Hugging Face's logo

xsestech
/
skilltext

Text2Text Generation
Transformers
Safetensors
t5
Generated from Trainer
text-generation-inference
Inference Endpoints
Model card Files Community
skilltext / README.md
xsestech's picture
xsestech
Model save
3e3e739 verified
preview code
|
raw
history blame
5.71 kB
metadata
base_model: ai-forever/ruT5-base
tags:
  - generated_from_trainer
metrics:
  - rouge
  - bleu
model-index:
  - name: skilltext
    results: []

skilltext

This model is a fine-tuned version of ai-forever/ruT5-base on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 1.1336
  • Rouge1: 36.6977
  • Rouge2: 18.3675
  • Rougel: 33.39
  • Rougelsum: 33.228
  • Bleu: 1.5692
  • Gen Len: 18.75

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: 1
  • eval_batch_size: 1
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 30
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Rouge1 Rouge2 Rougel Rougelsum Bleu Gen Len
No log 0.8065 50 1.9354 16.2951 4.8512 15.4376 15.441 0.6506 19.0
No log 1.6129 100 1.6656 19.5019 7.2001 18.8027 18.7638 0.594 19.0
No log 2.4194 150 1.5313 26.0869 11.2464 24.6085 24.0936 0.7185 18.4375
No log 3.2258 200 1.3688 26.6556 10.2516 25.4266 25.2205 0.8019 18.9375
No log 4.0323 250 1.3278 30.3107 16.1963 29.45 29.3516 0.9436 18.6875
No log 4.8387 300 1.2528 29.4018 13.6013 28.5142 28.3839 0.8982 18.8125
No log 5.6452 350 1.1710 30.4376 15.8148 28.1455 28.1251 1.0094 18.5625
No log 6.4516 400 1.1587 32.0952 15.6704 29.448 29.2287 0.9544 18.75
No log 7.2581 450 1.1462 34.8699 16.6354 32.2466 32.0099 1.1543 18.75
2.1867 8.0645 500 1.1531 38.1367 19.5983 35.6417 35.5354 1.156 18.625
2.1867 8.8710 550 1.1414 38.1785 19.9507 35.8596 35.754 1.0002 18.4375
2.1867 9.6774 600 1.1154 37.4513 20.4127 35.0672 35.1356 1.1571 18.375
2.1867 10.4839 650 1.1313 39.9692 22.8346 37.9243 37.8991 1.0607 18.375
2.1867 11.2903 700 1.1038 41.2595 26.6622 38.5816 38.0719 1.4158 18.625
2.1867 12.0968 750 1.1211 37.9702 20.308 35.0885 35.0968 1.1778 18.5625
2.1867 12.9032 800 1.1093 40.779 22.5908 38.9412 38.4138 1.4705 18.6875
2.1867 13.7097 850 1.0986 39.135 22.7832 37.1735 36.7976 1.4786 18.625
2.1867 14.5161 900 1.0948 39.023 23.3971 36.453 36.8176 1.6583 18.5625
2.1867 15.3226 950 1.0863 35.3105 20.0898 32.9446 33.4276 1.3792 18.625
0.9823 16.1290 1000 1.0708 36.8626 20.9323 34.8047 34.3997 1.326 18.75
0.9823 16.9355 1050 1.1206 35.2449 18.4541 33.2371 33.1517 1.4287 18.6875
0.9823 17.7419 1100 1.0607 36.4142 19.401 33.6263 33.3147 1.4083 18.75
0.9823 18.5484 1150 1.0700 37.1307 23.7712 36.1994 36.4324 1.5459 18.9375
0.9823 19.3548 1200 1.1096 36.0131 20.9223 34.8256 35.1558 1.5597 18.9375
0.9823 20.1613 1250 1.0649 37.1102 20.5373 34.4912 34.8616 1.5707 18.75
0.9823 20.9677 1300 1.0845 36.6058 20.0812 34.4313 34.7778 1.4728 18.75
0.9823 21.7742 1350 1.0907 36.1128 19.5435 33.9691 33.8883 1.5862 18.75
0.9823 22.5806 1400 1.1001 35.3522 17.4788 32.9141 32.6285 1.6144 18.75
0.9823 23.3871 1450 1.1312 37.2085 20.9517 34.6124 34.7443 1.5977 18.75
0.717 24.1935 1500 1.1230 35.2227 19.1712 33.0973 32.8891 1.662 18.75
0.717 25.0 1550 1.1096 35.3271 18.9454 33.1794 33.1112 1.4402 18.75
0.717 25.8065 1600 1.1325 36.9198 19.3878 34.0516 34.1622 1.4646 18.75
0.717 26.6129 1650 1.1274 37.309 19.8905 34.6163 34.7457 1.4478 18.75
0.717 27.4194 1700 1.1375 41.3137 21.9617 39.6297 39.1071 1.0891 18.6875
0.717 28.2258 1750 1.1307 40.4945 20.5615 38.2536 37.6927 1.2074 18.6875
0.717 29.0323 1800 1.1371 36.6977 18.3675 33.39 33.228 1.5783 18.75
0.717 29.8387 1850 1.1336 36.6977 18.3675 33.39 33.228 1.5692 18.75

Framework versions

  • Transformers 4.40.0
  • Pytorch 2.2.2
  • Datasets 2.12.0
  • Tokenizers 0.19.1