Improving Black-box Robustness with In-Context Rewriting
Collection
24 items
•
Updated
•
1
ag-news-twitter-9600-bert-base-uncased
This model is a fine-tuned version of bert-base-uncased on the ag_news dataset. It achieves the following results on the evaluation set:
- F1: 0.9162
- Acc: 0.9162
- Loss: 0.6033
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: 16
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 20
Training results
| Training Loss | Epoch | Step | F1 | Acc | Validation Loss |
|---|---|---|---|---|---|
| 0.8065 | 1.0 | 600 | 0.9060 | 0.9059 | 0.3013 |
| 0.2872 | 2.0 | 1200 | 0.9171 | 0.9170 | 0.2598 |
| 0.2156 | 3.0 | 1800 | 0.9178 | 0.9184 | 0.3117 |
| 0.1486 | 4.0 | 2400 | 0.9200 | 0.9197 | 0.3631 |
| 0.0683 | 5.0 | 3000 | 0.9202 | 0.9201 | 0.3782 |
| 0.045 | 6.0 | 3600 | 0.9186 | 0.9188 | 0.4846 |
| 0.0218 | 7.0 | 4200 | 0.9155 | 0.9155 | 0.5898 |
| 0.0245 | 8.0 | 4800 | 0.9162 | 0.9162 | 0.6033 |
Framework versions
- Transformers 4.35.0
- Pytorch 2.1.0+cu121
- Datasets 2.14.6
- Tokenizers 0.14.1
- Downloads last month
- 4
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.
Model tree for Kyle1668/ag-news-9600-bert-base-uncased
Base model
google-bert/bert-base-uncased