|
--- |
|
license: apache-2.0 |
|
base_model: google/vit-large-patch32-224-in21k |
|
tags: |
|
- image-classification |
|
- vision |
|
- generated_from_trainer |
|
metrics: |
|
- accuracy |
|
- precision |
|
- recall |
|
- f1 |
|
model-index: |
|
- name: vit-large-patch32-224-in21k-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-patch32-224-in21k-finetuned-galaxy10-decals |
|
|
|
This model is a fine-tuned version of [google/vit-large-patch32-224-in21k](https://huggingface.co/google/vit-large-patch32-224-in21k) on the matthieulel/galaxy10_decals dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.5849 |
|
- Accuracy: 0.8236 |
|
- Precision: 0.8257 |
|
- Recall: 0.8236 |
|
- F1: 0.8212 |
|
|
|
## 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: 32 |
|
- eval_batch_size: 32 |
|
- seed: 42 |
|
- gradient_accumulation_steps: 4 |
|
- total_train_batch_size: 128 |
|
- 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 | |
|
|:-------------:|:-----:|:----:|:---------------:|:--------:|:---------:|:------:|:------:| |
|
| 1.1583 | 0.99 | 124 | 1.0551 | 0.7069 | 0.6559 | 0.7069 | 0.6758 | |
|
| 0.8599 | 2.0 | 249 | 0.7914 | 0.7621 | 0.7717 | 0.7621 | 0.7557 | |
|
| 0.854 | 3.0 | 374 | 0.7115 | 0.7672 | 0.7850 | 0.7672 | 0.7642 | |
|
| 0.7282 | 4.0 | 499 | 0.6807 | 0.7683 | 0.7746 | 0.7683 | 0.7604 | |
|
| 0.6165 | 4.99 | 623 | 0.6208 | 0.8016 | 0.8088 | 0.8016 | 0.8015 | |
|
| 0.5946 | 6.0 | 748 | 0.5850 | 0.8044 | 0.8084 | 0.8044 | 0.8009 | |
|
| 0.6243 | 7.0 | 873 | 0.6090 | 0.7931 | 0.8037 | 0.7931 | 0.7935 | |
|
| 0.5429 | 8.0 | 998 | 0.5830 | 0.8021 | 0.8087 | 0.8021 | 0.8006 | |
|
| 0.558 | 8.99 | 1122 | 0.5725 | 0.8095 | 0.8191 | 0.8095 | 0.8081 | |
|
| 0.457 | 10.0 | 1247 | 0.5702 | 0.8123 | 0.8144 | 0.8123 | 0.8085 | |
|
| 0.4399 | 11.0 | 1372 | 0.5973 | 0.8021 | 0.8013 | 0.8021 | 0.7995 | |
|
| 0.4055 | 12.0 | 1497 | 0.5799 | 0.8157 | 0.8186 | 0.8157 | 0.8122 | |
|
| 0.417 | 12.99 | 1621 | 0.6006 | 0.8061 | 0.8175 | 0.8061 | 0.8066 | |
|
| 0.3843 | 14.0 | 1746 | 0.5849 | 0.8236 | 0.8257 | 0.8236 | 0.8212 | |
|
| 0.371 | 15.0 | 1871 | 0.5711 | 0.8196 | 0.8157 | 0.8196 | 0.8161 | |
|
| 0.3546 | 16.0 | 1996 | 0.6050 | 0.8140 | 0.8171 | 0.8140 | 0.8147 | |
|
| 0.2935 | 16.99 | 2120 | 0.6425 | 0.8106 | 0.8159 | 0.8106 | 0.8091 | |
|
| 0.2505 | 18.0 | 2245 | 0.6569 | 0.8112 | 0.8091 | 0.8112 | 0.8086 | |
|
| 0.3094 | 19.0 | 2370 | 0.6558 | 0.8162 | 0.8137 | 0.8162 | 0.8137 | |
|
| 0.2739 | 20.0 | 2495 | 0.7201 | 0.8067 | 0.8094 | 0.8067 | 0.8025 | |
|
| 0.2224 | 20.99 | 2619 | 0.7227 | 0.8140 | 0.8175 | 0.8140 | 0.8114 | |
|
| 0.2359 | 22.0 | 2744 | 0.6941 | 0.8157 | 0.8142 | 0.8157 | 0.8136 | |
|
| 0.2535 | 23.0 | 2869 | 0.7086 | 0.8157 | 0.8160 | 0.8157 | 0.8123 | |
|
| 0.2047 | 24.0 | 2994 | 0.7185 | 0.8236 | 0.8236 | 0.8236 | 0.8207 | |
|
| 0.2162 | 24.99 | 3118 | 0.7135 | 0.8219 | 0.8200 | 0.8219 | 0.8194 | |
|
| 0.2297 | 26.0 | 3243 | 0.7269 | 0.8213 | 0.8172 | 0.8213 | 0.8179 | |
|
| 0.2048 | 27.0 | 3368 | 0.7392 | 0.8145 | 0.8156 | 0.8145 | 0.8143 | |
|
| 0.2156 | 28.0 | 3493 | 0.7453 | 0.8207 | 0.8182 | 0.8207 | 0.8174 | |
|
| 0.1785 | 28.99 | 3617 | 0.7497 | 0.8168 | 0.8157 | 0.8168 | 0.8145 | |
|
| 0.1785 | 29.82 | 3720 | 0.7429 | 0.8202 | 0.8190 | 0.8202 | 0.8173 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.37.2 |
|
- Pytorch 2.3.0 |
|
- Datasets 2.19.1 |
|
- Tokenizers 0.15.1 |
|
|
