|
--- |
|
license: apache-2.0 |
|
tags: |
|
- simplification |
|
- generated_from_trainer |
|
metrics: |
|
- rouge |
|
- sari |
|
model-index: |
|
- name: mt5-small-clara-med |
|
results: [] |
|
datasets: |
|
- lcampillos/CLARA-MeD |
|
dataset: |
|
- lcampillos/CLARA-MeD |
|
language: |
|
- es |
|
--- |
|
|
|
<!-- 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. --> |
|
|
|
# mt5-small-clara-med |
|
|
|
This model is a fine-tuned version of [google/mt5-small](https://huggingface.co/google/mt5-small) on the [CLARA-MeD](https://huggingface.co/lcampillos/CLARA-MeD) dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 1.9850 |
|
- Rouge1: 33.0363 |
|
- Rouge2: 19.0613 |
|
- Rougel: 30.295 |
|
- Rougelsum: 30.2898 |
|
- SARI: 40.7094 |
|
|
|
## 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: 5.6e-05 |
|
- train_batch_size: 16 |
|
- eval_batch_size: 16 |
|
- 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 | |
|
|:-------------:|:-----:|:----:|:---------------:|:-------:|:-------:|:-------:|:---------:| |
|
| No log | 1.0 | 190 | 3.0286 | 18.0709 | 7.727 | 16.1995 | 16.3348 | |
|
| No log | 2.0 | 380 | 2.4754 | 24.9167 | 13.0501 | 22.3889 | 22.4724 | |
|
| 6.79 | 3.0 | 570 | 2.3542 | 29.9908 | 15.9829 | 26.3751 | 26.4343 | |
|
| 6.79 | 4.0 | 760 | 2.2894 | 30.4435 | 16.3176 | 27.1801 | 27.1926 | |
|
| 3.1288 | 5.0 | 950 | 2.2440 | 30.8602 | 16.8033 | 27.8195 | 27.8355 | |
|
| 3.1288 | 6.0 | 1140 | 2.1772 | 31.4202 | 17.3253 | 28.3394 | 28.3699 | |
|
| 3.1288 | 7.0 | 1330 | 2.1584 | 31.5591 | 17.7302 | 28.618 | 28.6189 | |
|
| 2.7919 | 8.0 | 1520 | 2.1286 | 31.6211 | 17.7423 | 28.7218 | 28.7462 | |
|
| 2.7919 | 9.0 | 1710 | 2.1031 | 31.9724 | 18.017 | 29.0754 | 29.0744 | |
|
| 2.6007 | 10.0 | 1900 | 2.0947 | 32.1588 | 18.2474 | 29.2957 | 29.2956 | |
|
| 2.6007 | 11.0 | 2090 | 2.0914 | 32.4959 | 18.4197 | 29.6052 | 29.609 | |
|
| 2.6007 | 12.0 | 2280 | 2.0726 | 32.6673 | 18.8962 | 29.9145 | 29.9122 | |
|
| 2.4911 | 13.0 | 2470 | 2.0487 | 32.4461 | 18.6804 | 29.6224 | 29.6274 | |
|
| 2.4911 | 14.0 | 2660 | 2.0436 | 32.8393 | 19.0315 | 30.1024 | 30.1097 | |
|
| 2.4168 | 15.0 | 2850 | 2.0229 | 32.8235 | 18.9549 | 30.0699 | 30.0605 | |
|
| 2.4168 | 16.0 | 3040 | 2.0253 | 32.8584 | 18.8602 | 30.0582 | 30.0712 | |
|
| 2.4168 | 17.0 | 3230 | 2.0177 | 32.7145 | 18.9059 | 30.0436 | 30.0771 | |
|
| 2.3452 | 18.0 | 3420 | 2.0151 | 32.6874 | 18.8339 | 29.9739 | 30.0004 | |
|
| 2.3452 | 19.0 | 3610 | 2.0138 | 32.516 | 18.6562 | 29.7823 | 29.7951 | |
|
| 2.302 | 20.0 | 3800 | 2.0085 | 32.8117 | 18.8208 | 30.0902 | 30.1282 | |
|
| 2.302 | 21.0 | 3990 | 2.0043 | 32.7633 | 18.8364 | 30.0619 | 30.0781 | |
|
| 2.302 | 22.0 | 4180 | 1.9972 | 32.9786 | 19.0354 | 30.2166 | 30.2286 | |
|
| 2.2641 | 23.0 | 4370 | 1.9927 | 33.0222 | 19.0501 | 30.2716 | 30.2951 | |
|
| 2.2641 | 24.0 | 4560 | 1.9905 | 32.9557 | 18.9958 | 30.1988 | 30.2004 | |
|
| 2.2366 | 25.0 | 4750 | 1.9897 | 33.0429 | 18.9806 | 30.2861 | 30.3012 | |
|
| 2.2366 | 26.0 | 4940 | 1.9850 | 33.047 | 19.118 | 30.3437 | 30.3368 | |
|
| 2.2366 | 27.0 | 5130 | 1.9860 | 33.0736 | 19.0805 | 30.3311 | 30.3476 | |
|
| 2.2157 | 28.0 | 5320 | 1.9870 | 33.0698 | 19.0649 | 30.2959 | 30.3093 | |
|
| 2.2157 | 29.0 | 5510 | 1.9844 | 33.0376 | 19.0397 | 30.299 | 30.2839 | |
|
| 2.2131 | 30.0 | 5700 | 1.9850 | 33.0363 | 19.0613 | 30.295 | 30.2898 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.25.1 |
|
- Pytorch 1.13.0 |
|
- Datasets 2.8.0 |
|
- Tokenizers 0.12.1 |
