFO | open_clip_train.train | Train Epoch: 26 [203194368/268959744 (76%)] Data (t): 0.665 Batch (t): 1.555, 21373.7/s, 2671.71/s/gpu LR: 0.000033 Logit Scale: 100.000 Contrastive_loss: 0.34176 (0.37990) Loss: 0.34176 (0.37990) Logit_scale: 100.00 (100.00)
2026-05-28,16:20:00 | INFO | open_clip_train.train | Train Epoch: 26 [206471168/268959744 (77%)] Data (t): 0.736 Batch (t): 1.552, 26347.7/s, 3293.47/s/gpu LR: 0.000033 Logit Scale: 100.000 Contrastive_loss: 0.36261 (0.37963) Loss: 0.36261 (0.37963) Logit_scale: 100.00 (100.00)
2026-05-28,16:22:34 | INFO | open_clip_train.train | Train Epoch: 26 [209747968/268959744 (78%)] Data (t): 0.788 Batch (t): 1.535, 24033.1/s, 3004.14/s/gpu LR: 0.000033 Logit Scale: 100.000 Contrastive_loss: 0.35310 (0.37922) Loss: 0.35310 (0.37922) Logit_scale: 100.00 (100.00)
2026-05-28,16:25:06 | INFO | open_clip_train.train | Train Epoch: 26 [213024768/268959744 (79%)] Data (t): 0.744 Batch (t): 1.524, 21219.1/s, 2652.39/s/gpu LR: 0.000032 Logit Scale: 100.000 Contrastive_loss: 0.37390 (0.37914) Loss: 0.37390 (0.37914) Logit_scale: 100.00 (100.00)
2026-05-28,16:27:39 | INFO | open_clip_train.train | Train Epoch: 26 [216301568/268959744 (80%)] Data (t): 0.735 Batch (t): 1.534, 21733.0/s, 2716.62/s/gpu LR: 0.000032 Logit Scale: 100.000 Contrastive_loss: 0.37034 (0.37901) Loss: 0.37034 (0.37901) Logit_scale: 100.00 (100.00)
2026-05-28,16:30:17 | INFO | open_clip_train.train | Train Epoch: 26 [219578368/268959744 (82%)] Data (t): 0.763 Batch (t): 1.577, 21660.0/s, 2707.50/s/gpu LR: 0.000032 Logit Scale: 100.000 Contrastive_loss: 0.34716 (0.37854) Loss: 0.34716 (0.37854) Logit_scale: 100.00 (100.00)
2026-05-28,16:32:55 | INFO | open_clip_train.train | Train Epoch: 26 [222855168/268959744 (83%)] Data (t): 0.713 Batch (t): 1.575, 20604.2/s, 2575.52/s/gpu LR: 0.000032 Logit Scale: 100.000 Contrastive_loss: 0.35657 (0.37822) Loss: 0.35657 (0.37822) Logit_scale: 100.00 (100.00)
2026-05-28,16:35:31 | INFO | open_clip_train.train | Train Epoch: 26 [226131968/268959744 (84%)] Data (t): 0.717 Batch (t): 1.559, 21338.8/s, 2667.35/s/gpu LR: 0.000032 Logit Scale: 100.000 Contrastive_loss: 0.35016 (0.37782) Loss: 0.35016 (0.37782) Logit_scale: 100.00 (100.00)
2026-05-28,16:38:07 | INFO | open_clip_train.train | Train Epoch: 26 [229408768/268959744 (85%)] Data (t): 0.724 Batch (t): 1.561, 20956.1/s, 2619.51/s/gpu LR: 0.000032 Logit Scale: 100.000 Contrastive_loss: 0.36410 (0.37763) Loss: 0.36410 (0.37763) Logit_scale: 100.00 (100.00)
2026-05-28,16:40:42 | INFO | open_clip_train.train | Train Epoch: 26 [232685568/268959744 (87%)] Data (t): 0.675 Batch (t): 1.556, 20345.6/s, 2543.20/s/gpu LR: 0.000032 Logit Scale: 100.000 Contrastive_loss: 0.36070 (0.37739) Loss: 0.36070 (0.37739) Logit_scale: 100.00 (100.00)
2026-05-28,16:43:17 | INFO | open_clip_train.train | Train Epoch: 26 [235962368/268959744 (88%)] Data (t): 0.692 Batch (t): 1.551, 20372.9/s, 2546.62/s/gpu LR: 0.000031 Logit Scale: 100.000 Contrastive_loss: 0.33190 (0.37677) Loss: 0.33190 (0.37677) Logit_scale: 100.00 (100.00)
2026-05-28,16:45:52 | INFO | open_clip_train.train | Train Epoch: 26 [239239168/268959744 (89%)] Data (t): 0.744 Batch (t): 1.548, 20862.5/s, 2607.82/s/gpu LR: 0.000031 Logit Scale: 100.000 Contrastive_loss: 0.35811 (0.37652) Loss: 0.35811 (0.37652) Logit_scale: 100.00 (100.00)
2026-05-28,16:48:29 | INFO | open_clip_train.train | Train Epoch: 26 [242515968/268959744 (90%)] Data (t): 0.728 Batch (t): 1.573, 21398.3/s, 2674.79/s/gpu LR: 0.000031 Logit Scale: 100.000 Contrastive_loss: 0.38767 (0.37667) Loss: 0.38767 (0.37667) Logit_scale: 100.00 (100.00)
2026-05-28,16:51:04 | INFO | open_clip_train.train | Train Epoch: 26 [245792768/268959744 (91%)] Data (t): 0.704 Batch (t): 1.544, 24094.3/s, 3011.79/s/gpu LR: 0.000031 Logit Scale: 100.000 Contrastive_loss: 0.41669 (0.37719) Loss: 0.41669 (0.37719) Logit_scale: 100.00 (100.00)
2026-05-28,16:53:38 | INFO | open_clip_train.train | Train Epoch: 26 [249069568/268959744 (93%)] Data (t): 0.715 Batch (t): 1.538, 20916.0/s, 2614.50/s/gpu LR: 0.000031 Logit Scale: 100.000 Contrastive_loss: 0.39156 (0.37738) Loss: 0.39156 (0.37738) Logit_scale: 100.00 (100.00)
2026-05-28,16:56:12 | INFO | open_clip_train.train | Train Epoch: 26 [252346368/268959744 (94%)] Data (t): 0.745 Batch (t): 1.544, 20364.8/s, 2545.60/s/gpu LR: 0.000031 Logit Scale: 100.000 Contrastive_loss: 0.36842 (0.37726) Loss: 0.36842 (0.37726) Logit_scale: 100.00 (100.00)
2026-05-28,16:58:41 | INFO | open_clip_train.train | Train Epoch: 26 [255623168/268959744 (95%)] Data (t): 0.729 Batch (t): 1.492, 26381.8/s, 3297.73/s/gpu LR: 0.000031 Logit Scale: 100.000 Contrastive_loss: 0.35805 (0.37702) Loss: 0.35805 (0.37702) Logit_scale: 100.00 (100.00)
2026-05-28,17:01:20 | INFO | open_clip_train.train | Train Epoch: 26 [258899968/268959744 (96%)] Data (t): 0.712 Batch (t): 1.588, 21491.4/s, 2686.42/s/gpu LR: 0.000030 Logit Scale: 99.999 Contrastive_loss: 0.36074 (0.37682) Loss: 0.36074 (0.37682) Logit_scale: 99.999 (100.00)
2026-05-28,17:03:46 | INFO | open_clip_train.train | Train Epoch: 26 [262176768/268959744 (97%)] Data (t): 0.728 Batch (t): 1.465, 21621.7/s, 2702.71/s/gpu LR: 0.000030 Logit Scale: 100.000 Contrastive_loss: 0.37989 (0.37686) Loss: 0.37989 (0.37686) Logit_scale: 100.00 (100.00)
2026-05-28,17:06:23 | INFO | open_clip_train.train | Train Epoch: 26 [265453568/268959744 (99%)] Data (t): 0.717 Batch (t): 1.569, 22693.8/s, 2836.73/s/gpu LR: 0.000030 Logit Scale: 100.000 Contrastive_loss: 0.41334 (0.37730) Loss: 0.41334 (0.37730) Logit_scale: 100.00 (100.00)
2026-05-28,17:08:54 | INFO | open_clip_train.train | Train Epoch: 26 [268730368/268959744 (100%)] Data (t): 0.633 Batch (t): 1.503, 28832.1/s, 3604.01/s/gpu LR: 0.000030 Logit Scale: 99.996 Contrastive_loss: 0.44483 (0.37811) Loss: 0.44483 (0.37811) Logit_scale: 99.996 (100.00)
2026-05-28,17:09:02 | INFO | open_clip_train.train | Train Epoch: 26 [268959744/268959744 (100%)] Data (t): 0.615 Batch (t): 1.146, 28519.3/s, 3564.91/s/gpu LR: 0.000030 Logit Scale: 99.990 Contrastive_loss: 0.37613 (0.37809) Loss: 0.37613 (0.37809) Logit_scale: 99.990 (100.00)
2026-05-28,17:09:06 | INFO | open_clip_train.main | Start epoch 27
2026-05-28,17:09:28 | INFO | open_clip_train.train | Train Epoch: 27 [ 32768/268959744 (0%)] Data (t): 18.582 Batch (t): 21.512, 1523.26/s, 190.407/s/gpu LR: 0.000030 Logit Scale: 99.990 Contrastive_loss: 0.37990 (0.37990) Loss: 0.37990 (0.37990) Logit_scale: 99.990 (99.990)
2026-05-28,17:12:09 | INFO | open_clip_train.train | Train Epoch: 27 [ 3309568/268959744 (1%)] Data (t): 0.696 Batch (t): 1.615, 20889.8/s, 2611.22/s/gpu LR: 0.000030 Logit Scale: 100.000 Contrastive_loss: 0.33924 (0.35957) Loss: 0.33924 (0.35957) Logit_scale: 100.00 (99.995)
2026-05-28,17:14:43 | INFO | open_clip_train.train | Train Epoch: 27 [ 6586368/268959744 (2%)] Data (t): 0.714 Batch (t): 1.536, 21710.8/s, 2713.85/s/gpu LR: 0.000030 Logit Scale: 100.000 Contrastive_loss: 0.35598 (0.35837) Loss: 0.35598 (0.35837) Logit_scale: 100.00 (99.997)
2026-05-28,17:17:10 | INFO | open_clip_train.train | Train Epoch: 27 [ 9863168/268959744 (4%)] Data (t): 0.744 Batch (t): 1.469, 22887.8/s, 2860.97/s/gpu LR: 0.000030 Logit Scale: 100.000 Contrastive_loss: 0.35724 (0.35809) Loss: 0.35724 (0.35809) Logit_scale: 100.00 (99.997)
2026-05-28,17:19:49 | INFO | open_clip_train.train | Train Epoch: 27 [ 13139968/268959744 (5%)] Data (t): 0.724 Batch (t): 1.595, 21175.1/s, 2646.89/s/gpu LR: 0.000029 Logit Scale: 100.000 Contrastive_loss: 0.36544 (0.35956) Loss: 0.36544 (0.35956) Logit_scale: 100.00 (99.998)
2026-05-28,17:22:23 | INFO | open_clip_train.train | Train Epoch: 27 [ 16416768/268959744 (6%)] Data (t): 0.677 Batch (t): 1.539, 21497.0/s, 2687.12/s/gpu LR: 0.000029 Logit Scale: 100.000 Contrastive_loss: 0.34555 (0.35722) Loss: 0.34555 (0.35722) Logit_scale: 100.00 (99.998)
2026-05-28,17:24:59 | INFO | open_clip_train.train | Train Epoch: 27 [ 19693568/268959744 (7%)] Data (t): 0.708 Batch (t): 1.562, 20664.8/s, 2583.10/s/gpu LR: 0.000029 Logit Scale: 100.000 Contrastive_loss: 0.34125 (0.35494) Loss: 0.34125 (0.35494) Logit_scale: 100.00 (99.999)
2026-05-28,17:27:33 | INFO | open_clip_train.train | Train Epoch: 27 [ 22970368/268959744 (9%)] Data (t): 0.738 Batch (t): 1.532, 21722.2/s, 2715.28/s/gpu LR: 0.000029 Logit Scale: 100.000 Contrastive_loss: 0.33849 (0.35289) Loss: 0.33849 (0.35289) Logit_scale: 100.00 (99.999)
2026-05-28,17:30:09 | INFO | open_clip_train.train | Train Epoch: 27 [ 26247168/268959744 (10%)] Data (t): 0.771 Batch (t): 1.563, 21585.6/s, 2698.19/s/gpu LR: 0.000029 Logit Scale: 100.000 Contrastive_loss: 0.35969 (0.35364) Loss: 0.35969 (0.35364) Logit_scale: 100.00 (99.999)
2026-05-28,17:32:41 | INFO | open_clip_train.train | Train Epoch: 27 [ 29523968/268959744 (11%)] Data (t): 0.744 Batch (t): 1.520, 21321.2/s, 2665.15/s/gpu LR: 0.000029 Logit Scale: 100.000 Contrastive_loss: 0.37568 (0.35585) Loss: 0.37568 (0.35585) Logit_scale: 100.00 (99.999)
2026-05-28,17:35:18 | INFO | open_clip_train.train | Train Epoch: 27 [ 32800768/268959744 (12%)] Data (t): 0.664 Batch (t): 1.568, 24409.9/s, 3051.23/s/gpu LR: 0.000029 Logit Scale: 100.000 Contrastive_loss: 0.34990 (0.35530) Loss: 0.34990 (0.35530) Logit_scale: 100.00 (99.999)
2026-05-28,17:37:52 | INFO | open_clip_train.train | Train Epoch: 27 [ 36077568/268959744 (13%)] Data (t): 0.711 Batch (t): 1.543, 21611.4/s, 2701.42/s/gpu LR: 0.000028 Logit Scale: 100.000 Contrastive_loss: 0.33887 (0.35393) Loss: 0.33887 (0.35393) Logit_scale: 100.00 (99.999)
2026-05-28,17:40:23 | INFO | open_clip_train.train | Train Epoch: 27 [ 39354368/268959744 (15%)] Data (t): 0.724 Batch (t): 1.509, 24202.3/s, 3025.28/s/gpu LR: 0.000028 Logit Scale: 100.000 Contrastive_loss: 0.35413 (0.35395) Loss: 0.35413 (0.35395) Logit_scale: 100.00 (99.999)
2026-05-28,17:42:56 | INFO | open_clip_train.train | Train Epoch: 27 [ 42631168/268959744 (16%)] Data (t): 0.697 Batch (t): 1.532, 21455.9/s, 2681.98/s/gpu LR: 0.000028 Logit Scale: 100.000 Contrastive_loss: 0.34357 (0.35321) Loss: 0.34357 (0.35321) Logit_scale: 100.00 (99.999)
2026-05-28,17:45:35 | INFO | open_clip_train.train | Train Epoch: 27 [ 45907968/268959744 (17%)] Data (t): 0.660 Batch (t): 1.592, 20619.4/s, 2577.43/s/gpu LR: 0.000028 Logit Scale: 100.000 Contrastive_loss: 0.38441 (0.35529) Loss: 0.38441 (0.35529) Logit_scale: 100.00 (99.999)
2026-05-28,17:48:12 | INFO | open_clip_train.train | Train Epoch: 27 [ 49184768/268959744 (18%)] Data (t): 0.653 Batch (t): 1.565, 20881.1/s, 2610.14/s/gpu LR: 0.000028 Logit Scale: 100.000 Contrastive_loss: 0.35749 (0.35543) Loss: 0.35749 (0.35543) Logit_scale: 100.00 (99.999)
2026-05-28,17:50:44 | INFO | open_clip_train.train | Train Epoch: 27 [ 52461568/268959744 (20%)] Data (t): 0.669 Batch (t): 1.525, 22840.0/s, 2855.00/s/gpu LR: 0.000028 Logit Scale: 100.000 Contrastive_loss: 0.35791 (0.35557) Loss: 0.35791 (0.35557) Logit_scale: 100.00 (99.999)
2026-05-28,17:53:12 | INFO | open_clip_train.train | Train Epoch: 27 [ 55738368/268959744 (21%)] Data (t): 0.643 Batch (t): 1.473, 20472.7/s, 2559.09/s/gpu LR: 0.000028 Logit Scale: 100.000 Contrastive_loss: 0.37171 (0.35647) Loss: 0.37171 (0.35647) Logit_scale: 100.00 (99.999)
2026-05-28,17:55:50 | INFO | open_clip_train.train | Train Epoch: 27 [ 59015168/268959744 (22%)] Data (t): 0.653 Batch (t): 1.587, 24072.7/s, 3009.09/s/gpu LR: 0.000027 Logit Scale: 100.000 Contrastive_loss: 0.34305 (0.35576) Loss: 0.34305 (0.35576) Logit_scale: 100.00 (99.999)
2026-05-28,17:58:24 | INFO | open_clip_train.train | Train Epoch: 27 [ 62291968/268959744 (23%)] Data (t): 0.685 Batch (t): 1.537, 21290.2/s, 2661.27/s/gpu LR: 0.000027 Logit Scale: 100.000 Contrastive_loss: 0.36192 (0.35607) Loss: 0.36192 (0.35607) Logit_scale: 100.00 (99.999)
2026-05-28,18:00:59 | INFO | open_clip_train.train | Train Epoch: 27 [ 65568768/268959744 (24%)] Data (t): 0.677 Batch (t): 1.546, 21136.3/s, 2642.04/s/gpu LR: 0.000027 Logit Scale: 100.000 Contrastive_loss: 0.37643 (0.35704) Loss: 0.37643 (0.35704) Logit_scale: 100.00 (99.999)
2026-05-28,18:03:30 | INFO | open_clip_train.train | Train Epoch: 27 [ 68845568/268959744 (26%)] Data (t): 0.716 Batch (t): 1.509, 26090.9/s, 3261.36/s/gpu LR: 0.000027 Logit Scale: 100.000 Contrastive_loss: 0.33660 (0.35611) Loss: 0.33660 (0.35611) Logit_scale: 100.00 (100.00)
2026-05-28,18:06:05 | INFO | open_clip_train.train | Train Epoch: 27 [ 72122368/268959744 (27%)] Data (t): 0.665 Batch (t): 1.552, 20094.7/s, 2511.84/s/gpu LR: 0.000027 Logit Scale: 100.000 Contrastive_loss: 0.35674 (0.35614) Loss: 0.35674 (0.35614) Logit_scale: 100.00 (100.00)
2026-05-28,18:08:45 | INFO | open_clip_train.train | Train Epoch: 27 [ 75399168/268959744 (28%)] Data (t): 0.691 Batch (t): 1.600, 24421.0/s, 3052.62/s/gpu LR: 0.000027 Logit Scale: 100.000 Contrastive_loss: 0.34341 (0.35561) Loss: 0.34341 (0.35561) Logit_scale: 100.00 (100.00)
2026-05-28,18:11:17 | INFO | open_clip_train.train | Train Epoch: 27 [ 78675968/268959744 (29%)] Data (t): 0.712 Batch (t): 1.525, 21817.9/s, 2727.24/s/gpu LR: 0.000027 Logit Scale: 100.000 Contrastive_loss: 0.35932 (0.35576) Loss: 0.35932 (0.35576) Logit_scale: 100.00 (100.00)
2026-05-28,18:13:47 | INFO | open_clip_train.train | Train Epoch: 27 [ 81952768/268959744 (30%)] Data (t): 0.692 Batch (t): 1.500, 21124.4/s, 2640.55/s/gpu LR: 0.000026 Logit Scale: 100.000 Contrastive_loss: 0.36679 (0.35618) Loss: 0.36679 (0.35618) Logit_scale: 100.00 (100.00)
2026-05-28,18:16:23 | INFO | open_clip_train.train | Train Epoch: 27 [ 85229568/268959744 (32%)] Data (t): 0.704 Batch (t): 1.555, 21082.3/s, 2635.28/s/gpu LR: 0.000026 Logit Scale: 100.000 Contrastive_loss: 0.35398 (0.35610) Loss: 0.35398 (0.35610) Logit_scale: 100.00 (100.00)
2026-05-28,18:19:01 | INFO | open_clip_train.train | Train Epoch: 27 [ 88506368/268959744 (33%)] Data (t): 0.648 Batch (t): 1.584, 19722.9/s, 2465.37/s/gpu LR: 0.000026 Logit Scale: 100.000 Contrastive_loss: 0.32893 (0.35513) Loss: 0.32893 (0.35513) Logit_scale: 100.00 (100.00)
2026-05-28,18:21:37 | INFO | open_clip_train.train | Train Epoch: 27 [ 91783168/268959744 (34%)] Data (t): 0.673 Batch (t): 1.556, 21067.0/s, 2633.37/s/gpu LR: 0.000026 Logit Scale: 100.000 Contrastive_loss: 0.33261 (0.35435) Loss: 0.33261 (0.35435) Logit_scale: 100.00 (100.00)
2026-05-28,18:24:12 | INFO | open_clip_train.train | Train Epoch: 27 [ 95059968/268959744 (35%)] Data (t): 0.713 Batch (t): 1.549, 21702.2/s, 2712.78/s/gpu LR: 0.000026 Logit Scale: 100.000 Contrastive_loss: 0.34757 (0.35413) Loss: 0.34757 (0.35413) Logit_scale: 100.00 (100.00)
2026-05-28,18:26:49 | INFO | open_clip_train.train | Train Epoch: 27 [ 98336768/268959744 (37%)] Data (t): 0.697 Batch (t): 1.576, 20061.4/s, 2507.68/s/gpu LR: 0.000026 Logit Scale: 100.000 Contrastive_loss: 0.33998 (0.35367) Loss: 0.33998 (0.35367) Logit_scale: 100.00 (100.00)
2026-05-28,18:29:26 | INFO | open_clip_train.train | Train Epoch: 27 [101613568/268959744 (38%)] Data (t): 0.718 Batch (t): 1.562, 21853.9/s, 2731.73/s/gpu LR: 0.000026 Logit Scale: 100.000 Contrastive_loss: 0.33299 (0.35302) Loss: 0.33299 (0.35302) Logit_scale: 100.00 (100.00)
2026-05-28,18:31:56 | INFO | open_clip_train.train | Train Epoch: 27 [104890368/268959744 (39%)] Data (t): 0.684 Batch (t): 1.505, 24158.0/s, 3019.75/s/gpu LR: 0.000026 Logit Scale: 100.000 Contrastive_loss: 0.34475 (0.35277) Loss: 0.34475 (0.35277) Logit_scale: 100.00 (100.00)
