trashbot
This model is a fine-tuned version of nvidia/mit-b5 on the mraottth/all_locations_pooled dataset. It achieves the following results on the evaluation set:
- Loss: 0.0191
- Mean Iou: 0.3997
- Mean Accuracy: 0.7995
- Overall Accuracy: 0.7995
- Accuracy Unlabeled: nan
- Accuracy Trash: 0.7995
- Iou Unlabeled: 0.0
- Iou Trash: 0.7995
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: 3
- eval_batch_size: 3
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10
Training results
Training Loss | Epoch | Step | Validation Loss | Mean Iou | Mean Accuracy | Overall Accuracy | Accuracy Unlabeled | Accuracy Trash | Iou Unlabeled | Iou Trash |
---|---|---|---|---|---|---|---|---|---|---|
0.0748 | 1.0 | 90 | 0.0386 | 0.3630 | 0.7259 | 0.7259 | nan | 0.7259 | 0.0 | 0.7259 |
0.039 | 2.0 | 180 | 0.0242 | 0.3803 | 0.7607 | 0.7607 | nan | 0.7607 | 0.0 | 0.7607 |
0.0194 | 3.0 | 270 | 0.0242 | 0.3605 | 0.7210 | 0.7210 | nan | 0.7210 | 0.0 | 0.7210 |
0.0112 | 4.0 | 360 | 0.0205 | 0.3995 | 0.7991 | 0.7991 | nan | 0.7991 | 0.0 | 0.7991 |
0.0169 | 5.0 | 450 | 0.0192 | 0.4000 | 0.8000 | 0.8000 | nan | 0.8000 | 0.0 | 0.8000 |
0.041 | 6.0 | 540 | 0.0196 | 0.3838 | 0.7677 | 0.7677 | nan | 0.7677 | 0.0 | 0.7677 |
0.0188 | 7.0 | 630 | 0.0191 | 0.4139 | 0.8277 | 0.8277 | nan | 0.8277 | 0.0 | 0.8277 |
0.0073 | 8.0 | 720 | 0.0190 | 0.4069 | 0.8138 | 0.8138 | nan | 0.8138 | 0.0 | 0.8138 |
0.025 | 9.0 | 810 | 0.0191 | 0.4087 | 0.8174 | 0.8174 | nan | 0.8174 | 0.0 | 0.8174 |
0.006 | 10.0 | 900 | 0.0191 | 0.3997 | 0.7995 | 0.7995 | nan | 0.7995 | 0.0 | 0.7995 |
Framework versions
- Transformers 4.26.0
- Pytorch 1.13.1+cu116
- Datasets 2.9.0
- Tokenizers 0.13.2
- Downloads last month
- 8
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.