<!-- 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. -->
molgpt_selfies_mosesonly
This model is a fine-tuned version of gpt2 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.5139
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: 0.0006
- 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: 1000
- num_epochs: 6
Training results
Training Loss | Epoch | Step | Validation Loss |
---|---|---|---|
1.1282 | 0.18 | 1000 | 0.7807 |
0.7302 | 0.36 | 2000 | 0.6754 |
0.6658 | 0.54 | 3000 | 0.6378 |
0.6381 | 0.72 | 4000 | 0.6180 |
0.6208 | 0.9 | 5000 | 0.6067 |
0.6072 | 1.08 | 6000 | 0.5968 |
0.5973 | 1.26 | 7000 | 0.5859 |
0.5897 | 1.44 | 8000 | 0.5795 |
0.5837 | 1.62 | 9000 | 0.5724 |
0.5778 | 1.79 | 10000 | 0.5683 |
0.5729 | 1.97 | 11000 | 0.5639 |
0.5664 | 2.15 | 12000 | 0.5613 |
0.5621 | 2.33 | 13000 | 0.5555 |
0.5592 | 2.51 | 14000 | 0.5520 |
0.5552 | 2.69 | 15000 | 0.5481 |
0.5524 | 2.87 | 16000 | 0.5449 |
0.5474 | 3.05 | 17000 | 0.5420 |
0.5426 | 3.23 | 18000 | 0.5397 |
0.5405 | 3.41 | 19000 | 0.5369 |
0.538 | 3.59 | 20000 | 0.5338 |
0.5353 | 3.77 | 21000 | 0.5307 |
0.5329 | 3.95 | 22000 | 0.5283 |
0.5266 | 4.13 | 23000 | 0.5264 |
0.5237 | 4.31 | 24000 | 0.5236 |
0.522 | 4.49 | 25000 | 0.5218 |
0.5206 | 4.67 | 26000 | 0.5198 |
0.5191 | 4.85 | 27000 | 0.5182 |
0.5165 | 5.03 | 28000 | 0.5168 |
0.5113 | 5.21 | 29000 | 0.5159 |
0.5104 | 5.38 | 30000 | 0.5150 |
0.5105 | 5.56 | 31000 | 0.5143 |
0.5098 | 5.74 | 32000 | 0.5140 |
0.5094 | 5.92 | 33000 | 0.5139 |
Framework versions
- Transformers 4.30.2
- Pytorch 2.0.0
- Datasets 2.1.0
- Tokenizers 0.13.3