Downloads · 30 days
3
3% of all-time downloads
CeLLaTe-bioformer-16L-tapt-adapted
OTAR3088/CeLLaTe-bioformer-16L-tapt-adapted
CeLLaTe-bioformer-16L-tapt-adapted is a fill-mask model from OTAR3088. Use it when you need the model to fill a missing word. It is set up for transformers.
should probably proofread and complete it, then remove this comment. --
All-time downloads
96
Public
Parameters
42.8M
1 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors171 MB · 99%
At a glance
- Task
- Fill-Mask
- Library
- transformers
- Model type
- bert
- Access
- Public
- Created
- Jul 15, 2026
- Updated
- Jul 16, 2026
- SHA
- 9dd46f8a
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. -->
CeLLaTe-bioformer-16L-tapt-adapted
This model is a fine-tuned version of Mardiyyah/bioformer-16l-tokenizer-adapted on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 2.6328
- Accuracy: 0.6251
- Perplexity: 13.9125
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: 32
- eval_batch_size: 32
- seed: 3407
- gradient_accumulation_steps: 2
- total_train_batch_size: 64
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-06 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.06
- num_epochs: 50
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Perplexity |
|---|---|---|---|---|---|
| No log | 1.0 | 15 | 9.5765 | 0.0008 | 14421.3859 |
| No log | 2.0 | 30 | 7.7367 | 0.0005 | 2290.8902 |
| No log | 3.0 | 45 | 6.8667 | 0.1422 | 959.7625 |
| No log | 4.0 | 60 | 5.8874 | 0.2321 | 360.4623 |
| No log | 5.0 | 75 | 5.2345 | 0.2851 | 187.6406 |
| No log | 6.0 | 90 | 4.7875 | 0.3268 | 119.9979 |
| 6.7557 | 7.0 | 105 | 4.4625 | 0.3565 | 86.6999 |
| 6.7557 | 8.0 | 120 | 4.1913 | 0.3941 | 66.1101 |
| 6.7557 | 9.0 | 135 | 4.0196 | 0.4181 | 55.6783 |
| 6.7557 | 10.0 | 150 | 3.8186 | 0.4525 | 45.5417 |
| 6.7557 | 11.0 | 165 | 3.6814 | 0.4725 | 39.7034 |
| 6.7557 | 12.0 | 180 | 3.5284 | 0.4950 | 34.0703 |
| 6.7557 | 13.0 | 195 | 3.4597 | 0.5131 | 31.8077 |
| 3.7563 | 14.0 | 210 | 3.3813 | 0.5219 | 29.4080 |
| 3.7563 | 15.0 | 225 | 3.2784 | 0.5396 | 26.5346 |
| 3.7563 | 16.0 | 240 | 3.1944 | 0.5526 | 24.3961 |
| 3.7563 | 17.0 | 255 | 3.1044 | 0.5656 | 22.2966 |
| 3.7563 | 18.0 | 270 | 3.1087 | 0.5623 | 22.3914 |
| 3.7563 | 19.0 | 285 | 2.9939 | 0.5856 | 19.9634 |
| 3.0567 | 20.0 | 300 | 3.0040 | 0.5806 | 20.1659 |
| 3.0567 | 21.0 | 315 | 2.9078 | 0.5942 | 18.3163 |
| 3.0567 | 22.0 | 330 | 2.9290 | 0.5936 | 18.7092 |
| 3.0567 | 23.0 | 345 | 2.8820 | 0.6016 | 17.8503 |
| 3.0567 | 24.0 | 360 | 2.8125 | 0.6057 | 16.6508 |
| 3.0567 | 25.0 | 375 | 2.8499 | 0.6026 | 17.2855 |
| 3.0567 | 26.0 | 390 | 2.7702 | 0.6179 | 15.9623 |
| 2.7526 | 27.0 | 405 | 2.7224 | 0.6209 | 15.2171 |
| 2.7526 | 28.0 | 420 | 2.6848 | 0.6209 | 14.6559 |
| 2.7526 | 29.0 | 435 | 2.7144 | 0.6192 | 15.0951 |
| 2.7526 | 30.0 | 450 | 2.6704 | 0.6236 | 14.4457 |
| 2.7526 | 31.0 | 465 | 2.6776 | 0.6191 | 14.5504 |
| 2.7526 | 32.0 | 480 | 2.6090 | 0.6336 | 13.5857 |
| 2.7526 | 33.0 | 495 | 2.6001 | 0.6337 | 13.4658 |
| 2.5645 | 34.0 | 510 | 2.6302 | 0.6269 | 13.8762 |
| 2.5645 | 35.0 | 525 | 2.6430 | 0.6254 | 14.0549 |
| 2.5645 | 36.0 | 540 | 2.6005 | 0.6343 | 13.4706 |
Framework versions
- Transformers 4.48.2
- Pytorch 2.4.1+cu121
- Datasets 3.0.2
- Tokenizers 0.21.0
- Task
- Fill-Mask
- Library
- transformers
- Type
- bert
- Created
- Jul 15, 2026
- Updated
- Jul 16, 2026