linker_v2
This model is a fine-tuned version of jarod0411/zinc10M_gpt2-medium_SMILES_step1 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.3096
- Accuracy: 0.8994
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: 8
- seed: 1
- distributed_type: multi-GPU
- num_devices: 8
- total_train_batch_size: 64
- total_eval_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10.0
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 0.3437 | 1.0 | 53844 | 0.3364 | 0.8920 |
| 0.3312 | 2.0 | 107688 | 0.3266 | 0.8947 |
| 0.3244 | 3.0 | 161532 | 0.3213 | 0.8961 |
| 0.32 | 4.0 | 215376 | 0.3178 | 0.8971 |
| 0.3162 | 5.0 | 269220 | 0.3153 | 0.8978 |
| 0.3133 | 6.0 | 323064 | 0.3134 | 0.8983 |
| 0.3113 | 7.0 | 376908 | 0.3120 | 0.8987 |
| 0.3094 | 8.0 | 430752 | 0.3108 | 0.8990 |
| 0.3077 | 9.0 | 484596 | 0.3100 | 0.8993 |
| 0.3072 | 10.0 | 538440 | 0.3096 | 0.8994 |
Framework versions
- Transformers 4.36.0.dev0
- Pytorch 2.1.1+cu121
- Datasets 2.15.0
- Tokenizers 0.15.0