Hugging Face's logo

pengold
/
t5-vietnamese-summarization

Summarization
Transformers
PyTorch
Vietnamese
t5
text2text-generation
text-generation-inference
Model card Files Community
3
t5-vietnamese-summarization / README.md
pengold's picture
pengold
End of training
7531d14
preview code
|
raw
history blame
10.3 kB
metadata
license: apache-2.0
base_model: t5-small
tags:
  - generated_from_trainer
metrics:
  - rouge
model-index:
  - name: t5-vietnamese-summarization
    results: []

t5-vietnamese-summarization

This model is a fine-tuned version of t5-small on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 5.2561
  • Rouge1: 0.4601
  • Rouge2: 0.1574
  • Rougel: 0.2977
  • Rougelsum: 0.2978
  • Gen Len: 18.806

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: 2e-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: 100

Training results

Training Loss Epoch Step Validation Loss Rouge1 Rouge2 Rougel Rougelsum Gen Len
6.9262 1.0 1250 6.6039 0.4025 0.1122 0.2573 0.2576 18.738
6.6661 2.0 2500 6.4575 0.4067 0.1184 0.2611 0.2617 18.868
6.5642 3.0 3750 6.3785 0.4176 0.1265 0.2722 0.2723 18.724
6.4586 4.0 5000 6.3006 0.4204 0.1286 0.2739 0.274 18.694
6.4089 5.0 6250 6.2724 0.426 0.1321 0.275 0.2752 18.784
6.328 6.0 7500 6.2053 0.4435 0.1409 0.2872 0.2873 18.818
6.2862 7.0 8750 6.1546 0.4397 0.1401 0.2845 0.2849 18.83
6.2394 8.0 10000 6.1136 0.443 0.1427 0.287 0.2874 18.816
6.2024 9.0 11250 6.0772 0.4438 0.1459 0.287 0.2874 18.858
6.1423 10.0 12500 6.0458 0.4455 0.1478 0.2891 0.2895 18.832
6.1265 11.0 13750 6.0011 0.4496 0.1474 0.2896 0.29 18.858
6.0642 12.0 15000 5.9603 0.4524 0.1488 0.2936 0.2942 18.884
6.0457 13.0 16250 5.9340 0.4479 0.1484 0.2903 0.2907 18.926
6.0059 14.0 17500 5.8934 0.4484 0.1458 0.2905 0.2908 18.892
5.9558 15.0 18750 5.8688 0.4506 0.1501 0.2917 0.292 18.896
5.9421 16.0 20000 5.8424 0.4502 0.1458 0.2863 0.2868 18.824
5.9049 17.0 21250 5.8208 0.448 0.1482 0.2889 0.2894 18.844
5.8713 18.0 22500 5.8003 0.449 0.1473 0.2892 0.2895 18.868
5.8481 19.0 23750 5.7704 0.4484 0.1486 0.2878 0.2883 18.87
5.8113 20.0 25000 5.7443 0.4513 0.152 0.2914 0.2919 18.822
5.7948 21.0 26250 5.7222 0.4485 0.1477 0.2894 0.2898 18.794
5.7728 22.0 27500 5.6990 0.4476 0.1506 0.289 0.2892 18.82
5.7476 23.0 28750 5.6802 0.4474 0.1493 0.2901 0.2904 18.846
5.7299 24.0 30000 5.6608 0.4514 0.1549 0.2924 0.2928 18.872
5.7021 25.0 31250 5.6533 0.451 0.1537 0.2921 0.2925 18.842
5.6861 26.0 32500 5.6371 0.4502 0.1534 0.291 0.2915 18.826
5.6833 27.0 33750 5.6241 0.4541 0.1542 0.2938 0.2941 18.876
5.6473 28.0 35000 5.6113 0.4509 0.1535 0.2932 0.2935 18.83
5.6248 29.0 36250 5.5896 0.454 0.1562 0.2934 0.2938 18.878
5.6126 30.0 37500 5.5768 0.4555 0.1563 0.2952 0.2954 18.924
5.6044 31.0 38750 5.5627 0.4526 0.1547 0.2929 0.2932 18.856
5.58 32.0 40000 5.5459 0.4482 0.1523 0.291 0.2914 18.898
5.5621 33.0 41250 5.5345 0.4524 0.1546 0.2936 0.294 18.88
5.5399 34.0 42500 5.5209 0.4554 0.1554 0.2939 0.2942 18.868
5.5272 35.0 43750 5.5011 0.4512 0.1562 0.2928 0.2931 18.858
5.5276 36.0 45000 5.5009 0.4504 0.1548 0.2926 0.2931 18.846
5.5355 37.0 46250 5.4912 0.4538 0.1552 0.2932 0.2936 18.874
