File size: 5,993 Bytes
dfe11e1 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 |
---
license: apache-2.0
tags:
- summarization
- generated_from_trainer
metrics:
- rouge
model-index:
- name: finetuned-tamil-text-summarization
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. -->
# finetuned-tamil-text-summarization
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: 0.9838
- Rouge1: 0.1323
- Rouge2: 0.0864
- Rougel: 0.13
- Rougelsum: 0.1323
## 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: 0.0005
- train_batch_size: 2
- eval_batch_size: 1
- seed: 42
- gradient_accumulation_steps: 16
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 90
- num_epochs: 10
### Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum |
|:-------------:|:-----:|:----:|:---------------:|:------:|:------:|:------:|:---------:|
| 0.6298 | 0.2 | 100 | 1.6144 | 0.0773 | 0.0611 | 0.0689 | 0.0833 |
| 0.4317 | 0.39 | 200 | 1.1760 | 0.1615 | 0.0699 | 0.1600 | 0.1600 |
| 0.2588 | 0.59 | 300 | 1.0634 | 0.08 | 0.0556 | 0.0841 | 0.0838 |
| 0.2665 | 0.79 | 400 | 1.0443 | 0.0631 | 0.0216 | 0.0631 | 0.0631 |
| 0.1972 | 0.99 | 500 | 1.0465 | 0.1909 | 0.1068 | 0.1793 | 0.1916 |
| 0.2041 | 1.19 | 600 | 0.9551 | 0.0904 | 0.0505 | 0.0968 | 0.0987 |
| 0.238 | 1.38 | 700 | 0.9423 | 0.1 | 0.0729 | 0.1 | 0.1033 |
| 0.275 | 1.58 | 800 | 0.9273 | 0.1467 | 0.1098 | 0.1504 | 0.1515 |
| 0.2379 | 1.78 | 900 | 0.9023 | 0.1 | 0.0833 | 0.1 | 0.1 |
| 0.2896 | 1.97 | 1000 | 0.9184 | 0.19 | 0.1 | 0.1889 | 0.1985 |
| 0.2663 | 2.17 | 1100 | 0.9003 | 0.0795 | 0.0678 | 0.0878 | 0.0833 |
| 0.237 | 2.37 | 1200 | 0.9139 | 0.1990 | 0.1029 | 0.1951 | 0.2062 |
| 0.2019 | 2.57 | 1300 | 0.9210 | 0.1128 | 0.0364 | 0.1128 | 0.1161 |
| 0.1794 | 2.77 | 1400 | 0.9038 | 0.1167 | 0.0864 | 0.1183 | 0.1206 |
| 0.1847 | 2.96 | 1500 | 0.8893 | 0.1434 | 0.1313 | 0.1438 | 0.1473 |
| 0.1436 | 3.16 | 1600 | 0.8872 | 0.1683 | 0.0583 | 0.1651 | 0.1729 |
| 0.138 | 3.36 | 1700 | 0.8929 | 0.2300 | 0.1249 | 0.2262 | 0.2312 |
| 0.1265 | 3.56 | 1800 | 0.9204 | 0.1745 | 0.0729 | 0.1700 | 0.1773 |
| 0.1828 | 3.75 | 1900 | 0.9094 | 0.18 | 0.1489 | 0.18 | 0.1862 |
| 0.1447 | 3.95 | 2000 | 0.8942 | 0.19 | 0.0989 | 0.1862 | 0.1962 |
| 0.099 | 4.15 | 2100 | 0.9297 | 0.2386 | 0.15 | 0.2352 | 0.2451 |
| 0.1366 | 4.35 | 2200 | 0.9124 | 0.12 | 0.0729 | 0.12 | 0.1245 |
| 0.1519 | 4.54 | 2300 | 0.9040 | 0.1873 | 0.0986 | 0.1833 | 0.1906 |
| 0.119 | 4.74 | 2400 | 0.9121 | 0.12 | 0.0458 | 0.1129 | 0.1229 |
| 0.1364 | 4.94 | 2500 | 0.9120 | 0.2090 | 0.1258 | 0.2067 | 0.2190 |
| 0.1 | 5.14 | 2600 | 0.9409 | 0.1251 | 0.0833 | 0.1240 | 0.1311 |
| 0.1683 | 5.34 | 2700 | 0.9423 | 0.1382 | 0.0951 | 0.1371 | 0.1417 |
| 0.1395 | 5.53 | 2800 | 0.9336 | 0.1612 | 0.1233 | 0.1600 | 0.1631 |
| 0.1067 | 5.73 | 2900 | 0.9290 | 0.2234 | 0.1316 | 0.2174 | 0.2169 |
| 0.1104 | 5.93 | 3000 | 0.9245 | 0.2 | 0.1 | 0.1915 | 0.1915 |
| 0.1474 | 6.13 | 3100 | 0.9423 | 0.2007 | 0.1030 | 0.1963 | 0.1985 |
| 0.1052 | 6.32 | 3200 | 0.9329 | 0.2023 | 0.1102 | 0.2000 | 0.2 |
| 0.1203 | 6.52 | 3300 | 0.9380 | 0.2023 | 0.1102 | 0.2000 | 0.2 |
| 0.1125 | 6.72 | 3400 | 0.9422 | 0.1896 | 0.0977 | 0.1862 | 0.19 |
| 0.1323 | 6.92 | 3500 | 0.9433 | 0.19 | 0.0977 | 0.1862 | 0.19 |
| 0.0949 | 7.11 | 3600 | 0.9529 | 0.1603 | 0.0945 | 0.1612 | 0.1599 |
| 0.1059 | 7.31 | 3700 | 0.9520 | 0.1383 | 0.0977 | 0.1419 | 0.1400 |
| 0.1482 | 7.51 | 3800 | 0.9514 | 0.2112 | 0.1205 | 0.2100 | 0.2073 |
| 0.1268 | 7.71 | 3900 | 0.9386 | 0.2038 | 0.1091 | 0.2015 | 0.2008 |
| 0.089 | 7.9 | 4000 | 0.9426 | 0.1508 | 0.1182 | 0.1562 | 0.1538 |
| 0.108 | 8.1 | 4100 | 0.9727 | 0.1383 | 0.1034 | 0.1445 | 0.1367 |
| 0.1292 | 8.3 | 4200 | 0.9640 | 0.2100 | 0.1256 | 0.2098 | 0.2098 |
| 0.0868 | 8.5 | 4300 | 0.9618 | 0.15 | 0.0943 | 0.1508 | 0.1465 |
| 0.1023 | 8.69 | 4400 | 0.9609 | 0.18 | 0.075 | 0.18 | 0.18 |
| 0.1102 | 8.89 | 4500 | 0.9644 | 0.1462 | 0.1 | 0.1512 | 0.145 |
| 0.1102 | 9.09 | 4600 | 0.9807 | 0.1262 | 0.0864 | 0.1362 | 0.1262 |
| 0.0942 | 9.29 | 4700 | 0.9866 | 0.1400 | 0.0977 | 0.1462 | 0.1400 |
| 0.129 | 9.49 | 4800 | 0.9853 | 0.1284 | 0.0864 | 0.1362 | 0.1295 |
| 0.0949 | 9.68 | 4900 | 0.9819 | 0.1911 | 0.0977 | 0.1962 | 0.1923 |
| 0.0852 | 9.88 | 5000 | 0.9852 | 0.1262 | 0.0864 | 0.1362 | 0.1262 |
### Framework versions
- Transformers 4.26.1
- Pytorch 1.13.1+cu116
- Tokenizers 0.13.2
|