This model is a fine-tuned version of aguxez/snap-plate-stack-instance-dfine-v5 on the aguxez/snap-plate-rim-marker-v6 dataset. It achieves the following results on the evaluation set:

Loss: 4.8775
Map: 0.213
Map 50: 0.5137
Map 75: 0.0035
Map Small: -1.0
Map Medium: 0.2593
Map Large: -1.0
Mar 1: 0.1
Mar 10: 0.25
Mar 100: 0.5
Mar Small: -1.0
Mar Medium: 0.5
Mar Large: -1.0
Map Plate Rim: 0.213
Mar 100 Plate Rim: 0.5

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: 7
optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 0.05
num_epochs: 8.0
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Map	Map 50	Map 75	Map Small	Map Medium	Map Large	Mar 1	Mar 10	Mar 100	Mar Small	Mar Medium	Mar Large	Map Plate Rim	Mar 100 Plate Rim
29.7278	1.0	8	7.8860	0.0003	0.0027	0.0	-1.0	0.0008	-1.0	0.0	0.0	0.025	-1.0	0.025	-1.0	0.0003	0.025
23.3235	2.0	16	6.4459	0.0052	0.0187	0.0	-1.0	0.0114	-1.0	0.0	0.0	0.25	-1.0	0.25	-1.0	0.0052	0.25
21.2134	3.0	24	5.7173	0.0126	0.0384	0.0	-1.0	0.022	-1.0	0.0	0.075	0.275	-1.0	0.275	-1.0	0.0126	0.275
20.3105	4.0	32	4.9907	0.0534	0.1645	0.0	-1.0	0.1002	-1.0	0.0	0.225	0.475	-1.0	0.475	-1.0	0.0534	0.475
18.5370	5.0	40	5.0561	0.0714	0.1992	0.0313	-1.0	0.1518	-1.0	0.0	0.25	0.55	-1.0	0.55	-1.0	0.0714	0.55
18.8839	6.0	48	4.8353	0.1103	0.2628	0.0384	-1.0	0.2293	-1.0	0.1	0.25	0.525	-1.0	0.525	-1.0	0.1103	0.525
18.7469	7.0	56	4.8062	0.1945	0.4755	0.0098	-1.0	0.2643	-1.0	0.1	0.25	0.55	-1.0	0.55	-1.0	0.1945	0.55
18.0282	8.0	64	4.8775	0.213	0.5137	0.0035	-1.0	0.2593	-1.0	0.1	0.25	0.5	-1.0	0.5	-1.0	0.213	0.5

Framework versions

Transformers 5.14.1
Pytorch 2.13.0+cu130
Datasets 5.0.1
Tokenizers 0.22.2

aguxez/snap-plate-rim-marker-dfine-v7

snap-plate-rim-marker-dfine-v7

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

aguxez/
snap-plate-rim-marker-dfine-v7