2026-05-28,18:34:36 | INFO | open_clip_train.train | Train Epoch: 27 [108167168/268959744 (40%)] Data (t): 0.698 Batch (t): 1.600, 20304.8/s, 2538.10/s/gpu LR: 0.000025 Logit Scale: 100.000 Contrastive_loss: 0.37519 (0.35343) Loss: 0.37519 (0.35343) Logit_scale: 100.00 (100.00)
2026-05-28,18:37:13 | INFO | open_clip_train.train | Train Epoch: 27 [111443968/268959744 (41%)] Data (t): 0.706 Batch (t): 1.564, 20483.7/s, 2560.46/s/gpu LR: 0.000025 Logit Scale: 100.000 Contrastive_loss: 0.34870 (0.35330) Loss: 0.34870 (0.35330) Logit_scale: 100.00 (100.00)
2026-05-28,18:39:50 | INFO | open_clip_train.train | Train Epoch: 27 [114720768/268959744 (43%)] Data (t): 0.741 Batch (t): 1.577, 20786.7/s, 2598.34/s/gpu LR: 0.000025 Logit Scale: 100.000 Contrastive_loss: 0.35383 (0.35331) Loss: 0.35383 (0.35331) Logit_scale: 100.00 (100.00)
2026-05-28,18:42:26 | INFO | open_clip_train.train | Train Epoch: 27 [117997568/268959744 (44%)] Data (t): 0.663 Batch (t): 1.555, 21337.4/s, 2667.17/s/gpu LR: 0.000025 Logit Scale: 100.000 Contrastive_loss: 0.32536 (0.35256) Loss: 0.32536 (0.35256) Logit_scale: 100.00 (100.00)
2026-05-28,18:45:00 | INFO | open_clip_train.train | Train Epoch: 27 [121274368/268959744 (45%)] Data (t): 0.786 Batch (t): 1.540, 24236.7/s, 3029.58/s/gpu LR: 0.000025 Logit Scale: 100.000 Contrastive_loss: 0.39265 (0.35361) Loss: 0.39265 (0.35361) Logit_scale: 100.00 (100.00)
2026-05-28,18:47:36 | INFO | open_clip_train.train | Train Epoch: 27 [124551168/268959744 (46%)] Data (t): 0.771 Batch (t): 1.562, 20177.4/s, 2522.17/s/gpu LR: 0.000025 Logit Scale: 100.000 Contrastive_loss: 0.33561 (0.35315) Loss: 0.33561 (0.35315) Logit_scale: 100.00 (100.00)
2026-05-28,18:50:09 | INFO | open_clip_train.train | Train Epoch: 27 [127827968/268959744 (48%)] Data (t): 0.756 Batch (t): 1.527, 20554.1/s, 2569.26/s/gpu LR: 0.000025 Logit Scale: 100.000 Contrastive_loss: 0.37116 (0.35360) Loss: 0.37116 (0.35360) Logit_scale: 100.00 (100.00)
2026-05-28,18:52:44 | INFO | open_clip_train.train | Train Epoch: 27 [131104768/268959744 (49%)] Data (t): 0.778 Batch (t): 1.554, 21457.2/s, 2682.15/s/gpu LR: 0.000025 Logit Scale: 100.000 Contrastive_loss: 0.34485 (0.35339) Loss: 0.34485 (0.35339) Logit_scale: 100.00 (100.00)
2026-05-28,18:55:19 | INFO | open_clip_train.train | Train Epoch: 27 [134381568/268959744 (50%)] Data (t): 0.675 Batch (t): 1.552, 20957.6/s, 2619.71/s/gpu LR: 0.000024 Logit Scale: 100.000 Contrastive_loss: 0.32628 (0.35274) Loss: 0.32628 (0.35274) Logit_scale: 100.00 (100.00)
2026-05-28,18:57:53 | INFO | open_clip_train.train | Train Epoch: 27 [137658368/268959744 (51%)] Data (t): 0.784 Batch (t): 1.539, 21151.7/s, 2643.96/s/gpu LR: 0.000024 Logit Scale: 100.000 Contrastive_loss: 0.34085 (0.35247) Loss: 0.34085 (0.35247) Logit_scale: 100.00 (100.00)
2026-05-28,19:00:22 | INFO | open_clip_train.train | Train Epoch: 27 [140935168/268959744 (52%)] Data (t): 0.766 Batch (t): 1.486, 24131.7/s, 3016.46/s/gpu LR: 0.000024 Logit Scale: 100.000 Contrastive_loss: 0.31119 (0.35153) Loss: 0.31119 (0.35153) Logit_scale: 100.00 (100.00)
2026-05-28,19:03:02 | INFO | open_clip_train.train | Train Epoch: 27 [144211968/268959744 (54%)] Data (t): 0.719 Batch (t): 1.602, 21216.2/s, 2652.02/s/gpu LR: 0.000024 Logit Scale: 100.000 Contrastive_loss: 0.32080 (0.35084) Loss: 0.32080 (0.35084) Logit_scale: 100.00 (100.00)
2026-05-28,19:05:38 | INFO | open_clip_train.train | Train Epoch: 27 [147488768/268959744 (55%)] Data (t): 0.698 Batch (t): 1.563, 19416.6/s, 2427.08/s/gpu LR: 0.000024 Logit Scale: 100.000 Contrastive_loss: 0.33989 (0.35061) Loss: 0.33989 (0.35061) Logit_scale: 100.00 (100.00)
2026-05-28,19:08:14 | INFO | open_clip_train.train | Train Epoch: 27 [150765568/268959744 (56%)] Data (t): 0.682 Batch (t): 1.561, 20625.5/s, 2578.19/s/gpu LR: 0.000024 Logit Scale: 100.000 Contrastive_loss: 0.35359 (0.35067) Loss: 0.35359 (0.35067) Logit_scale: 100.00 (100.00)
2026-05-28,19:10:51 | INFO | open_clip_train.train | Train Epoch: 27 [154042368/268959744 (57%)] Data (t): 0.754 Batch (t): 1.563, 20496.6/s, 2562.08/s/gpu LR: 0.000024 Logit Scale: 100.000 Contrastive_loss: 0.33712 (0.35039) Loss: 0.33712 (0.35039) Logit_scale: 100.00 (100.00)
2026-05-28,19:13:27 | INFO | open_clip_train.train | Train Epoch: 27 [157319168/268959744 (58%)] Data (t): 0.740 Batch (t): 1.565, 20904.4/s, 2613.06/s/gpu LR: 0.000023 Logit Scale: 100.000 Contrastive_loss: 0.32839 (0.34994) Loss: 0.32839 (0.34994) Logit_scale: 100.00 (100.00)
2026-05-28,19:16:03 | INFO | open_clip_train.train | Train Epoch: 27 [160595968/268959744 (60%)] Data (t): 0.720 Batch (t): 1.561, 20456.5/s, 2557.07/s/gpu LR: 0.000023 Logit Scale: 100.000 Contrastive_loss: 0.32477 (0.34943) Loss: 0.32477 (0.34943) Logit_scale: 100.00 (100.00)
2026-05-28,19:18:40 | INFO | open_clip_train.train | Train Epoch: 27 [163872768/268959744 (61%)] Data (t): 0.754 Batch (t): 1.568, 21767.0/s, 2720.88/s/gpu LR: 0.000023 Logit Scale: 100.000 Contrastive_loss: 0.33306 (0.34911) Loss: 0.33306 (0.34911) Logit_scale: 100.00 (100.00)
2026-05-28,19:21:15 | INFO | open_clip_train.train | Train Epoch: 27 [167149568/268959744 (62%)] Data (t): 0.766 Batch (t): 1.547, 20942.9/s, 2617.86/s/gpu LR: 0.000023 Logit Scale: 100.000 Contrastive_loss: 0.35106 (0.34915) Loss: 0.35106 (0.34915) Logit_scale: 100.00 (100.00)
2026-05-28,19:23:50 | INFO | open_clip_train.train | Train Epoch: 27 [170426368/268959744 (63%)] Data (t): 0.700 Batch (t): 1.555, 20648.9/s, 2581.11/s/gpu LR: 0.000023 Logit Scale: 100.000 Contrastive_loss: 0.33338 (0.34885) Loss: 0.33338 (0.34885) Logit_scale: 100.00 (100.00)
2026-05-28,19:26:28 | INFO | open_clip_train.train | Train Epoch: 27 [173703168/268959744 (65%)] Data (t): 0.764 Batch (t): 1.572, 20909.9/s, 2613.73/s/gpu LR: 0.000023 Logit Scale: 100.000 Contrastive_loss: 0.33648 (0.34862) Loss: 0.33648 (0.34862) Logit_scale: 100.00 (100.00)
2026-05-28,19:29:04 | INFO | open_clip_train.train | Train Epoch: 27 [176979968/268959744 (66%)] Data (t): 0.695 Batch (t): 1.560, 21389.7/s, 2673.72/s/gpu LR: 0.000023 Logit Scale: 100.000 Contrastive_loss: 0.32945 (0.34828) Loss: 0.32945 (0.34828) Logit_scale: 100.00 (100.00)
2026-05-28,19:31:39 | INFO | open_clip_train.train | Train Epoch: 27 [180256768/268959744 (67%)] Data (t): 0.733 Batch (t): 1.551, 21094.7/s, 2636.83/s/gpu LR: 0.000023 Logit Scale: 100.000 Contrastive_loss: 0.34474 (0.34821) Loss: 0.34474 (0.34821) Logit_scale: 100.00 (100.00)
2026-05-28,19:34:15 | INFO | open_clip_train.train | Train Epoch: 27 [183533568/268959744 (68%)] Data (t): 0.764 Batch (t): 1.559, 19728.4/s, 2466.05/s/gpu LR: 0.000022 Logit Scale: 100.000 Contrastive_loss: 0.33259 (0.34794) Loss: 0.33259 (0.34794) Logit_scale: 100.00 (100.00)
2026-05-28,19:36:50 | INFO | open_clip_train.train | Train Epoch: 27 [186810368/268959744 (69%)] Data (t): 0.837 Batch (t): 1.557, 20062.7/s, 2507.83/s/gpu LR: 0.000022 Logit Scale: 100.000 Contrastive_loss: 0.33431 (0.34770) Loss: 0.33431 (0.34770) Logit_scale: 100.00 (100.00)
2026-05-28,19:39:25 | INFO | open_clip_train.train | Train Epoch: 27 [190087168/268959744 (71%)] Data (t): 0.694 Batch (t): 1.552, 23200.6/s, 2900.08/s/gpu LR: 0.000022 Logit Scale: 100.000 Contrastive_loss: 0.34188 (0.34760) Loss: 0.34188 (0.34760) Logit_scale: 100.00 (100.00)
2026-05-28,19:41:54 | INFO | open_clip_train.train | Train Epoch: 27 [193363968/268959744 (72%)] Data (t): 0.748 Batch (t): 1.488, 24574.2/s, 3071.78/s/gpu LR: 0.000022 Logit Scale: 100.000 Contrastive_loss: 0.35147 (0.34767) Loss: 0.35147 (0.34767) Logit_scale: 100.00 (100.00)
2026-05-28,19:44:32 | INFO | open_clip_train.train | Train Epoch: 27 [196640768/268959744 (73%)] Data (t): 0.697 Batch (t): 1.576, 22571.3/s, 2821.41/s/gpu LR: 0.000022 Logit Scale: 100.000 Contrastive_loss: 0.34068 (0.34755) Loss: 0.34068 (0.34755) Logit_scale: 100.00 (100.00)
2026-05-28,19:47:08 | INFO | open_clip_train.train | Train Epoch: 27 [199917568/268959744 (74%)] Data (t): 0.664 Batch (t): 1.559, 23722.2/s, 2965.27/s/gpu LR: 0.000022 Logit Scale: 100.000 Contrastive_loss: 0.34842 (0.34757) Loss: 0.34842 (0.34757) Logit_scale: 100.00 (100.00)
2026-05-28,19:49:49 | INFO | open_clip_train.train | Train Epoch: 27 [203194368/268959744 (76%)] Data (t): 0.720 Batch (t): 1.608, 20772.6/s, 2596.58/s/gpu LR: 0.000022 Logit Scale: 100.000 Contrastive_loss: 0.34243 (0.34749) Loss: 0.34243 (0.34749) Logit_scale: 100.00 (100.00)
2026-05-28,19:52:16 | INFO | open_clip_train.train | Train Epoch: 27 [206471168/268959744 (77%)] Data (t): 0.686 Batch (t): 1.470, 24700.3/s, 3087.54/s/gpu LR: 0.000022 Logit Scale: 100.000 Contrastive_loss: 0.34334 (0.34742) Loss: 0.34334 (0.34742) Logit_scale: 100.00 (100.00)
2026-05-28,19:54:53 | INFO | open_clip_train.train | Train Epoch: 27 [209747968/268959744 (78%)] Data (t): 0.745 Batch (t): 1.573, 21576.0/s, 2697.00/s/gpu LR: 0.000021 Logit Scale: 100.000 Contrastive_loss: 0.31393 (0.34691) Loss: 0.31393 (0.34691) Logit_scale: 100.00 (100.00)
2026-05-28,19:57:30 | INFO | open_clip_train.train | Train Epoch: 27 [213024768/268959744 (79%)] Data (t): 0.797 Batch (t): 1.570, 20113.4/s, 2514.18/s/gpu LR: 0.000021 Logit Scale: 100.000 Contrastive_loss: 0.32027 (0.34650) Loss: 0.32027 (0.34650) Logit_scale: 100.00 (100.00)
2026-05-28,20:00:05 | INFO | open_clip_train.train | Train Epoch: 27 [216301568/268959744 (80%)] Data (t): 0.801 Batch (t): 1.555, 20762.3/s, 2595.29/s/gpu LR: 0.000021 Logit Scale: 100.000 Contrastive_loss: 0.32211 (0.34614) Loss: 0.32211 (0.34614) Logit_scale: 100.00 (100.00)
2026-05-28,20:02:39 | INFO | open_clip_train.train | Train Epoch: 27 [219578368/268959744 (82%)] Data (t): 0.827 Batch (t): 1.534, 21300.7/s, 2662.58/s/gpu LR: 0.000021 Logit Scale: 100.000 Contrastive_loss: 0.36122 (0.34636) Loss: 0.36122 (0.34636) Logit_scale: 100.00 (100.00)
2026-05-28,20:05:14 | INFO | open_clip_train.train | Train Epoch: 27 [222855168/268959744 (83%)] Data (t): 0.791 Batch (t): 1.555, 20597.3/s, 2574.66/s/gpu LR: 0.000021 Logit Scale: 100.000 Contrastive_loss: 0.33490 (0.34620) Loss: 0.33490 (0.34620) Logit_scale: 100.00 (100.00)
2026-05-28,20:07:50 | INFO | open_clip_train.train | Train Epoch: 27 [226131968/268959744 (84%)] Data (t): 0.806 Batch (t): 1.554, 20398.8/s, 2549.85/s/gpu LR: 0.000021 Logit Scale: 100.000 Contrastive_loss: 0.34392 (0.34616) Loss: 0.34392 (0.34616) Logit_scale: 100.00 (100.00)
2026-05-28,20:10:21 | INFO | open_clip_train.train | Train Epoch: 27 [229408768/268959744 (85%)] Data (t): 0.783 Batch (t): 1.513, 18558.5/s, 2319.82/s/gpu LR: 0.000021 Logit Scale: 100.000 Contrastive_loss: 0.33946 (0.34607) Loss: 0.33946 (0.34607) Logit_scale: 100.00 (100.00)
2026-05-28,20:12:55 | INFO | open_clip_train.train | Train Epoch: 27 [232685568/268959744 (87%)] Data (t): 0.734 Batch (t): 1.544, 21603.8/s, 2700.47/s/gpu LR: 0.000021 Logit Scale: 100.000 Contrastive_loss: 0.35148 (0.34614) Loss: 0.35148 (0.34614) Logit_scale: 100.00 (100.00)
2026-05-28,20:15:32 | INFO | open_clip_train.train | Train Epoch: 27 [235962368/268959744 (88%)] Data (t): 0.787 Batch (t): 1.569, 25188.4/s, 3148.55/s/gpu LR: 0.000021 Logit Scale: 100.000 Contrastive_loss: 0.32942 (0.34591) Loss: 0.32942 (0.34591) Logit_scale: 100.00 (100.00)
2026-05-28,20:18:10 | INFO | open_clip_train.train | Train Epoch: 27 [239239168/268959744 (89%)] Data (t): 0.707 Batch (t): 1.577, 20753.1/s, 2594.13/s/gpu LR: 0.000020 Logit Scale: 100.000 Contrastive_loss: 0.35613 (0.34605) Loss: 0.35613 (0.34605) Logit_scale: 100.00 (100.00)
2026-05-28,20:20:45 | INFO | open_clip_train.train | Train Epoch: 27 [242515968/268959744 (90%)] Data (t): 0.743 Batch (t): 1.551, 20942.3/s, 2617.79/s/gpu LR: 0.000020 Logit Scale: 100.000 Contrastive_loss: 0.34201 (0.34600) Loss: 0.34201 (0.34600) Logit_scale: 100.00 (100.00)
2026-05-28,20:23:14 | INFO | open_clip_train.train | Train Epoch: 27 [245792768/268959744 (91%)] Data (t): 0.802 Batch (t): 1.487, 23580.4/s, 2947.55/s/gpu LR: 0.000020 Logit Scale: 100.000 Contrastive_loss: 0.35538 (0.34612) Loss: 0.35538 (0.34612) Logit_scale: 100.00 (100.00)
2026-05-28,20:25:57 | INFO | open_clip_train.train | Train Epoch: 27 [249069568/268959744 (93%)] Data (t): 0.728 Batch (t): 1.636, 21237.8/s, 2654.72/s/gpu LR: 0.000020 Logit Scale: 100.000 Contrastive_loss: 0.34269 (0.34608) Loss: 0.34269 (0.34608) Logit_scale: 100.00 (100.00)
2026-05-28,20:28:34 | INFO | open_clip_train.train | Train Epoch: 27 [252346368/268959744 (94%)] Data (t): 0.779 Batch (t): 1.565, 20256.8/s, 2532.10/s/gpu LR: 0.000020 Logit Scale: 100.000 Contrastive_loss: 0.34958 (0.34612) Loss: 0.34958 (0.34612) Logit_scale: 100.00 (100.00)
2026-05-28,20:31:09 | INFO | open_clip_train.train | Train Epoch: 27 [255623168/268959744 (95%)] Data (t): 0.810 Batch (t): 1.551, 22535.0/s, 2816.88/s/gpu LR: 0.000020 Logit Scale: 100.000 Contrastive_loss: 0.38089 (0.34656) Loss: 0.38089 (0.34656) Logit_scale: 100.00 (100.00)
2026-05-28,20:33:40 | INFO | open_clip_train.train | Train Epoch: 27 [258899968/268959744 (96%)] Data (t): 0.822 Batch (t): 1.513, 21661.7/s, 2707.72/s/gpu LR: 0.000020 Logit Scale: 100.000 Contrastive_loss: 0.33562 (0.34643) Loss: 0.33562 (0.34643) Logit_scale: 100.00 (100.00)
2026-05-28,20:36:17 | INFO | open_clip_train.train | Train Epoch: 27 [262176768/268959744 (97%)] Data (t): 0.738 Batch (t): 1.567, 20379.0/s, 2547.38/s/gpu LR: 0.000020 Logit Scale: 100.000 Contrastive_loss: 0.35610 (0.34655) Loss: 0.35610 (0.34655) Logit_scale: 100.00 (100.00)
2026-05-28,20:38:57 | INFO | open_clip_train.train | Train Epoch: 27 [265453568/268959744 (99%)] Data (t): 0.734 Batch (t): 1.599, 20632.2/s, 2579.02/s/gpu LR: 0.000019 Logit Scale: 99.999 Contrastive_loss: 0.34152 (0.34648) Loss: 0.34152 (0.34648) Logit_scale: 99.999 (100.00)
2026-05-28,20:41:29 | INFO | open_clip_train.train | Train Epoch: 27 [268730368/268959744 (100%)] Data (t): 0.771 Batch (t): 1.519, 30288.0/s, 3786.00/s/gpu LR: 0.000019 Logit Scale: 99.987 Contrastive_loss: 0.39263 (0.34704) Loss: 0.39263 (0.34704) Logit_scale: 99.987 (100.00)
2026-05-28,20:41:37 | INFO | open_clip_train.train | Train Epoch: 27 [268959744/268959744 (100%)] Data (t): 0.629 Batch (t): 1.129, 28895.9/s, 3611.99/s/gpu LR: 0.000019 Logit Scale: 99.986 Contrastive_loss: 0.36221 (0.34722) Loss: 0.36221 (0.34722) Logit_scale: 99.986 (100.00)
2026-05-28,20:41:41 | INFO | open_clip_train.main | Start epoch 28
2026-05-28,20:42:04 | INFO | open_clip_train.train | Train Epoch: 28 [ 32768/268959744 (0%)] Data (t): 19.501 Batch (t): 22.609, 1449.33/s, 181.166/s/gpu LR: 0.000019 Logit Scale: 99.986 Contrastive_loss: 0.29557 (0.29557) Loss: 0.29557 (0.29557) Logit_scale: 99.986 (99.986)
2026-05-28,20:44:41 | INFO | open_clip_train.train | Train Epoch: 28 [ 3309568/268959744 (1%)] Data (t): 0.738 Batch (t): 1.571, 22519.1/s, 2814.89/s/gpu LR: 0.000019 Logit Scale: 100.000 Contrastive_loss: 0.33233 (0.31395) Loss: 0.33233 (0.31395) Logit_scale: 100.00 (99.993)
2026-05-28,20:47:18 | INFO | open_clip_train.train | Train Epoch: 28 [ 6586368/268959744 (2%)] Data (t): 0.734 Batch (t): 1.571, 21732.4/s, 2716.55/s/gpu LR: 0.000019 Logit Scale: 100.000 Contrastive_loss: 0.32651 (0.31814) Loss: 0.32651 (0.31814) Logit_scale: 100.00 (99.995)
2026-05-28,20:49:54 | INFO | open_clip_train.train | Train Epoch: 28 [ 9863168/268959744 (4%)] Data (t): 0.765 Batch (t): 1.561, 21118.1/s, 2639.76/s/gpu LR: 0.000019 Logit Scale: 100.000 Contrastive_loss: 0.32796 (0.32059) Loss: 0.32796 (0.32059) Logit_scale: 100.00 (99.997)
