|
--- |
|
base_model: google/mt5-base |
|
library_name: transformers |
|
license: apache-2.0 |
|
metrics: |
|
- rouge |
|
tags: |
|
- generated_from_trainer |
|
model-index: |
|
- name: mt5-rouge-durga-2 |
|
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. --> |
|
|
|
# mt5-rouge-durga-2 |
|
|
|
This model is a fine-tuned version of [google/mt5-base](https://huggingface.co/google/mt5-base) on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.0126 |
|
- Rouge1: 0.6270 |
|
- Rouge2: 0.6003 |
|
- Rougel: 0.6244 |
|
- Rougelsum: 0.6247 |
|
|
|
## 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: 2 |
|
- eval_batch_size: 2 |
|
- 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 | |
|
|:-------------:|:-----:|:----:|:---------------:|:------:|:------:|:------:|:---------:| |
|
| 4.989 | 1.0 | 85 | 2.8197 | 0.2164 | 0.0941 | 0.1882 | 0.1883 | |
|
| 3.116 | 2.0 | 170 | 2.0798 | 0.3122 | 0.1588 | 0.2604 | 0.2604 | |
|
| 2.8357 | 3.0 | 255 | 1.5681 | 0.3446 | 0.1935 | 0.2953 | 0.2955 | |
|
| 1.7776 | 4.0 | 340 | 1.1806 | 0.3324 | 0.1952 | 0.2895 | 0.2904 | |
|
| 1.1881 | 5.0 | 425 | 0.9407 | 0.3533 | 0.2228 | 0.3088 | 0.3091 | |
|
| 1.8511 | 6.0 | 510 | 0.6826 | 0.3971 | 0.2700 | 0.3644 | 0.3636 | |
|
| 1.7178 | 7.0 | 595 | 0.5128 | 0.4194 | 0.3120 | 0.3894 | 0.3891 | |
|
| 1.2772 | 8.0 | 680 | 0.3878 | 0.4590 | 0.3619 | 0.4311 | 0.4302 | |
|
| 1.3577 | 9.0 | 765 | 0.2709 | 0.4729 | 0.3881 | 0.4499 | 0.4497 | |
|
| 0.8291 | 10.0 | 850 | 0.2005 | 0.5006 | 0.4276 | 0.4748 | 0.4747 | |
|
| 0.6825 | 11.0 | 935 | 0.1616 | 0.5411 | 0.4732 | 0.5215 | 0.5224 | |
|
| 0.5006 | 12.0 | 1020 | 0.1182 | 0.5348 | 0.4782 | 0.5200 | 0.5196 | |
|
| 0.5193 | 13.0 | 1105 | 0.1027 | 0.5446 | 0.4910 | 0.5269 | 0.5286 | |
|
| 0.3933 | 14.0 | 1190 | 0.0881 | 0.5685 | 0.5200 | 0.5535 | 0.5548 | |
|
| 0.1584 | 15.0 | 1275 | 0.0708 | 0.5719 | 0.5327 | 0.5629 | 0.5645 | |
|
| 0.3657 | 16.0 | 1360 | 0.0646 | 0.5763 | 0.5315 | 0.5648 | 0.5659 | |
|
| 0.2731 | 17.0 | 1445 | 0.0525 | 0.5908 | 0.5500 | 0.5844 | 0.5844 | |
|
| 0.3466 | 18.0 | 1530 | 0.0511 | 0.5971 | 0.5596 | 0.5873 | 0.5886 | |
|
| 0.1892 | 19.0 | 1615 | 0.0384 | 0.6044 | 0.5675 | 0.5991 | 0.5995 | |
|
| 0.1684 | 20.0 | 1700 | 0.0328 | 0.6066 | 0.5744 | 0.6046 | 0.6050 | |
|
| 0.0691 | 21.0 | 1785 | 0.0295 | 0.6057 | 0.5726 | 0.6020 | 0.6027 | |
|
| 0.0326 | 22.0 | 1870 | 0.0243 | 0.6167 | 0.5872 | 0.6138 | 0.6146 | |
|
| 0.1872 | 23.0 | 1955 | 0.0195 | 0.6188 | 0.5899 | 0.6149 | 0.6160 | |
|
| 0.1372 | 24.0 | 2040 | 0.0183 | 0.6253 | 0.5961 | 0.6227 | 0.6233 | |
|
| 0.0621 | 25.0 | 2125 | 0.0166 | 0.6239 | 0.5957 | 0.6211 | 0.6225 | |
|
| 0.2539 | 26.0 | 2210 | 0.0161 | 0.6217 | 0.5926 | 0.6191 | 0.6200 | |
|
| 0.2532 | 27.0 | 2295 | 0.0166 | 0.6195 | 0.5910 | 0.6166 | 0.6173 | |
|
| 0.1158 | 28.0 | 2380 | 0.0145 | 0.6223 | 0.5943 | 0.6196 | 0.6202 | |
|
| 0.3496 | 29.0 | 2465 | 0.0132 | 0.6241 | 0.5957 | 0.6212 | 0.6217 | |
|
| 0.059 | 30.0 | 2550 | 0.0126 | 0.6270 | 0.6003 | 0.6244 | 0.6247 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.44.2 |
|
- Pytorch 2.4.0+cu121 |
|
- Datasets 3.0.0 |
|
- Tokenizers 0.19.1 |
|
|
