craa's picture
Upload folder using huggingface_hub
da85ac5 verified
metadata
library_name: transformers
tags:
  - generated_from_trainer
metrics:
  - accuracy
model-index:
  - name: exceptions_exp2_resemble_to_hit_frequency_5039
    results: []

Visualize in Weights & Biases

exceptions_exp2_resemble_to_hit_frequency_5039

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

  • Loss: 3.5570
  • Accuracy: 0.3699

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: 0.0006
  • train_batch_size: 16
  • eval_batch_size: 16
  • seed: 5039
  • gradient_accumulation_steps: 5
  • total_train_batch_size: 80
  • optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.98) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_steps: 100
  • num_epochs: 50.0
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Accuracy Validation Loss
4.8402 0.2914 1000 0.2550 4.7538
4.3353 0.5828 2000 0.3000 4.2827
4.1411 0.8741 3000 0.3160 4.0942
3.994 1.1655 4000 0.3256 3.9890
3.9303 1.4569 5000 0.3321 3.9130
3.8992 1.7483 6000 0.3375 3.8558
3.7487 2.0396 7000 0.3416 3.8128
3.7619 2.3310 8000 0.3448 3.7829
3.74 2.6224 9000 0.3477 3.7515
3.7152 2.9138 10000 0.3499 3.7270
3.6367 3.2051 11000 0.3517 3.7141
3.642 3.4965 12000 0.3535 3.6977
3.6416 3.7879 13000 0.3547 3.6786
3.5382 4.0793 14000 0.3562 3.6716
3.5647 4.3706 15000 0.3577 3.6587
3.5807 4.6620 16000 0.3585 3.6452
3.5693 4.9534 17000 0.3599 3.6329
3.5152 5.2448 18000 0.3607 3.6340
3.5148 5.5361 19000 0.3614 3.6241
3.5384 5.8275 20000 0.3626 3.6128
3.4338 6.1189 21000 0.3628 3.6156
3.4762 6.4103 22000 0.3631 3.6075
3.4815 6.7016 23000 0.3637 3.6009
3.4942 6.9930 24000 0.3649 3.5895
3.4237 7.2844 25000 0.3649 3.5999
3.4501 7.5758 26000 0.3657 3.5900
3.4529 7.8671 27000 0.3663 3.5784
3.3725 8.1585 28000 0.3663 3.5908
3.4032 8.4499 29000 0.3666 3.5825
3.4275 8.7413 30000 0.3674 3.5740
3.3293 9.0326 31000 0.3673 3.5798
3.3803 9.3240 32000 0.3675 3.5790
3.3931 9.6154 33000 0.3682 3.5727
3.4063 9.9068 34000 0.3688 3.5588
3.3379 10.1981 35000 0.3684 3.5747
3.3768 10.4895 36000 0.3689 3.5660
3.3929 10.7809 37000 0.3693 3.5597
3.2905 11.0723 38000 0.3693 3.5673
3.3375 11.3636 39000 0.3695 3.5638
3.352 11.6550 40000 0.3699 3.5570
3.3692 11.9464 41000 0.3707 3.5468
3.3088 12.2378 42000 0.3698 3.5638
3.3395 12.5291 43000 0.3704 3.5564
3.3436 12.8205 44000 0.3708 3.5493
3.2689 13.1119 45000 0.3706 3.5606
3.2973 13.4033 46000 0.3709 3.5571
3.3302 13.6946 47000 0.3710 3.5503
3.3507 13.9860 48000 0.3717 3.5406
3.2781 14.2774 49000 0.3713 3.5558
3.3064 14.5688 50000 0.3715 3.5498
3.318 14.8601 51000 0.3719 3.5419
3.2508 15.1515 52000 0.3715 3.5566
3.2845 15.4429 53000 0.3717 3.5524
3.2991 15.7343 54000 0.3725 3.5421
3.21 16.0256 55000 0.3720 3.5520
3.2507 16.3170 56000 0.3717 3.5529
3.2887 16.6084 57000 0.3727 3.5447
3.2842 16.8998 58000 0.3732 3.5341
3.2203 17.1911 59000 0.3722 3.5538
3.2473 17.4825 60000 0.3724 3.5490
3.2853 17.7739 61000 0.3733 3.5380
3.1926 18.0653 62000 0.3726 3.5528
3.2341 18.3566 63000 0.3730 3.5460
3.2576 18.6480 64000 0.3732 3.5389
3.2692 18.9394 65000 0.3735 3.5341
3.2072 19.2308 66000 0.3730 3.5510
3.2441 19.5221 67000 0.3733 3.5430
3.2584 19.8135 68000 0.3737 3.5368
3.1789 20.1049 69000 0.3730 3.5511
3.2145 20.3963 70000 0.3732 3.5492
3.2446 20.6876 71000 0.3736 3.5399
3.2494 20.9790 72000 0.3745 3.5293
3.1886 21.2704 73000 0.3735 3.5465
3.2211 21.5618 74000 0.3739 3.5401
3.2288 21.8531 75000 0.3745 3.5336
3.1628 22.1445 76000 0.3734 3.5492
3.205 22.4359 77000 0.3739 3.5424
3.2246 22.7273 78000 0.3746 3.5349
3.1374 23.0186 79000 0.3741 3.5473
3.1667 23.3100 80000 0.3739 3.5483
3.1871 23.6014 81000 3.5499 0.3739
3.2005 23.8928 82000 3.5442 0.3742
3.1587 24.1841 83000 3.5505 0.3737
3.1966 24.4755 84000 3.5427 0.3742
3.211 24.7669 85000 3.5338 0.3745
3.1336 25.0583 86000 3.5483 0.3742
3.1568 25.3497 87000 3.5447 0.3744
3.1885 25.6410 88000 3.5370 0.3746
3.2063 25.9324 89000 3.5297 0.3752
3.1457 26.2238 90000 3.5491 0.3741
3.1848 26.5152 91000 3.5441 0.3744
3.1945 26.8065 92000 3.5328 0.3750
3.1137 27.0979 93000 3.5511 0.3744
3.1564 27.3893 94000 3.5458 0.3746
3.1807 27.6807 95000 3.5402 0.3749
3.1905 27.9720 96000 3.5301 0.3753
3.1347 28.2634 97000 3.5479 0.3746
3.1577 28.5548 98000 3.5432 0.3748
3.16 28.8462 99000 3.5354 0.3755
3.1113 29.1375 100000 3.5508 0.3747

Framework versions

  • Transformers 4.55.2
  • Pytorch 2.8.0+cu128
  • Datasets 4.0.0
  • Tokenizers 0.21.4