Downloads · 30 days
19
6% of all-time downloads
reward-deberta-v3
Chat-Error/reward-deberta-v3
reward-deberta-v3 is a text classification model from Chat-Error. 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
336
Public
Repo size
1.5 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.bin738 MB · 98%
At a glance
- Task
- Text Classification
- Library
- transformers
- License
- mit
- Model type
- deberta-v2
- Access
- Public
- Created
- Oct 14, 2023
- Updated
- Oct 14, 2023
- SHA
- f08fcbaf
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. -->
out_2
This model is a fine-tuned version of microsoft/deberta-v3-base on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.6774
- F1: 0.7444
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: 6e-06
- train_batch_size: 3
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 16
- total_train_batch_size: 48
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- num_epochs: 5.0
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Accuracy | Validation Loss |
|---|---|---|---|---|
| 0.6448 | 0.21 | 500 | 0.6347 | 0.6498 |
| 0.6401 | 0.41 | 1000 | 0.6442 | 0.6312 |
| 0.6557 | 0.62 | 1500 | 0.6582 | 0.6314 |
| 0.5819 | 0.83 | 2000 | 0.6588 | 0.6320 |
| 0.6086 | 1.04 | 2500 | 0.6563 | 0.6343 |
| 0.6011 | 1.24 | 3000 | 0.6557 | 0.6165 |
| 0.5616 | 1.45 | 3500 | 0.6461 | 0.6376 |
| 0.5885 | 1.66 | 4000 | 0.6468 | 0.6304 |
| 0.6198 | 1.87 | 4500 | 0.6423 | 0.6448 |
| 0.5838 | 2.07 | 5000 | 0.6665 | 0.6320 |
| 0.5564 | 2.28 | 5500 | 0.6684 | 0.6428 |
| 0.5726 | 2.49 | 6000 | 0.6703 | 0.6401 |
| 0.5491 | 2.7 | 6500 | 0.6684 | 0.6455 |
| 0.5303 | 2.9 | 7000 | 0.6703 | 0.6339 |
| 0.497 | 3.11 | 7500 | 0.6607 | 0.6541 |
| 0.5041 | 3.32 | 8000 | 0.6760 | 0.6653 |
| 0.4978 | 3.53 | 8500 | 0.6696 | 0.6627 |
| 0.5272 | 3.73 | 9000 | 0.6677 | 0.6684 |
| 0.5487 | 3.94 | 9500 | 0.6760 | 0.6593 |
| 0.4998 | 4.15 | 10000 | 0.6747 | 0.6738 |
| 0.4626 | 4.36 | 10500 | 0.6753 | 0.6781 |
| 0.5202 | 4.56 | 11000 | 0.6722 | 0.6763 |
| 0.4623 | 4.77 | 11500 | 0.6728 | 0.6778 |
| 0.4383 | 4.98 | 12000 | 0.6741 | 0.6775 |
Framework versions
- Transformers 4.35.0.dev0
- Pytorch 2.0.1+cu117
- Datasets 2.14.4
- Tokenizers 0.14.1
- Task
- Text Classification
- Library
- transformers
- Type
- deberta-v2
- License
- mit
- Created
- Oct 14, 2023
- Updated
- Oct 14, 2023