Downloads · 30 days
5
1% of all-time downloads
AgitationTextV1
Dinithi/AgitationTextV1
AgitationTextV1 is a text classification model from Dinithi. Use it when you need a label for a piece of text. It is set up for transformers. The card lists the license as apache-2.0.
should probably proofread and complete it, then remove this comment. --
All-time downloads
397
Public
Repo size
536 MB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.bin268 MB · 100%
At a glance
- Task
- Text Classification
- Library
- transformers
- License
- apache-2.0
- Model type
- distilbert
- Access
- Public
- Created
- Sep 3, 2022
- Updated
- Sep 3, 2022
- SHA
- 2dc92b88
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. -->
AgitationTextV1
This model is a fine-tuned version of distilbert-base-uncased on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.5634
- Accuracy: 0.84
- Precision: 0.9054
- Recall: 0.8816
- F1: 0.8933
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: 1e-05
- train_batch_size: 16
- eval_batch_size: 16
- 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 | Accuracy | Precision | Recall | F1 |
|---|---|---|---|---|---|---|---|
| 0.558 | 1.0 | 50 | 0.5387 | 0.75 | 0.9048 | 0.75 | 0.8201 |
| 0.4431 | 2.0 | 100 | 0.4989 | 0.75 | 0.9474 | 0.7105 | 0.8120 |
| 0.3334 | 3.0 | 150 | 0.4703 | 0.79 | 0.9661 | 0.75 | 0.8444 |
| 0.2601 | 4.0 | 200 | 0.4391 | 0.85 | 0.9420 | 0.8553 | 0.8966 |
| 0.2081 | 5.0 | 250 | 0.4560 | 0.84 | 0.9167 | 0.8684 | 0.8919 |
| 0.1655 | 6.0 | 300 | 0.5424 | 0.84 | 0.8947 | 0.8947 | 0.8947 |
| 0.1427 | 7.0 | 350 | 0.4977 | 0.84 | 0.9286 | 0.8553 | 0.8904 |
| 0.1237 | 8.0 | 400 | 0.5642 | 0.85 | 0.9067 | 0.8947 | 0.9007 |
| 0.1122 | 9.0 | 450 | 0.6052 | 0.85 | 0.9067 | 0.8947 | 0.9007 |
| 0.0962 | 10.0 | 500 | 0.5634 | 0.84 | 0.9054 | 0.8816 | 0.8933 |
Framework versions
- Transformers 4.21.2
- Pytorch 1.12.1+cu113
- Datasets 2.4.0
- Tokenizers 0.12.1
- Task
- Text Classification
- Library
- transformers
- Type
- distilbert
- License
- apache-2.0
- Created
- Sep 3, 2022
- Updated
- Sep 3, 2022