Hugging Face's logo

haryoaw
/
scenario-KD-PR-MSV-EN-EN-D2_data-en-massive_all_1_166

PyTorch
deberta-v2
Generated from Trainer
Model card Files Community
scenario-KD-PR-MSV-EN-EN-D2_data-en-massive_all_1_166 / README.md
haryoaw's picture
haryoaw
Initial Commit
7735a7e verified
preview code
|
raw
history blame
9.29 kB
metadata
license: mit
base_model: haryoaw/scenario-MDBT-TCR_data-en-massive_all_1_1
tags:
  - generated_from_trainer
datasets:
  - massive
metrics:
  - accuracy
  - f1
model-index:
  - name: scenario-KD-PR-MSV-EN-EN-D2_data-en-massive_all_1_166
    results: []

scenario-KD-PR-MSV-EN-EN-D2_data-en-massive_all_1_166

This model is a fine-tuned version of haryoaw/scenario-MDBT-TCR_data-en-massive_all_1_1 on the massive dataset. It achieves the following results on the evaluation set:

  • Loss: 3.4492
  • Accuracy: 0.3818
  • F1: 0.3581

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: 5e-05
  • train_batch_size: 32
  • eval_batch_size: 32
  • seed: 66
  • 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 Accuracy F1
No log 0.28 100 3.8981 0.2168 0.1114
No log 0.56 200 3.5590 0.3252 0.2326
No log 0.83 300 3.6191 0.3248 0.2629
No log 1.11 400 3.5005 0.3542 0.2804
2.3711 1.39 500 3.4728 0.3569 0.2978
2.3711 1.67 600 3.4235 0.3693 0.3204
2.3711 1.94 700 3.5368 0.3430 0.3074
2.3711 2.22 800 3.5023 0.3635 0.3122
2.3711 2.5 900 3.3525 0.3822 0.3070
1.4454 2.78 1000 3.6046 0.3420 0.3023
1.4454 3.06 1100 3.4244 0.3775 0.3410
1.4454 3.33 1200 3.6073 0.3529 0.3042
1.4454 3.61 1300 3.6553 0.3534 0.3289
1.4454 3.89 1400 3.6924 0.3411 0.3067
1.2035 4.17 1500 3.4352 0.3786 0.3192
1.2035 4.44 1600 3.4389 0.3710 0.3269
1.2035 4.72 1700 3.5966 0.3514 0.3190
1.2035 5.0 1800 3.5810 0.3547 0.3168
1.2035 5.28 1900 3.3785 0.3915 0.3424
1.0761 5.56 2000 3.4477 0.3808 0.3318
1.0761 5.83 2100 3.3579 0.3894 0.3369
1.0761 6.11 2200 3.8225 0.3286 0.3007
1.0761 6.39 2300 3.6119 0.3583 0.3126
1.0761 6.67 2400 3.5649 0.3672 0.3298
1.0182 6.94 2500 3.9692 0.3115 0.3079
1.0182 7.22 2600 3.8932 0.3314 0.3159
1.0182 7.5 2700 3.7041 0.3464 0.3302
1.0182 7.78 2800 3.6965 0.3461 0.3072
1.0182 8.06 2900 3.8919 0.3247 0.3219
0.9596 8.33 3000 3.8834 0.3147 0.3119
0.9596 8.61 3100 3.6113 0.3597 0.3412
0.9596 8.89 3200 3.8647 0.3239 0.3064
0.9596 9.17 3300 3.5087 0.3684 0.3281
0.9596 9.44 3400 3.7126 0.3427 0.3211
0.9317 9.72 3500 3.5315 0.3774 0.3439
0.9317 10.0 3600 3.7289 0.3479 0.3339
0.9317 10.28 3700 3.6699 0.3538 0.3343
0.9317 10.56 3800 3.4747 0.3796 0.3378
0.9317 10.83 3900 3.6562 0.3548 0.3186
0.91 11.11 4000 3.4031 0.3896 0.3478
0.91 11.39 4100 3.5768 0.3557 0.3296
0.91 11.67 4200 3.5617 0.3642 0.3370
