metadata
license: mit
tags:
- generated_from_trainer
datasets:
- imagefolder
model-index:
- name: git-base-pokemon
results: []
git-base-pokemon
This model is a fine-tuned version of microsoft/git-base on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 0.0345
- Wer Score: 2.4097
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: 32
- eval_batch_size: 32
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 50
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer Score |
|---|---|---|---|---|
| 7.3695 | 4.17 | 50 | 4.5700 | 21.4160 |
| 2.3984 | 8.33 | 100 | 0.4696 | 10.9249 |
| 0.1439 | 12.5 | 150 | 0.0305 | 1.1692 |
| 0.02 | 16.67 | 200 | 0.0263 | 1.5229 |
| 0.0084 | 20.83 | 250 | 0.0295 | 2.6539 |
| 0.003 | 25.0 | 300 | 0.0324 | 3.2125 |
| 0.0018 | 29.17 | 350 | 0.0329 | 2.6628 |
| 0.0014 | 33.33 | 400 | 0.0336 | 2.5407 |
| 0.0013 | 37.5 | 450 | 0.0338 | 2.4008 |
| 0.0011 | 41.67 | 500 | 0.0344 | 2.5115 |
| 0.0011 | 45.83 | 550 | 0.0344 | 2.3766 |
| 0.0011 | 50.0 | 600 | 0.0345 | 2.4097 |
Framework versions
- Transformers 4.27.4
- Pytorch 1.13.1+cu116
- Datasets 2.11.0
- Tokenizers 0.13.2