<!-- 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-1
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: 0.5571
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: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 4
- num_warmup_steps: 20%
Training results
Training Loss | Epoch | Step | Validation Loss |
---|---|---|---|
2.2888 | 0.12 | 12 | 2.3097 |
2.122 | 0.25 | 24 | 2.2960 |
2.2792 | 0.38 | 36 | 2.2720 |
2.3256 | 0.5 | 48 | 2.2368 |
2.2339 | 0.62 | 60 | 2.1904 |
2.2148 | 0.75 | 72 | 2.1360 |
2.0433 | 0.88 | 84 | 2.0723 |
2.0242 | 1.0 | 96 | 2.0028 |
1.8316 | 1.12 | 108 | 1.9302 |
1.9608 | 1.25 | 120 | 1.8508 |
1.8278 | 1.38 | 132 | 1.7635 |
1.7828 | 1.5 | 144 | 1.6711 |
1.5522 | 1.62 | 156 | 1.5747 |
1.474 | 1.75 | 168 | 1.4774 |
1.4762 | 1.88 | 180 | 1.3790 |
1.3439 | 2.0 | 192 | 1.2820 |
1.1465 | 2.12 | 204 | 1.1896 |
1.1755 | 2.25 | 216 | 1.1024 |
1.0085 | 2.38 | 228 | 1.0212 |
1.0492 | 2.5 | 240 | 0.9492 |
0.8642 | 2.62 | 252 | 0.8858 |
0.8554 | 2.75 | 264 | 0.8291 |
0.8534 | 2.88 | 276 | 0.7792 |
0.7013 | 3.0 | 288 | 0.7364 |
0.6414 | 3.12 | 300 | 0.7023 |
0.681 | 3.25 | 312 | 0.6707 |
0.6045 | 3.38 | 324 | 0.6441 |
0.6374 | 3.5 | 336 | 0.6193 |
0.6192 | 3.62 | 348 | 0.5988 |
0.5478 | 3.75 | 360 | 0.5831 |
0.5891 | 3.88 | 372 | 0.5693 |
0.5411 | 4.0 | 384 | 0.5571 |
Framework versions
- Transformers 4.28.1
- Pytorch 1.13.0
- Datasets 2.1.0
- Tokenizers 0.13.2