learning_rate: 2e-06
train_batch_size: 2
eval_batch_size: 1
seed: 42
gradient_accumulation_steps: 16
total_train_batch_size: 32
optimizer: Use 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: 100
num_epochs: 10
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1
No log	1.0	145	0.0585	0.9332	0.9538	0.9434
No log	2.0	290	0.0367	0.9527	0.9727	0.9626
No log	3.0	435	0.0338	0.9776	0.9622	0.9698
1.3658	4.0	580	0.0265	0.9718	0.9769	0.9743
1.3658	5.0	725	0.0250	0.9708	0.9790	0.9749
1.3658	6.0	870	0.0323	0.9809	0.9695	0.9752
0.0959	7.0	1015	0.0285	0.9749	0.9800	0.9775
0.0959	8.0	1160	0.0319	0.9820	0.9727	0.9773
0.0959	9.0	1305	0.0313	0.9810	0.9737	0.9773
0.0959	10.0	1450	0.0294	0.9810	0.9748	0.9779

buddhadeb33/output_dinov2_large

output_dinov2_large