Hugging Face's logo

lengocquangLAB
/
T5-JSON-OM-IMP

Transformers
Safetensors
t5
text2text-generation
Generated from Trainer
text-generation-inference
Model card Files
xet
 Community
T5-JSON-OM-IMP / README.md
lengocquangLAB's picture
lengocquangLAB
End of training
e82483c verified
preview code
|
raw
history blame
17.4 kB
metadata 
library_name: transformers
license: apache-2.0
base_model: T5-small
tags:
  - generated_from_trainer
metrics:
  - bleu
  - rouge
model-index:
  - name: T5-JSON-OM-IMP
    results: []

T5-JSON-OM-IMP

This model is a fine-tuned version of T5-small on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 0.0061
  • Micro Precision: 0.4084
  • Micro Recall: 0.4806
  • Micro F1: 0.4415
  • Macro Precision: 0.4085
  • Macro Recall: 0.4869
  • Macro F1: 0.4443
  • Bleu: 75.5747
  • Rouge1: 0.7648
  • Rouge2: 0.5224

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: 8
  • eval_batch_size: 8
  • seed: 42
  • optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
  • lr_scheduler_type: linear
  • num_epochs: 10
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Micro Precision Micro Recall Micro F1 Macro Precision Macro Recall Macro F1 Bleu Rouge1 Rouge2
17.6561 0.1068 50 8.1745 0 0.0 0 0.0 0.0 0 0.0214 0.0192 0.0
3.6446 0.2137 100 0.1269 0 0.0 0 0.0 0.0 0 0.0030 0.0024 0.0
0.5555 0.3205 150 0.0917 0 0.0 0 0.0 0.0 0 0.0 0.0 0.0
0.1137 0.4274 200 0.0517 0 0.0 0 0.0 0.0 0 0.0 0.0 0.0
0.0641 0.5342 250 0.0316 0 0.0 0 0.0 0.0 0 0.0 0.0 0.0
0.0443 0.6410 300 0.0211 0 0.0 0 0.0 0.0 0 0.0000 0.0027 0.0005
0.0344 0.7479 350 0.0166 0.1667 0.0007 0.0015 0.125 0.0008 0.0015 0.6419 0.0312 0.0135
0.0284 0.8547 400 0.0139 0.4343 0.0315 0.0588 0.4115 0.0328 0.0607 1.9492 0.0924 0.0533
0.0239 0.9615 450 0.0115 0.4073 0.1387 0.2069 0.4033 0.1446 0.2129 18.6642 0.2914 0.1793
0.0209 1.0684 500 0.0095 0.3682 0.2869 0.3225 0.3682 0.2931 0.3264 48.1448 0.5384 0.3526
0.0186 1.1752 550 0.0086 0.3701 0.3764 0.3732 0.3700 0.3829 0.3764 63.7786 0.6727 0.4479
0.0171 1.2821 600 0.0076 0.3394 0.3830 0.3599 0.3393 0.3870 0.3616 70.2327 0.7395 0.4975
0.015 1.3889 650 0.0074 0.3674 0.4189 0.3915 0.3674 0.4229 0.3932 72.5827 0.7581 0.5092
0.0139 1.4957 700 0.0071 0.4025 0.4512 0.4255 0.4027 0.4562 0.4278 75.2921 0.7682 0.5221
0.0133 1.6026 750 0.0068 0.4035 0.4461 0.4237 0.4036 0.4507 0.4258 75.6009 0.7743 0.5314
