learning_rate: 1e-05
train_batch_size: 4
eval_batch_size: 4
seed: 42
optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
training_steps: 1000
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss
6.8509	0.0438	50	6.5878
6.5395	0.0876	100	6.0833
6.8232	0.1315	150	5.6343
4.8783	0.1753	200	5.0698
4.8238	0.2191	250	4.7423
4.3674	0.2629	300	4.3963
4.5075	0.3067	350	4.1901
4.6329	0.3506	400	4.0261
3.734	0.3944	450	3.8477
3.7449	0.4382	500	3.7750
4.0604	0.4820	550	3.6956
3.0591	0.5259	600	3.6346
4.2276	0.5697	650	3.6107
4.4193	0.6135	700	3.5955
4.5098	0.6573	750	3.5602
4.1579	0.7011	800	3.5349
4.0028	0.7450	850	3.5262
4.8916	0.7888	900	3.5146
4.5715	0.8326	950	3.5160
3.77	0.8764	1000	3.5093

Ahnj-Stability/detr-resnet-50-dc5-fashionpedia-finetuned

detr-resnet-50-dc5-fashionpedia-finetuned