metadata
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- imagefolder
metrics:
- accuracy
- f1
- precision
- recall
model-index:
- name: swinv2-tiny-patch4-window8-256-finetuned-eurosat
results:
- task:
name: Image Classification
type: image-classification
dataset:
name: imagefolder
type: imagefolder
config: default
split: train
args: default
metrics:
- name: Accuracy
type: accuracy
value: 0.9825925925925926
- name: F1
type: f1
value: 0.9825868474705166
- name: Precision
type: precision
value: 0.9828193476192771
- name: Recall
type: recall
value: 0.9825925925925926
swinv2-tiny-patch4-window8-256-finetuned-eurosat
This model is a fine-tuned version of microsoft/swinv2-tiny-patch4-window8-256 on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 0.0510
- Accuracy: 0.9826
- F1: 0.9826
- Precision: 0.9828
- Recall: 0.9826
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: 64
- eval_batch_size: 64
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 256
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.2
- num_epochs: 5
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 | Precision | Recall |
|---|---|---|---|---|---|---|---|
| 0.4479 | 1.0 | 95 | 0.1592 | 0.9478 | 0.9478 | 0.9500 | 0.9478 |
| 0.3078 | 2.0 | 190 | 0.0914 | 0.9685 | 0.9686 | 0.9695 | 0.9685 |
| 0.2307 | 3.0 | 285 | 0.0603 | 0.9785 | 0.9785 | 0.9790 | 0.9785 |
| 0.227 | 4.0 | 380 | 0.0531 | 0.9811 | 0.9811 | 0.9814 | 0.9811 |
| 0.1674 | 5.0 | 475 | 0.0510 | 0.9826 | 0.9826 | 0.9828 | 0.9826 |
Framework versions
- Transformers 4.22.1
- Pytorch 1.12.1+cu113
- Datasets 2.5.1
- Tokenizers 0.12.1