vit-convnext-mistral-SYDNEY-with-3-captioning
This model is a fine-tuned version of on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.4677
- Accuracy: 67.84
- Bleu-1: 0.8276
- Bleu-2: 0.7647
- Bleu-3: 0.7080
- Bleu-4: 0.6529
- Meteor: 0.7872
- Rouge-l: 0.7522
- Cider: 2.7295
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: 64
- eval_batch_size: 64
- seed: 50
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 0.1
- num_epochs: 128
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Bleu-1 | Bleu-2 | Bleu-3 | Bleu-4 | Meteor | Rouge-l | Cider |
|---|---|---|---|---|---|---|---|---|---|---|---|
| No log | 1.0 | 44 | 0.9229 | 65.29 | 0.5334 | 0.4164 | 0.3336 | 0.2624 | 0.4445 | 0.4631 | 0.8656 |
| No log | 2.0 | 88 | 0.5094 | 62.53 | 0.5480 | 0.4380 | 0.3590 | 0.2933 | 0.4270 | 0.4606 | 0.7265 |
| No log | 3.0 | 132 | 0.4333 | 67.07 | 0.6516 | 0.5537 | 0.4747 | 0.4075 | 0.5585 | 0.5842 | 1.5430 |
| No log | 4.0 | 176 | 0.4120 | 66.53 | 0.7884 | 0.6896 | 0.5943 | 0.5038 | 0.6917 | 0.6649 | 2.2437 |
| No log | 5.0 | 220 | 0.3894 | 66.55 | 0.7856 | 0.6987 | 0.6136 | 0.5245 | 0.7038 | 0.6536 | 2.1245 |
| No log | 6.0 | 264 | 0.3879 | 68.31 | 0.7695 | 0.6675 | 0.5711 | 0.4875 | 0.6892 | 0.6815 | 2.1739 |
| No log | 7.0 | 308 | 0.3940 | 68.41 | 0.7771 | 0.7067 | 0.6375 | 0.5753 | 0.7588 | 0.7246 | 2.4446 |
| No log | 8.0 | 352 | 0.3873 | 65.2 | 0.8230 | 0.7558 | 0.6975 | 0.6429 | 0.7782 | 0.7331 | 2.8550 |
| No log | 9.0 | 396 | 0.4036 | 67.78 | 0.8054 | 0.7255 | 0.6513 | 0.5856 | 0.7640 | 0.7412 | 2.5325 |
| No log | 10.0 | 440 | 0.4154 | 67.94 | 0.8138 | 0.7411 | 0.6724 | 0.6097 | 0.7607 | 0.7260 | 2.7450 |
| No log | 11.0 | 484 | 0.4043 | 69.18 | 0.8103 | 0.7487 | 0.6900 | 0.6370 | 0.7911 | 0.7598 | 2.8851 |
| No log | 12.0 | 528 | 0.4204 | 69.5 | 0.8225 | 0.7419 | 0.6646 | 0.5952 | 0.7517 | 0.7362 | 2.5103 |
| No log | 13.0 | 572 | 0.4245 | 66.61 | 0.8175 | 0.7467 | 0.6643 | 0.5809 | 0.7490 | 0.7098 | 2.4963 |
| No log | 14.0 | 616 | 0.4374 | 67.67 | 0.7974 | 0.7284 | 0.6683 | 0.6113 | 0.7553 | 0.7264 | 2.5928 |
| No log | 15.0 | 660 | 0.4389 | 69.27 | 0.8190 | 0.7337 | 0.6654 | 0.6027 | 0.7817 | 0.7489 | 2.5667 |
| No log | 16.0 | 704 | 0.4502 | 68.05 | 0.7931 | 0.7241 | 0.6659 | 0.6142 | 0.7625 | 0.7302 | 2.7312 |
| No log | 17.0 | 748 | 0.4592 | 67.88 | 0.8162 | 0.7518 | 0.6933 | 0.6382 | 0.7797 | 0.7362 | 2.7089 |
| No log | 18.0 | 792 | 0.4661 | 68.02 | 0.8298 | 0.7607 | 0.6962 | 0.6353 | 0.7886 | 0.7552 | 2.7675 |
| No log | 19.0 | 836 | 0.4602 | 68.15 | 0.8132 | 0.7418 | 0.6774 | 0.6170 | 0.7466 | 0.7108 | 2.4385 |
| No log | 20.0 | 880 | 0.4673 | 69.33 | 0.8171 | 0.7486 | 0.6851 | 0.6271 | 0.7725 | 0.7492 | 2.7374 |
| No log | 21.0 | 924 | 0.4677 | 67.84 | 0.8276 | 0.7647 | 0.7080 | 0.6529 | 0.7872 | 0.7522 | 2.7295 |
Framework versions
- Transformers 5.12.1
- Pytorch 2.12.1+cu130
- Datasets 5.0.0
- Tokenizers 0.22.2