|
--- |
|
license: apache-2.0 |
|
base_model: microsoft/swinv2-large-patch4-window12-192-22k |
|
tags: |
|
- image-classification |
|
- vision |
|
- generated_from_trainer |
|
metrics: |
|
- accuracy |
|
- precision |
|
- recall |
|
- f1 |
|
model-index: |
|
- name: swinv2-large-patch4-window12-192-22k-finetuned-galaxy10-decals |
|
results: [] |
|
--- |
|
|
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You |
|
should probably proofread and complete it, then remove this comment. --> |
|
|
|
# swinv2-large-patch4-window12-192-22k-finetuned-galaxy10-decals |
|
|
|
This model is a fine-tuned version of [microsoft/swinv2-large-patch4-window12-192-22k](https://huggingface.co/microsoft/swinv2-large-patch4-window12-192-22k) on the matthieulel/galaxy10_decals dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.4372 |
|
- Accuracy: 0.8568 |
|
- Precision: 0.8575 |
|
- Recall: 0.8568 |
|
- F1: 0.8550 |
|
|
|
## 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: 5e-05 |
|
- 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: 0.1 |
|
- num_epochs: 30 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1 | |
|
|:-------------:|:-----:|:----:|:---------------:|:--------:|:---------:|:------:|:------:| |
|
| 0.974 | 0.99 | 62 | 0.7350 | 0.7480 | 0.7464 | 0.7480 | 0.7365 | |
|
| 0.7716 | 2.0 | 125 | 0.6093 | 0.7982 | 0.8102 | 0.7982 | 0.7960 | |
|
| 0.6813 | 2.99 | 187 | 0.5034 | 0.8286 | 0.8301 | 0.8286 | 0.8254 | |
|
| 0.5998 | 4.0 | 250 | 0.4645 | 0.8433 | 0.8431 | 0.8433 | 0.8403 | |
|
| 0.5306 | 4.99 | 312 | 0.4889 | 0.8320 | 0.8377 | 0.8320 | 0.8336 | |
|
| 0.5234 | 6.0 | 375 | 0.5036 | 0.8309 | 0.8398 | 0.8309 | 0.8278 | |
|
| 0.4984 | 6.99 | 437 | 0.4482 | 0.8478 | 0.8484 | 0.8478 | 0.8461 | |
|
| 0.456 | 8.0 | 500 | 0.4370 | 0.8557 | 0.8573 | 0.8557 | 0.8557 | |
|
| 0.4672 | 8.99 | 562 | 0.4372 | 0.8568 | 0.8575 | 0.8568 | 0.8550 | |
|
| 0.4211 | 10.0 | 625 | 0.4428 | 0.8523 | 0.8513 | 0.8523 | 0.8505 | |
|
| 0.4228 | 10.99 | 687 | 0.4762 | 0.8433 | 0.8459 | 0.8433 | 0.8435 | |
|
| 0.3966 | 12.0 | 750 | 0.4943 | 0.8410 | 0.8434 | 0.8410 | 0.8404 | |
|
| 0.383 | 12.99 | 812 | 0.4885 | 0.8478 | 0.8503 | 0.8478 | 0.8463 | |
|
| 0.3899 | 14.0 | 875 | 0.5021 | 0.8472 | 0.8494 | 0.8472 | 0.8474 | |
|
| 0.3364 | 14.99 | 937 | 0.5107 | 0.8495 | 0.8488 | 0.8495 | 0.8486 | |
|
| 0.331 | 16.0 | 1000 | 0.5219 | 0.8484 | 0.8460 | 0.8484 | 0.8454 | |
|
| 0.288 | 16.99 | 1062 | 0.5696 | 0.8422 | 0.8429 | 0.8422 | 0.8410 | |
|
| 0.2867 | 18.0 | 1125 | 0.5529 | 0.8484 | 0.8474 | 0.8484 | 0.8473 | |
|
| 0.2889 | 18.99 | 1187 | 0.5613 | 0.8529 | 0.8522 | 0.8529 | 0.8520 | |
|
| 0.2809 | 20.0 | 1250 | 0.6093 | 0.8433 | 0.8378 | 0.8433 | 0.8391 | |
|
| 0.2684 | 20.99 | 1312 | 0.6096 | 0.8444 | 0.8409 | 0.8444 | 0.8419 | |
|
| 0.2809 | 22.0 | 1375 | 0.6100 | 0.8455 | 0.8453 | 0.8455 | 0.8445 | |
|
| 0.2661 | 22.99 | 1437 | 0.6161 | 0.8354 | 0.8378 | 0.8354 | 0.8359 | |
|
| 0.2435 | 24.0 | 1500 | 0.6540 | 0.8517 | 0.8512 | 0.8517 | 0.8512 | |
|
| 0.2593 | 24.99 | 1562 | 0.6644 | 0.8472 | 0.8462 | 0.8472 | 0.8456 | |
|
| 0.2343 | 26.0 | 1625 | 0.6655 | 0.8467 | 0.8441 | 0.8467 | 0.8449 | |
|
| 0.2281 | 26.99 | 1687 | 0.6759 | 0.8450 | 0.8438 | 0.8450 | 0.8440 | |
|
| 0.2334 | 28.0 | 1750 | 0.6836 | 0.8472 | 0.8445 | 0.8472 | 0.8451 | |
|
| 0.2129 | 28.99 | 1812 | 0.6731 | 0.8489 | 0.8466 | 0.8489 | 0.8471 | |
|
| 0.2252 | 29.76 | 1860 | 0.6773 | 0.8467 | 0.8440 | 0.8467 | 0.8449 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.37.2 |
|
- Pytorch 2.3.0 |
|
- Datasets 2.19.1 |
|
- Tokenizers 0.15.1 |
|
|
