|
--- |
|
license: apache-2.0 |
|
base_model: google/flan-t5-base |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- rouge |
|
model-index: |
|
- name: dialogsum |
|
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. --> |
|
|
|
# dialogsum |
|
|
|
This model is a fine-tuned version of [google/flan-t5-base](https://huggingface.co/google/flan-t5-base) on the None dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 1.0245 |
|
- Rouge1: 45.6581 |
|
- Rouge2: 15.9871 |
|
- Rougel: 43.5188 |
|
- Rougelsum: 43.913 |
|
- Gen Len: 19.0 |
|
|
|
## 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: 5 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len | |
|
|:-------------:|:-----:|:----:|:---------------:|:-------:|:-------:|:-------:|:---------:|:-------:| |
|
| 1.2634 | 1.0 | 786 | 1.1040 | 42.2483 | 13.4135 | 40.3622 | 40.5436 | 19.0 | |
|
| 1.1931 | 2.0 | 1572 | 1.0637 | 41.6604 | 14.0866 | 39.7604 | 39.9971 | 19.0 | |
|
| 1.1347 | 3.0 | 2358 | 1.0421 | 45.0247 | 14.5438 | 42.8164 | 43.1956 | 19.0 | |
|
| 1.1155 | 4.0 | 3144 | 1.0312 | 45.6977 | 16.2231 | 43.5457 | 43.8965 | 19.0 | |
|
| 1.0827 | 5.0 | 3930 | 1.0245 | 45.6581 | 15.9871 | 43.5188 | 43.913 | 19.0 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.35.2 |
|
- Pytorch 2.0.1+cu117 |
|
- Datasets 2.15.0 |
|
- Tokenizers 0.15.0 |
|
|