Downloads · 30 days
4
13% of all-time downloads
GQA_BERT
farid1088/GQA_BERT
GQA_BERT is a question answering model from farid1088. Use it when the input is a question plus a passage. It is set up for transformers.
should probably proofread and complete it, then remove this comment. --
All-time downloads
32
Public
Parameters
108M
5.2 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors434 MB · 100%
At a glance
- Task
- Question Answering
- Library
- transformers
- Model type
- bert
- Access
- Public
- Created
- Mar 2, 2024
- Updated
- Mar 2, 2024
- SHA
- c264706f
Try a prompt
Datasets
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. -->
GQA_BERT
This model was trained from scratch on the germanquad dataset. It achieves the following results on the evaluation set:
- Loss: 6.6310
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: 40
- eval_batch_size: 40
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 20
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| No log | 1.0 | 288 | 4.9619 |
| 4.5459 | 2.0 | 576 | 4.2345 |
| 4.5459 | 3.0 | 864 | 4.0881 |
| 3.6033 | 4.0 | 1152 | 4.0633 |
| 3.6033 | 5.0 | 1440 | 4.4137 |
| 2.9345 | 6.0 | 1728 | 4.4540 |
| 2.3507 | 7.0 | 2016 | 4.2502 |
| 2.3507 | 8.0 | 2304 | 4.8184 |
| 1.8315 | 9.0 | 2592 | 4.8832 |
| 1.8315 | 10.0 | 2880 | 4.9700 |
| 1.4566 | 11.0 | 3168 | 5.4482 |
| 1.4566 | 12.0 | 3456 | 5.6618 |
| 1.1586 | 13.0 | 3744 | 5.7003 |
| 0.9414 | 14.0 | 4032 | 6.2014 |
| 0.9414 | 15.0 | 4320 | 6.3395 |
| 0.7702 | 16.0 | 4608 | 6.2150 |
| 0.7702 | 17.0 | 4896 | 6.4060 |
| 0.6714 | 18.0 | 5184 | 6.4733 |
| 0.6714 | 19.0 | 5472 | 6.5721 |
| 0.5959 | 20.0 | 5760 | 6.6310 |
Framework versions
- Transformers 4.36.2
- Pytorch 2.1.2+cu121
- Datasets 2.14.7
- Tokenizers 0.15.0
- Task
- Question Answering
- Library
- transformers
- Type
- bert
- Created
- Mar 2, 2024
- Updated
- Mar 2, 2024