learning_rate: 1e-05
train_batch_size: 32
eval_batch_size: 32
seed: 42
optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 125
num_epochs: 50
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss
No log	2.0492	125	9.0295
No log	4.0984	250	8.3572
No log	6.1475	375	7.9395
7.7729	8.1967	500	7.5209
7.7729	10.2459	625	7.3253
7.7729	12.2951	750	7.4211
7.7729	14.3443	875	7.4970
4.8593	16.3934	1000	7.3769
4.8593	18.4426	1125	7.3811
4.8593	20.4918	1250	7.3434
4.8593	22.5410	1375	7.3463
3.9879	24.5902	1500	7.4509
3.9879	26.6393	1625	7.3332
3.9879	28.6885	1750	7.2705
3.9879	30.7377	1875	7.1949
3.7833	32.7869	2000	7.3268
3.7833	34.8361	2125	7.2636
3.7833	36.8852	2250	7.2168
3.7833	38.9344	2375	7.2514
3.6853	40.9836	2500	7.1643
3.6853	43.0328	2625	7.2198
3.6853	45.0820	2750	7.1966
3.6853	47.1311	2875	7.1910
3.6303	49.1803	3000	7.1502

poglicious/algae-detr-dinov3

algae-detr-dinov3