File size: 2,244 Bytes
0ce906d |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 |
---
license: apache-2.0
base_model: bert-base-uncased
tags:
- generated_from_trainer
datasets:
- wikitext
model-index:
- name: run_2
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_2
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: 0.9502
## 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: 0.0005
- 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: 5
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 8.4559 | 0.27 | 50 | 7.1236 |
| 6.8523 | 0.55 | 100 | 6.6676 |
| 6.6103 | 0.82 | 150 | 6.5582 |
| 6.2417 | 1.1 | 200 | 5.6994 |
| 4.9738 | 1.37 | 250 | 4.3440 |
| 4.1043 | 1.65 | 300 | 3.7804 |
| 3.4265 | 1.92 | 350 | 3.0136 |
| 2.7667 | 2.2 | 400 | 2.5318 |
| 2.3538 | 2.47 | 450 | 2.0903 |
| 1.9591 | 2.75 | 500 | 1.7367 |
| 1.6652 | 3.02 | 550 | 1.5016 |
| 1.4318 | 3.29 | 600 | 1.3162 |
| 1.275 | 3.57 | 650 | 1.1657 |
| 1.1553 | 3.84 | 700 | 1.0655 |
| 1.0629 | 4.12 | 750 | 1.0029 |
| 1.0029 | 4.39 | 800 | 0.9683 |
| 0.9881 | 4.67 | 850 | 0.9536 |
| 0.9779 | 4.94 | 900 | 0.9502 |
### Framework versions
- Transformers 4.33.1
- Pytorch 1.12.1
- Datasets 2.14.6
- Tokenizers 0.13.3
|