Downloads · 30 days
19
45% of all-time downloads
BERT_Dec-6_tokenized
Mozart-coder/BERT_Dec-6_tokenized
BERT_Dec-6_tokenized is a fill-mask model from Mozart-coder. 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
42
Public
Parameters
89.2M
1.1 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.bin357 MB · 50%
Parameter types
How the weights are stored.
F3289.2M · 100%
Try a prompt
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. -->
BERT_Dec-6_tokenized
This model is a fine-tuned version of armheb/DNA_bert_6 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.0372
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: 64
- eval_batch_size: 64
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.0625 | 1.0 | 273 | 0.0376 |
| 0.039 | 2.0 | 546 | 0.0375 |
| 0.0385 | 3.0 | 819 | 0.0358 |
| 0.0375 | 4.0 | 1092 | 0.0380 |
| 0.0374 | 5.0 | 1365 | 0.0387 |
| 0.0358 | 6.0 | 1638 | 0.0378 |
| 0.0363 | 7.0 | 1911 | 0.0381 |
| 0.0373 | 8.0 | 2184 | 0.0377 |
| 0.0362 | 9.0 | 2457 | 0.0373 |
| 0.037 | 10.0 | 2730 | 0.0380 |
Framework versions
- Transformers 4.23.1
- Pytorch 1.12.1+cu113
- Datasets 2.6.1
- Tokenizers 0.13.1
- Task
- Fill-Mask
- Library
- transformers
- Type
- bert
- Created
- Oct 25, 2022
- Updated
- Jul 8, 2023