File size: 4,432 Bytes
acce1f1 ae16f21 acce1f1 ae16f21 acce1f1 ae16f21 acce1f1 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 |
---
license: apache-2.0
base_model: google/vit-large-patch32-384
tags:
- image-classification
- vision
- generated_from_trainer
metrics:
- accuracy
- precision
- recall
- f1
model-index:
- name: vit-large-patch32-384-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. -->
# vit-large-patch32-384-finetuned-galaxy10-decals
This model is a fine-tuned version of [google/vit-large-patch32-384](https://huggingface.co/google/vit-large-patch32-384) on the matthieulel/galaxy10_decals dataset.
It achieves the following results on the evaluation set:
- Loss: 0.6766
- Accuracy: 0.8371
- Precision: 0.8374
- Recall: 0.8371
- F1: 0.8357
## 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: 128
- eval_batch_size: 128
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 512
- 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 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:---------:|:------:|:------:|
| 1.3342 | 0.99 | 31 | 1.0491 | 0.6313 | 0.6077 | 0.6313 | 0.6052 |
| 0.7979 | 1.98 | 62 | 0.6901 | 0.7672 | 0.7717 | 0.7672 | 0.7652 |
| 0.7197 | 2.98 | 93 | 0.6200 | 0.7785 | 0.7716 | 0.7785 | 0.7705 |
| 0.6321 | 4.0 | 125 | 0.5693 | 0.8061 | 0.8035 | 0.8061 | 0.7957 |
| 0.5768 | 4.99 | 156 | 0.5501 | 0.8112 | 0.8213 | 0.8112 | 0.8134 |
| 0.5173 | 5.98 | 187 | 0.5165 | 0.8213 | 0.8306 | 0.8213 | 0.8202 |
| 0.4781 | 6.98 | 218 | 0.5220 | 0.8106 | 0.8161 | 0.8106 | 0.8090 |
| 0.451 | 8.0 | 250 | 0.5133 | 0.8185 | 0.8227 | 0.8185 | 0.8153 |
| 0.4373 | 8.99 | 281 | 0.5118 | 0.8303 | 0.8325 | 0.8303 | 0.8288 |
| 0.3826 | 9.98 | 312 | 0.5280 | 0.8258 | 0.8269 | 0.8258 | 0.8243 |
| 0.378 | 10.98 | 343 | 0.5477 | 0.8174 | 0.8156 | 0.8174 | 0.8142 |
| 0.3509 | 12.0 | 375 | 0.5437 | 0.8281 | 0.8292 | 0.8281 | 0.8244 |
| 0.3358 | 12.99 | 406 | 0.5627 | 0.8258 | 0.8268 | 0.8258 | 0.8241 |
| 0.3027 | 13.98 | 437 | 0.5558 | 0.8326 | 0.8341 | 0.8326 | 0.8310 |
| 0.3027 | 14.98 | 468 | 0.5703 | 0.8326 | 0.8358 | 0.8326 | 0.8295 |
| 0.2786 | 16.0 | 500 | 0.5791 | 0.8281 | 0.8268 | 0.8281 | 0.8249 |
| 0.2379 | 16.99 | 531 | 0.5864 | 0.8275 | 0.8264 | 0.8275 | 0.8251 |
| 0.2426 | 17.98 | 562 | 0.5984 | 0.8320 | 0.8320 | 0.8320 | 0.8305 |
| 0.2325 | 18.98 | 593 | 0.6217 | 0.8264 | 0.8281 | 0.8264 | 0.8252 |
| 0.2208 | 20.0 | 625 | 0.6166 | 0.8258 | 0.8230 | 0.8258 | 0.8236 |
| 0.2196 | 20.99 | 656 | 0.6308 | 0.8286 | 0.8280 | 0.8286 | 0.8259 |
| 0.2077 | 21.98 | 687 | 0.6242 | 0.8326 | 0.8307 | 0.8326 | 0.8305 |
| 0.2048 | 22.98 | 718 | 0.6801 | 0.8275 | 0.8303 | 0.8275 | 0.8263 |
| 0.1886 | 24.0 | 750 | 0.6615 | 0.8264 | 0.8280 | 0.8264 | 0.8256 |
| 0.2007 | 24.99 | 781 | 0.6847 | 0.8275 | 0.8280 | 0.8275 | 0.8267 |
| 0.1815 | 25.98 | 812 | 0.6669 | 0.8326 | 0.8311 | 0.8326 | 0.8305 |
| 0.1958 | 26.98 | 843 | 0.6766 | 0.8371 | 0.8374 | 0.8371 | 0.8357 |
| 0.1806 | 28.0 | 875 | 0.6679 | 0.8360 | 0.8353 | 0.8360 | 0.8342 |
| 0.1835 | 28.99 | 906 | 0.6767 | 0.8348 | 0.8334 | 0.8348 | 0.8328 |
| 0.1796 | 29.76 | 930 | 0.6787 | 0.8343 | 0.8336 | 0.8343 | 0.8326 |
### Framework versions
- Transformers 4.37.2
- Pytorch 2.3.0
- Datasets 2.19.1
- Tokenizers 0.15.1
|