|
--- |
|
license: apache-2.0 |
|
base_model: bert-base-uncased |
|
tags: |
|
- generated_from_trainer |
|
datasets: |
|
- wikitext |
|
model-index: |
|
- name: run_opt |
|
results: [] |
|
--- |
|
|
|
<!-- 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. --> |
|
|
|
# run_opt |
|
|
|
This model is a fine-tuned version of [bert-base-uncased](https://huggingface.co/bert-base-uncased) on the wikitext dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.2727 |
|
|
|
## 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: 0.0001 |
|
- train_batch_size: 64 |
|
- eval_batch_size: 64 |
|
- seed: 42 |
|
- gradient_accumulation_steps: 8 |
|
- total_train_batch_size: 512 |
|
- 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 | |
|
|:-------------:|:-----:|:----:|:---------------:| |
|
| 9.1471 | 0.55 | 50 | 7.9703 | |
|
| 7.1714 | 1.1 | 100 | 6.6558 | |
|
| 6.4707 | 1.65 | 150 | 6.2924 | |
|
| 6.072 | 2.19 | 200 | 5.8048 | |
|
| 5.1389 | 2.74 | 250 | 3.8826 | |
|
| 3.1897 | 3.29 | 300 | 2.3133 | |
|
| 1.9697 | 3.84 | 350 | 1.4230 | |
|
| 1.2783 | 4.39 | 400 | 0.9488 | |
|
| 0.8952 | 4.94 | 450 | 0.6810 | |
|
| 0.6593 | 5.49 | 500 | 0.5228 | |
|
| 0.5278 | 6.04 | 550 | 0.4249 | |
|
| 0.4339 | 6.58 | 600 | 0.3630 | |
|
| 0.3809 | 7.13 | 650 | 0.3237 | |
|
| 0.3443 | 7.68 | 700 | 0.2991 | |
|
| 0.3212 | 8.23 | 750 | 0.2843 | |
|
| 0.3094 | 8.78 | 800 | 0.2765 | |
|
| 0.3033 | 9.33 | 850 | 0.2734 | |
|
| 0.3 | 9.88 | 900 | 0.2727 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.33.1 |
|
- Pytorch 1.12.1 |
|
- Datasets 2.14.6 |
|
- Tokenizers 0.13.3 |
|
|