2026-05-28,20:52:29 | INFO | open_clip_train.train | Train Epoch: 28 [ 13139968/268959744 (5%)] Data (t): 0.779 Batch (t): 1.547, 21287.1/s, 2660.89/s/gpu LR: 0.000019 Logit Scale: 100.000 Contrastive_loss: 0.31855 (0.32018) Loss: 0.31855 (0.32018) Logit_scale: 100.00 (99.997)
2026-05-28,20:55:04 | INFO | open_clip_train.train | Train Epoch: 28 [ 16416768/268959744 (6%)] Data (t): 0.704 Batch (t): 1.547, 22134.6/s, 2766.83/s/gpu LR: 0.000019 Logit Scale: 100.000 Contrastive_loss: 0.34074 (0.32361) Loss: 0.34074 (0.32361) Logit_scale: 100.00 (99.998)
2026-05-28,20:57:37 | INFO | open_clip_train.train | Train Epoch: 28 [ 19693568/268959744 (7%)] Data (t): 0.821 Batch (t): 1.534, 21714.1/s, 2714.27/s/gpu LR: 0.000019 Logit Scale: 100.000 Contrastive_loss: 0.33277 (0.32492) Loss: 0.33277 (0.32492) Logit_scale: 100.00 (99.998)
2026-05-28,21:00:07 | INFO | open_clip_train.train | Train Epoch: 28 [ 22970368/268959744 (9%)] Data (t): 0.744 Batch (t): 1.494, 21035.6/s, 2629.45/s/gpu LR: 0.000019 Logit Scale: 100.000 Contrastive_loss: 0.32974 (0.32552) Loss: 0.32974 (0.32552) Logit_scale: 100.00 (99.998)
2026-05-28,21:02:42 | INFO | open_clip_train.train | Train Epoch: 28 [ 26247168/268959744 (10%)] Data (t): 0.753 Batch (t): 1.553, 21582.6/s, 2697.83/s/gpu LR: 0.000018 Logit Scale: 100.000 Contrastive_loss: 0.31936 (0.32484) Loss: 0.31936 (0.32484) Logit_scale: 100.00 (99.998)
2026-05-28,21:05:14 | INFO | open_clip_train.train | Train Epoch: 28 [ 29523968/268959744 (11%)] Data (t): 0.776 Batch (t): 1.522, 21672.0/s, 2709.00/s/gpu LR: 0.000018 Logit Scale: 100.000 Contrastive_loss: 0.31340 (0.32369) Loss: 0.31340 (0.32369) Logit_scale: 100.00 (99.999)
2026-05-28,21:07:52 | INFO | open_clip_train.train | Train Epoch: 28 [ 32800768/268959744 (12%)] Data (t): 0.728 Batch (t): 1.580, 20170.6/s, 2521.32/s/gpu LR: 0.000018 Logit Scale: 100.000 Contrastive_loss: 0.32936 (0.32421) Loss: 0.32936 (0.32421) Logit_scale: 100.00 (99.999)
2026-05-28,21:10:27 | INFO | open_clip_train.train | Train Epoch: 28 [ 36077568/268959744 (13%)] Data (t): 0.770 Batch (t): 1.550, 21218.7/s, 2652.34/s/gpu LR: 0.000018 Logit Scale: 100.000 Contrastive_loss: 0.30847 (0.32290) Loss: 0.30847 (0.32290) Logit_scale: 100.00 (99.999)
2026-05-28,21:13:03 | INFO | open_clip_train.train | Train Epoch: 28 [ 39354368/268959744 (15%)] Data (t): 0.748 Batch (t): 1.557, 21061.7/s, 2632.71/s/gpu LR: 0.000018 Logit Scale: 100.000 Contrastive_loss: 0.29344 (0.32063) Loss: 0.29344 (0.32063) Logit_scale: 100.00 (99.999)
2026-05-28,21:15:40 | INFO | open_clip_train.train | Train Epoch: 28 [ 42631168/268959744 (16%)] Data (t): 0.706 Batch (t): 1.573, 20302.0/s, 2537.75/s/gpu LR: 0.000018 Logit Scale: 100.000 Contrastive_loss: 0.31549 (0.32026) Loss: 0.31549 (0.32026) Logit_scale: 100.00 (99.999)
2026-05-28,21:18:15 | INFO | open_clip_train.train | Train Epoch: 28 [ 45907968/268959744 (17%)] Data (t): 0.758 Batch (t): 1.544, 21277.8/s, 2659.73/s/gpu LR: 0.000018 Logit Scale: 100.000 Contrastive_loss: 0.31067 (0.31962) Loss: 0.31067 (0.31962) Logit_scale: 100.00 (99.999)
2026-05-28,21:20:50 | INFO | open_clip_train.train | Train Epoch: 28 [ 49184768/268959744 (18%)] Data (t): 0.750 Batch (t): 1.550, 21403.2/s, 2675.40/s/gpu LR: 0.000018 Logit Scale: 100.000 Contrastive_loss: 0.31021 (0.31904) Loss: 0.31021 (0.31904) Logit_scale: 100.00 (99.999)
2026-05-28,21:23:24 | INFO | open_clip_train.train | Train Epoch: 28 [ 52461568/268959744 (20%)] Data (t): 0.759 Batch (t): 1.546, 20649.0/s, 2581.13/s/gpu LR: 0.000018 Logit Scale: 100.000 Contrastive_loss: 0.32170 (0.31919) Loss: 0.32170 (0.31919) Logit_scale: 100.00 (99.999)
2026-05-28,21:26:00 | INFO | open_clip_train.train | Train Epoch: 28 [ 55738368/268959744 (21%)] Data (t): 0.758 Batch (t): 1.562, 21823.1/s, 2727.89/s/gpu LR: 0.000017 Logit Scale: 100.000 Contrastive_loss: 0.29455 (0.31782) Loss: 0.29455 (0.31782) Logit_scale: 100.00 (99.999)
2026-05-28,21:28:36 | INFO | open_clip_train.train | Train Epoch: 28 [ 59015168/268959744 (22%)] Data (t): 0.770 Batch (t): 1.552, 24264.6/s, 3033.07/s/gpu LR: 0.000017 Logit Scale: 100.000 Contrastive_loss: 0.33202 (0.31857) Loss: 0.33202 (0.31857) Logit_scale: 100.00 (99.999)
2026-05-28,21:31:05 | INFO | open_clip_train.train | Train Epoch: 28 [ 62291968/268959744 (23%)] Data (t): 0.809 Batch (t): 1.492, 23788.4/s, 2973.55/s/gpu LR: 0.000017 Logit Scale: 100.000 Contrastive_loss: 0.35694 (0.32049) Loss: 0.35694 (0.32049) Logit_scale: 100.00 (99.999)
2026-05-28,21:33:46 | INFO | open_clip_train.train | Train Epoch: 28 [ 65568768/268959744 (24%)] Data (t): 0.735 Batch (t): 1.611, 21758.2/s, 2719.78/s/gpu LR: 0.000017 Logit Scale: 100.000 Contrastive_loss: 0.30752 (0.31987) Loss: 0.30752 (0.31987) Logit_scale: 100.00 (99.999)
2026-05-28,21:36:22 | INFO | open_clip_train.train | Train Epoch: 28 [ 68845568/268959744 (26%)] Data (t): 0.739 Batch (t): 1.563, 20653.2/s, 2581.65/s/gpu LR: 0.000017 Logit Scale: 100.000 Contrastive_loss: 0.33386 (0.32051) Loss: 0.33386 (0.32051) Logit_scale: 100.00 (99.999)
2026-05-28,21:38:55 | INFO | open_clip_train.train | Train Epoch: 28 [ 72122368/268959744 (27%)] Data (t): 0.799 Batch (t): 1.527, 20689.8/s, 2586.22/s/gpu LR: 0.000017 Logit Scale: 100.000 Contrastive_loss: 0.31605 (0.32031) Loss: 0.31605 (0.32031) Logit_scale: 100.00 (99.999)
2026-05-28,21:41:29 | INFO | open_clip_train.train | Train Epoch: 28 [ 75399168/268959744 (28%)] Data (t): 0.793 Batch (t): 1.543, 20791.8/s, 2598.98/s/gpu LR: 0.000017 Logit Scale: 100.000 Contrastive_loss: 0.32515 (0.32052) Loss: 0.32515 (0.32052) Logit_scale: 100.00 (99.999)
2026-05-28,21:44:06 | INFO | open_clip_train.train | Train Epoch: 28 [ 78675968/268959744 (29%)] Data (t): 0.728 Batch (t): 1.569, 20692.5/s, 2586.57/s/gpu LR: 0.000017 Logit Scale: 100.000 Contrastive_loss: 0.34390 (0.32145) Loss: 0.34390 (0.32145) Logit_scale: 100.00 (99.999)
2026-05-28,21:46:41 | INFO | open_clip_train.train | Train Epoch: 28 [ 81952768/268959744 (30%)] Data (t): 0.765 Batch (t): 1.551, 20531.3/s, 2566.42/s/gpu LR: 0.000017 Logit Scale: 100.000 Contrastive_loss: 0.30060 (0.32065) Loss: 0.30060 (0.32065) Logit_scale: 100.00 (99.999)
2026-05-28,21:49:14 | INFO | open_clip_train.train | Train Epoch: 28 [ 85229568/268959744 (32%)] Data (t): 0.713 Batch (t): 1.524, 21100.5/s, 2637.56/s/gpu LR: 0.000016 Logit Scale: 100.000 Contrastive_loss: 0.32223 (0.32071) Loss: 0.32223 (0.32071) Logit_scale: 100.00 (99.999)
2026-05-28,21:51:50 | INFO | open_clip_train.train | Train Epoch: 28 [ 88506368/268959744 (33%)] Data (t): 0.773 Batch (t): 1.562, 24118.7/s, 3014.84/s/gpu LR: 0.000016 Logit Scale: 100.000 Contrastive_loss: 0.31019 (0.32033) Loss: 0.31019 (0.32033) Logit_scale: 100.00 (100.00)
2026-05-28,21:54:26 | INFO | open_clip_train.train | Train Epoch: 28 [ 91783168/268959744 (34%)] Data (t): 0.767 Batch (t): 1.558, 20935.7/s, 2616.96/s/gpu LR: 0.000016 Logit Scale: 100.000 Contrastive_loss: 0.31461 (0.32013) Loss: 0.31461 (0.32013) Logit_scale: 100.00 (100.00)
2026-05-28,21:57:01 | INFO | open_clip_train.train | Train Epoch: 28 [ 95059968/268959744 (35%)] Data (t): 0.697 Batch (t): 1.558, 19452.9/s, 2431.62/s/gpu LR: 0.000016 Logit Scale: 100.000 Contrastive_loss: 0.29835 (0.31941) Loss: 0.29835 (0.31941) Logit_scale: 100.00 (100.00)
2026-05-28,21:59:35 | INFO | open_clip_train.train | Train Epoch: 28 [ 98336768/268959744 (37%)] Data (t): 0.794 Batch (t): 1.537, 20483.5/s, 2560.44/s/gpu LR: 0.000016 Logit Scale: 100.000 Contrastive_loss: 0.31501 (0.31927) Loss: 0.31501 (0.31927) Logit_scale: 100.00 (100.00)
2026-05-28,22:02:09 | INFO | open_clip_train.train | Train Epoch: 28 [101613568/268959744 (38%)] Data (t): 0.786 Batch (t): 1.544, 22529.9/s, 2816.24/s/gpu LR: 0.000016 Logit Scale: 100.000 Contrastive_loss: 0.32975 (0.31959) Loss: 0.32975 (0.31959) Logit_scale: 100.00 (100.00)
2026-05-28,22:04:46 | INFO | open_clip_train.train | Train Epoch: 28 [104890368/268959744 (39%)] Data (t): 0.893 Batch (t): 1.566, 21548.6/s, 2693.58/s/gpu LR: 0.000016 Logit Scale: 100.000 Contrastive_loss: 0.30464 (0.31914) Loss: 0.30464 (0.31914) Logit_scale: 100.00 (100.00)
2026-05-28,22:07:22 | INFO | open_clip_train.train | Train Epoch: 28 [108167168/268959744 (40%)] Data (t): 0.871 Batch (t): 1.564, 20380.9/s, 2547.61/s/gpu LR: 0.000016 Logit Scale: 100.000 Contrastive_loss: 0.28594 (0.31816) Loss: 0.28594 (0.31816) Logit_scale: 100.00 (100.00)
2026-05-28,22:09:59 | INFO | open_clip_train.train | Train Epoch: 28 [111443968/268959744 (41%)] Data (t): 0.878 Batch (t): 1.568, 19653.3/s, 2456.66/s/gpu LR: 0.000016 Logit Scale: 100.000 Contrastive_loss: 0.30364 (0.31775) Loss: 0.30364 (0.31775) Logit_scale: 100.00 (100.00)
2026-05-28,22:12:35 | INFO | open_clip_train.train | Train Epoch: 28 [114720768/268959744 (43%)] Data (t): 0.826 Batch (t): 1.557, 20655.8/s, 2581.98/s/gpu LR: 0.000015 Logit Scale: 100.000 Contrastive_loss: 0.32312 (0.31790) Loss: 0.32312 (0.31790) Logit_scale: 100.00 (100.00)
2026-05-28,22:15:11 | INFO | open_clip_train.train | Train Epoch: 28 [117997568/268959744 (44%)] Data (t): 0.797 Batch (t): 1.557, 22109.8/s, 2763.72/s/gpu LR: 0.000015 Logit Scale: 100.000 Contrastive_loss: 0.29564 (0.31730) Loss: 0.29564 (0.31730) Logit_scale: 100.00 (100.00)
2026-05-28,22:17:47 | INFO | open_clip_train.train | Train Epoch: 28 [121274368/268959744 (45%)] Data (t): 0.807 Batch (t): 1.563, 22100.5/s, 2762.57/s/gpu LR: 0.000015 Logit Scale: 100.000 Contrastive_loss: 0.30939 (0.31709) Loss: 0.30939 (0.31709) Logit_scale: 100.00 (100.00)
2026-05-28,22:20:23 | INFO | open_clip_train.train | Train Epoch: 28 [124551168/268959744 (46%)] Data (t): 0.784 Batch (t): 1.562, 20706.8/s, 2588.35/s/gpu LR: 0.000015 Logit Scale: 100.000 Contrastive_loss: 0.29494 (0.31652) Loss: 0.29494 (0.31652) Logit_scale: 100.00 (100.00)
2026-05-28,22:22:59 | INFO | open_clip_train.train | Train Epoch: 28 [127827968/268959744 (48%)] Data (t): 0.827 Batch (t): 1.559, 20152.9/s, 2519.11/s/gpu LR: 0.000015 Logit Scale: 100.000 Contrastive_loss: 0.29042 (0.31587) Loss: 0.29042 (0.31587) Logit_scale: 100.00 (100.00)
2026-05-28,22:25:34 | INFO | open_clip_train.train | Train Epoch: 28 [131104768/268959744 (49%)] Data (t): 0.816 Batch (t): 1.549, 21375.1/s, 2671.89/s/gpu LR: 0.000015 Logit Scale: 100.000 Contrastive_loss: 0.32193 (0.31602) Loss: 0.32193 (0.31602) Logit_scale: 100.00 (100.00)
2026-05-28,22:28:09 | INFO | open_clip_train.train | Train Epoch: 28 [134381568/268959744 (50%)] Data (t): 0.835 Batch (t): 1.550, 19281.4/s, 2410.18/s/gpu LR: 0.000015 Logit Scale: 100.000 Contrastive_loss: 0.30516 (0.31576) Loss: 0.30516 (0.31576) Logit_scale: 100.00 (100.00)
2026-05-28,22:30:43 | INFO | open_clip_train.train | Train Epoch: 28 [137658368/268959744 (51%)] Data (t): 0.791 Batch (t): 1.545, 21114.0/s, 2639.24/s/gpu LR: 0.000015 Logit Scale: 100.000 Contrastive_loss: 0.32242 (0.31591) Loss: 0.32242 (0.31591) Logit_scale: 100.00 (100.00)
2026-05-28,22:33:20 | INFO | open_clip_train.train | Train Epoch: 28 [140935168/268959744 (52%)] Data (t): 0.871 Batch (t): 1.570, 20619.7/s, 2577.46/s/gpu LR: 0.000015 Logit Scale: 100.000 Contrastive_loss: 0.26272 (0.31470) Loss: 0.26272 (0.31470) Logit_scale: 100.00 (100.00)
2026-05-28,22:35:57 | INFO | open_clip_train.train | Train Epoch: 28 [144211968/268959744 (54%)] Data (t): 0.840 Batch (t): 1.563, 20877.6/s, 2609.70/s/gpu LR: 0.000015 Logit Scale: 100.000 Contrastive_loss: 0.32149 (0.31485) Loss: 0.32149 (0.31485) Logit_scale: 100.00 (100.00)
2026-05-28,22:38:33 | INFO | open_clip_train.train | Train Epoch: 28 [147488768/268959744 (55%)] Data (t): 0.907 Batch (t): 1.567, 23481.8/s, 2935.23/s/gpu LR: 0.000014 Logit Scale: 100.000 Contrastive_loss: 0.29658 (0.31446) Loss: 0.29658 (0.31446) Logit_scale: 100.00 (100.00)
2026-05-28,22:41:10 | INFO | open_clip_train.train | Train Epoch: 28 [150765568/268959744 (56%)] Data (t): 0.855 Batch (t): 1.563, 21199.1/s, 2649.89/s/gpu LR: 0.000014 Logit Scale: 100.000 Contrastive_loss: 0.29721 (0.31409) Loss: 0.29721 (0.31409) Logit_scale: 100.00 (100.00)
2026-05-28,22:43:45 | INFO | open_clip_train.train | Train Epoch: 28 [154042368/268959744 (57%)] Data (t): 0.860 Batch (t): 1.556, 20377.7/s, 2547.21/s/gpu LR: 0.000014 Logit Scale: 100.000 Contrastive_loss: 0.30468 (0.31389) Loss: 0.30468 (0.31389) Logit_scale: 100.00 (100.00)
2026-05-28,22:46:22 | INFO | open_clip_train.train | Train Epoch: 28 [157319168/268959744 (58%)] Data (t): 0.880 Batch (t): 1.570, 21405.4/s, 2675.67/s/gpu LR: 0.000014 Logit Scale: 100.000 Contrastive_loss: 0.32397 (0.31410) Loss: 0.32397 (0.31410) Logit_scale: 100.00 (100.00)
2026-05-28,22:48:58 | INFO | open_clip_train.train | Train Epoch: 28 [160595968/268959744 (60%)] Data (t): 0.808 Batch (t): 1.552, 20953.3/s, 2619.16/s/gpu LR: 0.000014 Logit Scale: 100.000 Contrastive_loss: 0.29660 (0.31375) Loss: 0.29660 (0.31375) Logit_scale: 100.00 (100.00)
2026-05-28,22:51:29 | INFO | open_clip_train.train | Train Epoch: 28 [163872768/268959744 (61%)] Data (t): 0.797 Batch (t): 1.511, 21741.6/s, 2717.70/s/gpu LR: 0.000014 Logit Scale: 100.000 Contrastive_loss: 0.30401 (0.31356) Loss: 0.30401 (0.31356) Logit_scale: 100.00 (100.00)
2026-05-28,22:54:01 | INFO | open_clip_train.train | Train Epoch: 28 [167149568/268959744 (62%)] Data (t): 0.797 Batch (t): 1.525, 21580.5/s, 2697.57/s/gpu LR: 0.000014 Logit Scale: 100.000 Contrastive_loss: 0.32118 (0.31371) Loss: 0.32118 (0.31371) Logit_scale: 100.00 (100.00)
2026-05-28,22:56:38 | INFO | open_clip_train.train | Train Epoch: 28 [170426368/268959744 (63%)] Data (t): 0.855 Batch (t): 1.572, 20659.8/s, 2582.47/s/gpu LR: 0.000014 Logit Scale: 100.000 Contrastive_loss: 0.32049 (0.31383) Loss: 0.32049 (0.31383) Logit_scale: 100.00 (100.00)
2026-05-28,22:59:16 | INFO | open_clip_train.train | Train Epoch: 28 [173703168/268959744 (65%)] Data (t): 0.814 Batch (t): 1.577, 21096.1/s, 2637.02/s/gpu LR: 0.000014 Logit Scale: 100.000 Contrastive_loss: 0.31734 (0.31390) Loss: 0.31734 (0.31390) Logit_scale: 100.00 (100.00)
2026-05-28,23:01:59 | INFO | open_clip_train.train | Train Epoch: 28 [176979968/268959744 (66%)] Data (t): 0.849 Batch (t): 1.629, 21619.4/s, 2702.42/s/gpu LR: 0.000014 Logit Scale: 100.000 Contrastive_loss: 0.31883 (0.31399) Loss: 0.31883 (0.31399) Logit_scale: 100.00 (100.00)
2026-05-28,23:04:34 | INFO | open_clip_train.train | Train Epoch: 28 [180256768/268959744 (67%)] Data (t): 0.878 Batch (t): 1.546, 20838.2/s, 2604.78/s/gpu LR: 0.000013 Logit Scale: 100.000 Contrastive_loss: 0.33219 (0.31431) Loss: 0.33219 (0.31431) Logit_scale: 100.00 (100.00)
2026-05-28,23:07:06 | INFO | open_clip_train.train | Train Epoch: 28 [183533568/268959744 (68%)] Data (t): 0.814 Batch (t): 1.528, 21684.4/s, 2710.54/s/gpu LR: 0.000013 Logit Scale: 100.000 Contrastive_loss: 0.31705 (0.31436) Loss: 0.31705 (0.31436) Logit_scale: 100.00 (100.00)
2026-05-28,23:09:40 | INFO | open_clip_train.train | Train Epoch: 28 [186810368/268959744 (69%)] Data (t): 0.792 Batch (t): 1.541, 20610.9/s, 2576.36/s/gpu LR: 0.000013 Logit Scale: 100.000 Contrastive_loss: 0.31027 (0.31429) Loss: 0.31027 (0.31429) Logit_scale: 100.00 (100.00)
2026-05-28,23:12:15 | INFO | open_clip_train.train | Train Epoch: 28 [190087168/268959744 (71%)] Data (t): 0.893 Batch (t): 1.542, 20823.1/s, 2602.89/s/gpu LR: 0.000013 Logit Scale: 100.000 Contrastive_loss: 0.29012 (0.31388) Loss: 0.29012 (0.31388) Logit_scale: 100.00 (100.00)
