Downloads · 30 days
5
3% of all-time downloads
mbart_sanlishTObangla-normalsplit
thunderboltc/mbart_sanlishTObangla-normalsplit
mbart_sanlishTObangla-normalsplit is a machine learning model from thunderboltc. Use it for the machine learning task on the model card, and read the license before you ship it in a product. It is set up for transformers.
should probably proofread and complete it, then remove this comment. --
All-time downloads
193
Public
Parameters
611M
105 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors4.9 GB · 50%
At a glance
- Library
- transformers
- Model type
- mbart
- Access
- Public
- Created
- Aug 23, 2026
- Updated
- Aug 23, 2026
- SHA
- 189ea078
Base models
Model card
From the Hugging Face model README
<!-- 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. -->
mbart_sanlishTObangla-normalsplit
This model is a fine-tuned version of facebook/mbart-large-50-many-to-many-mmt on the None dataset. It achieves the following results on the evaluation set:
- Loss: 3.4547
- Bleu: 12.5668
- Chrf: 39.6162
- Meteor: 0.3268
- Bertscore: 0.8438
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: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Use OptimizerNames.ADAMW_TORCH 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: 50
- num_epochs: 25
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Bleu | Chrf | Meteor | Bertscore |
|---|---|---|---|---|---|---|---|
| 3.2539 | 1.0 | 170 | 3.1423 | 1.7392 | 19.3777 | 0.1436 | 0.7790 |
| 2.0844 | 2.0 | 340 | 2.6069 | 5.3533 | 26.9762 | 0.2192 | 0.8135 |
| 1.1583 | 3.0 | 510 | 2.7966 | 4.8498 | 30.3510 | 0.2648 | 0.8215 |
| 0.5838 | 4.0 | 680 | 2.9800 | 7.2103 | 32.2728 | 0.2667 | 0.8284 |
| 0.3718 | 5.0 | 850 | 3.1945 | 8.7638 | 33.0685 | 0.2727 | 0.8277 |
| 0.2664 | 6.0 | 1020 | 3.2909 | 6.7010 | 32.7344 | 0.2783 | 0.8275 |
| 0.2057 | 7.0 | 1190 | 3.3111 | 10.9847 | 37.0257 | 0.3015 | 0.8379 |
| 0.1559 | 8.0 | 1360 | 3.3645 | 8.5493 | 34.5928 | 0.2845 | 0.8297 |
| 0.1136 | 9.0 | 1530 | 3.2491 | 10.5297 | 37.7148 | 0.3097 | 0.8394 |
| 0.0892 | 10.0 | 1700 | 3.3594 | 7.3390 | 37.6466 | 0.2966 | 0.8328 |
| 0.055 | 11.0 | 1870 | 3.3290 | 10.6032 | 37.3186 | 0.3051 | 0.8370 |
| 0.0596 | 12.0 | 2040 | 3.3866 | 11.3514 | 38.0836 | 0.2981 | 0.8338 |
| 0.0381 | 13.0 | 2210 | 3.3468 | 10.5465 | 38.0382 | 0.3241 | 0.8401 |
| 0.0289 | 14.0 | 2380 | 3.3628 | 12.0275 | 38.5699 | 0.3161 | 0.8392 |
| 0.0213 | 15.0 | 2550 | 3.3908 | 11.5866 | 39.1137 | 0.3349 | 0.8430 |
| 0.0221 | 16.0 | 2720 | 3.4049 | 9.3194 | 38.4373 | 0.3109 | 0.8368 |
| 0.0206 | 17.0 | 2890 | 3.4186 | 11.3289 | 39.1458 | 0.3226 | 0.8392 |
| 0.015 | 18.0 | 3060 | 3.4406 | 10.9030 | 38.1423 | 0.3098 | 0.8386 |
| 0.0221 | 19.0 | 3230 | 3.4122 | 11.9456 | 39.4578 | 0.3222 | 0.8431 |
| 0.0109 | 20.0 | 3400 | 3.4199 | 12.3585 | 39.7991 | 0.3297 | 0.8447 |
| 0.0147 | 21.0 | 3570 | 3.4121 | 11.8312 | 38.9662 | 0.3233 | 0.8422 |
| 0.0106 | 22.0 | 3740 | 3.4329 | 12.0815 | 38.8514 | 0.3198 | 0.8421 |
| 0.0078 | 23.0 | 3910 | 3.4413 | 12.8363 | 40.1073 | 0.3291 | 0.8448 |
| 0.0086 | 24.0 | 4080 | 3.4546 | 12.5969 | 39.6882 | 0.3260 | 0.8437 |
| 0.006 | 25.0 | 4250 | 3.4547 | 12.5668 | 39.6162 | 0.3268 | 0.8438 |
Framework versions
- Transformers 4.46.3
- Pytorch 2.11.0+cu128
- Datasets 4.0.0
- Tokenizers 0.20.3
- Library
- transformers
- Type
- mbart
- Created
- Aug 23, 2026
- Updated
- Aug 23, 2026