<!-- 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. -->
tulio-gpt-neo
This model is a fine-tuned version of EleutherAI/gpt-neo-2.7B on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.6494
- Accuracy: 0.6498
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: 1
- eval_batch_size: 8
- seed: 13
- distributed_type: multi-GPU
- num_devices: 2
- gradient_accumulation_steps: 2
- total_train_batch_size: 4
- total_eval_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 1.0
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 2.0707 | 0.06 | 2000 | 1.8760 | 0.6169 |
| 2.0158 | 0.12 | 4000 | 1.8242 | 0.6244 |
| 1.9811 | 0.17 | 6000 | 1.7930 | 0.6289 |
| 1.9313 | 0.23 | 8000 | 1.7695 | 0.6320 |
| 1.9437 | 0.29 | 10000 | 1.7510 | 0.6349 |
| 1.9157 | 0.35 | 12000 | 1.7363 | 0.6371 |
| 1.9071 | 0.41 | 14000 | 1.7227 | 0.6388 |
| 1.871 | 0.47 | 16000 | 1.7100 | 0.6409 |
| 1.8679 | 0.52 | 18000 | 1.7012 | 0.6419 |
| 1.848 | 0.58 | 20000 | 1.6904 | 0.6440 |
| 1.8353 | 0.64 | 22000 | 1.6836 | 0.6449 |
| 1.8603 | 0.7 | 24000 | 1.6768 | 0.6460 |
| 1.844 | 0.76 | 26000 | 1.6699 | 0.6468 |
| 1.8396 | 0.81 | 28000 | 1.6699 | 0.6464 |
| 1.8095 | 0.87 | 30000 | 1.6592 | 0.6483 |
| 1.8149 | 0.93 | 32000 | 1.6553 | 0.6488 |
| 1.8098 | 0.99 | 34000 | 1.6494 | 0.6498 |
Framework versions
- Transformers 4.26.0
- Pytorch 1.13.0+cu116
- Datasets 2.7.1
- Tokenizers 0.13.2
