<!-- 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. -->
disaster-tweet-distilbert
This model is a fine-tuned version of distilbert-base-uncased-finetuned-sst-2-english on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.4605
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: 3e-05
- train_batch_size: 64
- eval_batch_size: 64
- seed: 7
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 4
- num_warmup_steps: 10%
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 2.2877 | 0.12 | 12 | 2.3051 |
| 2.1129 | 0.25 | 24 | 2.2778 |
| 2.2514 | 0.38 | 36 | 2.2299 |
| 2.2691 | 0.5 | 48 | 2.1606 |
| 2.1401 | 0.62 | 60 | 2.0706 |
| 2.075 | 0.75 | 72 | 1.9672 |
| 1.8594 | 0.88 | 84 | 1.8498 |
| 1.7927 | 1.0 | 96 | 1.7257 |
| 1.5639 | 1.12 | 108 | 1.6010 |
| 1.6001 | 1.25 | 120 | 1.4670 |
| 1.4207 | 1.38 | 132 | 1.3314 |
| 1.3183 | 1.5 | 144 | 1.1993 |
| 1.0767 | 1.62 | 156 | 1.0798 |
| 0.9672 | 1.75 | 168 | 0.9742 |
| 0.9523 | 1.88 | 180 | 0.8821 |
| 0.813 | 2.0 | 192 | 0.8027 |
| 0.7004 | 2.12 | 204 | 0.7424 |
| 0.7044 | 2.25 | 216 | 0.6904 |
| 0.6218 | 2.38 | 228 | 0.6495 |
| 0.6472 | 2.5 | 240 | 0.6158 |
| 0.5585 | 2.62 | 252 | 0.5896 |
| 0.5613 | 2.75 | 264 | 0.5685 |
| 0.5911 | 2.88 | 276 | 0.5499 |
| 0.5062 | 3.0 | 288 | 0.5357 |
| 0.4806 | 3.12 | 300 | 0.5257 |
| 0.4862 | 3.25 | 312 | 0.5091 |
| 0.4433 | 3.38 | 324 | 0.4997 |
| 0.486 | 3.5 | 336 | 0.4892 |
| 0.4746 | 3.62 | 348 | 0.4802 |
| 0.4317 | 3.75 | 360 | 0.4759 |
| 0.4874 | 3.88 | 372 | 0.4670 |
| 0.4411 | 4.0 | 384 | 0.4605 |
Framework versions
- Transformers 4.28.1
- Pytorch 1.13.0
- Datasets 2.1.0
- Tokenizers 0.13.2
