small-mlm-imdb
This model is a fine-tuned version of google/bert_uncased_L-4_H-512_A-8 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 2.3673
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 |
|---|---|---|---|
| 2.7542 | 0.16 | 500 | 2.5445 |
| 2.6734 | 0.32 | 1000 | 2.5191 |
| 2.6552 | 0.48 | 1500 | 2.4976 |
| 2.6481 | 0.64 | 2000 | 2.4866 |
| 2.6291 | 0.8 | 2500 | 2.4599 |
| 2.6134 | 0.96 | 3000 | 2.4585 |
| 2.5627 | 1.12 | 3500 | 2.4476 |
| 2.5564 | 1.28 | 4000 | 2.4340 |
| 2.5493 | 1.44 | 4500 | 2.4354 |
| 2.5435 | 1.6 | 5000 | 2.4307 |
| 2.5352 | 1.76 | 5500 | 2.4224 |
| 2.5445 | 1.92 | 6000 | 2.4167 |
| 2.5191 | 2.08 | 6500 | 2.4175 |
| 2.5143 | 2.24 | 7000 | 2.4149 |
| 2.5059 | 2.4 | 7500 | 2.4117 |
| 2.4865 | 2.56 | 8000 | 2.4063 |
| 2.5113 | 2.72 | 8500 | 2.3976 |
| 2.5115 | 2.88 | 9000 | 2.3959 |
| 2.485 | 3.04 | 9500 | 2.3917 |
| 2.4652 | 3.2 | 10000 | 2.3908 |
| 2.4569 | 3.36 | 10500 | 2.3877 |
| 2.4706 | 3.52 | 11000 | 2.3836 |
| 2.4375 | 3.68 | 11500 | 2.3870 |
| 2.4556 | 3.84 | 12000 | 2.3819 |
| 2.4487 | 4.0 | 12500 | 2.3842 |
| 2.4233 | 4.16 | 13000 | 2.3731 |
| 2.4238 | 4.32 | 13500 | 2.3801 |
| 2.4051 | 4.48 | 14000 | 2.3809 |
| 2.432 | 4.64 | 14500 | 2.3641 |
| 2.428 | 4.8 | 15000 | 2.3686 |
| 2.4248 | 4.96 | 15500 | 2.3741 |
| 2.4109 | 5.12 | 16000 | 2.3673 |
Framework versions
- Transformers 4.25.1
- Pytorch 1.12.1
- Datasets 2.7.1
- Tokenizers 0.13.2