0.0121 1.7094 800 0.0066 0.3972 0.4549 0.4241 0.3974 0.4609 0.4268 75.4443 0.7705 0.5265
0.0117 1.8162 850 0.0065 0.4113 0.4593 0.4340 0.4119 0.4650 0.4368 75.7436 0.7670 0.5204
0.0113 1.9231 900 0.0065 0.4077 0.4637 0.4339 0.4082 0.4696 0.4368 75.7085 0.7675 0.5223
0.0109 2.0299 950 0.0064 0.4016 0.4674 0.4320 0.4018 0.4730 0.4345 75.2841 0.7655 0.5199
0.01 2.1368 1000 0.0064 0.4058 0.4696 0.4354 0.4061 0.4755 0.4381 75.5945 0.7661 0.5224
0.01 2.2436 1050 0.0063 0.4034 0.4718 0.4349 0.4035 0.4779 0.4376 75.3710 0.7636 0.5183
0.0099 2.3504 1100 0.0065 0.4092 0.4681 0.4367 0.4096 0.4741 0.4395 75.6263 0.7650 0.5195
0.0096 2.4573 1150 0.0063 0.4047 0.4718 0.4356 0.4048 0.4778 0.4383 75.3904 0.7646 0.5198
0.0104 2.5641 1200 0.0063 0.4035 0.4696 0.4340 0.4037 0.4754 0.4366 75.2097 0.7623 0.5147
0.0093 2.6709 1250 0.0062 0.3927 0.4629 0.4249 0.3927 0.4687 0.4274 74.7695 0.7647 0.5188
0.0089 2.7778 1300 0.0063 0.4034 0.4747 0.4361 0.4035 0.4809 0.4388 75.3197 0.7633 0.5179
0.0091 2.8846 1350 0.0062 0.4061 0.4762 0.4384 0.4063 0.4826 0.4412 75.4966 0.7643 0.5202
0.0088 2.9915 1400 0.0063 0.4072 0.4703 0.4365 0.4075 0.4763 0.4392 75.4925 0.7639 0.5187
0.0088 3.0983 1450 0.0062 0.4032 0.4754 0.4364 0.4033 0.4817 0.4390 75.2804 0.7628 0.5172
0.0087 3.2051 1500 0.0063 0.4075 0.4688 0.4360 0.4078 0.4748 0.4388 75.5317 0.7643 0.5194
0.0088 3.3120 1550 0.0062 0.4058 0.4740 0.4372 0.4060 0.4801 0.4399 75.4846 0.7645 0.5203
0.0082 3.4188 1600 0.0062 0.4032 0.4754 0.4364 0.4033 0.4817 0.4390 75.1861 0.7613 0.5147
0.0084 3.5256 1650 0.0062 0.4052 0.4769 0.4382 0.4054 0.4831 0.4408 75.4375 0.7643 0.5200
0.0084 3.6325 1700 0.0062 0.4063 0.4754 0.4381 0.4064 0.4817 0.4409 75.4375 0.7639 0.5188
0.0079 3.7393 1750 0.0063 0.4083 0.4688 0.4365 0.4086 0.4748 0.4393 75.5317 0.7641 0.5187
0.0081 3.8462 1800 0.0062 0.4045 0.4769 0.4377 0.4046 0.4831 0.4404 75.3433 0.7632 0.5181
0.008 3.9530 1850 0.0062 0.4045 0.4769 0.4377 0.4046 0.4831 0.4404 75.3433 0.7632 0.5181
0.0082 4.0598 1900 0.0063 0.4070 0.4718 0.4370 0.4072 0.4779 0.4397 75.3668 0.7625 0.5156
0.0082 4.1667 1950 0.0062 0.4052 0.4769 0.4382 0.4054 0.4831 0.4408 75.3433 0.7630 0.5174
0.0078 4.2735 2000 0.0062 0.4047 0.4754 0.4372 0.4049 0.4817 0.4399 75.2883 0.7623 0.5161
0.0079 4.3803 2050 0.0061 0.4030 0.4769 0.4368 0.4031 0.4831 0.4395 75.1861 0.7614 0.5153
0.0078 4.4872 2100 0.0061 0.4018 0.4791 0.4371 0.4019 0.4855 0.4397 75.1663 0.7617 0.5166
