|
--- |
|
library_name: transformers |
|
license: apache-2.0 |
|
base_model: google/flan-t5-base |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- rouge |
|
model-index: |
|
- name: flan-t5-rouge-squad-qg |
|
results: [] |
|
--- |
|
|
|
<!-- 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-rouge-squad-qg |
|
|
|
This model is a fine-tuned version of [google/flan-t5-base](https://huggingface.co/google/flan-t5-base) on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.3225 |
|
- Rouge1: 0.4603 |
|
- Rouge2: 0.1838 |
|
- Rougel: 0.4115 |
|
- Rougelsum: 0.4478 |
|
|
|
## 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: 0.0003 |
|
- train_batch_size: 24 |
|
- eval_batch_size: 24 |
|
- seed: 42 |
|
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments |
|
- lr_scheduler_type: linear |
|
- num_epochs: 30 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | |
|
|:-------------:|:-----:|:----:|:---------------:|:------:|:------:|:------:|:---------:| |
|
| 16.5081 | 1.0 | 3 | 15.3312 | 0.2528 | 0.1156 | 0.2032 | 0.2053 | |
|
| 8.4005 | 2.0 | 6 | 4.6577 | 0.2538 | 0.1115 | 0.1987 | 0.2066 | |
|
| 4.0936 | 3.0 | 9 | 4.0634 | 0.2540 | 0.1245 | 0.2186 | 0.2184 | |
|
| 3.6521 | 4.0 | 12 | 3.6769 | 0.2898 | 0.1640 | 0.2539 | 0.2607 | |
|
| 3.1589 | 5.0 | 15 | 3.1339 | 0.3017 | 0.1714 | 0.2645 | 0.2717 | |
|
| 2.5217 | 6.0 | 18 | 2.3186 | 0.3017 | 0.1714 | 0.2645 | 0.2717 | |
|
| 1.5472 | 7.0 | 21 | 0.9929 | 0.3164 | 0.1571 | 0.2498 | 0.2563 | |
|
| 1.5284 | 8.0 | 24 | 0.6101 | 0.2596 | 0.1178 | 0.2018 | 0.2043 | |
|
| 1.2359 | 9.0 | 27 | 0.4815 | 0.2596 | 0.1178 | 0.2018 | 0.2043 | |
|
| 1.1669 | 10.0 | 30 | 0.3870 | 0.5138 | 0.2379 | 0.4589 | 0.4960 | |
|
| 0.6514 | 11.0 | 33 | 0.3166 | 0.5138 | 0.2379 | 0.4589 | 0.4960 | |
|
| 0.3465 | 12.0 | 36 | 0.2860 | 0.5138 | 0.2379 | 0.4589 | 0.4960 | |
|
| 0.2721 | 13.0 | 39 | 0.2776 | 0.5138 | 0.2379 | 0.4589 | 0.4960 | |
|
| 0.4964 | 14.0 | 42 | 0.2746 | 0.5152 | 0.2444 | 0.4655 | 0.4940 | |
|
| 0.2382 | 15.0 | 45 | 0.2761 | 0.4675 | 0.1901 | 0.4208 | 0.4483 | |
|
| 0.5074 | 16.0 | 48 | 0.2801 | 0.4675 | 0.1901 | 0.4208 | 0.4483 | |
|
| 0.353 | 17.0 | 51 | 0.2858 | 0.4675 | 0.1901 | 0.4208 | 0.4483 | |
|
| 0.1808 | 18.0 | 54 | 0.2916 | 0.4675 | 0.1901 | 0.4208 | 0.4483 | |
|
| 0.2506 | 19.0 | 57 | 0.2961 | 0.4665 | 0.1838 | 0.4145 | 0.4506 | |
|
| 0.2296 | 20.0 | 60 | 0.3027 | 0.5138 | 0.2379 | 0.4589 | 0.4960 | |
|
| 0.3785 | 21.0 | 63 | 0.3073 | 0.5138 | 0.2379 | 0.4589 | 0.4960 | |
|
| 0.2014 | 22.0 | 66 | 0.3107 | 0.5138 | 0.2379 | 0.4589 | 0.4960 | |
|
| 0.4355 | 23.0 | 69 | 0.3131 | 0.5138 | 0.2379 | 0.4589 | 0.4960 | |
|
| 0.1727 | 24.0 | 72 | 0.3159 | 0.5138 | 0.2379 | 0.4589 | 0.4960 | |
|
| 0.1342 | 25.0 | 75 | 0.3173 | 0.5138 | 0.2379 | 0.4589 | 0.4960 | |
|
| 0.193 | 26.0 | 78 | 0.3200 | 0.5138 | 0.2379 | 0.4589 | 0.4960 | |
|
| 0.1703 | 27.0 | 81 | 0.3211 | 0.5138 | 0.2379 | 0.4589 | 0.4960 | |
|
| 0.0676 | 28.0 | 84 | 0.3216 | 0.4603 | 0.1838 | 0.4115 | 0.4478 | |
|
| 0.1207 | 29.0 | 87 | 0.3222 | 0.4603 | 0.1838 | 0.4115 | 0.4478 | |
|
| 0.1002 | 30.0 | 90 | 0.3225 | 0.4603 | 0.1838 | 0.4115 | 0.4478 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.47.0 |
|
- Pytorch 2.5.1+cu121 |
|
- Datasets 3.1.0 |
|
- Tokenizers 0.21.0 |
|
|
