flan-t5-base-finetuned-python_qa_v2
This model is a fine-tuned version of google/flan-t5-base on the Python Questions from Stack Overflow dataset.
It achieves the following results on the evaluation set:
- Loss: 1.9023
- Rouge1: 0.1919
- Rouge2: 0.0535
- Rougel: 0.1492
- Rougelsum: 0.1655
Model description
More information needed
Intended uses & limitations
- Question answering
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 0.0001
- train_batch_size: 8
- eval_batch_size: 4
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum |
|---|---|---|---|---|---|---|---|
| 2.0314 | 1.0 | 2000 | 1.9083 | 0.1876 | 0.0546 | 0.1485 | 0.1640 |
| 1.9586 | 2.0 | 4000 | 1.9031 | 0.1896 | 0.0531 | 0.1485 | 0.1643 |
| 1.923 | 3.0 | 6000 | 1.9023 | 0.1919 | 0.0535 | 0.1492 | 0.1655 |
Framework versions
- Transformers 4.34.1
- Pytorch 2.1.0+cu118
- Datasets 2.14.5
- Tokenizers 0.14.1
