<!-- 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
This model is a fine-tuned version of gpt2 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.5522
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 |
---|---|---|---|
1.3326 | 0.1 | 1000 | 0.9231 |
0.8619 | 0.2 | 2000 | 0.8089 |
0.789 | 0.3 | 3000 | 0.7596 |
0.7543 | 0.39 | 4000 | 0.7328 |
0.7308 | 0.49 | 5000 | 0.7119 |
0.7143 | 0.59 | 6000 | 0.6979 |
0.7023 | 0.69 | 7000 | 0.6885 |
0.6912 | 0.79 | 8000 | 0.6776 |
0.683 | 0.89 | 9000 | 0.6705 |
0.6758 | 0.98 | 10000 | 0.6619 |
0.668 | 1.08 | 11000 | 0.6625 |
0.662 | 1.18 | 12000 | 0.6526 |
0.6574 | 1.28 | 13000 | 0.6470 |
0.6535 | 1.38 | 14000 | 0.6435 |
0.6494 | 1.48 | 15000 | 0.6384 |
0.6458 | 1.58 | 16000 | 0.6346 |
0.6416 | 1.67 | 17000 | 0.6318 |
0.6386 | 1.77 | 18000 | 0.6291 |
0.6356 | 1.87 | 19000 | 0.6258 |
0.6325 | 1.97 | 20000 | 0.6221 |
0.6277 | 2.07 | 21000 | 0.6194 |
0.6244 | 2.17 | 22000 | 0.6178 |
0.6218 | 2.27 | 23000 | 0.6147 |
0.6194 | 2.36 | 24000 | 0.6138 |
0.6176 | 2.46 | 25000 | 0.6104 |
0.6158 | 2.56 | 26000 | 0.6076 |
0.6132 | 2.66 | 27000 | 0.6054 |
0.6108 | 2.76 | 28000 | 0.6030 |
0.6087 | 2.86 | 29000 | 0.6002 |
0.6062 | 2.95 | 30000 | 0.5985 |
0.6017 | 3.05 | 31000 | 0.5965 |
0.598 | 3.15 | 32000 | 0.5931 |
0.5961 | 3.25 | 33000 | 0.5907 |
0.595 | 3.35 | 34000 | 0.5890 |
0.5925 | 3.45 | 35000 | 0.5873 |
0.5908 | 3.55 | 36000 | 0.5846 |
0.5884 | 3.64 | 37000 | 0.5818 |
0.5867 | 3.74 | 38000 | 0.5806 |
0.5851 | 3.84 | 39000 | 0.5780 |
0.5829 | 3.94 | 40000 | 0.5757 |
0.5784 | 4.04 | 41000 | 0.5744 |
0.5737 | 4.14 | 42000 | 0.5719 |
0.572 | 4.23 | 43000 | 0.5700 |
0.5709 | 4.33 | 44000 | 0.5684 |
0.5693 | 4.43 | 45000 | 0.5662 |
0.5676 | 4.53 | 46000 | 0.5644 |
0.5662 | 4.63 | 47000 | 0.5628 |
0.5646 | 4.73 | 48000 | 0.5611 |
0.5629 | 4.83 | 49000 | 0.5595 |
0.5612 | 4.92 | 50000 | 0.5582 |
0.5591 | 5.02 | 51000 | 0.5572 |
0.5531 | 5.12 | 52000 | 0.5562 |
0.5527 | 5.22 | 53000 | 0.5553 |
0.552 | 5.32 | 54000 | 0.5544 |
0.5519 | 5.42 | 55000 | 0.5537 |
0.5509 | 5.52 | 56000 | 0.5532 |
0.5509 | 5.61 | 57000 | 0.5528 |
0.5507 | 5.71 | 58000 | 0.5525 |
0.5502 | 5.81 | 59000 | 0.5523 |
0.5504 | 5.91 | 60000 | 0.5522 |
Framework versions
- Transformers 4.30.2
- Pytorch 2.0.0
- Datasets 2.1.0
- Tokenizers 0.13.3