2026-05-28,23:14:48 | INFO | open_clip_train.train | Train Epoch: 28 [193363968/268959744 (72%)] Data (t): 0.882 Batch (t): 1.533, 20870.4/s, 2608.81/s/gpu LR: 0.000013 Logit Scale: 100.000 Contrastive_loss: 0.30627 (0.31375) Loss: 0.30627 (0.31375) Logit_scale: 100.00 (100.00)
2026-05-28,23:17:24 | INFO | open_clip_train.train | Train Epoch: 28 [196640768/268959744 (73%)] Data (t): 0.920 Batch (t): 1.561, 21303.2/s, 2662.90/s/gpu LR: 0.000013 Logit Scale: 100.000 Contrastive_loss: 0.31573 (0.31379) Loss: 0.31573 (0.31379) Logit_scale: 100.00 (100.00)
2026-05-28,23:20:00 | INFO | open_clip_train.train | Train Epoch: 28 [199917568/268959744 (74%)] Data (t): 0.783 Batch (t): 1.561, 20306.9/s, 2538.37/s/gpu LR: 0.000013 Logit Scale: 100.000 Contrastive_loss: 0.30033 (0.31357) Loss: 0.30033 (0.31357) Logit_scale: 100.00 (100.00)
2026-05-28,23:22:35 | INFO | open_clip_train.train | Train Epoch: 28 [203194368/268959744 (76%)] Data (t): 0.903 Batch (t): 1.549, 19696.2/s, 2462.03/s/gpu LR: 0.000013 Logit Scale: 100.000 Contrastive_loss: 0.32110 (0.31369) Loss: 0.32110 (0.31369) Logit_scale: 100.00 (100.00)
2026-05-28,23:25:12 | INFO | open_clip_train.train | Train Epoch: 28 [206471168/268959744 (77%)] Data (t): 0.895 Batch (t): 1.569, 21174.5/s, 2646.82/s/gpu LR: 0.000013 Logit Scale: 100.000 Contrastive_loss: 0.29932 (0.31346) Loss: 0.29932 (0.31346) Logit_scale: 100.00 (100.00)
2026-05-28,23:27:46 | INFO | open_clip_train.train | Train Epoch: 28 [209747968/268959744 (78%)] Data (t): 0.859 Batch (t): 1.541, 21170.1/s, 2646.26/s/gpu LR: 0.000013 Logit Scale: 100.000 Contrastive_loss: 0.32438 (0.31363) Loss: 0.32438 (0.31363) Logit_scale: 100.00 (100.00)
2026-05-28,23:30:21 | INFO | open_clip_train.train | Train Epoch: 28 [213024768/268959744 (79%)] Data (t): 0.917 Batch (t): 1.552, 21242.3/s, 2655.28/s/gpu LR: 0.000012 Logit Scale: 100.000 Contrastive_loss: 0.32067 (0.31374) Loss: 0.32067 (0.31374) Logit_scale: 100.00 (100.00)
2026-05-28,23:32:57 | INFO | open_clip_train.train | Train Epoch: 28 [216301568/268959744 (80%)] Data (t): 0.903 Batch (t): 1.562, 21037.5/s, 2629.69/s/gpu LR: 0.000012 Logit Scale: 100.000 Contrastive_loss: 0.30879 (0.31367) Loss: 0.30879 (0.31367) Logit_scale: 100.00 (100.00)
2026-05-28,23:35:33 | INFO | open_clip_train.train | Train Epoch: 28 [219578368/268959744 (82%)] Data (t): 0.881 Batch (t): 1.558, 20680.9/s, 2585.11/s/gpu LR: 0.000012 Logit Scale: 100.000 Contrastive_loss: 0.30029 (0.31347) Loss: 0.30029 (0.31347) Logit_scale: 100.00 (100.00)
2026-05-28,23:38:10 | INFO | open_clip_train.train | Train Epoch: 28 [222855168/268959744 (83%)] Data (t): 0.875 Batch (t): 1.569, 21043.3/s, 2630.41/s/gpu LR: 0.000012 Logit Scale: 100.000 Contrastive_loss: 0.31432 (0.31348) Loss: 0.31432 (0.31348) Logit_scale: 100.00 (100.00)
2026-05-28,23:40:44 | INFO | open_clip_train.train | Train Epoch: 28 [226131968/268959744 (84%)] Data (t): 0.927 Batch (t): 1.543, 21805.0/s, 2725.63/s/gpu LR: 0.000012 Logit Scale: 100.000 Contrastive_loss: 0.28523 (0.31308) Loss: 0.28523 (0.31308) Logit_scale: 100.00 (100.00)
2026-05-28,23:43:18 | INFO | open_clip_train.train | Train Epoch: 28 [229408768/268959744 (85%)] Data (t): 0.863 Batch (t): 1.534, 19807.5/s, 2475.94/s/gpu LR: 0.000012 Logit Scale: 100.000 Contrastive_loss: 0.30317 (0.31294) Loss: 0.30317 (0.31294) Logit_scale: 100.00 (100.00)
2026-05-28,23:45:53 | INFO | open_clip_train.train | Train Epoch: 28 [232685568/268959744 (87%)] Data (t): 0.941 Batch (t): 1.547, 24636.0/s, 3079.50/s/gpu LR: 0.000012 Logit Scale: 100.000 Contrastive_loss: 0.27926 (0.31247) Loss: 0.27926 (0.31247) Logit_scale: 100.00 (100.00)
2026-05-28,23:48:29 | INFO | open_clip_train.train | Train Epoch: 28 [235962368/268959744 (88%)] Data (t): 0.881 Batch (t): 1.567, 20337.5/s, 2542.19/s/gpu LR: 0.000012 Logit Scale: 100.000 Contrastive_loss: 0.28837 (0.31214) Loss: 0.28837 (0.31214) Logit_scale: 100.00 (100.00)
2026-05-28,23:51:05 | INFO | open_clip_train.train | Train Epoch: 28 [239239168/268959744 (89%)] Data (t): 0.882 Batch (t): 1.558, 20319.9/s, 2539.99/s/gpu LR: 0.000012 Logit Scale: 100.000 Contrastive_loss: 0.31010 (0.31211) Loss: 0.31010 (0.31211) Logit_scale: 100.00 (100.00)
2026-05-28,23:53:42 | INFO | open_clip_train.train | Train Epoch: 28 [242515968/268959744 (90%)] Data (t): 0.902 Batch (t): 1.565, 20527.9/s, 2565.99/s/gpu LR: 0.000012 Logit Scale: 100.000 Contrastive_loss: 0.28814 (0.31179) Loss: 0.28814 (0.31179) Logit_scale: 100.00 (100.00)
2026-05-28,23:56:14 | INFO | open_clip_train.train | Train Epoch: 28 [245792768/268959744 (91%)] Data (t): 0.877 Batch (t): 1.529, 21415.2/s, 2676.90/s/gpu LR: 0.000012 Logit Scale: 100.000 Contrastive_loss: 0.31628 (0.31185) Loss: 0.31628 (0.31185) Logit_scale: 100.00 (100.00)
2026-05-28,23:58:49 | INFO | open_clip_train.train | Train Epoch: 28 [249069568/268959744 (93%)] Data (t): 0.904 Batch (t): 1.545, 20898.1/s, 2612.27/s/gpu LR: 0.000011 Logit Scale: 100.000 Contrastive_loss: 0.34596 (0.31230) Loss: 0.34596 (0.31230) Logit_scale: 100.00 (100.00)
2026-05-29,00:01:23 | INFO | open_clip_train.train | Train Epoch: 28 [252346368/268959744 (94%)] Data (t): 0.801 Batch (t): 1.536, 20492.6/s, 2561.58/s/gpu LR: 0.000011 Logit Scale: 100.000 Contrastive_loss: 0.29659 (0.31209) Loss: 0.29659 (0.31209) Logit_scale: 100.00 (100.00)
2026-05-29,00:03:56 | INFO | open_clip_train.train | Train Epoch: 28 [255623168/268959744 (95%)] Data (t): 0.831 Batch (t): 1.531, 20479.3/s, 2559.91/s/gpu LR: 0.000011 Logit Scale: 100.000 Contrastive_loss: 0.33418 (0.31237) Loss: 0.33418 (0.31237) Logit_scale: 100.00 (100.00)
2026-05-29,00:06:31 | INFO | open_clip_train.train | Train Epoch: 28 [258899968/268959744 (96%)] Data (t): 0.848 Batch (t): 1.555, 22043.6/s, 2755.45/s/gpu LR: 0.000011 Logit Scale: 99.999 Contrastive_loss: 0.31750 (0.31244) Loss: 0.31750 (0.31244) Logit_scale: 99.999 (100.00)
2026-05-29,00:09:08 | INFO | open_clip_train.train | Train Epoch: 28 [262176768/268959744 (97%)] Data (t): 0.858 Batch (t): 1.565, 20637.3/s, 2579.66/s/gpu LR: 0.000011 Logit Scale: 99.999 Contrastive_loss: 0.30409 (0.31233) Loss: 0.30409 (0.31233) Logit_scale: 99.999 (100.00)
2026-05-29,00:11:43 | INFO | open_clip_train.train | Train Epoch: 28 [265453568/268959744 (99%)] Data (t): 0.811 Batch (t): 1.555, 20532.6/s, 2566.57/s/gpu LR: 0.000011 Logit Scale: 99.987 Contrastive_loss: 0.34429 (0.31272) Loss: 0.34429 (0.31272) Logit_scale: 99.987 (100.00)
2026-05-29,00:14:14 | INFO | open_clip_train.train | Train Epoch: 28 [268730368/268959744 (100%)] Data (t): 0.798 Batch (t): 1.507, 29389.8/s, 3673.73/s/gpu LR: 0.000011 Logit Scale: 99.971 Contrastive_loss: 0.39996 (0.31378) Loss: 0.39996 (0.31378) Logit_scale: 99.971 (99.999)
2026-05-29,00:14:22 | INFO | open_clip_train.train | Train Epoch: 28 [268959744/268959744 (100%)] Data (t): 0.617 Batch (t): 1.137, 28535.4/s, 3566.92/s/gpu LR: 0.000011 Logit Scale: 99.972 Contrastive_loss: 0.40491 (0.31486) Loss: 0.40491 (0.31486) Logit_scale: 99.972 (99.999)
2026-05-29,00:14:26 | INFO | open_clip_train.main | Start epoch 29
2026-05-29,00:14:48 | INFO | open_clip_train.train | Train Epoch: 29 [ 32768/268959744 (0%)] Data (t): 21.362 Batch (t): 21.810, 1502.45/s, 187.806/s/gpu LR: 0.000011 Logit Scale: 99.971 Contrastive_loss: 0.29542 (0.29542) Loss: 0.29542 (0.29542) Logit_scale: 99.971 (99.971)
2026-05-29,00:17:25 | INFO | open_clip_train.train | Train Epoch: 29 [ 3309568/268959744 (1%)] Data (t): 0.871 Batch (t): 1.569, 21764.6/s, 2720.58/s/gpu LR: 0.000011 Logit Scale: 100.000 Contrastive_loss: 0.30716 (0.30129) Loss: 0.30716 (0.30129) Logit_scale: 100.00 (99.986)
2026-05-29,00:20:01 | INFO | open_clip_train.train | Train Epoch: 29 [ 6586368/268959744 (2%)] Data (t): 0.811 Batch (t): 1.557, 22342.2/s, 2792.78/s/gpu LR: 0.000011 Logit Scale: 100.000 Contrastive_loss: 0.29767 (0.30008) Loss: 0.29767 (0.30008) Logit_scale: 100.00 (99.990)
2026-05-29,00:22:39 | INFO | open_clip_train.train | Train Epoch: 29 [ 9863168/268959744 (4%)] Data (t): 0.857 Batch (t): 1.581, 20614.3/s, 2576.79/s/gpu LR: 0.000011 Logit Scale: 100.000 Contrastive_loss: 0.28974 (0.29750) Loss: 0.28974 (0.29750) Logit_scale: 100.00 (99.993)
2026-05-29,00:25:14 | INFO | open_clip_train.train | Train Epoch: 29 [ 13139968/268959744 (5%)] Data (t): 0.782 Batch (t): 1.551, 20788.3/s, 2598.54/s/gpu LR: 0.000011 Logit Scale: 100.000 Contrastive_loss: 0.30079 (0.29816) Loss: 0.30079 (0.29816) Logit_scale: 100.00 (99.994)
2026-05-29,00:27:50 | INFO | open_clip_train.train | Train Epoch: 29 [ 16416768/268959744 (6%)] Data (t): 0.808 Batch (t): 1.557, 20668.7/s, 2583.59/s/gpu LR: 0.000010 Logit Scale: 100.000 Contrastive_loss: 0.27863 (0.29490) Loss: 0.27863 (0.29490) Logit_scale: 100.00 (99.995)
2026-05-29,00:30:25 | INFO | open_clip_train.train | Train Epoch: 29 [ 19693568/268959744 (7%)] Data (t): 0.906 Batch (t): 1.555, 20465.5/s, 2558.19/s/gpu LR: 0.000010 Logit Scale: 100.000 Contrastive_loss: 0.28111 (0.29293) Loss: 0.28111 (0.29293) Logit_scale: 100.00 (99.996)
2026-05-29,00:33:01 | INFO | open_clip_train.train | Train Epoch: 29 [ 22970368/268959744 (9%)] Data (t): 0.904 Batch (t): 1.559, 21150.2/s, 2643.77/s/gpu LR: 0.000010 Logit Scale: 100.000 Contrastive_loss: 0.29515 (0.29321) Loss: 0.29515 (0.29321) Logit_scale: 100.00 (99.996)
2026-05-29,00:35:38 | INFO | open_clip_train.train | Train Epoch: 29 [ 26247168/268959744 (10%)] Data (t): 0.869 Batch (t): 1.564, 21027.3/s, 2628.41/s/gpu LR: 0.000010 Logit Scale: 100.000 Contrastive_loss: 0.29430 (0.29333) Loss: 0.29430 (0.29333) Logit_scale: 100.00 (99.997)
2026-05-29,00:38:14 | INFO | open_clip_train.train | Train Epoch: 29 [ 29523968/268959744 (11%)] Data (t): 0.851 Batch (t): 1.564, 21388.6/s, 2673.57/s/gpu LR: 0.000010 Logit Scale: 100.000 Contrastive_loss: 0.27842 (0.29184) Loss: 0.27842 (0.29184) Logit_scale: 100.00 (99.997)
2026-05-29,00:40:50 | INFO | open_clip_train.train | Train Epoch: 29 [ 32800768/268959744 (12%)] Data (t): 0.904 Batch (t): 1.561, 20833.3/s, 2604.17/s/gpu LR: 0.000010 Logit Scale: 100.000 Contrastive_loss: 0.29652 (0.29226) Loss: 0.29652 (0.29226) Logit_scale: 100.00 (99.997)
2026-05-29,00:43:26 | INFO | open_clip_train.train | Train Epoch: 29 [ 36077568/268959744 (13%)] Data (t): 0.884 Batch (t): 1.563, 20489.2/s, 2561.15/s/gpu LR: 0.000010 Logit Scale: 100.000 Contrastive_loss: 0.29644 (0.29261) Loss: 0.29644 (0.29261) Logit_scale: 100.00 (99.998)
2026-05-29,00:46:03 | INFO | open_clip_train.train | Train Epoch: 29 [ 39354368/268959744 (15%)] Data (t): 0.930 Batch (t): 1.568, 20045.6/s, 2505.70/s/gpu LR: 0.000010 Logit Scale: 100.000 Contrastive_loss: 0.31374 (0.29424) Loss: 0.31374 (0.29424) Logit_scale: 100.00 (99.998)
2026-05-29,00:48:40 | INFO | open_clip_train.train | Train Epoch: 29 [ 42631168/268959744 (16%)] Data (t): 0.881 Batch (t): 1.566, 20966.4/s, 2620.80/s/gpu LR: 0.000010 Logit Scale: 100.000 Contrastive_loss: 0.32103 (0.29615) Loss: 0.32103 (0.29615) Logit_scale: 100.00 (99.998)
2026-05-29,00:51:16 | INFO | open_clip_train.train | Train Epoch: 29 [ 45907968/268959744 (17%)] Data (t): 0.913 Batch (t): 1.561, 20961.7/s, 2620.21/s/gpu LR: 0.000010 Logit Scale: 100.000 Contrastive_loss: 0.27626 (0.29482) Loss: 0.27626 (0.29482) Logit_scale: 100.00 (99.998)
2026-05-29,00:53:53 | INFO | open_clip_train.train | Train Epoch: 29 [ 49184768/268959744 (18%)] Data (t): 0.877 Batch (t): 1.569, 20902.9/s, 2612.86/s/gpu LR: 0.000010 Logit Scale: 100.000 Contrastive_loss: 0.29967 (0.29513) Loss: 0.29967 (0.29513) Logit_scale: 100.00 (99.998)
2026-05-29,00:56:28 | INFO | open_clip_train.train | Train Epoch: 29 [ 52461568/268959744 (20%)] Data (t): 0.918 Batch (t): 1.550, 23615.0/s, 2951.87/s/gpu LR: 0.000010 Logit Scale: 100.000 Contrastive_loss: 0.27342 (0.29385) Loss: 0.27342 (0.29385) Logit_scale: 100.00 (99.998)
2026-05-29,00:59:04 | INFO | open_clip_train.train | Train Epoch: 29 [ 55738368/268959744 (21%)] Data (t): 0.886 Batch (t): 1.563, 21314.9/s, 2664.37/s/gpu LR: 0.000009 Logit Scale: 100.000 Contrastive_loss: 0.27069 (0.29256) Loss: 0.27069 (0.29256) Logit_scale: 100.00 (99.998)
2026-05-29,01:01:40 | INFO | open_clip_train.train | Train Epoch: 29 [ 59015168/268959744 (22%)] Data (t): 0.861 Batch (t): 1.556, 21058.4/s, 2632.30/s/gpu LR: 0.000009 Logit Scale: 100.000 Contrastive_loss: 0.27148 (0.29145) Loss: 0.27148 (0.29145) Logit_scale: 100.00 (99.998)
2026-05-29,01:04:18 | INFO | open_clip_train.train | Train Epoch: 29 [ 62291968/268959744 (23%)] Data (t): 0.887 Batch (t): 1.578, 22447.3/s, 2805.91/s/gpu LR: 0.000009 Logit Scale: 100.000 Contrastive_loss: 0.27975 (0.29087) Loss: 0.27975 (0.29087) Logit_scale: 100.00 (99.999)
2026-05-29,01:06:54 | INFO | open_clip_train.train | Train Epoch: 29 [ 65568768/268959744 (24%)] Data (t): 0.868 Batch (t): 1.569, 19795.7/s, 2474.46/s/gpu LR: 0.000009 Logit Scale: 100.000 Contrastive_loss: 0.29018 (0.29084) Loss: 0.29018 (0.29084) Logit_scale: 100.00 (99.999)
2026-05-29,01:09:31 | INFO | open_clip_train.train | Train Epoch: 29 [ 68845568/268959744 (26%)] Data (t): 0.886 Batch (t): 1.568, 20372.6/s, 2546.58/s/gpu LR: 0.000009 Logit Scale: 100.000 Contrastive_loss: 0.29367 (0.29096) Loss: 0.29367 (0.29096) Logit_scale: 100.00 (99.999)
2026-05-29,01:12:09 | INFO | open_clip_train.train | Train Epoch: 29 [ 72122368/268959744 (27%)] Data (t): 0.934 Batch (t): 1.575, 20872.7/s, 2609.09/s/gpu LR: 0.000009 Logit Scale: 100.000 Contrastive_loss: 0.29161 (0.29099) Loss: 0.29161 (0.29099) Logit_scale: 100.00 (99.999)
2026-05-29,01:14:45 | INFO | open_clip_train.train | Train Epoch: 29 [ 75399168/268959744 (28%)] Data (t): 0.908 Batch (t): 1.558, 21409.2/s, 2676.15/s/gpu LR: 0.000009 Logit Scale: 100.000 Contrastive_loss: 0.30433 (0.29155) Loss: 0.30433 (0.29155) Logit_scale: 100.00 (99.999)
2026-05-29,01:17:21 | INFO | open_clip_train.train | Train Epoch: 29 [ 78675968/268959744 (29%)] Data (t): 0.868 Batch (t): 1.565, 20942.0/s, 2617.75/s/gpu LR: 0.000009 Logit Scale: 100.000 Contrastive_loss: 0.30471 (0.29208) Loss: 0.30471 (0.29208) Logit_scale: 100.00 (99.999)
2026-05-29,01:19:57 | INFO | open_clip_train.train | Train Epoch: 29 [ 81952768/268959744 (30%)] Data (t): 0.951 Batch (t): 1.563, 22251.2/s, 2781.40/s/gpu LR: 0.000009 Logit Scale: 100.000 Contrastive_loss: 0.31112 (0.29281) Loss: 0.31112 (0.29281) Logit_scale: 100.00 (99.999)
2026-05-29,01:22:34 | INFO | open_clip_train.train | Train Epoch: 29 [ 85229568/268959744 (32%)] Data (t): 0.842 Batch (t): 1.569, 21331.9/s, 2666.48/s/gpu LR: 0.000009 Logit Scale: 100.000 Contrastive_loss: 0.29520 (0.29290) Loss: 0.29520 (0.29290) Logit_scale: 100.00 (99.999)
2026-05-29,01:25:10 | INFO | open_clip_train.train | Train Epoch: 29 [ 88506368/268959744 (33%)] Data (t): 0.860 Batch (t): 1.553, 21274.1/s, 2659.26/s/gpu LR: 0.000009 Logit Scale: 100.000 Contrastive_loss: 0.29340 (0.29291) Loss: 0.29340 (0.29291) Logit_scale: 100.00 (99.999)
2026-05-29,01:27:42 | INFO | open_clip_train.train | Train Epoch: 29 [ 91783168/268959744 (34%)] Data (t): 0.894 Batch (t): 1.529, 20440.8/s, 2555.09/s/gpu LR: 0.000009 Logit Scale: 100.000 Contrastive_loss: 0.27129 (0.29217) Loss: 0.27129 (0.29217) Logit_scale: 100.00 (99.999)
2026-05-29,01:30:19 | INFO | open_clip_train.train | Train Epoch: 29 [ 95059968/268959744 (35%)] Data (t): 0.869 Batch (t): 1.568, 20957.6/s, 2619.69/s/gpu LR: 0.000009 Logit Scale: 100.000 Contrastive_loss: 0.29752 (0.29235) Loss: 0.29752 (0.29235) Logit_scale: 100.00 (99.999)
