File size: 2,058 Bytes
9efed4a |
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 |
---
license: apache-2.0
base_model: google/vit-base-patch16-224-in21k
tags:
- generated_from_trainer
metrics:
- accuracy
model-index:
- name: ViTuned_buildings
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. -->
# ViTuned_buildings
This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.0432
- Accuracy: 0.9931
## 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: 0.0002
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 4
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
| 1.1985 | 0.33 | 100 | 1.1271 | 0.9726 |
| 0.4085 | 0.67 | 200 | 0.3959 | 0.9743 |
| 0.186 | 1.0 | 300 | 0.1963 | 0.9846 |
| 0.1066 | 1.34 | 400 | 0.2404 | 0.9417 |
| 0.1117 | 1.67 | 500 | 0.1423 | 0.9726 |
| 0.0923 | 2.01 | 600 | 0.1076 | 0.9794 |
| 0.0315 | 2.34 | 700 | 0.0656 | 0.9846 |
| 0.0263 | 2.68 | 800 | 0.0645 | 0.9880 |
| 0.0542 | 3.01 | 900 | 0.0458 | 0.9949 |
| 0.0203 | 3.34 | 1000 | 0.0444 | 0.9931 |
| 0.0189 | 3.68 | 1100 | 0.0432 | 0.9931 |
### Framework versions
- Transformers 4.38.2
- Pytorch 2.1.2
- Datasets 2.1.0
- Tokenizers 0.15.2
|