<!-- 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-qag-ep6
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: 1.0075
- Rouge1: 39.0569
- Rouge2: 17.6997
- Rougel: 35.5578
- Rougelsum: 35.5512
- F1: 18.3373
- Exact Match: 13.014
- Gen Len: 18.4509
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: 84
- eval_batch_size: 168
- seed: 1799
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 6
Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | F1 | Exact Match | Gen Len |
|---|---|---|---|---|---|---|---|---|---|---|
| 1.2679 | 0.89 | 200 | 1.0379 | 38.2769 | 16.8906 | 34.7944 | 34.8178 | 16.7543 | 11.7078 | 18.5225 |
| 1.151 | 1.78 | 400 | 1.0222 | 38.7208 | 17.2857 | 35.2522 | 35.2879 | 17.8568 | 12.8205 | 18.4756 |
| 1.1243 | 2.67 | 600 | 1.0141 | 38.677 | 17.2919 | 35.207 | 35.2374 | 17.8534 | 12.9173 | 18.4741 |
| 1.1183 | 3.56 | 800 | 1.0109 | 38.5725 | 17.1551 | 35.0545 | 35.0774 | 18.0599 | 13.2559 | 18.4722 |
| 1.0949 | 4.44 | 1000 | 1.0096 | 38.9484 | 17.6335 | 35.4819 | 35.4888 | 18.0128 | 13.2075 | 18.464 |
| 1.0983 | 5.33 | 1200 | 1.0075 | 39.0569 | 17.6997 | 35.5578 | 35.5512 | 18.3373 | 13.014 | 18.4509 |
Framework versions
- Transformers 4.18.0
- Pytorch 1.11.0+cu113
- Datasets 2.5.1
- Tokenizers 0.12.1
