results
This model is a fine-tuned version of google/flan-t5-base on question/answer pairs scraped from Quora. It achieves the following results on the evaluation set:
- Loss: 2.9379
Intended uses & limitations
Closed book question answering. Prefix the inputs with "answer the question: "
Training and evaluation data
50,000 question/answer pairs scraped from Quora
Training procedure
Seq2Seq training described here: https://huggingface.co/docs/evaluate/transformers_integrations
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 0.0003
- 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: 2
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 3.158 | 1.0 | 5641 | 2.9785 |
| 3.0283 | 2.0 | 11282 | 2.9379 |
Framework versions
- Transformers 4.32.0
- Pytorch 2.0.0
- Datasets 2.14.4
- Tokenizers 0.13.3
- Downloads last month
- 3