medium-mlm-tweet-target-imdb
This model is a fine-tuned version of muhtasham/medium-mlm-tweet on the imdb dataset.
It achieves the following results on the evaluation set:
- Loss: 0.3691
- Accuracy: 0.9363
- F1: 0.9671
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: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: constant
- num_epochs: 200
Training results
| Training Loss |
Epoch |
Step |
Validation Loss |
Accuracy |
F1 |
| 0.3135 |
0.64 |
500 |
0.2323 |
0.9056 |
0.9505 |
| 0.2094 |
1.28 |
1000 |
0.2166 |
0.9187 |
0.9576 |
| 0.1622 |
1.92 |
1500 |
0.2011 |
0.9206 |
0.9587 |
| 0.112 |
2.56 |
2000 |
0.3647 |
0.9032 |
0.9491 |
| 0.093 |
3.2 |
2500 |
0.5445 |
0.8788 |
0.9355 |
| 0.0692 |
3.84 |
3000 |
0.2071 |
0.9452 |
0.9718 |
| 0.0545 |
4.48 |
3500 |
0.2308 |
0.9548 |
0.9769 |
| 0.0482 |
5.12 |
4000 |
0.3297 |
0.9373 |
0.9676 |
| 0.0464 |
5.75 |
4500 |
0.3698 |
0.926 |
0.9616 |
| 0.0308 |
6.39 |
5000 |
0.3691 |
0.9363 |
0.9671 |
Framework versions
- Transformers 4.25.1
- Pytorch 1.12.1
- Datasets 2.7.1
- Tokenizers 0.13.2
