Downloads · 30 days
6
2% of all-time downloads
deberta_sentcls
kkmkorea/deberta_sentcls
deberta_sentcls is a text classification model from kkmkorea. Use it when you need a label for a piece of text. 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
399
Public
Repo size
2.2 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.bin738 MB · 100%
At a glance
- Task
- Text Classification
- Library
- transformers
- License
- mit
- Model type
- deberta-v2
- Access
- Public
- Created
- Sep 12, 2023
- Updated
- Nov 17, 2023
- SHA
- 7f3ec04a
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. -->
deberta_sentcls
This model is a fine-tuned version of kisti/korscideberta on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.4937
- Accuracy: 0.8409
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: 1.5e-05
- train_batch_size: 32
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 5
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 0.9366 | 1.03 | 500 | 0.6700 | 0.7659 |
| 0.472 | 2.05 | 1000 | 0.5110 | 0.8261 |
| 0.3567 | 3.08 | 1500 | 0.4979 | 0.8371 |
| 0.2626 | 4.11 | 2000 | 0.4937 | 0.8409 |
Framework versions
- Transformers 4.12.5
- Pytorch 1.10.1+cu102
- Datasets 2.14.6
- Tokenizers 0.10.3
- Task
- Text Classification
- Library
- transformers
- Type
- deberta-v2
- License
- mit
- Created
- Sep 12, 2023
- Updated
- Nov 17, 2023