Chitti-Base-model-for-GPT-v11
This model is a fine-tuned version of Dhahlan2000/Chitti-Large-model-for-GPT-v11 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.4147
- Bleu: 34.6359
- Gen Len: 4.4545
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: 2e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 6
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Bleu | Gen Len |
|---|---|---|---|---|---|
| No log | 1.0 | 65 | 1.6932 | 12.1428 | 5.0909 |
| No log | 2.0 | 130 | 1.5630 | 28.3754 | 4.6364 |
| No log | 3.0 | 195 | 1.4900 | 27.9945 | 4.6364 |
| No log | 4.0 | 260 | 1.4456 | 27.9945 | 4.6364 |
| No log | 5.0 | 325 | 1.4220 | 34.6359 | 4.4545 |
| No log | 6.0 | 390 | 1.4147 | 34.6359 | 4.4545 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.3.0+cu121
- Datasets 2.20.0
- Tokenizers 0.19.1