2026-05-29,01:32:55 | INFO | open_clip_train.train | Train Epoch: 29 [ 98336768/268959744 (37%)] Data (t): 0.934 Batch (t): 1.559, 19939.1/s, 2492.39/s/gpu LR: 0.000008 Logit Scale: 100.000 Contrastive_loss: 0.28622 (0.29215) Loss: 0.28622 (0.29215) Logit_scale: 100.00 (99.999)
2026-05-29,01:35:32 | INFO | open_clip_train.train | Train Epoch: 29 [101613568/268959744 (38%)] Data (t): 0.842 Batch (t): 1.572, 24028.1/s, 3003.51/s/gpu LR: 0.000008 Logit Scale: 100.000 Contrastive_loss: 0.26375 (0.29126) Loss: 0.26375 (0.29126) Logit_scale: 100.00 (99.999)
2026-05-29,01:38:09 | INFO | open_clip_train.train | Train Epoch: 29 [104890368/268959744 (39%)] Data (t): 0.894 Batch (t): 1.567, 21601.8/s, 2700.23/s/gpu LR: 0.000008 Logit Scale: 100.000 Contrastive_loss: 0.28232 (0.29099) Loss: 0.28232 (0.29099) Logit_scale: 100.00 (99.999)
2026-05-29,01:40:46 | INFO | open_clip_train.train | Train Epoch: 29 [108167168/268959744 (40%)] Data (t): 0.922 Batch (t): 1.565, 21487.5/s, 2685.94/s/gpu LR: 0.000008 Logit Scale: 100.000 Contrastive_loss: 0.30828 (0.29150) Loss: 0.30828 (0.29150) Logit_scale: 100.00 (99.999)
2026-05-29,01:43:22 | INFO | open_clip_train.train | Train Epoch: 29 [111443968/268959744 (41%)] Data (t): 0.822 Batch (t): 1.564, 20808.7/s, 2601.09/s/gpu LR: 0.000008 Logit Scale: 100.000 Contrastive_loss: 0.30285 (0.29182) Loss: 0.30285 (0.29182) Logit_scale: 100.00 (99.999)
2026-05-29,01:45:58 | INFO | open_clip_train.train | Train Epoch: 29 [114720768/268959744 (43%)] Data (t): 0.837 Batch (t): 1.562, 20644.4/s, 2580.55/s/gpu LR: 0.000008 Logit Scale: 100.000 Contrastive_loss: 0.28617 (0.29167) Loss: 0.28617 (0.29167) Logit_scale: 100.00 (99.999)
2026-05-29,01:48:35 | INFO | open_clip_train.train | Train Epoch: 29 [117997568/268959744 (44%)] Data (t): 0.885 Batch (t): 1.565, 21750.0/s, 2718.75/s/gpu LR: 0.000008 Logit Scale: 100.000 Contrastive_loss: 0.28405 (0.29146) Loss: 0.28405 (0.29146) Logit_scale: 100.00 (99.999)
2026-05-29,01:51:09 | INFO | open_clip_train.train | Train Epoch: 29 [121274368/268959744 (45%)] Data (t): 0.879 Batch (t): 1.547, 21475.8/s, 2684.48/s/gpu LR: 0.000008 Logit Scale: 100.000 Contrastive_loss: 0.30488 (0.29181) Loss: 0.30488 (0.29181) Logit_scale: 100.00 (99.999)
2026-05-29,01:53:43 | INFO | open_clip_train.train | Train Epoch: 29 [124551168/268959744 (46%)] Data (t): 0.887 Batch (t): 1.542, 21635.6/s, 2704.45/s/gpu LR: 0.000008 Logit Scale: 100.000 Contrastive_loss: 0.27678 (0.29143) Loss: 0.27678 (0.29143) Logit_scale: 100.00 (99.999)
2026-05-29,01:56:17 | INFO | open_clip_train.train | Train Epoch: 29 [127827968/268959744 (48%)] Data (t): 0.875 Batch (t): 1.539, 22062.2/s, 2757.78/s/gpu LR: 0.000008 Logit Scale: 100.000 Contrastive_loss: 0.27142 (0.29093) Loss: 0.27142 (0.29093) Logit_scale: 100.00 (99.999)
2026-05-29,01:58:48 | INFO | open_clip_train.train | Train Epoch: 29 [131104768/268959744 (49%)] Data (t): 0.823 Batch (t): 1.512, 22972.3/s, 2871.53/s/gpu LR: 0.000008 Logit Scale: 100.000 Contrastive_loss: 0.27374 (0.29051) Loss: 0.27374 (0.29051) Logit_scale: 100.00 (99.999)
2026-05-29,02:01:24 | INFO | open_clip_train.train | Train Epoch: 29 [134381568/268959744 (50%)] Data (t): 0.793 Batch (t): 1.555, 21289.7/s, 2661.21/s/gpu LR: 0.000008 Logit Scale: 100.000 Contrastive_loss: 0.31845 (0.29117) Loss: 0.31845 (0.29117) Logit_scale: 100.00 (99.999)
2026-05-29,02:03:59 | INFO | open_clip_train.train | Train Epoch: 29 [137658368/268959744 (51%)] Data (t): 0.904 Batch (t): 1.553, 22195.8/s, 2774.47/s/gpu LR: 0.000008 Logit Scale: 100.000 Contrastive_loss: 0.30378 (0.29147) Loss: 0.30378 (0.29147) Logit_scale: 100.00 (99.999)
2026-05-29,02:06:33 | INFO | open_clip_train.train | Train Epoch: 29 [140935168/268959744 (52%)] Data (t): 0.907 Batch (t): 1.535, 20672.9/s, 2584.11/s/gpu LR: 0.000007 Logit Scale: 100.000 Contrastive_loss: 0.28817 (0.29139) Loss: 0.28817 (0.29139) Logit_scale: 100.00 (99.999)
2026-05-29,02:09:09 | INFO | open_clip_train.train | Train Epoch: 29 [144211968/268959744 (54%)] Data (t): 0.814 Batch (t): 1.566, 21249.6/s, 2656.20/s/gpu LR: 0.000007 Logit Scale: 100.000 Contrastive_loss: 0.29614 (0.29150) Loss: 0.29614 (0.29150) Logit_scale: 100.00 (99.999)
2026-05-29,02:11:50 | INFO | open_clip_train.train | Train Epoch: 29 [147488768/268959744 (55%)] Data (t): 0.861 Batch (t): 1.601, 20970.7/s, 2621.34/s/gpu LR: 0.000007 Logit Scale: 100.000 Contrastive_loss: 0.26524 (0.29093) Loss: 0.26524 (0.29093) Logit_scale: 100.00 (99.999)
2026-05-29,02:14:25 | INFO | open_clip_train.train | Train Epoch: 29 [150765568/268959744 (56%)] Data (t): 0.874 Batch (t): 1.555, 21130.3/s, 2641.29/s/gpu LR: 0.000007 Logit Scale: 100.000 Contrastive_loss: 0.26728 (0.29042) Loss: 0.26728 (0.29042) Logit_scale: 100.00 (99.999)
2026-05-29,02:16:59 | INFO | open_clip_train.train | Train Epoch: 29 [154042368/268959744 (57%)] Data (t): 0.926 Batch (t): 1.540, 20807.7/s, 2600.96/s/gpu LR: 0.000007 Logit Scale: 100.000 Contrastive_loss: 0.25806 (0.28975) Loss: 0.25806 (0.28975) Logit_scale: 100.00 (99.999)
2026-05-29,02:19:30 | INFO | open_clip_train.train | Train Epoch: 29 [157319168/268959744 (58%)] Data (t): 0.813 Batch (t): 1.513, 21942.3/s, 2742.79/s/gpu LR: 0.000007 Logit Scale: 100.000 Contrastive_loss: 0.31009 (0.29016) Loss: 0.31009 (0.29016) Logit_scale: 100.00 (99.999)
2026-05-29,02:22:06 | INFO | open_clip_train.train | Train Epoch: 29 [160595968/268959744 (60%)] Data (t): 0.792 Batch (t): 1.562, 21238.9/s, 2654.86/s/gpu LR: 0.000007 Logit Scale: 100.000 Contrastive_loss: 0.28274 (0.29002) Loss: 0.28274 (0.29002) Logit_scale: 100.00 (99.999)
2026-05-29,02:24:42 | INFO | open_clip_train.train | Train Epoch: 29 [163872768/268959744 (61%)] Data (t): 0.899 Batch (t): 1.551, 21513.3/s, 2689.17/s/gpu LR: 0.000007 Logit Scale: 100.000 Contrastive_loss: 0.25844 (0.28940) Loss: 0.25844 (0.28940) Logit_scale: 100.00 (99.999)
2026-05-29,02:27:18 | INFO | open_clip_train.train | Train Epoch: 29 [167149568/268959744 (62%)] Data (t): 0.897 Batch (t): 1.559, 20675.0/s, 2584.37/s/gpu LR: 0.000007 Logit Scale: 100.000 Contrastive_loss: 0.28045 (0.28922) Loss: 0.28045 (0.28922) Logit_scale: 100.00 (99.999)
2026-05-29,02:29:51 | INFO | open_clip_train.train | Train Epoch: 29 [170426368/268959744 (63%)] Data (t): 0.846 Batch (t): 1.535, 21863.2/s, 2732.90/s/gpu LR: 0.000007 Logit Scale: 100.000 Contrastive_loss: 0.30468 (0.28952) Loss: 0.30468 (0.28952) Logit_scale: 100.00 (99.999)
2026-05-29,02:32:24 | INFO | open_clip_train.train | Train Epoch: 29 [173703168/268959744 (65%)] Data (t): 0.913 Batch (t): 1.534, 21099.5/s, 2637.44/s/gpu LR: 0.000007 Logit Scale: 100.000 Contrastive_loss: 0.30336 (0.28977) Loss: 0.30336 (0.28977) Logit_scale: 100.00 (99.999)
2026-05-29,02:35:01 | INFO | open_clip_train.train | Train Epoch: 29 [176979968/268959744 (66%)] Data (t): 0.905 Batch (t): 1.568, 20857.9/s, 2607.24/s/gpu LR: 0.000007 Logit Scale: 100.000 Contrastive_loss: 0.29438 (0.28986) Loss: 0.29438 (0.28986) Logit_scale: 100.00 (99.999)
2026-05-29,02:37:37 | INFO | open_clip_train.train | Train Epoch: 29 [180256768/268959744 (67%)] Data (t): 0.922 Batch (t): 1.562, 22364.3/s, 2795.54/s/gpu LR: 0.000007 Logit Scale: 100.000 Contrastive_loss: 0.28782 (0.28982) Loss: 0.28782 (0.28982) Logit_scale: 100.00 (99.999)
2026-05-29,02:40:13 | INFO | open_clip_train.train | Train Epoch: 29 [183533568/268959744 (68%)] Data (t): 0.950 Batch (t): 1.558, 20990.6/s, 2623.82/s/gpu LR: 0.000007 Logit Scale: 100.000 Contrastive_loss: 0.29147 (0.28985) Loss: 0.29147 (0.28985) Logit_scale: 100.00 (99.999)
2026-05-29,02:42:49 | INFO | open_clip_train.train | Train Epoch: 29 [186810368/268959744 (69%)] Data (t): 0.886 Batch (t): 1.552, 21282.1/s, 2660.26/s/gpu LR: 0.000006 Logit Scale: 100.000 Contrastive_loss: 0.30034 (0.29003) Loss: 0.30034 (0.29003) Logit_scale: 100.00 (100.00)
2026-05-29,02:45:22 | INFO | open_clip_train.train | Train Epoch: 29 [190087168/268959744 (71%)] Data (t): 0.874 Batch (t): 1.534, 23716.3/s, 2964.53/s/gpu LR: 0.000006 Logit Scale: 100.000 Contrastive_loss: 0.30574 (0.29030) Loss: 0.30574 (0.29030) Logit_scale: 100.00 (100.00)
2026-05-29,02:47:57 | INFO | open_clip_train.train | Train Epoch: 29 [193363968/268959744 (72%)] Data (t): 0.787 Batch (t): 1.552, 20126.1/s, 2515.77/s/gpu LR: 0.000006 Logit Scale: 100.000 Contrastive_loss: 0.28875 (0.29027) Loss: 0.28875 (0.29027) Logit_scale: 100.00 (100.00)
2026-05-29,02:50:34 | INFO | open_clip_train.train | Train Epoch: 29 [196640768/268959744 (73%)] Data (t): 0.918 Batch (t): 1.565, 20582.6/s, 2572.82/s/gpu LR: 0.000006 Logit Scale: 100.000 Contrastive_loss: 0.30345 (0.29049) Loss: 0.30345 (0.29049) Logit_scale: 100.00 (100.00)
2026-05-29,02:53:09 | INFO | open_clip_train.train | Train Epoch: 29 [199917568/268959744 (74%)] Data (t): 0.914 Batch (t): 1.551, 19798.5/s, 2474.81/s/gpu LR: 0.000006 Logit Scale: 100.000 Contrastive_loss: 0.26757 (0.29012) Loss: 0.26757 (0.29012) Logit_scale: 100.00 (100.00)
2026-05-29,02:55:44 | INFO | open_clip_train.train | Train Epoch: 29 [203194368/268959744 (76%)] Data (t): 0.859 Batch (t): 1.552, 22732.4/s, 2841.55/s/gpu LR: 0.000006 Logit Scale: 100.000 Contrastive_loss: 0.27367 (0.28986) Loss: 0.27367 (0.28986) Logit_scale: 100.00 (100.00)
2026-05-29,02:58:20 | INFO | open_clip_train.train | Train Epoch: 29 [206471168/268959744 (77%)] Data (t): 0.881 Batch (t): 1.558, 22216.5/s, 2777.06/s/gpu LR: 0.000006 Logit Scale: 100.000 Contrastive_loss: 0.28887 (0.28984) Loss: 0.28887 (0.28984) Logit_scale: 100.00 (100.00)
2026-05-29,03:00:56 | INFO | open_clip_train.train | Train Epoch: 29 [209747968/268959744 (78%)] Data (t): 0.876 Batch (t): 1.566, 21210.0/s, 2651.25/s/gpu LR: 0.000006 Logit Scale: 100.000 Contrastive_loss: 0.29586 (0.28993) Loss: 0.29586 (0.28993) Logit_scale: 100.00 (100.00)
2026-05-29,03:03:31 | INFO | open_clip_train.train | Train Epoch: 29 [213024768/268959744 (79%)] Data (t): 0.883 Batch (t): 1.548, 20713.5/s, 2589.19/s/gpu LR: 0.000006 Logit Scale: 100.000 Contrastive_loss: 0.27313 (0.28968) Loss: 0.27313 (0.28968) Logit_scale: 100.00 (100.00)
2026-05-29,03:06:07 | INFO | open_clip_train.train | Train Epoch: 29 [216301568/268959744 (80%)] Data (t): 0.896 Batch (t): 1.560, 20965.6/s, 2620.69/s/gpu LR: 0.000006 Logit Scale: 100.000 Contrastive_loss: 0.29062 (0.28969) Loss: 0.29062 (0.28969) Logit_scale: 100.00 (100.00)
2026-05-29,03:08:43 | INFO | open_clip_train.train | Train Epoch: 29 [219578368/268959744 (82%)] Data (t): 0.848 Batch (t): 1.559, 20442.8/s, 2555.35/s/gpu LR: 0.000006 Logit Scale: 100.000 Contrastive_loss: 0.28539 (0.28963) Loss: 0.28539 (0.28963) Logit_scale: 100.00 (100.00)
2026-05-29,03:11:20 | INFO | open_clip_train.train | Train Epoch: 29 [222855168/268959744 (83%)] Data (t): 0.872 Batch (t): 1.565, 21245.6/s, 2655.70/s/gpu LR: 0.000006 Logit Scale: 100.000 Contrastive_loss: 0.28091 (0.28950) Loss: 0.28091 (0.28950) Logit_scale: 100.00 (100.00)
2026-05-29,03:13:55 | INFO | open_clip_train.train | Train Epoch: 29 [226131968/268959744 (84%)] Data (t): 0.848 Batch (t): 1.554, 20821.3/s, 2602.66/s/gpu LR: 0.000006 Logit Scale: 100.000 Contrastive_loss: 0.30094 (0.28967) Loss: 0.30094 (0.28967) Logit_scale: 100.00 (100.00)
2026-05-29,03:16:31 | INFO | open_clip_train.train | Train Epoch: 29 [229408768/268959744 (85%)] Data (t): 0.856 Batch (t): 1.559, 18969.8/s, 2371.23/s/gpu LR: 0.000006 Logit Scale: 100.000 Contrastive_loss: 0.28110 (0.28955) Loss: 0.28110 (0.28955) Logit_scale: 100.00 (100.00)
2026-05-29,03:19:07 | INFO | open_clip_train.train | Train Epoch: 29 [232685568/268959744 (87%)] Data (t): 0.887 Batch (t): 1.562, 21129.4/s, 2641.17/s/gpu LR: 0.000006 Logit Scale: 100.000 Contrastive_loss: 0.24543 (0.28893) Loss: 0.24543 (0.28893) Logit_scale: 100.00 (100.00)
2026-05-29,03:21:42 | INFO | open_clip_train.train | Train Epoch: 29 [235962368/268959744 (88%)] Data (t): 0.897 Batch (t): 1.547, 20437.6/s, 2554.70/s/gpu LR: 0.000005 Logit Scale: 100.000 Contrastive_loss: 0.25029 (0.28840) Loss: 0.25029 (0.28840) Logit_scale: 100.00 (100.00)
2026-05-29,03:24:17 | INFO | open_clip_train.train | Train Epoch: 29 [239239168/268959744 (89%)] Data (t): 0.917 Batch (t): 1.556, 22426.1/s, 2803.26/s/gpu LR: 0.000005 Logit Scale: 100.000 Contrastive_loss: 0.26320 (0.28806) Loss: 0.26320 (0.28806) Logit_scale: 100.00 (100.00)
2026-05-29,03:26:54 | INFO | open_clip_train.train | Train Epoch: 29 [242515968/268959744 (90%)] Data (t): 0.828 Batch (t): 1.568, 21486.0/s, 2685.75/s/gpu LR: 0.000005 Logit Scale: 100.000 Contrastive_loss: 0.30529 (0.28829) Loss: 0.30529 (0.28829) Logit_scale: 100.00 (100.00)
2026-05-29,03:29:30 | INFO | open_clip_train.train | Train Epoch: 29 [245792768/268959744 (91%)] Data (t): 0.900 Batch (t): 1.557, 21507.2/s, 2688.40/s/gpu LR: 0.000005 Logit Scale: 100.000 Contrastive_loss: 0.29693 (0.28841) Loss: 0.29693 (0.28841) Logit_scale: 100.00 (100.00)
2026-05-29,03:32:04 | INFO | open_clip_train.train | Train Epoch: 29 [249069568/268959744 (93%)] Data (t): 0.916 Batch (t): 1.543, 20698.4/s, 2587.30/s/gpu LR: 0.000005 Logit Scale: 100.000 Contrastive_loss: 0.29346 (0.28847) Loss: 0.29346 (0.28847) Logit_scale: 100.00 (100.00)
2026-05-29,03:34:41 | INFO | open_clip_train.train | Train Epoch: 29 [252346368/268959744 (94%)] Data (t): 0.826 Batch (t): 1.563, 21601.6/s, 2700.20/s/gpu LR: 0.000005 Logit Scale: 100.000 Contrastive_loss: 0.27933 (0.28835) Loss: 0.27933 (0.28835) Logit_scale: 100.00 (100.00)
2026-05-29,03:37:15 | INFO | open_clip_train.train | Train Epoch: 29 [255623168/268959744 (95%)] Data (t): 0.851 Batch (t): 1.547, 25279.0/s, 3159.87/s/gpu LR: 0.000005 Logit Scale: 100.000 Contrastive_loss: 0.32008 (0.28876) Loss: 0.32008 (0.28876) Logit_scale: 100.00 (100.00)
2026-05-29,03:39:52 | INFO | open_clip_train.train | Train Epoch: 29 [258899968/268959744 (96%)] Data (t): 0.876 Batch (t): 1.563, 21662.5/s, 2707.82/s/gpu LR: 0.000005 Logit Scale: 99.997 Contrastive_loss: 0.33472 (0.28933) Loss: 0.33472 (0.28933) Logit_scale: 99.997 (100.00)
2026-05-29,03:42:28 | INFO | open_clip_train.train | Train Epoch: 29 [262176768/268959744 (97%)] Data (t): 0.938 Batch (t): 1.564, 21035.3/s, 2629.41/s/gpu LR: 0.000005 Logit Scale: 99.997 Contrastive_loss: 0.31087 (0.28960) Loss: 0.31087 (0.28960) Logit_scale: 99.997 (100.00)
2026-05-29,03:45:01 | INFO | open_clip_train.train | Train Epoch: 29 [265453568/268959744 (99%)] Data (t): 0.914 Batch (t): 1.534, 21284.8/s, 2660.61/s/gpu LR: 0.000005 Logit Scale: 99.983 Contrastive_loss: 0.39631 (0.29090) Loss: 0.39631 (0.29090) Logit_scale: 99.983 (99.999)
2026-05-29,03:47:31 | INFO | open_clip_train.train | Train Epoch: 29 [268730368/268959744 (100%)] Data (t): 0.841 Batch (t): 1.495, 29697.2/s, 3712.14/s/gpu LR: 0.000005 Logit Scale: 99.963 Contrastive_loss: 0.34592 (0.29156) Loss: 0.34592 (0.29156) Logit_scale: 99.963 (99.999)
2026-05-29,03:47:39 | INFO | open_clip_train.train | Train Epoch: 29 [268959744/268959744 (100%)] Data (t): 0.637 Batch (t): 1.127, 28924.2/s, 3615.53/s/gpu LR: 0.000005 Logit Scale: 99.961 Contrastive_loss: 0.39294 (0.29277) Loss: 0.39294 (0.29277) Logit_scale: 99.961 (99.998)
2026-05-29,03:47:43 | INFO | open_clip_train.main | Start epoch 30
2026-05-29,03:48:05 | INFO | open_clip_train.train | Train Epoch: 30 [ 32768/268959744 (0%)] Data (t): 18.811 Batch (t): 21.534, 1521.69/s, 190.211/s/gpu LR: 0.000005 Logit Scale: 99.960 Contrastive_loss: 0.26481 (0.26481) Loss: 0.26481 (0.26481) Logit_scale: 99.960 (99.960)
