metadata
library_name: transformers
license: other
base_model: facebook/mask2former-swin-tiny-coco-instance
tags:
- image-segmentation
- instance-segmentation
- vision
- generated_from_trainer
model-index:
- name: finetune-instance-segmentation-ade20k-mini-mask2former
results: []
finetune-instance-segmentation-ade20k-mini-mask2former
This model is a fine-tuned version of facebook/mask2former-swin-tiny-coco-instance on the qubvel-hf/ade20k-mini dataset. It achieves the following results on the evaluation set:
- Loss: 28.4481
- Map: 0.2172
- Map 50: 0.4234
- Map 75: 0.2041
- Map Small: 0.1458
- Map Medium: 0.6353
- Map Large: 0.8076
- Mar 1: 0.0953
- Mar 10: 0.254
- Mar 100: 0.2903
- Mar Small: 0.2169
- Mar Medium: 0.7113
- Mar Large: 0.8594
- Map Person: 0.1476
- Mar 100 Person: 0.205
- Map Car: 0.2867
- Mar 100 Car: 0.3755
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: 1e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: constant
- num_epochs: 4.0
- 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 Person | Mar 100 Person | Map Car | Mar 100 Car |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 33.5412 | 1.0 | 100 | 31.5328 | 0.1929 | 0.3918 | 0.1737 | 0.1281 | 0.6122 | 0.7895 | 0.0904 | 0.2473 | 0.2836 | 0.2105 | 0.7063 | 0.8229 | 0.13 | 0.2001 | 0.2558 | 0.3672 |
| 27.9471 | 2.0 | 200 | 29.7181 | 0.2053 | 0.4151 | 0.1851 | 0.1387 | 0.6192 | 0.8018 | 0.093 | 0.2507 | 0.2872 | 0.2142 | 0.7079 | 0.8323 | 0.1364 | 0.2029 | 0.2741 | 0.3714 |
| 26.4855 | 3.0 | 300 | 28.9786 | 0.2134 | 0.4219 | 0.1945 | 0.1451 | 0.6255 | 0.8047 | 0.0944 | 0.2543 | 0.2918 | 0.2198 | 0.7045 | 0.8594 | 0.143 | 0.2059 | 0.2837 | 0.3777 |
| 25.4746 | 4.0 | 400 | 28.4481 | 0.2172 | 0.4234 | 0.2041 | 0.1458 | 0.6353 | 0.8076 | 0.0953 | 0.254 | 0.2903 | 0.2169 | 0.7113 | 0.8594 | 0.1476 | 0.205 | 0.2867 | 0.3755 |
Framework versions
- Transformers 4.48.0.dev0
- Pytorch 2.5.0+cu121
- Datasets 2.19.1
- Tokenizers 0.21.0