metadata
library_name: transformers
license: other
base_model: nvidia/mit-b0
tags:
- vision
- image-segmentation
- generated_from_trainer
model-index:
- name: segformer-b0-finetuned-segments-sidewalk-oct-22
results: []
segformer-b0-finetuned-segments-sidewalk-oct-22
This model is a fine-tuned version of nvidia/mit-b0 on the adsjnajefwnb/di dataset. It achieves the following results on the evaluation set:
- Loss: 1.1425
- Mean Iou: 0.4608
- Mean Accuracy: 0.6144
- Overall Accuracy: 0.9524
- Accuracy Unlabeled: nan
- Accuracy Anqvan: nan
- Accuracy Baohuxiang: nan
- Accuracy Biaozhi: nan
- Accuracy Dianlan: 0.9591
- Accuracy Miehuoqi: 0.8841
- Accuracy Zhaoming: 0.0
- Iou Unlabeled: 0.0
- Iou Anqvan: nan
- Iou Baohuxiang: nan
- Iou Biaozhi: nan
- Iou Dianlan: 0.9591
- Iou Miehuoqi: 0.8841
- Iou Zhaoming: 0.0
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 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 Unlabeled | Accuracy Anqvan | Accuracy Baohuxiang | Accuracy Biaozhi | Accuracy Dianlan | Accuracy Miehuoqi | Accuracy Zhaoming | Iou Unlabeled | Iou Anqvan | Iou Baohuxiang | Iou Biaozhi | Iou Dianlan | Iou Miehuoqi | Iou Zhaoming |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1.5348 | 10.0 | 20 | 1.8080 | 0.3181 | 0.6370 | 0.9720 | nan | nan | nan | nan | 0.9776 | 0.9334 | 0.0 | 0.0 | nan | 0.0 | 0.0 | 0.9759 | 0.9328 | 0.0 |
| 1.5003 | 20.0 | 40 | 1.5287 | 0.3865 | 0.6442 | 0.9726 | nan | nan | nan | nan | 0.9773 | 0.9552 | 0.0 | 0.0 | nan | nan | 0.0 | 0.9773 | 0.9552 | 0.0 |
| 1.4909 | 30.0 | 60 | 1.2543 | 0.4770 | 0.6359 | 0.9695 | nan | nan | nan | nan | 0.9750 | 0.9328 | 0.0 | 0.0 | nan | nan | nan | 0.9750 | 0.9328 | 0.0 |
| 1.331 | 40.0 | 80 | 1.1463 | 0.4671 | 0.6229 | 0.9613 | nan | nan | nan | nan | 0.9677 | 0.9009 | 0.0 | 0.0 | nan | nan | nan | 0.9677 | 0.9009 | 0.0 |
| 1.0843 | 50.0 | 100 | 1.1425 | 0.4608 | 0.6144 | 0.9524 | nan | nan | nan | nan | 0.9591 | 0.8841 | 0.0 | 0.0 | nan | nan | nan | 0.9591 | 0.8841 | 0.0 |
Framework versions
- Transformers 4.46.3
- Pytorch 1.12.1+cu113
- Datasets 3.1.0
- Tokenizers 0.20.3