2026-05-29,03:50:43 | INFO | open_clip_train.train | Train Epoch: 30 [ 3309568/268959744 (1%)] Data (t): 0.785 Batch (t): 1.574, 22565.8/s, 2820.73/s/gpu LR: 0.000005 Logit Scale: 99.999 Contrastive_loss: 0.27093 (0.26787) Loss: 0.27093 (0.26787) Logit_scale: 99.999 (99.980)
2026-05-29,03:53:20 | INFO | open_clip_train.train | Train Epoch: 30 [ 6586368/268959744 (2%)] Data (t): 0.840 Batch (t): 1.572, 20698.9/s, 2587.37/s/gpu LR: 0.000005 Logit Scale: 100.000 Contrastive_loss: 0.30553 (0.28043) Loss: 0.30553 (0.28043) Logit_scale: 100.00 (99.986)
2026-05-29,03:55:56 | INFO | open_clip_train.train | Train Epoch: 30 [ 9863168/268959744 (4%)] Data (t): 0.782 Batch (t): 1.561, 20634.9/s, 2579.37/s/gpu LR: 0.000005 Logit Scale: 100.000 Contrastive_loss: 0.27109 (0.27809) Loss: 0.27109 (0.27809) Logit_scale: 100.00 (99.990)
2026-05-29,03:58:31 | INFO | open_clip_train.train | Train Epoch: 30 [ 13139968/268959744 (5%)] Data (t): 0.906 Batch (t): 1.546, 22141.4/s, 2767.67/s/gpu LR: 0.000005 Logit Scale: 100.000 Contrastive_loss: 0.27797 (0.27807) Loss: 0.27797 (0.27807) Logit_scale: 100.00 (99.992)
2026-05-29,04:01:06 | INFO | open_clip_train.train | Train Epoch: 30 [ 16416768/268959744 (6%)] Data (t): 0.908 Batch (t): 1.559, 20899.1/s, 2612.38/s/gpu LR: 0.000005 Logit Scale: 100.000 Contrastive_loss: 0.27012 (0.27674) Loss: 0.27012 (0.27674) Logit_scale: 100.00 (99.993)
2026-05-29,04:03:41 | INFO | open_clip_train.train | Train Epoch: 30 [ 19693568/268959744 (7%)] Data (t): 0.895 Batch (t): 1.549, 21668.5/s, 2708.56/s/gpu LR: 0.000005 Logit Scale: 100.000 Contrastive_loss: 0.29317 (0.27909) Loss: 0.29317 (0.27909) Logit_scale: 100.00 (99.994)
2026-05-29,04:06:16 | INFO | open_clip_train.train | Train Epoch: 30 [ 22970368/268959744 (9%)] Data (t): 0.882 Batch (t): 1.547, 20911.9/s, 2613.98/s/gpu LR: 0.000004 Logit Scale: 100.000 Contrastive_loss: 0.30374 (0.28217) Loss: 0.30374 (0.28217) Logit_scale: 100.00 (99.995)
2026-05-29,04:08:51 | INFO | open_clip_train.train | Train Epoch: 30 [ 26247168/268959744 (10%)] Data (t): 0.865 Batch (t): 1.553, 21221.0/s, 2652.63/s/gpu LR: 0.000004 Logit Scale: 100.000 Contrastive_loss: 0.27559 (0.28144) Loss: 0.27559 (0.28144) Logit_scale: 100.00 (99.995)
2026-05-29,04:11:25 | INFO | open_clip_train.train | Train Epoch: 30 [ 29523968/268959744 (11%)] Data (t): 0.873 Batch (t): 1.535, 20659.5/s, 2582.43/s/gpu LR: 0.000004 Logit Scale: 100.000 Contrastive_loss: 0.24692 (0.27799) Loss: 0.24692 (0.27799) Logit_scale: 100.00 (99.996)
2026-05-29,04:14:00 | INFO | open_clip_train.train | Train Epoch: 30 [ 32800768/268959744 (12%)] Data (t): 0.951 Batch (t): 1.553, 21359.9/s, 2669.99/s/gpu LR: 0.000004 Logit Scale: 100.000 Contrastive_loss: 0.27672 (0.27787) Loss: 0.27672 (0.27787) Logit_scale: 100.00 (99.996)
2026-05-29,04:16:35 | INFO | open_clip_train.train | Train Epoch: 30 [ 36077568/268959744 (13%)] Data (t): 0.856 Batch (t): 1.553, 20496.0/s, 2562.00/s/gpu LR: 0.000004 Logit Scale: 100.000 Contrastive_loss: 0.28107 (0.27814) Loss: 0.28107 (0.27814) Logit_scale: 100.00 (99.997)
2026-05-29,04:19:10 | INFO | open_clip_train.train | Train Epoch: 30 [ 39354368/268959744 (15%)] Data (t): 0.832 Batch (t): 1.542, 21135.6/s, 2641.95/s/gpu LR: 0.000004 Logit Scale: 100.000 Contrastive_loss: 0.29791 (0.27966) Loss: 0.29791 (0.27966) Logit_scale: 100.00 (99.997)
2026-05-29,04:21:46 | INFO | open_clip_train.train | Train Epoch: 30 [ 42631168/268959744 (16%)] Data (t): 0.879 Batch (t): 1.561, 20749.8/s, 2593.73/s/gpu LR: 0.000004 Logit Scale: 100.000 Contrastive_loss: 0.25847 (0.27815) Loss: 0.25847 (0.27815) Logit_scale: 100.00 (99.997)
2026-05-29,04:24:20 | INFO | open_clip_train.train | Train Epoch: 30 [ 45907968/268959744 (17%)] Data (t): 0.903 Batch (t): 1.544, 21582.5/s, 2697.82/s/gpu LR: 0.000004 Logit Scale: 100.000 Contrastive_loss: 0.25917 (0.27688) Loss: 0.25917 (0.27688) Logit_scale: 100.00 (99.997)
2026-05-29,04:26:54 | INFO | open_clip_train.train | Train Epoch: 30 [ 49184768/268959744 (18%)] Data (t): 0.902 Batch (t): 1.543, 22523.7/s, 2815.46/s/gpu LR: 0.000004 Logit Scale: 100.000 Contrastive_loss: 0.27096 (0.27651) Loss: 0.27096 (0.27651) Logit_scale: 100.00 (99.997)
2026-05-29,04:29:29 | INFO | open_clip_train.train | Train Epoch: 30 [ 52461568/268959744 (20%)] Data (t): 0.872 Batch (t): 1.547, 21297.9/s, 2662.23/s/gpu LR: 0.000004 Logit Scale: 100.000 Contrastive_loss: 0.29983 (0.27788) Loss: 0.29983 (0.27788) Logit_scale: 100.00 (99.998)
2026-05-29,04:32:02 | INFO | open_clip_train.train | Train Epoch: 30 [ 55738368/268959744 (21%)] Data (t): 0.830 Batch (t): 1.524, 21623.3/s, 2702.91/s/gpu LR: 0.000004 Logit Scale: 100.000 Contrastive_loss: 0.28580 (0.27832) Loss: 0.28580 (0.27832) Logit_scale: 100.00 (99.998)
2026-05-29,04:34:36 | INFO | open_clip_train.train | Train Epoch: 30 [ 59015168/268959744 (22%)] Data (t): 0.862 Batch (t): 1.546, 20648.6/s, 2581.08/s/gpu LR: 0.000004 Logit Scale: 100.000 Contrastive_loss: 0.28341 (0.27859) Loss: 0.28341 (0.27859) Logit_scale: 100.00 (99.998)
2026-05-29,04:37:12 | INFO | open_clip_train.train | Train Epoch: 30 [ 62291968/268959744 (23%)] Data (t): 0.819 Batch (t): 1.559, 21487.2/s, 2685.90/s/gpu LR: 0.000004 Logit Scale: 100.000 Contrastive_loss: 0.28253 (0.27879) Loss: 0.28253 (0.27879) Logit_scale: 100.00 (99.998)
2026-05-29,04:39:48 | INFO | open_clip_train.train | Train Epoch: 30 [ 65568768/268959744 (24%)] Data (t): 0.881 Batch (t): 1.556, 20904.0/s, 2613.00/s/gpu LR: 0.000004 Logit Scale: 100.000 Contrastive_loss: 0.26653 (0.27820) Loss: 0.26653 (0.27820) Logit_scale: 100.00 (99.998)
2026-05-29,04:42:24 | INFO | open_clip_train.train | Train Epoch: 30 [ 68845568/268959744 (26%)] Data (t): 0.854 Batch (t): 1.562, 21876.9/s, 2734.61/s/gpu LR: 0.000004 Logit Scale: 100.000 Contrastive_loss: 0.26102 (0.27742) Loss: 0.26102 (0.27742) Logit_scale: 100.00 (99.998)
2026-05-29,04:44:58 | INFO | open_clip_train.train | Train Epoch: 30 [ 72122368/268959744 (27%)] Data (t): 0.836 Batch (t): 1.545, 22101.9/s, 2762.74/s/gpu LR: 0.000004 Logit Scale: 100.000 Contrastive_loss: 0.26669 (0.27696) Loss: 0.26669 (0.27696) Logit_scale: 100.00 (99.998)
2026-05-29,04:47:33 | INFO | open_clip_train.train | Train Epoch: 30 [ 75399168/268959744 (28%)] Data (t): 0.878 Batch (t): 1.543, 20356.2/s, 2544.53/s/gpu LR: 0.000004 Logit Scale: 100.000 Contrastive_loss: 0.27624 (0.27693) Loss: 0.27624 (0.27693) Logit_scale: 100.00 (99.998)
2026-05-29,04:50:09 | INFO | open_clip_train.train | Train Epoch: 30 [ 78675968/268959744 (29%)] Data (t): 0.858 Batch (t): 1.561, 23423.6/s, 2927.94/s/gpu LR: 0.000004 Logit Scale: 100.000 Contrastive_loss: 0.31240 (0.27834) Loss: 0.31240 (0.27834) Logit_scale: 100.00 (99.998)
2026-05-29,04:52:42 | INFO | open_clip_train.train | Train Epoch: 30 [ 81952768/268959744 (30%)] Data (t): 0.889 Batch (t): 1.536, 21010.2/s, 2626.28/s/gpu LR: 0.000004 Logit Scale: 100.000 Contrastive_loss: 0.26544 (0.27785) Loss: 0.26544 (0.27785) Logit_scale: 100.00 (99.998)
2026-05-29,04:55:16 | INFO | open_clip_train.train | Train Epoch: 30 [ 85229568/268959744 (32%)] Data (t): 0.838 Batch (t): 1.533, 20237.2/s, 2529.64/s/gpu LR: 0.000003 Logit Scale: 100.000 Contrastive_loss: 0.27264 (0.27766) Loss: 0.27264 (0.27766) Logit_scale: 100.00 (99.998)
2026-05-29,04:57:48 | INFO | open_clip_train.train | Train Epoch: 30 [ 88506368/268959744 (33%)] Data (t): 0.914 Batch (t): 1.527, 23826.7/s, 2978.34/s/gpu LR: 0.000003 Logit Scale: 100.000 Contrastive_loss: 0.28810 (0.27803) Loss: 0.28810 (0.27803) Logit_scale: 100.00 (99.999)
2026-05-29,05:00:20 | INFO | open_clip_train.train | Train Epoch: 30 [ 91783168/268959744 (34%)] Data (t): 0.917 Batch (t): 1.520, 21953.6/s, 2744.20/s/gpu LR: 0.000003 Logit Scale: 100.000 Contrastive_loss: 0.28089 (0.27813) Loss: 0.28089 (0.27813) Logit_scale: 100.00 (99.999)
2026-05-29,05:02:59 | INFO | open_clip_train.train | Train Epoch: 30 [ 95059968/268959744 (35%)] Data (t): 0.887 Batch (t): 1.585, 21092.2/s, 2636.53/s/gpu LR: 0.000003 Logit Scale: 100.000 Contrastive_loss: 0.26327 (0.27763) Loss: 0.26327 (0.27763) Logit_scale: 100.00 (99.999)
2026-05-29,05:05:36 | INFO | open_clip_train.train | Train Epoch: 30 [ 98336768/268959744 (37%)] Data (t): 0.853 Batch (t): 1.569, 20295.4/s, 2536.93/s/gpu LR: 0.000003 Logit Scale: 100.000 Contrastive_loss: 0.25948 (0.27705) Loss: 0.25948 (0.27705) Logit_scale: 100.00 (99.999)
2026-05-29,05:08:13 | INFO | open_clip_train.train | Train Epoch: 30 [101613568/268959744 (38%)] Data (t): 0.910 Batch (t): 1.569, 21602.0/s, 2700.25/s/gpu LR: 0.000003 Logit Scale: 100.000 Contrastive_loss: 0.27638 (0.27703) Loss: 0.27638 (0.27703) Logit_scale: 100.00 (99.999)
2026-05-29,05:10:48 | INFO | open_clip_train.train | Train Epoch: 30 [104890368/268959744 (39%)] Data (t): 0.865 Batch (t): 1.557, 22563.7/s, 2820.46/s/gpu LR: 0.000003 Logit Scale: 100.000 Contrastive_loss: 0.26648 (0.27671) Loss: 0.26648 (0.27671) Logit_scale: 100.00 (99.999)
2026-05-29,05:13:22 | INFO | open_clip_train.train | Train Epoch: 30 [108167168/268959744 (40%)] Data (t): 0.913 Batch (t): 1.538, 22901.7/s, 2862.71/s/gpu LR: 0.000003 Logit Scale: 100.000 Contrastive_loss: 0.28047 (0.27682) Loss: 0.28047 (0.27682) Logit_scale: 100.00 (99.999)
2026-05-29,05:15:57 | INFO | open_clip_train.train | Train Epoch: 30 [111443968/268959744 (41%)] Data (t): 0.905 Batch (t): 1.550, 20816.2/s, 2602.03/s/gpu LR: 0.000003 Logit Scale: 100.000 Contrastive_loss: 0.29817 (0.27743) Loss: 0.29817 (0.27743) Logit_scale: 100.00 (99.999)
2026-05-29,05:18:35 | INFO | open_clip_train.train | Train Epoch: 30 [114720768/268959744 (43%)] Data (t): 0.882 Batch (t): 1.575, 19340.8/s, 2417.60/s/gpu LR: 0.000003 Logit Scale: 100.000 Contrastive_loss: 0.27770 (0.27743) Loss: 0.27770 (0.27743) Logit_scale: 100.00 (99.999)
2026-05-29,05:21:11 | INFO | open_clip_train.train | Train Epoch: 30 [117997568/268959744 (44%)] Data (t): 0.870 Batch (t): 1.562, 22711.6/s, 2838.95/s/gpu LR: 0.000003 Logit Scale: 100.000 Contrastive_loss: 0.29491 (0.27791) Loss: 0.29491 (0.27791) Logit_scale: 100.00 (99.999)
2026-05-29,05:23:41 | INFO | open_clip_train.train | Train Epoch: 30 [121274368/268959744 (45%)] Data (t): 0.815 Batch (t): 1.497, 21932.3/s, 2741.54/s/gpu LR: 0.000003 Logit Scale: 100.000 Contrastive_loss: 0.27675 (0.27788) Loss: 0.27675 (0.27788) Logit_scale: 100.00 (99.999)
2026-05-29,05:26:14 | INFO | open_clip_train.train | Train Epoch: 30 [124551168/268959744 (46%)] Data (t): 0.807 Batch (t): 1.534, 18952.6/s, 2369.07/s/gpu LR: 0.000003 Logit Scale: 100.000 Contrastive_loss: 0.26911 (0.27765) Loss: 0.26911 (0.27765) Logit_scale: 100.00 (99.999)
2026-05-29,05:28:50 | INFO | open_clip_train.train | Train Epoch: 30 [127827968/268959744 (48%)] Data (t): 0.858 Batch (t): 1.556, 20358.6/s, 2544.83/s/gpu LR: 0.000003 Logit Scale: 100.000 Contrastive_loss: 0.27229 (0.27752) Loss: 0.27229 (0.27752) Logit_scale: 100.00 (99.999)
2026-05-29,05:31:27 | INFO | open_clip_train.train | Train Epoch: 30 [131104768/268959744 (49%)] Data (t): 0.930 Batch (t): 1.571, 21481.5/s, 2685.19/s/gpu LR: 0.000003 Logit Scale: 100.000 Contrastive_loss: 0.26467 (0.27720) Loss: 0.26467 (0.27720) Logit_scale: 100.00 (99.999)
2026-05-29,05:34:02 | INFO | open_clip_train.train | Train Epoch: 30 [134381568/268959744 (50%)] Data (t): 0.901 Batch (t): 1.549, 21173.6/s, 2646.70/s/gpu LR: 0.000003 Logit Scale: 100.000 Contrastive_loss: 0.27237 (0.27709) Loss: 0.27237 (0.27709) Logit_scale: 100.00 (99.999)
2026-05-29,05:36:37 | INFO | open_clip_train.train | Train Epoch: 30 [137658368/268959744 (51%)] Data (t): 0.888 Batch (t): 1.556, 20981.2/s, 2622.65/s/gpu LR: 0.000003 Logit Scale: 100.000 Contrastive_loss: 0.26505 (0.27681) Loss: 0.26505 (0.27681) Logit_scale: 100.00 (99.999)
2026-05-29,05:39:13 | INFO | open_clip_train.train | Train Epoch: 30 [140935168/268959744 (52%)] Data (t): 0.894 Batch (t): 1.553, 21484.7/s, 2685.59/s/gpu LR: 0.000003 Logit Scale: 100.000 Contrastive_loss: 0.27369 (0.27674) Loss: 0.27369 (0.27674) Logit_scale: 100.00 (99.999)
2026-05-29,05:41:48 | INFO | open_clip_train.train | Train Epoch: 30 [144211968/268959744 (54%)] Data (t): 0.851 Batch (t): 1.550, 23163.8/s, 2895.47/s/gpu LR: 0.000003 Logit Scale: 100.000 Contrastive_loss: 0.27367 (0.27667) Loss: 0.27367 (0.27667) Logit_scale: 100.00 (99.999)
2026-05-29,05:44:23 | INFO | open_clip_train.train | Train Epoch: 30 [147488768/268959744 (55%)] Data (t): 0.884 Batch (t): 1.552, 19436.9/s, 2429.62/s/gpu LR: 0.000003 Logit Scale: 100.000 Contrastive_loss: 0.25546 (0.27621) Loss: 0.25546 (0.27621) Logit_scale: 100.00 (99.999)
2026-05-29,05:46:55 | INFO | open_clip_train.train | Train Epoch: 30 [150765568/268959744 (56%)] Data (t): 0.834 Batch (t): 1.524, 21449.6/s, 2681.20/s/gpu LR: 0.000003 Logit Scale: 100.000 Contrastive_loss: 0.27736 (0.27623) Loss: 0.27736 (0.27623) Logit_scale: 100.00 (99.999)
2026-05-29,05:49:30 | INFO | open_clip_train.train | Train Epoch: 30 [154042368/268959744 (57%)] Data (t): 0.881 Batch (t): 1.552, 20420.0/s, 2552.50/s/gpu LR: 0.000002 Logit Scale: 100.000 Contrastive_loss: 0.27105 (0.27613) Loss: 0.27105 (0.27613) Logit_scale: 100.00 (99.999)
2026-05-29,05:52:06 | INFO | open_clip_train.train | Train Epoch: 30 [157319168/268959744 (58%)] Data (t): 0.794 Batch (t): 1.560, 20891.4/s, 2611.42/s/gpu LR: 0.000002 Logit Scale: 100.000 Contrastive_loss: 0.31234 (0.27686) Loss: 0.31234 (0.27686) Logit_scale: 100.00 (99.999)
2026-05-29,05:54:43 | INFO | open_clip_train.train | Train Epoch: 30 [160595968/268959744 (60%)] Data (t): 0.886 Batch (t): 1.563, 20232.0/s, 2529.00/s/gpu LR: 0.000002 Logit Scale: 100.000 Contrastive_loss: 0.26626 (0.27665) Loss: 0.26626 (0.27665) Logit_scale: 100.00 (99.999)
2026-05-29,05:57:19 | INFO | open_clip_train.train | Train Epoch: 30 [163872768/268959744 (61%)] Data (t): 0.789 Batch (t): 1.565, 19214.0/s, 2401.75/s/gpu LR: 0.000002 Logit Scale: 100.000 Contrastive_loss: 0.26736 (0.27647) Loss: 0.26736 (0.27647) Logit_scale: 100.00 (99.999)
2026-05-29,05:59:54 | INFO | open_clip_train.train | Train Epoch: 30 [167149568/268959744 (62%)] Data (t): 0.902 Batch (t): 1.552, 21157.8/s, 2644.72/s/gpu LR: 0.000002 Logit Scale: 100.000 Contrastive_loss: 0.26501 (0.27625) Loss: 0.26501 (0.27625) Logit_scale: 100.00 (99.999)
2026-05-29,06:02:30 | INFO | open_clip_train.train | Train Epoch: 30 [170426368/268959744 (63%)] Data (t): 0.828 Batch (t): 1.559, 20823.3/s, 2602.91/s/gpu LR: 0.000002 Logit Scale: 100.000 Contrastive_loss: 0.29429 (0.27659) Loss: 0.29429 (0.27659) Logit_scale: 100.00 (99.999)
2026-05-29,06:05:07 | INFO | open_clip_train.train | Train Epoch: 30 [173703168/268959744 (65%)] Data (t): 0.885 Batch (t): 1.567, 19673.2/s, 2459.15/s/gpu LR: 0.000002 Logit Scale: 100.000 Contrastive_loss: 0.28757 (0.27679) Loss: 0.28757 (0.27679) Logit_scale: 100.00 (99.999)
2026-05-29,06:07:41 | INFO | open_clip_train.train | Train Epoch: 30 [176979968/268959744 (66%)] Data (t): 0.929 Batch (t): 1.544, 21299.1/s, 2662.39/s/gpu LR: 0.000002 Logit Scale: 100.000 Contrastive_loss: 0.29532 (0.27713) Loss: 0.29532 (0.27713) Logit_scale: 100.00 (99.999)
2026-05-29,06:10:17 | INFO | open_clip_train.train | Train Epoch: 30 [180256768/268959744 (67%)] Data (t): 0.924 Batch (t): 1.560, 21323.1/s, 2665.39/s/gpu LR: 0.000002 Logit Scale: 100.000 Contrastive_loss: 0.27235 (0.27704) Loss: 0.27235 (0.27704) Logit_scale: 100.00 (99.999)
2026-05-29,06:12:54 | INFO | open_clip_train.train | Train Epoch: 30 [183533568/268959744 (68%)] Data (t): 0.831 Batch (t): 1.565, 20869.7/s, 2608.71/s/gpu LR: 0.000002 Logit Scale: 100.000 Contrastive_loss: 0.27986 (0.27709) Loss: 0.27986 (0.27709) Logit_scale: 100.00 (99.999)
