|
--- |
|
license: apache-2.0 |
|
base_model: microsoft/beit-large-patch16-224-pt22k |
|
tags: |
|
- image-classification |
|
- vision |
|
- generated_from_trainer |
|
metrics: |
|
- accuracy |
|
- precision |
|
- recall |
|
- f1 |
|
model-index: |
|
- name: beit-large-patch16-224-pt22k-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. --> |
|
|
|
# beit-large-patch16-224-pt22k-finetuned-galaxy10-decals |
|
|
|
This model is a fine-tuned version of [microsoft/beit-large-patch16-224-pt22k](https://huggingface.co/microsoft/beit-large-patch16-224-pt22k) on the matthieulel/galaxy10_decals dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.5038 |
|
- Accuracy: 0.8794 |
|
- Precision: 0.8781 |
|
- Recall: 0.8794 |
|
- F1: 0.8780 |
|
|
|
## 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 | |
|
|:-------------:|:-----:|:----:|:---------------:|:--------:|:---------:|:------:|:------:| |
|
| 1.5123 | 0.99 | 62 | 1.2940 | 0.5276 | 0.5208 | 0.5276 | 0.5021 | |
|
| 0.9691 | 2.0 | 125 | 0.7947 | 0.7272 | 0.7161 | 0.7272 | 0.7095 | |
|
| 0.7326 | 2.99 | 187 | 0.5790 | 0.8010 | 0.7979 | 0.8010 | 0.7970 | |
|
| 0.6346 | 4.0 | 250 | 0.6230 | 0.7931 | 0.7984 | 0.7931 | 0.7883 | |
|
| 0.5945 | 4.99 | 312 | 0.5042 | 0.8360 | 0.8390 | 0.8360 | 0.8349 | |
|
| 0.5607 | 6.0 | 375 | 0.4401 | 0.8455 | 0.8464 | 0.8455 | 0.8421 | |
|
| 0.5137 | 6.99 | 437 | 0.4689 | 0.8506 | 0.8533 | 0.8506 | 0.8449 | |
|
| 0.4842 | 8.0 | 500 | 0.4586 | 0.8484 | 0.8560 | 0.8484 | 0.8498 | |
|
| 0.4816 | 8.99 | 562 | 0.4310 | 0.8534 | 0.8548 | 0.8534 | 0.8518 | |
|
| 0.4538 | 10.0 | 625 | 0.4380 | 0.8529 | 0.8528 | 0.8529 | 0.8493 | |
|
| 0.4334 | 10.99 | 687 | 0.4288 | 0.8625 | 0.8628 | 0.8625 | 0.8617 | |
|
| 0.4086 | 12.0 | 750 | 0.4904 | 0.8608 | 0.8627 | 0.8608 | 0.8592 | |
|
| 0.4143 | 12.99 | 812 | 0.4148 | 0.8675 | 0.8697 | 0.8675 | 0.8663 | |
|
| 0.4164 | 14.0 | 875 | 0.4477 | 0.8647 | 0.8676 | 0.8647 | 0.8649 | |
|
| 0.3464 | 14.99 | 937 | 0.4843 | 0.8512 | 0.8534 | 0.8512 | 0.8500 | |
|
| 0.3654 | 16.0 | 1000 | 0.4632 | 0.8625 | 0.8631 | 0.8625 | 0.8619 | |
|
| 0.2933 | 16.99 | 1062 | 0.4811 | 0.8596 | 0.8605 | 0.8596 | 0.8574 | |
|
| 0.3299 | 18.0 | 1125 | 0.4574 | 0.8664 | 0.8664 | 0.8664 | 0.8656 | |
|
| 0.3178 | 18.99 | 1187 | 0.4504 | 0.8703 | 0.8697 | 0.8703 | 0.8687 | |
|
| 0.2976 | 20.0 | 1250 | 0.5002 | 0.8636 | 0.8619 | 0.8636 | 0.8610 | |
|
| 0.2982 | 20.99 | 1312 | 0.4977 | 0.8720 | 0.8701 | 0.8720 | 0.8701 | |
|
| 0.3092 | 22.0 | 1375 | 0.4820 | 0.8703 | 0.8710 | 0.8703 | 0.8687 | |
|
| 0.2835 | 22.99 | 1437 | 0.4671 | 0.8715 | 0.8711 | 0.8715 | 0.8709 | |
|
| 0.2596 | 24.0 | 1500 | 0.5075 | 0.8732 | 0.8737 | 0.8732 | 0.8729 | |
|
| 0.2669 | 24.99 | 1562 | 0.4963 | 0.8732 | 0.8719 | 0.8732 | 0.8716 | |
|
| 0.2409 | 26.0 | 1625 | 0.4955 | 0.8766 | 0.8749 | 0.8766 | 0.8754 | |
|
| 0.2409 | 26.99 | 1687 | 0.4988 | 0.8777 | 0.8783 | 0.8777 | 0.8776 | |
|
| 0.2683 | 28.0 | 1750 | 0.5038 | 0.8794 | 0.8781 | 0.8794 | 0.8780 | |
|
| 0.2299 | 28.99 | 1812 | 0.5038 | 0.8771 | 0.8760 | 0.8771 | 0.8759 | |
|
| 0.2394 | 29.76 | 1860 | 0.5048 | 0.8788 | 0.8779 | 0.8788 | 0.8775 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.37.2 |
|
- Pytorch 2.3.0 |
|
- Datasets 2.19.1 |
|
- Tokenizers 0.15.1 |
|
|
