<!-- 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. -->
gpt2-kl_1_07-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: 0.5387
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 |
---|---|---|---|
74.3927 | 0.02 | 10 | 69.5741 |
46.8756 | 0.04 | 20 | 32.9634 |
14.3637 | 0.06 | 30 | 10.6462 |
7.5685 | 0.08 | 40 | 4.2441 |
4.4164 | 0.1 | 50 | 2.0652 |
2.2355 | 0.12 | 60 | 1.1276 |
1.7411 | 0.14 | 70 | 0.8881 |
1.5599 | 0.16 | 80 | 0.7747 |
1.5609 | 0.18 | 90 | 0.6453 |
1.4701 | 0.2 | 100 | 0.6145 |
1.3214 | 0.22 | 110 | 0.6367 |
1.3795 | 0.24 | 120 | 0.6098 |
1.3714 | 0.26 | 130 | 0.5903 |
1.5096 | 0.28 | 140 | 0.5764 |
1.4983 | 0.3 | 150 | 0.5695 |
1.3707 | 0.32 | 160 | 0.5712 |
1.2352 | 0.34 | 170 | 0.5689 |
1.4272 | 0.36 | 180 | 0.5650 |
1.4578 | 0.38 | 190 | 0.5599 |
1.2548 | 0.4 | 200 | 0.5577 |
1.488 | 0.42 | 210 | 0.5575 |
1.3988 | 0.44 | 220 | 0.5518 |
1.1963 | 0.46 | 230 | 0.5536 |
1.2992 | 0.48 | 240 | 0.5519 |
1.1852 | 0.5 | 250 | 0.5517 |
1.2691 | 0.52 | 260 | 0.5473 |
1.2183 | 0.54 | 270 | 0.5433 |
1.3085 | 0.56 | 280 | 0.5443 |
1.1756 | 0.58 | 290 | 0.5422 |
1.1152 | 0.6 | 300 | 0.5420 |
1.2707 | 0.62 | 310 | 0.5391 |
1.3634 | 0.64 | 320 | 0.5376 |
1.1368 | 0.66 | 330 | 0.5378 |
1.1393 | 0.68 | 340 | 0.5378 |
1.1335 | 0.7 | 350 | 0.5387 |
Framework versions
- Transformers 4.28.0
- Pytorch 1.11.0+cu113
- Datasets 2.11.0
- Tokenizers 0.12.1