2026-05-29,06:15:30 | INFO | open_clip_train.train | Train Epoch: 30 [186810368/268959744 (69%)] Data (t): 0.898 Batch (t): 1.560, 20780.0/s, 2597.51/s/gpu LR: 0.000002 Logit Scale: 100.000 Contrastive_loss: 0.26519 (0.27689) Loss: 0.26519 (0.27689) Logit_scale: 100.00 (99.999)
2026-05-29,06:18:05 | INFO | open_clip_train.train | Train Epoch: 30 [190087168/268959744 (71%)] Data (t): 0.847 Batch (t): 1.552, 21006.6/s, 2625.82/s/gpu LR: 0.000002 Logit Scale: 100.000 Contrastive_loss: 0.26837 (0.27674) Loss: 0.26837 (0.27674) Logit_scale: 100.00 (99.999)
2026-05-29,06:20:40 | INFO | open_clip_train.train | Train Epoch: 30 [193363968/268959744 (72%)] Data (t): 0.863 Batch (t): 1.545, 20882.2/s, 2610.28/s/gpu LR: 0.000002 Logit Scale: 100.000 Contrastive_loss: 0.27157 (0.27666) Loss: 0.27157 (0.27666) Logit_scale: 100.00 (99.999)
2026-05-29,06:23:14 | INFO | open_clip_train.train | Train Epoch: 30 [196640768/268959744 (73%)] Data (t): 0.852 Batch (t): 1.545, 20160.4/s, 2520.05/s/gpu LR: 0.000002 Logit Scale: 100.000 Contrastive_loss: 0.25632 (0.27632) Loss: 0.25632 (0.27632) Logit_scale: 100.00 (99.999)
2026-05-29,06:25:47 | INFO | open_clip_train.train | Train Epoch: 30 [199917568/268959744 (74%)] Data (t): 0.866 Batch (t): 1.533, 21119.9/s, 2639.99/s/gpu LR: 0.000002 Logit Scale: 100.000 Contrastive_loss: 0.25913 (0.27605) Loss: 0.25913 (0.27605) Logit_scale: 100.00 (99.999)
2026-05-29,06:28:20 | INFO | open_clip_train.train | Train Epoch: 30 [203194368/268959744 (76%)] Data (t): 0.866 Batch (t): 1.527, 21325.6/s, 2665.70/s/gpu LR: 0.000002 Logit Scale: 100.000 Contrastive_loss: 0.26264 (0.27583) Loss: 0.26264 (0.27583) Logit_scale: 100.00 (99.999)
2026-05-29,06:30:59 | INFO | open_clip_train.train | Train Epoch: 30 [206471168/268959744 (77%)] Data (t): 0.856 Batch (t): 1.589, 20456.5/s, 2557.06/s/gpu LR: 0.000002 Logit Scale: 100.000 Contrastive_loss: 0.29025 (0.27606) Loss: 0.29025 (0.27606) Logit_scale: 100.00 (99.999)
2026-05-29,06:33:35 | INFO | open_clip_train.train | Train Epoch: 30 [209747968/268959744 (78%)] Data (t): 0.842 Batch (t): 1.563, 21171.1/s, 2646.38/s/gpu LR: 0.000002 Logit Scale: 100.000 Contrastive_loss: 0.26823 (0.27594) Loss: 0.26823 (0.27594) Logit_scale: 100.00 (99.999)
2026-05-29,06:36:11 | INFO | open_clip_train.train | Train Epoch: 30 [213024768/268959744 (79%)] Data (t): 0.899 Batch (t): 1.558, 20848.7/s, 2606.08/s/gpu LR: 0.000002 Logit Scale: 100.000 Contrastive_loss: 0.26616 (0.27579) Loss: 0.26616 (0.27579) Logit_scale: 100.00 (99.999)
2026-05-29,06:38:46 | INFO | open_clip_train.train | Train Epoch: 30 [216301568/268959744 (80%)] Data (t): 0.835 Batch (t): 1.552, 22986.5/s, 2873.31/s/gpu LR: 0.000002 Logit Scale: 100.000 Contrastive_loss: 0.24995 (0.27541) Loss: 0.24995 (0.27541) Logit_scale: 100.00 (99.999)
2026-05-29,06:41:19 | INFO | open_clip_train.train | Train Epoch: 30 [219578368/268959744 (82%)] Data (t): 0.886 Batch (t): 1.523, 22997.5/s, 2874.69/s/gpu LR: 0.000002 Logit Scale: 100.000 Contrastive_loss: 0.27305 (0.27537) Loss: 0.27305 (0.27537) Logit_scale: 100.00 (99.999)
2026-05-29,06:43:55 | INFO | open_clip_train.train | Train Epoch: 30 [222855168/268959744 (83%)] Data (t): 0.888 Batch (t): 1.565, 21394.5/s, 2674.31/s/gpu LR: 0.000002 Logit Scale: 100.000 Contrastive_loss: 0.27694 (0.27539) Loss: 0.27694 (0.27539) Logit_scale: 100.00 (99.999)
2026-05-29,06:46:32 | INFO | open_clip_train.train | Train Epoch: 30 [226131968/268959744 (84%)] Data (t): 0.864 Batch (t): 1.564, 20581.3/s, 2572.66/s/gpu LR: 0.000002 Logit Scale: 100.000 Contrastive_loss: 0.26340 (0.27522) Loss: 0.26340 (0.27522) Logit_scale: 100.00 (99.999)
2026-05-29,06:49:07 | INFO | open_clip_train.train | Train Epoch: 30 [229408768/268959744 (85%)] Data (t): 0.857 Batch (t): 1.558, 20632.0/s, 2579.00/s/gpu LR: 0.000002 Logit Scale: 100.000 Contrastive_loss: 0.27690 (0.27525) Loss: 0.27690 (0.27525) Logit_scale: 100.00 (99.999)
2026-05-29,06:51:42 | INFO | open_clip_train.train | Train Epoch: 30 [232685568/268959744 (87%)] Data (t): 0.929 Batch (t): 1.548, 20648.6/s, 2581.07/s/gpu LR: 0.000002 Logit Scale: 100.000 Contrastive_loss: 0.26747 (0.27514) Loss: 0.26747 (0.27514) Logit_scale: 100.00 (99.999)
2026-05-29,06:54:18 | INFO | open_clip_train.train | Train Epoch: 30 [235962368/268959744 (88%)] Data (t): 0.939 Batch (t): 1.555, 21807.0/s, 2725.87/s/gpu LR: 0.000002 Logit Scale: 100.000 Contrastive_loss: 0.26129 (0.27495) Loss: 0.26129 (0.27495) Logit_scale: 100.00 (99.999)
2026-05-29,06:56:49 | INFO | open_clip_train.train | Train Epoch: 30 [239239168/268959744 (89%)] Data (t): 0.853 Batch (t): 1.515, 26796.4/s, 3349.55/s/gpu LR: 0.000002 Logit Scale: 100.000 Contrastive_loss: 0.30116 (0.27530) Loss: 0.30116 (0.27530) Logit_scale: 100.00 (99.999)
2026-05-29,06:59:25 | INFO | open_clip_train.train | Train Epoch: 30 [242515968/268959744 (90%)] Data (t): 0.881 Batch (t): 1.556, 23959.0/s, 2994.87/s/gpu LR: 0.000001 Logit Scale: 100.000 Contrastive_loss: 0.28394 (0.27542) Loss: 0.28394 (0.27542) Logit_scale: 100.00 (99.999)
2026-05-29,07:01:59 | INFO | open_clip_train.train | Train Epoch: 30 [245792768/268959744 (91%)] Data (t): 0.861 Batch (t): 1.542, 20246.2/s, 2530.78/s/gpu LR: 0.000001 Logit Scale: 100.000 Contrastive_loss: 0.31754 (0.27597) Loss: 0.31754 (0.27597) Logit_scale: 100.00 (99.999)
2026-05-29,07:04:33 | INFO | open_clip_train.train | Train Epoch: 30 [249069568/268959744 (93%)] Data (t): 0.918 Batch (t): 1.542, 20702.3/s, 2587.79/s/gpu LR: 0.000001 Logit Scale: 100.000 Contrastive_loss: 0.27682 (0.27598) Loss: 0.27682 (0.27598) Logit_scale: 100.00 (99.999)
2026-05-29,07:07:08 | INFO | open_clip_train.train | Train Epoch: 30 [252346368/268959744 (94%)] Data (t): 0.925 Batch (t): 1.548, 20926.2/s, 2615.77/s/gpu LR: 0.000001 Logit Scale: 100.000 Contrastive_loss: 0.27445 (0.27596) Loss: 0.27445 (0.27596) Logit_scale: 100.00 (99.999)
2026-05-29,07:09:40 | INFO | open_clip_train.train | Train Epoch: 30 [255623168/268959744 (95%)] Data (t): 0.868 Batch (t): 1.524, 24345.7/s, 3043.22/s/gpu LR: 0.000001 Logit Scale: 99.998 Contrastive_loss: 0.27223 (0.27592) Loss: 0.27223 (0.27592) Logit_scale: 99.998 (99.999)
2026-05-29,07:12:17 | INFO | open_clip_train.train | Train Epoch: 30 [258899968/268959744 (96%)] Data (t): 0.875 Batch (t): 1.563, 21134.0/s, 2641.74/s/gpu LR: 0.000001 Logit Scale: 99.995 Contrastive_loss: 0.30883 (0.27633) Loss: 0.30883 (0.27633) Logit_scale: 99.995 (99.999)
2026-05-29,07:14:54 | INFO | open_clip_train.train | Train Epoch: 30 [262176768/268959744 (97%)] Data (t): 0.850 Batch (t): 1.572, 21684.8/s, 2710.61/s/gpu LR: 0.000001 Logit Scale: 99.989 Contrastive_loss: 0.27334 (0.27629) Loss: 0.27334 (0.27629) Logit_scale: 99.989 (99.999)
2026-05-29,07:17:30 | INFO | open_clip_train.train | Train Epoch: 30 [265453568/268959744 (99%)] Data (t): 0.932 Batch (t): 1.560, 21153.7/s, 2644.21/s/gpu LR: 0.000001 Logit Scale: 99.980 Contrastive_loss: 0.31869 (0.27681) Loss: 0.31869 (0.27681) Logit_scale: 99.980 (99.999)
2026-05-29,07:20:00 | INFO | open_clip_train.train | Train Epoch: 30 [268730368/268959744 (100%)] Data (t): 0.833 Batch (t): 1.502, 29803.8/s, 3725.47/s/gpu LR: 0.000001 Logit Scale: 99.970 Contrastive_loss: 0.33318 (0.27749) Loss: 0.33318 (0.27749) Logit_scale: 99.970 (99.999)
2026-05-29,07:20:08 | INFO | open_clip_train.train | Train Epoch: 30 [268959744/268959744 (100%)] Data (t): 0.642 Batch (t): 1.121, 29568.2/s, 3696.02/s/gpu LR: 0.000001 Logit Scale: 99.970 Contrastive_loss: 0.37144 (0.27861) Loss: 0.37144 (0.27861) Logit_scale: 99.970 (99.998)
2026-05-29,07:20:13 | INFO | open_clip_train.main | Start epoch 31
2026-05-29,07:20:34 | INFO | open_clip_train.train | Train Epoch: 31 [ 32768/268959744 (0%)] Data (t): 18.798 Batch (t): 21.663, 1512.62/s, 189.077/s/gpu LR: 0.000001 Logit Scale: 99.970 Contrastive_loss: 0.24836 (0.24836) Loss: 0.24836 (0.24836) Logit_scale: 99.970 (99.970)
2026-05-29,07:23:07 | INFO | open_clip_train.train | Train Epoch: 31 [ 3309568/268959744 (1%)] Data (t): 0.694 Batch (t): 1.529, 18513.7/s, 2314.22/s/gpu LR: 0.000001 Logit Scale: 99.980 Contrastive_loss: 0.26939 (0.25887) Loss: 0.26939 (0.25887) Logit_scale: 99.980 (99.975)
2026-05-29,07:25:47 | INFO | open_clip_train.train | Train Epoch: 31 [ 6586368/268959744 (2%)] Data (t): 0.802 Batch (t): 1.593, 21333.7/s, 2666.72/s/gpu LR: 0.000001 Logit Scale: 99.990 Contrastive_loss: 0.25326 (0.25700) Loss: 0.25326 (0.25700) Logit_scale: 99.990 (99.980)
2026-05-29,07:28:23 | INFO | open_clip_train.train | Train Epoch: 31 [ 9863168/268959744 (4%)] Data (t): 0.891 Batch (t): 1.561, 20626.5/s, 2578.31/s/gpu LR: 0.000001 Logit Scale: 99.999 Contrastive_loss: 0.27960 (0.26265) Loss: 0.27960 (0.26265) Logit_scale: 99.999 (99.985)
2026-05-29,07:30:59 | INFO | open_clip_train.train | Train Epoch: 31 [ 13139968/268959744 (5%)] Data (t): 0.898 Batch (t): 1.561, 21847.9/s, 2730.98/s/gpu LR: 0.000001 Logit Scale: 100.000 Contrastive_loss: 0.26488 (0.26310) Loss: 0.26488 (0.26310) Logit_scale: 100.00 (99.988)
2026-05-29,07:33:31 | INFO | open_clip_train.train | Train Epoch: 31 [ 16416768/268959744 (6%)] Data (t): 0.918 Batch (t): 1.521, 21659.1/s, 2707.39/s/gpu LR: 0.000001 Logit Scale: 100.000 Contrastive_loss: 0.28182 (0.26622) Loss: 0.28182 (0.26622) Logit_scale: 100.00 (99.990)
2026-05-29,07:36:06 | INFO | open_clip_train.train | Train Epoch: 31 [ 19693568/268959744 (7%)] Data (t): 0.823 Batch (t): 1.549, 20187.1/s, 2523.38/s/gpu LR: 0.000001 Logit Scale: 100.000 Contrastive_loss: 0.24181 (0.26273) Loss: 0.24181 (0.26273) Logit_scale: 100.00 (99.991)
2026-05-29,07:38:41 | INFO | open_clip_train.train | Train Epoch: 31 [ 22970368/268959744 (9%)] Data (t): 0.846 Batch (t): 1.552, 21558.4/s, 2694.80/s/gpu LR: 0.000001 Logit Scale: 100.000 Contrastive_loss: 0.26258 (0.26271) Loss: 0.26258 (0.26271) Logit_scale: 100.00 (99.992)
2026-05-29,07:41:17 | INFO | open_clip_train.train | Train Epoch: 31 [ 26247168/268959744 (10%)] Data (t): 0.861 Batch (t): 1.563, 21499.7/s, 2687.46/s/gpu LR: 0.000001 Logit Scale: 100.000 Contrastive_loss: 0.26761 (0.26326) Loss: 0.26761 (0.26326) Logit_scale: 100.00 (99.993)
2026-05-29,07:43:53 | INFO | open_clip_train.train | Train Epoch: 31 [ 29523968/268959744 (11%)] Data (t): 0.895 Batch (t): 1.557, 20805.1/s, 2600.64/s/gpu LR: 0.000001 Logit Scale: 100.000 Contrastive_loss: 0.26566 (0.26350) Loss: 0.26566 (0.26350) Logit_scale: 100.00 (99.994)
2026-05-29,07:46:26 | INFO | open_clip_train.train | Train Epoch: 31 [ 32800768/268959744 (12%)] Data (t): 0.842 Batch (t): 1.536, 22575.5/s, 2821.94/s/gpu LR: 0.000001 Logit Scale: 100.000 Contrastive_loss: 0.29009 (0.26591) Loss: 0.29009 (0.26591) Logit_scale: 100.00 (99.994)
2026-05-29,07:48:58 | INFO | open_clip_train.train | Train Epoch: 31 [ 36077568/268959744 (13%)] Data (t): 0.855 Batch (t): 1.519, 21519.8/s, 2689.98/s/gpu LR: 0.000001 Logit Scale: 100.000 Contrastive_loss: 0.26489 (0.26583) Loss: 0.26489 (0.26583) Logit_scale: 100.00 (99.995)
2026-05-29,07:51:30 | INFO | open_clip_train.train | Train Epoch: 31 [ 39354368/268959744 (15%)] Data (t): 0.886 Batch (t): 1.514, 20194.2/s, 2524.27/s/gpu LR: 0.000001 Logit Scale: 100.000 Contrastive_loss: 0.23806 (0.26369) Loss: 0.23806 (0.26369) Logit_scale: 100.00 (99.995)
2026-05-29,07:54:05 | INFO | open_clip_train.train | Train Epoch: 31 [ 42631168/268959744 (16%)] Data (t): 0.873 Batch (t): 1.549, 22847.4/s, 2855.93/s/gpu LR: 0.000001 Logit Scale: 100.000 Contrastive_loss: 0.27141 (0.26424) Loss: 0.27141 (0.26424) Logit_scale: 100.00 (99.996)
2026-05-29,07:56:41 | INFO | open_clip_train.train | Train Epoch: 31 [ 45907968/268959744 (17%)] Data (t): 0.893 Batch (t): 1.566, 20855.3/s, 2606.91/s/gpu LR: 0.000001 Logit Scale: 100.000 Contrastive_loss: 0.22740 (0.26179) Loss: 0.22740 (0.26179) Logit_scale: 100.00 (99.996)
2026-05-29,07:59:17 | INFO | open_clip_train.train | Train Epoch: 31 [ 49184768/268959744 (18%)] Data (t): 0.855 Batch (t): 1.555, 21130.6/s, 2641.32/s/gpu LR: 0.000001 Logit Scale: 100.000 Contrastive_loss: 0.27593 (0.26267) Loss: 0.27593 (0.26267) Logit_scale: 100.00 (99.996)
2026-05-29,08:01:53 | INFO | open_clip_train.train | Train Epoch: 31 [ 52461568/268959744 (20%)] Data (t): 0.916 Batch (t): 1.561, 20964.4/s, 2620.55/s/gpu LR: 0.000001 Logit Scale: 100.000 Contrastive_loss: 0.25457 (0.26220) Loss: 0.25457 (0.26220) Logit_scale: 100.00 (99.996)
2026-05-29,08:04:27 | INFO | open_clip_train.train | Train Epoch: 31 [ 55738368/268959744 (21%)] Data (t): 0.842 Batch (t): 1.541, 19996.2/s, 2499.52/s/gpu LR: 0.000001 Logit Scale: 100.000 Contrastive_loss: 0.25734 (0.26193) Loss: 0.25734 (0.26193) Logit_scale: 100.00 (99.997)
2026-05-29,08:07:03 | INFO | open_clip_train.train | Train Epoch: 31 [ 59015168/268959744 (22%)] Data (t): 0.889 Batch (t): 1.561, 19866.1/s, 2483.26/s/gpu LR: 0.000001 Logit Scale: 100.000 Contrastive_loss: 0.28068 (0.26291) Loss: 0.28068 (0.26291) Logit_scale: 100.00 (99.997)
2026-05-29,08:09:38 | INFO | open_clip_train.train | Train Epoch: 31 [ 62291968/268959744 (23%)] Data (t): 0.786 Batch (t): 1.554, 22394.4/s, 2799.30/s/gpu LR: 0.000001 Logit Scale: 100.000 Contrastive_loss: 0.28255 (0.26390) Loss: 0.28255 (0.26390) Logit_scale: 100.00 (99.997)
2026-05-29,08:12:14 | INFO | open_clip_train.train | Train Epoch: 31 [ 65568768/268959744 (24%)] Data (t): 0.889 Batch (t): 1.554, 20799.2/s, 2599.90/s/gpu LR: 0.000001 Logit Scale: 100.000 Contrastive_loss: 0.24844 (0.26316) Loss: 0.24844 (0.26316) Logit_scale: 100.00 (99.997)
2026-05-29,08:14:50 | INFO | open_clip_train.train | Train Epoch: 31 [ 68845568/268959744 (26%)] Data (t): 0.885 Batch (t): 1.557, 22521.4/s, 2815.18/s/gpu LR: 0.000001 Logit Scale: 100.000 Contrastive_loss: 0.25351 (0.26272) Loss: 0.25351 (0.26272) Logit_scale: 100.00 (99.997)
2026-05-29,08:17:21 | INFO | open_clip_train.train | Train Epoch: 31 [ 72122368/268959744 (27%)] Data (t): 0.912 Batch (t): 1.516, 22170.2/s, 2771.27/s/gpu LR: 0.000001 Logit Scale: 100.000 Contrastive_loss: 0.26285 (0.26273) Loss: 0.26285 (0.26273) Logit_scale: 100.00 (99.997)
2026-05-29,08:19:56 | INFO | open_clip_train.train | Train Epoch: 31 [ 75399168/268959744 (28%)] Data (t): 0.832 Batch (t): 1.545, 20648.6/s, 2581.07/s/gpu LR: 0.000001 Logit Scale: 100.000 Contrastive_loss: 0.26829 (0.26296) Loss: 0.26829 (0.26296) Logit_scale: 100.00 (99.997)
2026-05-29,08:22:31 | INFO | open_clip_train.train | Train Epoch: 31 [ 78675968/268959744 (29%)] Data (t): 0.872 Batch (t): 1.554, 19923.5/s, 2490.43/s/gpu LR: 0.000001 Logit Scale: 100.000 Contrastive_loss: 0.24994 (0.26244) Loss: 0.24994 (0.26244) Logit_scale: 100.00 (99.998)
2026-05-29,08:25:08 | INFO | open_clip_train.train | Train Epoch: 31 [ 81952768/268959744 (30%)] Data (t): 0.873 Batch (t): 1.565, 20428.6/s, 2553.57/s/gpu LR: 0.000001 Logit Scale: 100.000 Contrastive_loss: 0.25505 (0.26215) Loss: 0.25505 (0.26215) Logit_scale: 100.00 (99.998)
2026-05-29,08:27:42 | INFO | open_clip_train.train | Train Epoch: 31 [ 85229568/268959744 (32%)] Data (t): 0.856 Batch (t): 1.549, 20632.5/s, 2579.06/s/gpu LR: 0.000001 Logit Scale: 100.000 Contrastive_loss: 0.26365 (0.26221) Loss: 0.26365 (0.26221) Logit_scale: 100.00 (99.998)
2026-05-29,08:30:17 | INFO | open_clip_train.train | Train Epoch: 31 [ 88506368/268959744 (33%)] Data (t): 0.863 Batch (t): 1.550, 21170.9/s, 2646.36/s/gpu LR: 0.000001 Logit Scale: 100.000 Contrastive_loss: 0.29151 (0.26326) Loss: 0.29151 (0.26326) Logit_scale: 100.00 (99.998)
