umaxtools-b0-v01-finetuned-segments-outputs

This model is a fine-tuned version of nvidia/mit-b0 on the jenniferlumeng/umaxtools dataset. It achieves the following results on the evaluation set:

• Loss: 0.4120
• Mean Iou: 0.4577
• Mean Accuracy: 0.9154
• Overall Accuracy: 0.9154
• Accuracy Background: nan
• Accuracy Object: 0.9154
• Iou Background: 0.0
• Iou Object: 0.9154

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: 6e-05
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Mean Iou	Mean Accuracy	Overall Accuracy	Accuracy Background	Accuracy Object	Iou Background	Iou Object
0.4422	5.0	10	0.5913	0.4820	0.9640	0.9640	nan	0.9640	0.0	0.9640
0.3684	10.0	20	0.5141	0.4821	0.9643	0.9643	nan	0.9643	0.0	0.9643
0.294	15.0	30	0.4940	0.4325	0.8650	0.8650	nan	0.8650	0.0	0.8650
0.2614	20.0	40	0.4563	0.4451	0.8901	0.8901	nan	0.8901	0.0	0.8901
0.2981	25.0	50	0.4381	0.4500	0.8999	0.8999	nan	0.8999	0.0	0.8999
0.245	30.0	60	0.4189	0.4588	0.9176	0.9176	nan	0.9176	0.0	0.9176
0.2356	35.0	70	0.4185	0.4502	0.9004	0.9004	nan	0.9004	0.0	0.9004
0.2228	40.0	80	0.4165	0.4583	0.9166	0.9166	nan	0.9166	0.0	0.9166
0.2162	45.0	90	0.4128	0.4613	0.9226	0.9226	nan	0.9226	0.0	0.9226
0.2288	50.0	100	0.4120	0.4577	0.9154	0.9154	nan	0.9154	0.0	0.9154

Framework versions

• Transformers 4.52.2
• Pytorch 2.6.0+cu124
• Datasets 2.16.1
• Tokenizers 0.21.1