target_hold_hands

This model is a fine-tuned version of facebook/detr-resnet-50 on the c14kevincardenas/beta_caller_284_target_hold_hands dataset. It achieves the following results on the evaluation set:

• Loss: 0.6006
• Iou: 0.0000

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: 64
• eval_batch_size: 64
• seed: 2014
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 250
• num_epochs: 20.0
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Iou
0.957	1.0	52	0.8557	0.0
0.7261	2.0	104	0.6803	0.0000
0.6689	3.0	156	0.6545	0.0000
0.7126	4.0	208	0.7020	0.0000
0.6688	5.0	260	0.6712	0.0000
0.7126	6.0	312	0.6633	0.0000
0.6633	7.0	364	0.6083	0.0000
0.6113	8.0	416	0.6061	0.0000
0.6101	9.0	468	0.6027	0.0000
0.6028	10.0	520	0.6007	0.0
0.5983	11.0	572	0.6019	0.0000
0.6014	12.0	624	0.6006	0.0000
0.5968	13.0	676	0.6014	0.0000
0.5932	14.0	728	0.6021	0.0000
0.592	15.0	780	0.6047	0.0
0.5935	16.0	832	0.6020	0.0000
0.5873	17.0	884	0.6026	0.0000
0.5853	18.0	936	0.6115	0.0000
0.5835	19.0	988	0.6087	0.0000
0.5868	20.0	1040	0.6076	0.0000

Framework versions

• Transformers 4.45.2
• Pytorch 2.5.0+cu124
• Datasets 3.0.1
• Tokenizers 0.20.1