2026-05-29,08:32:53 | INFO | open_clip_train.train | Train Epoch: 31 [ 91783168/268959744 (34%)] Data (t): 0.818 Batch (t): 1.554, 21654.0/s, 2706.75/s/gpu LR: 0.000001 Logit Scale: 100.000 Contrastive_loss: 0.25719 (0.26305) Loss: 0.25719 (0.26305) Logit_scale: 100.00 (99.998)
2026-05-29,08:35:28 | INFO | open_clip_train.train | Train Epoch: 31 [ 95059968/268959744 (35%)] Data (t): 0.833 Batch (t): 1.550, 21263.6/s, 2657.95/s/gpu LR: 0.000001 Logit Scale: 100.000 Contrastive_loss: 0.25879 (0.26290) Loss: 0.25879 (0.26290) Logit_scale: 100.00 (99.998)
2026-05-29,08:38:03 | INFO | open_clip_train.train | Train Epoch: 31 [ 98336768/268959744 (37%)] Data (t): 0.876 Batch (t): 1.554, 20424.9/s, 2553.12/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.27336 (0.26324) Loss: 0.27336 (0.26324) Logit_scale: 100.00 (99.998)
2026-05-29,08:40:38 | INFO | open_clip_train.train | Train Epoch: 31 [101613568/268959744 (38%)] Data (t): 0.871 Batch (t): 1.551, 20870.8/s, 2608.85/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.26026 (0.26315) Loss: 0.26026 (0.26315) Logit_scale: 100.00 (99.998)
2026-05-29,08:43:15 | INFO | open_clip_train.train | Train Epoch: 31 [104890368/268959744 (39%)] Data (t): 0.896 Batch (t): 1.564, 20292.9/s, 2536.61/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.28192 (0.26372) Loss: 0.28192 (0.26372) Logit_scale: 100.00 (99.998)
2026-05-29,08:45:49 | INFO | open_clip_train.train | Train Epoch: 31 [108167168/268959744 (40%)] Data (t): 0.940 Batch (t): 1.545, 24892.5/s, 3111.57/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.25563 (0.26348) Loss: 0.25563 (0.26348) Logit_scale: 100.00 (99.998)
2026-05-29,08:48:24 | INFO | open_clip_train.train | Train Epoch: 31 [111443968/268959744 (41%)] Data (t): 0.906 Batch (t): 1.552, 20886.4/s, 2610.80/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.22717 (0.26244) Loss: 0.22717 (0.26244) Logit_scale: 100.00 (99.998)
2026-05-29,08:50:58 | INFO | open_clip_train.train | Train Epoch: 31 [114720768/268959744 (43%)] Data (t): 0.876 Batch (t): 1.535, 20343.6/s, 2542.95/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.25108 (0.26213) Loss: 0.25108 (0.26213) Logit_scale: 100.00 (99.998)
2026-05-29,08:53:31 | INFO | open_clip_train.train | Train Epoch: 31 [117997568/268959744 (44%)] Data (t): 0.815 Batch (t): 1.527, 21998.5/s, 2749.81/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.26277 (0.26214) Loss: 0.26277 (0.26214) Logit_scale: 100.00 (99.998)
2026-05-29,08:56:07 | INFO | open_clip_train.train | Train Epoch: 31 [121274368/268959744 (45%)] Data (t): 0.875 Batch (t): 1.563, 19978.1/s, 2497.26/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.26263 (0.26216) Loss: 0.26263 (0.26216) Logit_scale: 100.00 (99.998)
2026-05-29,08:58:42 | INFO | open_clip_train.train | Train Epoch: 31 [124551168/268959744 (46%)] Data (t): 0.907 Batch (t): 1.550, 21848.2/s, 2731.03/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.25929 (0.26208) Loss: 0.25929 (0.26208) Logit_scale: 100.00 (99.998)
2026-05-29,09:01:18 | INFO | open_clip_train.train | Train Epoch: 31 [127827968/268959744 (48%)] Data (t): 0.910 Batch (t): 1.562, 21403.2/s, 2675.40/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.26767 (0.26222) Loss: 0.26767 (0.26222) Logit_scale: 100.00 (99.998)
2026-05-29,09:03:54 | INFO | open_clip_train.train | Train Epoch: 31 [131104768/268959744 (49%)] Data (t): 0.868 Batch (t): 1.563, 20689.8/s, 2586.22/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.23416 (0.26154) Loss: 0.23416 (0.26154) Logit_scale: 100.00 (99.999)
2026-05-29,09:06:31 | INFO | open_clip_train.train | Train Epoch: 31 [134381568/268959744 (50%)] Data (t): 0.908 Batch (t): 1.561, 20662.1/s, 2582.76/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.26657 (0.26166) Loss: 0.26657 (0.26166) Logit_scale: 100.00 (99.999)
2026-05-29,09:09:05 | INFO | open_clip_train.train | Train Epoch: 31 [137658368/268959744 (51%)] Data (t): 0.874 Batch (t): 1.543, 21400.3/s, 2675.03/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.26145 (0.26165) Loss: 0.26145 (0.26165) Logit_scale: 100.00 (99.999)
2026-05-29,09:11:33 | INFO | open_clip_train.train | Train Epoch: 31 [140935168/268959744 (52%)] Data (t): 0.838 Batch (t): 1.484, 21188.2/s, 2648.53/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.26475 (0.26172) Loss: 0.26475 (0.26172) Logit_scale: 100.00 (99.999)
2026-05-29,09:14:11 | INFO | open_clip_train.train | Train Epoch: 31 [144211968/268959744 (54%)] Data (t): 0.840 Batch (t): 1.573, 21547.6/s, 2693.45/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.26611 (0.26182) Loss: 0.26611 (0.26182) Logit_scale: 100.00 (99.999)
2026-05-29,09:16:47 | INFO | open_clip_train.train | Train Epoch: 31 [147488768/268959744 (55%)] Data (t): 0.824 Batch (t): 1.567, 19914.7/s, 2489.34/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.27369 (0.26208) Loss: 0.27369 (0.26208) Logit_scale: 100.00 (99.999)
2026-05-29,09:19:24 | INFO | open_clip_train.train | Train Epoch: 31 [150765568/268959744 (56%)] Data (t): 0.823 Batch (t): 1.565, 21111.5/s, 2638.94/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.23173 (0.26143) Loss: 0.23173 (0.26143) Logit_scale: 100.00 (99.999)
2026-05-29,09:22:00 | INFO | open_clip_train.train | Train Epoch: 31 [154042368/268959744 (57%)] Data (t): 0.929 Batch (t): 1.562, 20500.1/s, 2562.51/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.26739 (0.26156) Loss: 0.26739 (0.26156) Logit_scale: 100.00 (99.999)
2026-05-29,09:24:34 | INFO | open_clip_train.train | Train Epoch: 31 [157319168/268959744 (58%)] Data (t): 0.867 Batch (t): 1.542, 22274.1/s, 2784.27/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.25541 (0.26143) Loss: 0.25541 (0.26143) Logit_scale: 100.00 (99.999)
2026-05-29,09:27:09 | INFO | open_clip_train.train | Train Epoch: 31 [160595968/268959744 (60%)] Data (t): 0.900 Batch (t): 1.549, 20815.2/s, 2601.90/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.24300 (0.26106) Loss: 0.24300 (0.26106) Logit_scale: 100.00 (99.999)
2026-05-29,09:29:44 | INFO | open_clip_train.train | Train Epoch: 31 [163872768/268959744 (61%)] Data (t): 0.894 Batch (t): 1.553, 21780.3/s, 2722.54/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.26957 (0.26123) Loss: 0.26957 (0.26123) Logit_scale: 100.00 (99.999)
2026-05-29,09:32:21 | INFO | open_clip_train.train | Train Epoch: 31 [167149568/268959744 (62%)] Data (t): 0.887 Batch (t): 1.569, 21121.6/s, 2640.20/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.24918 (0.26100) Loss: 0.24918 (0.26100) Logit_scale: 100.00 (99.999)
2026-05-29,09:34:57 | INFO | open_clip_train.train | Train Epoch: 31 [170426368/268959744 (63%)] Data (t): 0.881 Batch (t): 1.555, 20299.3/s, 2537.42/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.25885 (0.26096) Loss: 0.25885 (0.26096) Logit_scale: 100.00 (99.999)
2026-05-29,09:37:33 | INFO | open_clip_train.train | Train Epoch: 31 [173703168/268959744 (65%)] Data (t): 0.881 Batch (t): 1.567, 20828.3/s, 2603.53/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.24881 (0.26073) Loss: 0.24881 (0.26073) Logit_scale: 100.00 (99.999)
2026-05-29,09:40:09 | INFO | open_clip_train.train | Train Epoch: 31 [176979968/268959744 (66%)] Data (t): 0.860 Batch (t): 1.559, 21628.7/s, 2703.59/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.26007 (0.26072) Loss: 0.26007 (0.26072) Logit_scale: 100.00 (99.999)
2026-05-29,09:42:45 | INFO | open_clip_train.train | Train Epoch: 31 [180256768/268959744 (67%)] Data (t): 0.829 Batch (t): 1.561, 20309.0/s, 2538.63/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.26422 (0.26078) Loss: 0.26422 (0.26078) Logit_scale: 100.00 (99.999)
2026-05-29,09:45:22 | INFO | open_clip_train.train | Train Epoch: 31 [183533568/268959744 (68%)] Data (t): 0.882 Batch (t): 1.566, 20462.9/s, 2557.87/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.28144 (0.26115) Loss: 0.28144 (0.26115) Logit_scale: 100.00 (99.999)
2026-05-29,09:47:58 | INFO | open_clip_train.train | Train Epoch: 31 [186810368/268959744 (69%)] Data (t): 0.852 Batch (t): 1.564, 23812.6/s, 2976.58/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.26024 (0.26113) Loss: 0.26024 (0.26113) Logit_scale: 100.00 (99.999)
2026-05-29,09:50:33 | INFO | open_clip_train.train | Train Epoch: 31 [190087168/268959744 (71%)] Data (t): 0.835 Batch (t): 1.553, 20352.6/s, 2544.08/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.24361 (0.26083) Loss: 0.24361 (0.26083) Logit_scale: 100.00 (99.999)
2026-05-29,09:53:08 | INFO | open_clip_train.train | Train Epoch: 31 [193363968/268959744 (72%)] Data (t): 0.917 Batch (t): 1.546, 21352.6/s, 2669.08/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.26262 (0.26086) Loss: 0.26262 (0.26086) Logit_scale: 100.00 (99.999)
2026-05-29,09:55:44 | INFO | open_clip_train.train | Train Epoch: 31 [196640768/268959744 (73%)] Data (t): 0.878 Batch (t): 1.562, 22210.7/s, 2776.34/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.26680 (0.26096) Loss: 0.26680 (0.26096) Logit_scale: 100.00 (99.999)
2026-05-29,09:58:20 | INFO | open_clip_train.train | Train Epoch: 31 [199917568/268959744 (74%)] Data (t): 0.863 Batch (t): 1.554, 20306.3/s, 2538.29/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.26422 (0.26101) Loss: 0.26422 (0.26101) Logit_scale: 100.00 (99.999)
2026-05-29,10:00:56 | INFO | open_clip_train.train | Train Epoch: 31 [203194368/268959744 (76%)] Data (t): 0.854 Batch (t): 1.567, 19625.1/s, 2453.13/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.26235 (0.26103) Loss: 0.26235 (0.26103) Logit_scale: 100.00 (99.999)
2026-05-29,10:03:33 | INFO | open_clip_train.train | Train Epoch: 31 [206471168/268959744 (77%)] Data (t): 0.871 Batch (t): 1.563, 20923.2/s, 2615.40/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.24417 (0.26077) Loss: 0.24417 (0.26077) Logit_scale: 100.00 (99.999)
2026-05-29,10:06:09 | INFO | open_clip_train.train | Train Epoch: 31 [209747968/268959744 (78%)] Data (t): 0.859 Batch (t): 1.560, 20748.9/s, 2593.62/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.27297 (0.26096) Loss: 0.27297 (0.26096) Logit_scale: 100.00 (99.999)
2026-05-29,10:08:45 | INFO | open_clip_train.train | Train Epoch: 31 [213024768/268959744 (79%)] Data (t): 0.835 Batch (t): 1.566, 20482.5/s, 2560.31/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.30153 (0.26157) Loss: 0.30153 (0.26157) Logit_scale: 100.00 (99.999)
2026-05-29,10:11:18 | INFO | open_clip_train.train | Train Epoch: 31 [216301568/268959744 (80%)] Data (t): 0.907 Batch (t): 1.532, 21819.1/s, 2727.38/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.24780 (0.26137) Loss: 0.24780 (0.26137) Logit_scale: 100.00 (99.999)
2026-05-29,10:13:52 | INFO | open_clip_train.train | Train Epoch: 31 [219578368/268959744 (82%)] Data (t): 0.863 Batch (t): 1.538, 23769.8/s, 2971.23/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.27508 (0.26157) Loss: 0.27508 (0.26157) Logit_scale: 100.00 (99.999)
2026-05-29,10:16:27 | INFO | open_clip_train.train | Train Epoch: 31 [222855168/268959744 (83%)] Data (t): 0.876 Batch (t): 1.548, 20135.1/s, 2516.89/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.27481 (0.26176) Loss: 0.27481 (0.26176) Logit_scale: 100.00 (99.999)
2026-05-29,10:19:02 | INFO | open_clip_train.train | Train Epoch: 31 [226131968/268959744 (84%)] Data (t): 0.914 Batch (t): 1.551, 22458.1/s, 2807.26/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.27973 (0.26202) Loss: 0.27973 (0.26202) Logit_scale: 100.00 (99.999)
2026-05-29,10:21:38 | INFO | open_clip_train.train | Train Epoch: 31 [229408768/268959744 (85%)] Data (t): 0.819 Batch (t): 1.553, 20809.3/s, 2601.16/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.30684 (0.26265) Loss: 0.30684 (0.26265) Logit_scale: 100.00 (99.999)
2026-05-29,10:24:12 | INFO | open_clip_train.train | Train Epoch: 31 [232685568/268959744 (87%)] Data (t): 0.900 Batch (t): 1.547, 22116.5/s, 2764.57/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.25859 (0.26259) Loss: 0.25859 (0.26259) Logit_scale: 100.00 (99.999)
2026-05-29,10:26:49 | INFO | open_clip_train.train | Train Epoch: 31 [235962368/268959744 (88%)] Data (t): 0.843 Batch (t): 1.566, 20322.4/s, 2540.30/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.26307 (0.26260) Loss: 0.26307 (0.26260) Logit_scale: 100.00 (99.999)
2026-05-29,10:29:24 | INFO | open_clip_train.train | Train Epoch: 31 [239239168/268959744 (89%)] Data (t): 0.775 Batch (t): 1.557, 20704.3/s, 2588.04/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.25191 (0.26245) Loss: 0.25191 (0.26245) Logit_scale: 100.00 (99.999)
2026-05-29,10:32:00 | INFO | open_clip_train.train | Train Epoch: 31 [242515968/268959744 (90%)] Data (t): 0.853 Batch (t): 1.558, 20506.6/s, 2563.32/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.28485 (0.26275) Loss: 0.28485 (0.26275) Logit_scale: 100.00 (99.999)
2026-05-29,10:34:34 | INFO | open_clip_train.train | Train Epoch: 31 [245792768/268959744 (91%)] Data (t): 0.928 Batch (t): 1.541, 20411.6/s, 2551.45/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.28180 (0.26300) Loss: 0.28180 (0.26300) Logit_scale: 100.00 (99.999)
2026-05-29,10:37:10 | INFO | open_clip_train.train | Train Epoch: 31 [249069568/268959744 (93%)] Data (t): 0.873 Batch (t): 1.558, 20969.7/s, 2621.22/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.26146 (0.26298) Loss: 0.26146 (0.26298) Logit_scale: 100.00 (99.999)
2026-05-29,10:39:46 | INFO | open_clip_train.train | Train Epoch: 31 [252346368/268959744 (94%)] Data (t): 0.852 Batch (t): 1.557, 20015.2/s, 2501.90/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.27320 (0.26311) Loss: 0.27320 (0.26311) Logit_scale: 100.00 (99.999)
2026-05-29,10:42:22 | INFO | open_clip_train.train | Train Epoch: 31 [255623168/268959744 (95%)] Data (t): 0.916 Batch (t): 1.557, 18941.7/s, 2367.71/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.29904 (0.26357) Loss: 0.29904 (0.26357) Logit_scale: 100.00 (99.999)
2026-05-29,10:44:58 | INFO | open_clip_train.train | Train Epoch: 31 [258899968/268959744 (96%)] Data (t): 0.846 Batch (t): 1.563, 22143.3/s, 2767.91/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.30720 (0.26411) Loss: 0.30720 (0.26411) Logit_scale: 100.00 (99.999)
2026-05-29,10:47:34 | INFO | open_clip_train.train | Train Epoch: 31 [262176768/268959744 (97%)] Data (t): 0.833 Batch (t): 1.563, 21783.8/s, 2722.98/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.34136 (0.26507) Loss: 0.34136 (0.26507) Logit_scale: 100.00 (99.999)
2026-05-29,10:50:10 | INFO | open_clip_train.train | Train Epoch: 31 [265453568/268959744 (99%)] Data (t): 0.895 Batch (t): 1.554, 21180.9/s, 2647.61/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.32555 (0.26581) Loss: 0.32555 (0.26581) Logit_scale: 100.00 (99.999)
2026-05-29,10:52:40 | INFO | open_clip_train.train | Train Epoch: 31 [268730368/268959744 (100%)] Data (t): 0.860 Batch (t): 1.505, 29378.2/s, 3672.28/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.28223 (0.26600) Loss: 0.28223 (0.26600) Logit_scale: 100.00 (99.999)
2026-05-29,10:52:48 | INFO | open_clip_train.train | Train Epoch: 31 [268959744/268959744 (100%)] Data (t): 0.634 Batch (t): 1.131, 28806.3/s, 3600.78/s/gpu LR: 0.000000 Logit Scale: 100.000 Contrastive_loss: 0.41558 (0.26778) Loss: 0.41558 (0.26778) Logit_scale: 100.00 (99.999)