|
--- |
|
license: apache-2.0 |
|
base_model: google/flan-t5-base |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- rouge |
|
model-index: |
|
- name: t5-flan-semantic |
|
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. --> |
|
|
|
# t5-flan-semantic |
|
|
|
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: 0.0553 |
|
- Rouge1: 0.8952 |
|
- Rouge2: 0.8673 |
|
- Rougel: 0.8952 |
|
- Rougelsum: 0.8952 |
|
|
|
## 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: 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: 30 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | |
|
|:-------------:|:-----:|:----:|:---------------:|:------:|:------:|:------:|:---------:| |
|
| 0.0194 | 1.0 | 2 | 0.0082 | 0.9333 | 0.9158 | 0.9333 | 0.9333 | |
|
| 0.0042 | 2.0 | 4 | 0.0408 | 0.9238 | 0.9031 | 0.9238 | 0.9238 | |
|
| 0.0002 | 3.0 | 6 | 0.0647 | 0.9238 | 0.9031 | 0.9238 | 0.9238 | |
|
| 0.0002 | 4.0 | 8 | 0.1117 | 0.9238 | 0.9031 | 0.9238 | 0.9238 | |
|
| 0.0007 | 5.0 | 10 | 0.1404 | 0.9238 | 0.9031 | 0.9238 | 0.9238 | |
|
| 0.0006 | 6.0 | 12 | 0.0987 | 0.9238 | 0.9031 | 0.9238 | 0.9238 | |
|
| 0.0005 | 7.0 | 14 | 0.0587 | 0.9238 | 0.9031 | 0.9238 | 0.9238 | |
|
| 0.0005 | 8.0 | 16 | 0.0251 | 0.9238 | 0.9031 | 0.9238 | 0.9238 | |
|
| 0.0022 | 9.0 | 18 | 0.0128 | 0.9095 | 0.8852 | 0.9095 | 0.9095 | |
|
| 0.0002 | 10.0 | 20 | 0.0228 | 0.8952 | 0.8622 | 0.8952 | 0.8952 | |
|
| 0.0003 | 11.0 | 22 | 0.0351 | 0.8952 | 0.8622 | 0.8952 | 0.8952 | |
|
| 0.0008 | 12.0 | 24 | 0.0374 | 0.9190 | 0.8980 | 0.9190 | 0.9190 | |
|
| 0.0004 | 13.0 | 26 | 0.0462 | 0.8952 | 0.8673 | 0.8952 | 0.8952 | |
|
| 0.0 | 14.0 | 28 | 0.0610 | 0.8952 | 0.8673 | 0.8952 | 0.8952 | |
|
| 0.0001 | 15.0 | 30 | 0.0737 | 0.8952 | 0.8673 | 0.8952 | 0.8952 | |
|
| 0.0005 | 16.0 | 32 | 0.0839 | 0.8952 | 0.8673 | 0.8952 | 0.8952 | |
|
| 0.0002 | 17.0 | 34 | 0.0917 | 0.8952 | 0.8673 | 0.8952 | 0.8952 | |
|
| 0.0009 | 18.0 | 36 | 0.1001 | 0.8952 | 0.8673 | 0.8952 | 0.8952 | |
|
| 0.0005 | 19.0 | 38 | 0.1054 | 0.8952 | 0.8673 | 0.8952 | 0.8952 | |
|
| 0.0012 | 20.0 | 40 | 0.1079 | 0.8952 | 0.8673 | 0.8952 | 0.8952 | |
|
| 0.0 | 21.0 | 42 | 0.1085 | 0.8952 | 0.8673 | 0.8952 | 0.8952 | |
|
| 0.0 | 22.0 | 44 | 0.1015 | 0.8952 | 0.8673 | 0.8952 | 0.8952 | |
|
| 0.0018 | 23.0 | 46 | 0.0862 | 0.8952 | 0.8673 | 0.8952 | 0.8952 | |
|
| 0.0001 | 24.0 | 48 | 0.0752 | 0.8952 | 0.8673 | 0.8952 | 0.8952 | |
|
| 0.0004 | 25.0 | 50 | 0.0675 | 0.8952 | 0.8673 | 0.8952 | 0.8952 | |
|
| 0.0001 | 26.0 | 52 | 0.0623 | 0.8952 | 0.8673 | 0.8952 | 0.8952 | |
|
| 0.0 | 27.0 | 54 | 0.0589 | 0.8952 | 0.8673 | 0.8952 | 0.8952 | |
|
| 0.0005 | 28.0 | 56 | 0.0568 | 0.8952 | 0.8673 | 0.8952 | 0.8952 | |
|
| 0.0002 | 29.0 | 58 | 0.0557 | 0.8952 | 0.8673 | 0.8952 | 0.8952 | |
|
| 0.0002 | 30.0 | 60 | 0.0553 | 0.8952 | 0.8673 | 0.8952 | 0.8952 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.38.2 |
|
- Pytorch 2.2.1+cu121 |
|
- Datasets 2.18.0 |
|
- Tokenizers 0.15.2 |
|
|
