<!-- 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. -->
flan-t5-base-SQuAD-qg-ep10
This model is a fine-tuned version of google/flan-t5-base on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.3457
- Rouge1: 40.1205
- Rouge2: 18.6383
- Rougel: 36.6147
- Rougelsum: 36.6007
- Gen Len: 13.6362
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: 72
- eval_batch_size: 144
- seed: 1799
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
|---|---|---|---|---|---|---|---|---|
| 1.6188 | 0.76 | 200 | 1.3784 | 38.118 | 16.6537 | 34.5034 | 34.5096 | 14.0914 |
| 1.5475 | 1.52 | 400 | 1.3669 | 38.4347 | 16.9434 | 34.832 | 34.8234 | 14.0755 |
| 1.531 | 2.28 | 600 | 1.3597 | 38.7086 | 17.293 | 35.1629 | 35.1641 | 13.8447 |
| 1.4996 | 3.04 | 800 | 1.3532 | 39.684 | 18.3222 | 36.162 | 36.1706 | 13.7223 |
| 1.4857 | 3.8 | 1000 | 1.3521 | 39.518 | 18.1439 | 35.9424 | 35.9616 | 13.7654 |
| 1.4661 | 4.56 | 1200 | 1.3499 | 39.8722 | 18.5417 | 36.3954 | 36.4133 | 13.6062 |
| 1.4591 | 5.32 | 1400 | 1.3494 | 40.1145 | 18.6701 | 36.581 | 36.5954 | 13.6018 |
| 1.4488 | 6.08 | 1600 | 1.3470 | 40.0079 | 18.6909 | 36.5312 | 36.5373 | 13.701 |
| 1.437 | 6.84 | 1800 | 1.3471 | 40.355 | 18.8814 | 36.8203 | 36.8338 | 13.6488 |
| 1.4389 | 7.6 | 2000 | 1.3461 | 40.189 | 18.7211 | 36.6556 | 36.6487 | 13.6744 |
| 1.4146 | 8.37 | 2200 | 1.3464 | 40.3692 | 18.8247 | 36.7903 | 36.7986 | 13.6589 |
| 1.421 | 9.13 | 2400 | 1.3457 | 40.1205 | 18.6383 | 36.6147 | 36.6007 | 13.6362 |
| 1.4163 | 9.89 | 2600 | 1.3458 | 40.223 | 18.6957 | 36.6983 | 36.6998 | 13.657 |
Framework versions
- Transformers 4.18.0
- Pytorch 1.10.0+cu111
- Datasets 2.5.1
- Tokenizers 0.12.1