0.008 4.5940 2150 0.0061 0.4030 0.4769 0.4368 0.4031 0.4831 0.4395 75.2490 0.7624 0.5168
0.0078 4.7009 2200 0.0061 0.4021 0.4776 0.4366 0.4022 0.4840 0.4393 75.2804 0.7635 0.5185
0.0077 4.8077 2250 0.0061 0.4037 0.4769 0.4373 0.4038 0.4831 0.4399 75.3433 0.7634 0.5188
0.0078 4.9145 2300 0.0062 0.4060 0.4769 0.4386 0.4061 0.4831 0.4413 75.3433 0.7627 0.5168
0.0078 5.0214 2350 0.0062 0.4060 0.4769 0.4386 0.4061 0.4831 0.4413 75.4375 0.7640 0.5194
0.0077 5.1282 2400 0.0062 0.4037 0.4769 0.4373 0.4038 0.4831 0.4399 75.2490 0.7622 0.5162
0.0077 5.2350 2450 0.0061 0.4015 0.4769 0.4359 0.4015 0.4831 0.4386 75.1861 0.7621 0.5165
0.0078 5.3419 2500 0.0061 0.4030 0.4769 0.4368 0.4031 0.4831 0.4395 75.2490 0.7624 0.5168
0.0076 5.4487 2550 0.0062 0.4035 0.4784 0.4377 0.4035 0.4846 0.4404 75.2920 0.7628 0.5184
0.0076 5.5556 2600 0.0062 0.4057 0.4784 0.4391 0.4058 0.4846 0.4417 75.3863 0.7632 0.5191
0.0075 5.6624 2650 0.0062 0.4035 0.4784 0.4377 0.4035 0.4846 0.4404 75.2920 0.7628 0.5184
0.0075 5.7692 2700 0.0062 0.4053 0.4806 0.4397 0.4054 0.4869 0.4424 75.3863 0.7636 0.5197
0.0076 5.8761 2750 0.0061 0.4025 0.4798 0.4378 0.4025 0.4862 0.4404 75.2176 0.7626 0.5169
0.0074 5.9829 2800 0.0061 0.4037 0.4813 0.4391 0.4037 0.4877 0.4418 75.2606 0.7626 0.5179
0.0075 6.0897 2850 0.0061 0.4057 0.4784 0.4391 0.4058 0.4846 0.4417 75.3863 0.7632 0.5191
0.0073 6.1966 2900 0.0062 0.4068 0.4806 0.4406 0.4069 0.4869 0.4434 75.4805 0.7642 0.5211
0.0073 6.3034 2950 0.0061 0.4076 0.4806 0.4411 0.4077 0.4869 0.4438 75.4805 0.7638 0.5205
0.0074 6.4103 3000 0.0062 0.4052 0.4769 0.4382 0.4054 0.4831 0.4408 75.3433 0.7627 0.5176
0.0073 6.5171 3050 0.0061 0.4038 0.4806 0.4389 0.4039 0.4869 0.4415 75.2920 0.7628 0.5184
0.0073 6.6239 3100 0.0061 0.4028 0.4820 0.4389 0.4028 0.4886 0.4416 75.1663 0.7615 0.5165
0.0073 6.7308 3150 0.0061 0.4025 0.4798 0.4378 0.4025 0.4862 0.4404 75.2176 0.7626 0.5169
0.0074 6.8376 3200 0.0061 0.4061 0.4806 0.4402 0.4062 0.4869 0.4429 75.3863 0.7632 0.5191
0.0074 6.9444 3250 0.0061 0.4061 0.4806 0.4402 0.4062 0.4869 0.4429 75.3863 0.7632 0.5191
0.0074 7.0513 3300 0.0062 0.4080 0.4784 0.4404 0.4082 0.4846 0.4431 75.4805 0.7636 0.5198
0.0076 7.1581 3350 0.0061 0.4053 0.4806 0.4397 0.4054 0.4869 0.4424 75.2920 0.7622 0.5172
0.0071 7.2650 3400 0.0061 0.4068 0.4806 0.4406 0.4069 0.4869 0.4434 75.3863 0.7628 0.5185
