metadata
license: mit
tags:
- generated_from_trainer
model-index:
- name: gpt2-kl_1_04-hs_cn_decay
results: []
gpt2-kl_1_04-hs_cn_decay
This model is a fine-tuned version of gpt2-medium on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.5463
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: 5e-05
- train_batch_size: 8
- eval_batch_size: 4
- seed: 21
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 3.0
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 73.8246 | 0.02 | 10 | 69.5789 |
| 46.3793 | 0.04 | 20 | 33.0123 |
| 13.8067 | 0.06 | 30 | 10.6405 |
| 7.1214 | 0.08 | 40 | 4.2678 |
| 3.1606 | 0.1 | 50 | 2.0640 |
| 1.6919 | 0.12 | 60 | 1.0966 |
| 1.2735 | 0.14 | 70 | 0.9815 |
| 1.181 | 0.16 | 80 | 0.7299 |
| 1.0336 | 0.18 | 90 | 0.6517 |
| 0.9677 | 0.2 | 100 | 0.6244 |
| 0.8425 | 0.22 | 110 | 0.6201 |
| 0.8964 | 0.24 | 120 | 0.6004 |
| 0.8844 | 0.26 | 130 | 0.5775 |
| 0.8976 | 0.28 | 140 | 0.5743 |
| 0.9151 | 0.3 | 150 | 0.5743 |
| 0.8358 | 0.32 | 160 | 0.5741 |
| 0.7451 | 0.34 | 170 | 0.5741 |
| 0.9072 | 0.36 | 180 | 0.5639 |
| 0.9161 | 0.38 | 190 | 0.5593 |
| 0.7969 | 0.4 | 200 | 0.5564 |
| 0.9549 | 0.42 | 210 | 0.5581 |
| 0.9 | 0.44 | 220 | 0.5553 |
| 0.7417 | 0.46 | 230 | 0.5567 |
| 0.8002 | 0.48 | 240 | 0.5522 |
| 0.7666 | 0.5 | 250 | 0.5543 |
| 0.7831 | 0.52 | 260 | 0.5513 |
| 0.7815 | 0.54 | 270 | 0.5493 |
| 0.8298 | 0.56 | 280 | 0.5460 |
| 0.7784 | 0.58 | 290 | 0.5488 |
| 0.7235 | 0.6 | 300 | 0.5433 |
| 0.7862 | 0.62 | 310 | 0.5426 |
| 0.9031 | 0.64 | 320 | 0.5499 |
| 0.711 | 0.66 | 330 | 0.5495 |
| 0.7145 | 0.68 | 340 | 0.5463 |
Framework versions
- Transformers 4.28.0
- Pytorch 1.11.0+cu113
- Datasets 2.11.0
- Tokenizers 0.12.1