detr-r50-cd45rb-8ah-6l-gelu-corrected

This model is a fine-tuned version of facebook/detr-resnet-50 on the cd45rb dataset. It achieves the following results on the evaluation set:

• Loss: 1.5861

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
3.0438	1.0	4606	1.9413
2.3933	2.0	9212	1.8238
2.2782	3.0	13818	1.7718
2.2383	4.0	18424	1.7528
2.2046	5.0	23030	1.7265
2.1659	6.0	27636	1.7125
2.1457	7.0	32242	1.6760
2.1111	8.0	36848	1.6622
2.0959	9.0	41454	1.6467
2.0826	10.0	46060	1.6392
2.1132	11.0	50666	1.6875
2.1145	12.0	55272	1.6863
2.0947	13.0	59878	1.6528
2.0777	14.0	64484	1.6669
2.0551	15.0	69090	1.6241
2.0567	16.0	73696	1.6241
2.042	17.0	78302	1.6171
2.0306	18.0	82908	1.6062
2.015	19.0	87514	1.5989
2.0206	20.0	92120	1.6168
2.026	21.0	96726	1.6022
2.0109	22.0	101332	1.5996
2.0133	23.0	105938	1.5983
2.0081	24.0	110544	1.5888
1.9975	25.0	115150	1.5861

Framework versions

• Transformers 4.28.0
• Pytorch 2.0.1
• Datasets 2.12.0
• Tokenizers 0.13.3