0.0073 7.3718 3450 0.0061 0.4053 0.4806 0.4397 0.4054 0.4869 0.4424 75.2920 0.7622 0.5172
0.0074 7.4786 3500 0.0061 0.4061 0.4806 0.4402 0.4062 0.4869 0.4429 75.3863 0.7632 0.5191
0.0073 7.5855 3550 0.0061 0.4076 0.4806 0.4411 0.4077 0.4869 0.4438 75.4805 0.7638 0.5204
0.0074 7.6923 3600 0.0061 0.4084 0.4806 0.4415 0.4085 0.4869 0.4443 75.5747 0.7648 0.5224
0.0072 7.7991 3650 0.0061 0.4046 0.4806 0.4393 0.4046 0.4869 0.4420 75.2920 0.7626 0.5177
0.0075 7.9060 3700 0.0061 0.4084 0.4806 0.4415 0.4085 0.4869 0.4443 75.5747 0.7648 0.5224
0.0073 8.0128 3750 0.0061 0.4084 0.4806 0.4415 0.4085 0.4869 0.4443 75.5747 0.7648 0.5224
0.0075 8.1197 3800 0.0061 0.4084 0.4806 0.4415 0.4085 0.4869 0.4443 75.5747 0.7648 0.5224
0.0073 8.2265 3850 0.0061 0.4063 0.4791 0.4397 0.4065 0.4855 0.4425 75.4375 0.7638 0.5195
0.0075 8.3333 3900 0.0061 0.4061 0.4806 0.4402 0.4062 0.4869 0.4429 75.3863 0.7632 0.5191
0.0071 8.4402 3950 0.0061 0.4059 0.4813 0.4404 0.4060 0.4877 0.4431 75.3549 0.7629 0.5186
0.0072 8.5470 4000 0.0061 0.4044 0.4813 0.4395 0.4045 0.4877 0.4422 75.2920 0.7624 0.5183
0.0074 8.6538 4050 0.0061 0.4068 0.4806 0.4406 0.4069 0.4869 0.4434 75.4805 0.7642 0.5210
0.0069 8.7607 4100 0.0061 0.4068 0.4806 0.4406 0.4069 0.4869 0.4434 75.4805 0.7642 0.5210
0.0074 8.8675 4150 0.0061 0.4084 0.4806 0.4415 0.4085 0.4869 0.4443 75.5747 0.7648 0.5224
0.0072 8.9744 4200 0.0061 0.4091 0.4806 0.4420 0.4093 0.4869 0.4447 75.6688 0.7659 0.5243
0.0072 9.0812 4250 0.0061 0.4091 0.4806 0.4420 0.4093 0.4869 0.4447 75.6688 0.7659 0.5243
0.0071 9.1880 4300 0.0061 0.4091 0.4806 0.4420 0.4093 0.4869 0.4447 75.6688 0.7659 0.5243
0.0072 9.2949 4350 0.0061 0.4091 0.4806 0.4420 0.4093 0.4869 0.4447 75.6688 0.7659 0.5243
0.0073 9.4017 4400 0.0061 0.4091 0.4806 0.4420 0.4093 0.4869 0.4447 75.6688 0.7659 0.5243
0.0073 9.5085 4450 0.0061 0.4084 0.4806 0.4415 0.4085 0.4869 0.4443 75.5747 0.7648 0.5224
0.0072 9.6154 4500 0.0061 0.4084 0.4806 0.4415 0.4085 0.4869 0.4443 75.5747 0.7648 0.5224
0.0073 9.7222 4550 0.0061 0.4084 0.4806 0.4415 0.4085 0.4869 0.4443 75.5747 0.7648 0.5224
0.0071 9.8291 4600 0.0061 0.4084 0.4806 0.4415 0.4085 0.4869 0.4443 75.5747 0.7648 0.5224
0.0071 9.9359 4650 0.0061 0.4084 0.4806 0.4415 0.4085 0.4869 0.4443 75.5747 0.7648 0.5224

Framework versions

  • Transformers 4.47.0
  • Pytorch 2.5.1+cu121
  • Datasets 3.3.1
  • Tokenizers 0.21.0