|
--- |
|
license: apache-2.0 |
|
base_model: google/flan-t5-small |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- rouge |
|
model-index: |
|
- name: QA-flan-t5-small |
|
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. --> |
|
|
|
# QA-flan-t5-small |
|
|
|
This model is a fine-tuned version of [google/flan-t5-small](https://huggingface.co/google/flan-t5-small) on the None dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.4875 |
|
- Rouge1: 0.6697 |
|
- Rouge2: 0.4916 |
|
- Rougel: 0.6692 |
|
- Rougelsum: 0.6694 |
|
|
|
## 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: 5e-05 |
|
- train_batch_size: 8 |
|
- eval_batch_size: 8 |
|
- seed: 42 |
|
- 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 | |
|
|:-------------:|:-----:|:------:|:---------------:|:------:|:------:|:------:|:---------:| |
|
| 0.7836 | 1.0 | 30468 | 0.5139 | 0.6377 | 0.4690 | 0.6373 | 0.6371 | |
|
| 0.6817 | 2.0 | 60936 | 0.4990 | 0.6428 | 0.4674 | 0.6421 | 0.6422 | |
|
| 0.6913 | 3.0 | 91404 | 0.4999 | 0.6504 | 0.4706 | 0.6496 | 0.6496 | |
|
| 0.5839 | 4.0 | 121872 | 0.4808 | 0.6680 | 0.4918 | 0.6675 | 0.6676 | |
|
| 0.58 | 5.0 | 152340 | 0.4871 | 0.6702 | 0.4919 | 0.6697 | 0.6698 | |
|
| 0.5921 | 6.0 | 182808 | 0.4875 | 0.6697 | 0.4916 | 0.6692 | 0.6694 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.38.1 |
|
- Pytorch 2.2.1+cu121 |
|
- Datasets 2.17.1 |
|
- Tokenizers 0.15.2 |
|
|