5.4894 38.0 47500 5.4792 0.455 0.1591 0.2932 0.2937 18.872
5.4872 39.0 48750 5.4692 0.4558 0.1556 0.2918 0.2923 18.864
5.4716 40.0 50000 5.4585 0.4564 0.159 0.2964 0.2966 18.844
5.4461 41.0 51250 5.4532 0.4591 0.1604 0.2961 0.2964 18.85
5.4423 42.0 52500 5.4420 0.4557 0.1577 0.295 0.2952 18.862
5.4259 43.0 53750 5.4341 0.4534 0.1565 0.2929 0.2929 18.83
5.4125 44.0 55000 5.4303 0.4543 0.1579 0.2935 0.2936 18.854
5.4101 45.0 56250 5.4062 0.457 0.1594 0.2945 0.2948 18.836
5.4027 46.0 57500 5.4094 0.4539 0.1553 0.2934 0.2937 18.822
5.3947 47.0 58750 5.4018 0.4567 0.1555 0.2944 0.2949 18.79
5.3905 48.0 60000 5.4001 0.4557 0.1554 0.295 0.2952 18.802
5.3798 49.0 61250 5.3843 0.4549 0.156 0.2949 0.2952 18.818
5.3556 50.0 62500 5.3866 0.4578 0.1581 0.2948 0.2951 18.828
5.3796 51.0 63750 5.3794 0.4564 0.1577 0.2963 0.2967 18.81
5.341 52.0 65000 5.3720 0.4573 0.1578 0.2959 0.2964 18.796
5.3461 53.0 66250 5.3592 0.4579 0.1571 0.2955 0.2956 18.812
5.3385 54.0 67500 5.3622 0.4567 0.1562 0.2954 0.2957 18.756
5.3163 55.0 68750 5.3548 0.4591 0.155 0.2956 0.2959 18.824
5.3222 56.0 70000 5.3542 0.4585 0.1564 0.2955 0.2959 18.836
5.3232 57.0 71250 5.3478 0.4577 0.1567 0.2959 0.2961 18.82
5.2974 58.0 72500 5.3366 0.4538 0.1545 0.2932 0.2934 18.81
5.284 59.0 73750 5.3386 0.4578 0.1557 0.2955 0.2959 18.79
5.3004 60.0 75000 5.3349 0.4569 0.1568 0.2957 0.2959 18.794
5.259 61.0 76250 5.3238 0.4607 0.1566 0.2987 0.2991 18.822
5.2885 62.0 77500 5.3232 0.4607 0.1591 0.2981 0.2986 18.81
5.2857 63.0 78750 5.3139 0.4594 0.1574 0.2957 0.2959 18.77
5.274 64.0 80000 5.3202 0.4601 0.1558 0.2971 0.2972 18.824
5.2665 65.0 81250 5.3123 0.4599 0.1575 0.2967 0.2969 18.884
5.2482 66.0 82500 5.3004 0.4601 0.1572 0.2985 0.2985 18.812
5.2429 67.0 83750 5.2976 0.4572 0.1535 0.2957 0.2958 18.792
5.2407 68.0 85000 5.2985 0.4591 0.1581 0.2966 0.2966 18.828
5.2371 69.0 86250 5.2896 0.4604 0.1584 0.2982 0.2984 18.828
5.2341 70.0 87500 5.2917 0.4612 0.1605 0.2988 0.2991 18.83
5.2311 71.0 88750 5.2882 0.4594 0.1574 0.2977 0.298 18.778
5.2395 72.0 90000 5.2811 0.4609 0.1587 0.2974 0.2975 18.854
5.2187 73.0 91250 5.2836 0.4606 0.1599 0.2986 0.2986 18.804
5.2158 74.0 92500 5.2781 0.4603 0.1584 0.2974 0.2976 18.8
5.2153 75.0 93750 5.2802 0.4603 0.1577 0.2973 0.2975 18.806
5.2153 76.0 95000 5.2771 0.4596 0.1563 0.2954 0.2957 18.816
5.1939 77.0 96250 5.2771 0.4594 0.1566 0.2966 0.2968 18.828
5.2215 78.0 97500 5.2725 0.4578 0.1576 0.2973 0.2975 18.78
5.1974 79.0 98750 5.2704 0.458 0.1578 0.2974 0.2976 18.77
5.2068 80.0 100000 5.2657 0.4612 0.1575 0.2991 0.2994 18.786
5.2018 81.0 101250 5.2643 0.4592 0.157 0.2971 0.2971 18.812
5.21 82.0 102500 5.2608 0.459 0.1577 0.2979 0.2979 18.792
5.2032 83.0 103750 5.2612 0.4604 0.1574 0.2978 0.298 18.798
5.1908 84.0 105000 5.2594 0.4593 0.1587 0.2983 0.2985 18.822
5.194 85.0 106250 5.2548 0.4581 0.1576 0.2981 0.2981 18.792
5.1821 86.0 107500 5.2580 0.459 0.1568 0.2971 0.2971 18.802
5.1901 87.0 108750 5.2561 0.4601 0.1574 0.2977 0.2978 18.806

Framework versions

  • Transformers 4.33.2
  • Pytorch 2.0.1+cu117
  • Datasets 2.14.5
  • Tokenizers 0.13.3