0.91 11.94 4300 3.7519 0.3351 0.3163
0.91 12.22 4400 3.5107 0.3755 0.3402
0.892 12.5 4500 3.6299 0.3625 0.3322
0.892 12.78 4600 3.5303 0.3773 0.3394
0.892 13.06 4700 3.6772 0.3513 0.3279
0.892 13.33 4800 3.6287 0.3550 0.3322
0.892 13.61 4900 3.6046 0.3608 0.3363
0.8782 13.89 5000 3.5809 0.3728 0.3401
0.8782 14.17 5100 3.6615 0.3544 0.3327
0.8782 14.44 5200 3.4584 0.3782 0.3471
0.8782 14.72 5300 3.6412 0.3705 0.3456
0.8782 15.0 5400 3.6166 0.3647 0.3482
0.8675 15.28 5500 3.7989 0.3411 0.3259
0.8675 15.56 5600 3.5574 0.3703 0.3332
0.8675 15.83 5700 3.5888 0.3649 0.3332
0.8675 16.11 5800 3.3744 0.3900 0.3450
0.8675 16.39 5900 3.6122 0.3645 0.3442
0.862 16.67 6000 3.3953 0.3876 0.3457
0.862 16.94 6100 3.3995 0.3945 0.3594
0.862 17.22 6200 3.4168 0.3880 0.3463
0.862 17.5 6300 3.6119 0.3668 0.3461
0.862 17.78 6400 3.5063 0.3735 0.3325
0.8544 18.06 6500 3.6581 0.3539 0.3344
0.8544 18.33 6600 3.5380 0.3673 0.3352
0.8544 18.61 6700 3.5699 0.3613 0.3399
0.8544 18.89 6800 3.4977 0.3703 0.3441
0.8544 19.17 6900 3.5746 0.3664 0.3401
0.8494 19.44 7000 3.3279 0.4027 0.3671
0.8494 19.72 7100 3.6689 0.3596 0.3504
0.8494 20.0 7200 3.5632 0.3626 0.3439
0.8494 20.28 7300 3.5577 0.3693 0.3394
0.8494 20.56 7400 3.5795 0.3634 0.3458
0.8452 20.83 7500 3.4764 0.3766 0.3444
0.8452 21.11 7600 3.3944 0.3893 0.3569
0.8452 21.39 7700 3.4161 0.3913 0.3588
0.8452 21.67 7800 3.5015 0.3791 0.3527
0.8452 21.94 7900 3.5177 0.3766 0.3493
0.8413 22.22 8000 3.4390 0.3803 0.3547
0.8413 22.5 8100 3.4736 0.3765 0.3537
0.8413 22.78 8200 3.6093 0.3602 0.3438
0.8413 23.06 8300 3.3350 0.3965 0.3576
0.8413 23.33 8400 3.5186 0.3725 0.3478
0.8393 23.61 8500 3.4701 0.3836 0.3639
0.8393 23.89 8600 3.5562 0.3667 0.3485
0.8393 24.17 8700 3.5092 0.3765 0.3536
0.8393 24.44 8800 3.5955 0.3642 0.3451
0.8393 24.72 8900 3.5135 0.3728 0.3542
0.8363 25.0 9000 3.4253 0.3870 0.3583
0.8363 25.28 9100 3.4027 0.3868 0.3585
0.8363 25.56 9200 3.4741 0.3803 0.3568
0.8363 25.83 9300 3.4929 0.3790 0.3585
0.8363 26.11 9400 3.4803 0.3782 0.3549
0.8328 26.39 9500 3.4915 0.3757 0.3557
0.8328 26.67 9600 3.4388 0.3839 0.3603
0.8328 26.94 9700 3.5595 0.3679 0.3510
0.8328 27.22 9800 3.5496 0.3679 0.3530
0.8328 27.5 9900 3.4764 0.3767 0.3540
0.833 27.78 10000 3.5036 0.3759 0.3553
0.833 28.06 10100 3.5016 0.3769 0.3546
0.833 28.33 10200 3.5109 0.3752 0.3561
0.833 28.61 10300 3.4405 0.3844 0.3602
0.833 28.89 10400 3.4607 0.3835 0.3616
0.8304 29.17 10500 3.4678 0.3791 0.3591
0.8304 29.44 10600 3.4306 0.3858 0.3608
0.8304 29.72 10700 3.4509 0.3818 0.3604
0.8304 30.0 10800 3.4492 0.3818 0.3581

Framework versions

  • Transformers 4.33.3
  • Pytorch 2.1.1+cu121
  • Datasets 2.14.5
  • Tokenizers 0.13.3