<!-- 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. -->
small-mlm-wikitext-custom-tokenizer
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: 5.9013
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 |
---|---|---|---|
7.443 | 0.4 | 500 | 7.1092 |
6.9585 | 0.8 | 1000 | 6.9474 |
6.8416 | 1.2 | 1500 | nan |
6.7094 | 1.6 | 2000 | 6.6852 |
6.6686 | 2.0 | 2500 | 6.6998 |
6.5422 | 2.4 | 3000 | 6.6184 |
6.5449 | 2.8 | 3500 | 6.6460 |
6.4334 | 3.2 | 4000 | nan |
6.3839 | 3.6 | 4500 | 6.4225 |
6.3278 | 4.0 | 5000 | 6.4024 |
6.264 | 4.4 | 5500 | 6.4076 |
6.1794 | 4.8 | 6000 | 6.3197 |
6.1637 | 5.2 | 6500 | 6.3163 |
6.091 | 5.6 | 7000 | 6.2943 |
6.0679 | 6.0 | 7500 | nan |
6.0375 | 6.4 | 8000 | 6.1639 |
5.9742 | 6.8 | 8500 | 6.1480 |
5.9223 | 7.2 | 9000 | 6.1752 |
5.9037 | 7.6 | 9500 | 6.0949 |
5.8416 | 8.0 | 10000 | 6.0662 |
5.8246 | 8.4 | 10500 | 6.0426 |
5.7151 | 8.8 | 11000 | 5.9719 |
5.6765 | 9.2 | 11500 | 5.9907 |
5.7426 | 9.6 | 12000 | 5.9008 |
5.6745 | 10.0 | 12500 | 5.9193 |
5.6063 | 10.4 | 13000 | 5.9051 |
5.5536 | 10.8 | 13500 | 5.9013 |
Framework versions
- Transformers 4.26.0.dev0
- Pytorch 1.13.0+cu116
- Datasets 2.8.1.dev0
- Tokenizers 0.13.2