SER_model_xapiens / README.md
Nugrahasetyaardi's picture
Model save
e59dba6 verified
|
raw
history blame
No virus
7.77 kB
metadata
license: apache-2.0
base_model: facebook/wav2vec2-base
tags:
  - generated_from_trainer
metrics:
  - accuracy
model-index:
  - name: SER_model_xapiens
    results: []

SER_model_xapiens

This model is a fine-tuned version of facebook/wav2vec2-base on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 2.6028
  • Accuracy: 0.2857

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: 3e-05
  • train_batch_size: 64
  • eval_batch_size: 64
  • seed: 42
  • gradient_accumulation_steps: 4
  • total_train_batch_size: 256
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_ratio: 0.1
  • num_epochs: 100

Training results

Training Loss Epoch Step Validation Loss Accuracy
No log 1.0 9 1.6904 0.4050
1.7287 2.0 18 1.6896 0.4050
1.7165 3.0 27 1.7053 0.3932
1.7139 4.0 36 1.6924 0.4063
1.7197 5.0 45 1.6880 0.4089
1.7001 6.0 54 1.7197 0.3879
1.7189 7.0 63 1.6830 0.4142
1.7085 8.0 72 1.6863 0.4155
1.7037 9.0 81 1.6922 0.4024
1.6801 10.0 90 1.7362 0.3906
1.6801 11.0 99 1.6845 0.3997
1.6703 12.0 108 1.7386 0.3840
1.6618 13.0 117 1.7161 0.3644
1.6553 14.0 126 1.7323 0.3893
1.6675 15.0 135 1.7140 0.3984
1.6092 16.0 144 1.7814 0.3499
1.6335 17.0 153 1.7173 0.4115
1.5963 18.0 162 1.7237 0.3945
1.5934 19.0 171 1.7524 0.3893
1.5604 20.0 180 1.7808 0.3696
1.5604 21.0 189 1.7854 0.3696
1.5403 22.0 198 1.7852 0.3604
1.5435 23.0 207 1.8085 0.3578
1.509 24.0 216 1.9569 0.3145
1.4945 25.0 225 1.7921 0.3499
1.474 26.0 234 1.9064 0.3329
1.4117 27.0 243 1.8558 0.3539
1.4322 28.0 252 1.8654 0.3499
1.3891 29.0 261 1.8813 0.3394
1.3702 30.0 270 1.8723 0.3578
1.3702 31.0 279 1.8935 0.3381
1.3541 32.0 288 2.0439 0.2923
1.3233 33.0 297 1.9074 0.3394
1.3015 34.0 306 1.9286 0.3460
1.319 35.0 315 2.0303 0.3119
1.2758 36.0 324 1.9812 0.3408
1.2442 37.0 333 1.9786 0.3460
1.2629 38.0 342 1.9383 0.3565
1.2735 39.0 351 1.9711 0.3617
1.2826 40.0 360 2.0502 0.3303
1.2826 41.0 369 2.0452 0.3342
1.2416 42.0 378 2.0793 0.3342
1.1714 43.0 387 2.1191 0.3250
1.1593 44.0 396 2.0360 0.3342
1.1263 45.0 405 2.0652 0.3526
1.143 46.0 414 2.1089 0.3080
1.0774 47.0 423 2.1723 0.3132
1.0518 48.0 432 2.1094 0.3316
1.0804 49.0 441 2.1944 0.3028
1.0304 50.0 450 2.1495 0.3028
1.0304 51.0 459 2.1812 0.3198
1.01 52.0 468 2.1290 0.3447
0.9972 53.0 477 2.2778 0.3028
0.9883 54.0 486 2.2143 0.3211
0.9725 55.0 495 2.2285 0.3421
0.9637 56.0 504 2.1786 0.3342
0.9545 57.0 513 2.1761 0.3552
0.9102 58.0 522 2.2526 0.3342
0.9037 59.0 531 2.2311 0.3093
0.867 60.0 540 2.3681 0.2896
0.867 61.0 549 2.3130 0.3093
0.8681 62.0 558 2.2619 0.3132
0.8157 63.0 567 2.2726 0.3080
0.8166 64.0 576 2.2933 0.3263
0.8126 65.0 585 2.2754 0.3381
0.8223 66.0 594 2.3631 0.3014
0.7593 67.0 603 2.3403 0.3080
0.774 68.0 612 2.3941 0.3001
0.7347 69.0 621 2.3866 0.2896
0.7397 70.0 630 2.4285 0.2896
0.7397 71.0 639 2.3924 0.2936
0.7556 72.0 648 2.3677 0.3067
0.7019 73.0 657 2.4436 0.2844
0.7186 74.0 666 2.4368 0.2883
0.686 75.0 675 2.4293 0.2870
0.6895 76.0 684 2.4057 0.2936
0.6719 77.0 693 2.4716 0.3001
0.6507 78.0 702 2.4922 0.2936
0.6741 79.0 711 2.4093 0.3132
0.6265 80.0 720 2.5002 0.3145
0.6265 81.0 729 2.5818 0.2661
0.617 82.0 738 2.4953 0.2883
0.6441 83.0 747 2.5105 0.2923
0.5921 84.0 756 2.5884 0.2792
0.6288 85.0 765 2.5165 0.2962
0.5871 86.0 774 2.5537 0.3067
0.5974 87.0 783 2.5740 0.2805
0.5942 88.0 792 2.5240 0.2936
0.5782 89.0 801 2.6014 0.2910
0.5717 90.0 810 2.5036 0.3054
0.5717 91.0 819 2.5913 0.2831
0.5761 92.0 828 2.5776 0.2910
0.5672 93.0 837 2.6038 0.2844
0.535 94.0 846 2.5620 0.3014
0.5591 95.0 855 2.6176 0.2936
0.558 96.0 864 2.5990 0.2988
0.5319 97.0 873 2.5890 0.2936
0.5266 98.0 882 2.6031 0.2844
0.5337 99.0 891 2.6090 0.2805
0.5252 100.0 900 2.6028 0.2857

Framework versions

  • Transformers 4.42.0.dev0
  • Pytorch 2.3.0
  • Datasets 2.19.2.dev0
  • Tokenizers 0.19.1