|
--- |
|
license: apache-2.0 |
|
base_model: facebook/deit-tiny-distilled-patch16-224 |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- accuracy |
|
model-index: |
|
- name: quickdraw-DeiT-tiny-c |
|
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. --> |
|
|
|
# quickdraw-DeiT-tiny-c |
|
|
|
This model is a fine-tuned version of [facebook/deit-tiny-distilled-patch16-224](https://huggingface.co/facebook/deit-tiny-distilled-patch16-224) on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.8784 |
|
- Accuracy: 0.7849 |
|
|
|
## 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: 0.0008 |
|
- train_batch_size: 512 |
|
- eval_batch_size: 512 |
|
- seed: 42 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: cosine |
|
- lr_scheduler_warmup_steps: 10000 |
|
- num_epochs: 8 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Accuracy | |
|
|:-------------:|:------:|:-----:|:---------------:|:--------:| |
|
| 1.2697 | 0.5688 | 5000 | 1.2368 | 0.6883 | |
|
| 1.1262 | 1.1377 | 10000 | 1.1299 | 0.7127 | |
|
| 1.0215 | 1.7065 | 15000 | 1.0110 | 0.7403 | |
|
| 0.939 | 2.2753 | 20000 | 0.9628 | 0.7521 | |
|
| 0.9129 | 2.8441 | 25000 | 0.9281 | 0.7606 | |
|
| 0.8507 | 3.4130 | 30000 | 0.8973 | 0.7687 | |
|
| 0.8354 | 3.9818 | 35000 | 0.8696 | 0.7752 | |
|
| 0.7773 | 4.5506 | 40000 | 0.8575 | 0.7791 | |
|
| 0.7011 | 5.1195 | 45000 | 0.8497 | 0.7829 | |
|
| 0.6989 | 5.6883 | 50000 | 0.8350 | 0.7860 | |
|
| 0.624 | 6.2571 | 55000 | 0.8524 | 0.7857 | |
|
| 0.6245 | 6.8259 | 60000 | 0.8499 | 0.7874 | |
|
| 0.565 | 7.3948 | 65000 | 0.8795 | 0.7849 | |
|
| 0.5663 | 7.9636 | 70000 | 0.8784 | 0.7849 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.41.0 |
|
- Pytorch 2.2.1 |
|
- Datasets 2.19.1 |
|
- Tokenizers 0.19.1 |
|
|
