Configuration Parsing Warning:In adapter_config.json: "peft.task_type" must be a string
bigearthnet-vit
This model is a fine-tuned version of google/vit-base-patch16-224 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.0968
- F1 Micro: 0.6503
- F1 Macro: 0.2662
- Map: 0.3291
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: 0.0001
- train_batch_size: 512
- eval_batch_size: 512
- seed: 42
- 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: cosine_with_restarts
- lr_scheduler_warmup_steps: 500
- num_epochs: 2
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | F1 Macro | F1 Micro | Validation Loss | Map |
|---|---|---|---|---|---|---|
| 0.1646 | 0.9488 | 500 | 0.2670 | 0.6567 | 0.0970 | 0.3339 |
| 0.1245 | 1.8975 | 1000 | 0.0971 | 0.6492 | 0.2668 | 0.3284 |
| 0.1245 | 2.0 | 1054 | 0.0968 | 0.6503 | 0.2662 | 0.3291 |
Framework versions
- PEFT 0.19.1
- Transformers 5.8.1
- Pytorch 2.10.0+cu128
- Datasets 4.8.5
- Tokenizers 0.22.2
- Downloads last month
- 13
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support
Model tree for Ganymede981/bigearthnet-vit
Base model
google/vit-base-patch16-224