detr-fashion-4cat

This model is a fine-tuned version of facebook/detr-resnet-50 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.0440

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 5e-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
• num_epochs: 25

Training results

Training Loss	Epoch	Step	Validation Loss
1.7384	1.0	300	1.6380
1.3931	2.0	600	1.4684
1.3187	3.0	900	1.3165
1.3685	4.0	1200	1.2539
1.2105	5.0	1500	1.1905
1.2659	6.0	1800	1.1981
1.1682	7.0	2100	1.1874
1.1649	8.0	2400	1.1931
1.1522	9.0	2700	1.1604
1.1317	10.0	3000	1.1260
1.0819	11.0	3300	1.1122
1.0025	12.0	3600	1.1030
0.985	13.0	3900	1.0857
0.9731	14.0	4200	1.0724
1.0007	15.0	4500	1.0754
0.9768	16.0	4800	1.0502
1.0687	17.0	5100	1.0470
1.0269	18.0	5400	1.0400
1.0892	19.0	5700	1.0273
0.9363	20.0	6000	1.0302
1.0073	21.0	6300	1.0330
1.002	22.0	6600	1.0335
0.9418	23.0	6900	1.0266
0.9004	24.0	7200	1.0192
0.9459	25.0	7500	1.0213

Framework versions

• Transformers 4.51.3
• Pytorch 2.6.0+cu124
• Datasets 3.5.0
• Tokenizers 0.21.1