Downloads · 30 days
8
13% of all-time downloads
mmbert-small-tamil-pos-tagging
saiyiram/mmbert-small-tamil-pos-tagging
mmbert-small-tamil-pos-tagging is a token classification model from saiyiram. Use it when you need labels on individual words, such as names. It is set up for transformers. The card lists the license as mit.
should probably proofread and complete it, then remove this comment. --
All-time downloads
63
Public
Parameters
141M
1.7 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors563 MB · 94%
At a glance
- Task
- Token Classification
- Library
- transformers
- License
- mit
- Model type
- modernbert
- Access
- Public
- Created
- Sep 6, 2026
- Updated
- Sep 6, 2026
- SHA
- 281a837c
Try a prompt
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. -->
mmbert-small-tamil-pos-tagging
This model is a fine-tuned version of jhu-clsp/mmBERT-small on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.5282
- Accuracy: 0.8504
- F1 Macro: 0.8048
- F1 Weighted: 0.8546
- Precision Macro: 0.8371
- Recall Macro: 0.8181
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: 2e-05
- train_batch_size: 8
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- 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: 0.1
- num_epochs: 5
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 Macro | F1 Weighted | Precision Macro | Recall Macro |
|---|---|---|---|---|---|---|---|---|
| 10.6821 | 1.0 | 25 | 1.6184 | 0.5091 | 0.2767 | 0.4661 | 0.3090 | 0.2915 |
| 1.8548 | 2.0 | 50 | 0.6751 | 0.8100 | 0.7240 | 0.8144 | 0.7741 | 0.7266 |
| 0.7878 | 3.0 | 75 | 0.5300 | 0.8266 | 0.7610 | 0.8303 | 0.8156 | 0.7647 |
| 0.4396 | 4.0 | 100 | 0.4915 | 0.8575 | 0.8044 | 0.8606 | 0.8349 | 0.8145 |
| 0.2771 | 5.0 | 125 | 0.5282 | 0.8504 | 0.8048 | 0.8546 | 0.8371 | 0.8181 |
Framework versions
- Transformers 5.0.0
- Pytorch 2.10.0+cu128
- Datasets 5.0.0
- Tokenizers 0.22.2
- Task
- Token Classification
- Library
- transformers
- Type
- modernbert
- License
- mit
- Created
- Sep 6, 2026
- Updated
- Sep 6, 2026