detr-r50-mist1-bg-8ah-6l

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: 2.1031

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: 1e-05
• train_batch_size: 4
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 25
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
3.6794	1.0	115	2.8222
3.0269	2.0	230	2.8120
2.8681	3.0	345	2.7980
2.752	4.0	460	2.4853
2.7715	5.0	575	2.4140
2.6846	6.0	690	2.4715
2.6236	7.0	805	2.4614
2.5318	8.0	920	2.3441
2.5224	9.0	1035	2.2837
2.4661	10.0	1150	2.2510
2.4313	11.0	1265	2.3339
2.4125	12.0	1380	2.2957
2.4113	13.0	1495	2.2358
2.3784	14.0	1610	2.2635
2.3199	15.0	1725	2.2320
2.3321	16.0	1840	2.2250
2.3305	17.0	1955	2.2020
2.2932	18.0	2070	2.1826
2.2952	19.0	2185	2.1626
2.2663	20.0	2300	2.1573
2.2916	21.0	2415	2.1653
2.2703	22.0	2530	2.1444
2.2431	23.0	2645	2.1374
2.2243	24.0	2760	2.1276
2.2413	25.0	2875	2.1031

Framework versions

• Transformers 4.35.0
• Pytorch 2.0.0
• Datasets 2.1.0
• Tokenizers 0.14.1