detr-r50-cd45rb-1ah-6l-corrected

This model is a fine-tuned version of facebook/detr-resnet-50 on the cd45rb_nan_xywh dataset. It achieves the following results on the evaluation set:

• Loss: 1.9721

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
1.9229	1.0	4606	1.3205
1.703	2.0	9212	1.2364
1.5976	3.0	13818	1.1736
1.5323	4.0	18424	1.1464
1.483	5.0	23030	1.1084
1.4431	6.0	27636	1.1037
1.4153	7.0	32242	1.0743
1.3898	8.0	36848	1.0755
1.3712	9.0	41454	1.0604
1.3616	10.0	46060	1.0556
2.807	11.0	50666	2.1286
2.6791	12.0	55272	2.1132
2.6331	13.0	59878	2.0794
2.6108	14.0	64484	2.0570
2.5818	15.0	69090	2.0455
2.5693	16.0	73696	2.0330
2.5446	17.0	78302	2.0361
2.5298	18.0	82908	2.0160
2.5098	19.0	87514	2.0395
2.4951	20.0	92120	2.0063
2.4913	21.0	96726	2.0016
2.4703	22.0	101332	1.9800
2.4643	23.0	105938	1.9811
2.4611	24.0	110544	1.9807
2.4568	25.0	115150	1.9721

Framework versions

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