metadata
tags:
- generated_from_trainer
model-index:
name: Persian-Image-Captioning
Persian-Image-Captioning
This model is a fine-tuned version of Vision Encoder Decoder on coco-flickr-farsi. It achieves the following results on the evaluation set:
- Loss: 2.4300
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: 4e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 2
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 3.4861 | 0.24 | 3500 | 2.9843 |
| 2.8854 | 0.47 | 7000 | 2.7656 |
| 2.701 | 0.71 | 10500 | 2.6337 |
| 2.6279 | 0.95 | 14000 | 2.5680 |
| 2.4782 | 1.19 | 17500 | 2.5179 |
| 2.4321 | 1.42 | 21000 | 2.4838 |
| 2.3876 | 1.66 | 24500 | 2.4513 |
| 2.3854 | 1.9 | 28000 | 2.4300 |
Framework versions
- Transformers 4.12.5
- Pytorch 1.9.1
- Datasets 1.16.1
- Tokenizers 0.10.3