metadata
license: apache-2.0
library_name: peft
tags:
- generated_from_trainer
datasets:
- medmnist-v2
metrics:
- accuracy
- precision
- recall
- f1
base_model: microsoft/swin-large-patch4-window7-224-in22k
model-index:
- name: blood-swin-base-finetuned-wandb
results: []
blood-swin-base-finetuned-wandb
This model is a fine-tuned version of microsoft/swin-large-patch4-window7-224-in22k on the medmnist-v2 dataset. It achieves the following results on the evaluation set:
- Loss: 0.1036
- Accuracy: 0.9649
- Precision: 0.9627
- Recall: 0.9616
- F1: 0.9619
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: 0.005
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1 |
|---|---|---|---|---|---|---|---|
| 0.5141 | 1.0 | 187 | 0.2833 | 0.9065 | 0.8954 | 0.8949 | 0.8873 |
| 0.4176 | 2.0 | 374 | 0.1986 | 0.9311 | 0.9243 | 0.9209 | 0.9182 |
| 0.3454 | 3.0 | 561 | 0.1567 | 0.9504 | 0.9427 | 0.9397 | 0.9403 |
| 0.3228 | 4.0 | 748 | 0.1849 | 0.9357 | 0.9232 | 0.9426 | 0.9283 |
| 0.3382 | 5.0 | 935 | 0.1627 | 0.9398 | 0.9302 | 0.9397 | 0.9321 |
| 0.3363 | 6.0 | 1122 | 0.1414 | 0.9509 | 0.9498 | 0.9442 | 0.9456 |
| 0.2981 | 7.0 | 1309 | 0.1117 | 0.9544 | 0.9458 | 0.9542 | 0.9480 |
| 0.2214 | 8.0 | 1496 | 0.1131 | 0.9650 | 0.9642 | 0.9584 | 0.9610 |
| 0.1928 | 9.0 | 1683 | 0.0966 | 0.9650 | 0.9632 | 0.9628 | 0.9624 |
| 0.1901 | 10.0 | 1870 | 0.0775 | 0.9714 | 0.9690 | 0.9699 | 0.9692 |
Framework versions
- PEFT 0.9.0
- Transformers 4.38.2
- Pytorch 2.2.1+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2