metadata
license: apache-2.0
base_model: bert-base-uncased
tags:
- generated_from_trainer
datasets:
- wikitext
model-index:
- name: run_opt
results: []
run_opt
This model is a fine-tuned version of bert-base-uncased on the wikitext dataset. It achieves the following results on the evaluation set:
- Loss: 0.0107
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: 0.0003
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- gradient_accumulation_steps: 8
- total_train_batch_size: 256
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 100
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 7.6252 | 0.55 | 100 | 6.4113 |
| 4.839 | 1.1 | 200 | 2.0385 |
| 0.9137 | 1.65 | 300 | 0.3108 |
| 0.171 | 2.2 | 400 | 0.0877 |
| 0.0542 | 2.75 | 500 | 0.0396 |
| 0.025 | 3.29 | 600 | 0.0242 |
| 0.0148 | 3.84 | 700 | 0.0180 |
| 0.0098 | 4.39 | 800 | 0.0148 |
| 0.0077 | 4.94 | 900 | 0.0130 |
| 0.006 | 5.49 | 1000 | 0.0121 |
| 0.0053 | 6.04 | 1100 | 0.0115 |
| 0.0045 | 6.59 | 1200 | 0.0112 |
| 0.0042 | 7.14 | 1300 | 0.0110 |
| 0.0039 | 7.69 | 1400 | 0.0109 |
| 0.0038 | 8.24 | 1500 | 0.0108 |
| 0.0037 | 8.79 | 1600 | 0.0107 |
| 0.0037 | 9.33 | 1700 | 0.0107 |
| 0.0036 | 9.88 | 1800 | 0.0107 |
Framework versions
- Transformers 4.33.1
- Pytorch 1.12.1
- Datasets 2.14.6
- Tokenizers 0.13.3