metadata
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- squad_v2_modified_for_t5_qg
widget:
- text: >-
generate question: Python is developed by Guido Van Rossum and released in
1991.</s>
model-index:
- name: t5-end2end-questions-generation-squadV2
results: []
t5-end2end-questions-generation-squadV2
This model is a fine-tuned version of t5-base on the squad_v2_modified_for_t5_qg dataset. It achieves the following results on the evaluation set:
- Loss: 1.6425
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: 0.0001
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 16
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 2.6214 | 0.34 | 100 | 1.9787 |
| 1.9902 | 0.67 | 200 | 1.7963 |
| 1.8772 | 1.01 | 300 | 1.7362 |
| 1.7767 | 1.35 | 400 | 1.7150 |
| 1.7448 | 1.68 | 500 | 1.6903 |
| 1.736 | 2.02 | 600 | 1.6708 |
| 1.6759 | 2.36 | 700 | 1.6758 |
| 1.6587 | 2.69 | 800 | 1.6682 |
| 1.6544 | 3.03 | 900 | 1.6550 |
| 1.6185 | 3.37 | 1000 | 1.6511 |
| 1.6116 | 3.7 | 1100 | 1.6527 |
| 1.6158 | 4.04 | 1200 | 1.6543 |
| 1.5811 | 4.38 | 1300 | 1.6415 |
| 1.5831 | 4.71 | 1400 | 1.6425 |
Framework versions
- Transformers 4.20.1
- Pytorch 1.11.0+cu113
- Datasets 2.3.2
- Tokenizers 0.12.1