fydhfzh's picture
End of training
d038e92 verified
|
raw
history blame
16.7 kB
metadata
license: apache-2.0
base_model: facebook/hubert-base-ls960
tags:
  - generated_from_trainer
metrics:
  - accuracy
  - precision
  - recall
  - f1
model-index:
  - name: hubert-classifier-aug-fold-2
    results: []

hubert-classifier-aug-fold-2

This model is a fine-tuned version of facebook/hubert-base-ls960 on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 0.7004
  • Accuracy: 0.8895
  • Precision: 0.9016
  • Recall: 0.8895
  • F1: 0.8901
  • Binary: 0.9228

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: 0.0001
  • train_batch_size: 32
  • eval_batch_size: 32
  • seed: 42
  • gradient_accumulation_steps: 4
  • total_train_batch_size: 128
  • 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 Accuracy Precision Recall F1 Binary
No log 0.13 50 3.8323 0.0877 0.0522 0.0877 0.0397 0.3479
No log 0.27 100 3.2624 0.1808 0.1370 0.1808 0.1221 0.4188
No log 0.4 150 2.8729 0.2955 0.2056 0.2955 0.2056 0.5031
No log 0.54 200 2.3928 0.4157 0.3408 0.4157 0.3318 0.5904
No log 0.67 250 2.0739 0.4548 0.4212 0.4548 0.3875 0.6186
No log 0.81 300 1.7722 0.5547 0.5132 0.5547 0.4923 0.6887
No log 0.94 350 1.5220 0.5951 0.5844 0.5951 0.5489 0.7157
2.8578 1.08 400 1.4062 0.6154 0.5903 0.6154 0.5690 0.7301
2.8578 1.21 450 1.1712 0.7126 0.7203 0.7126 0.6873 0.7978
2.8578 1.35 500 1.0957 0.6815 0.6716 0.6815 0.6441 0.7764
2.8578 1.48 550 1.0095 0.7395 0.7600 0.7395 0.7227 0.8165
2.8578 1.62 600 0.9283 0.7571 0.7774 0.7571 0.7454 0.8301
2.8578 1.75 650 0.9876 0.7544 0.7818 0.7544 0.7451 0.8270
2.8578 1.89 700 0.7728 0.8016 0.8095 0.8016 0.7934 0.8601
1.2835 2.02 750 0.8472 0.7827 0.7946 0.7827 0.7714 0.8457
1.2835 2.16 800 0.7331 0.7989 0.8183 0.7989 0.7962 0.8575
1.2835 2.29 850 0.8126 0.7814 0.7986 0.7814 0.7745 0.8441
1.2835 2.43 900 0.7898 0.7814 0.8120 0.7814 0.7749 0.8463
1.2835 2.56 950 0.7014 0.8111 0.8311 0.8111 0.8082 0.8671
1.2835 2.7 1000 0.6225 0.8300 0.8532 0.8300 0.8277 0.8808
1.2835 2.83 1050 0.7096 0.8178 0.8356 0.8178 0.8155 0.8715
1.2835 2.96 1100 0.6304 0.8232 0.8458 0.8232 0.8202 0.8773
0.8295 3.1 1150 0.5950 0.8435 0.8601 0.8435 0.8428 0.8907
0.8295 3.23 1200 0.6140 0.8421 0.8579 0.8421 0.8410 0.8892
0.8295 3.37 1250 0.6443 0.8327 0.8578 0.8327 0.8293 0.8826
0.8295 3.5 1300 0.6662 0.8205 0.8456 0.8205 0.8158 0.8737
0.8295 3.64 1350 0.6056 0.8502 0.8638 0.8502 0.8483 0.8949
0.8295 3.77 1400 0.5968 0.8448 0.8591 0.8448 0.8421 0.8907
0.8295 3.91 1450 0.5734 0.8354 0.8542 0.8354 0.8329 0.8845
0.6277 4.04 1500 0.6580 0.8340 0.8521 0.8340 0.8331 0.8830
0.6277 4.18 1550 0.6149 0.8529 0.8664 0.8529 0.8523 0.8981
0.6277 4.31 1600 0.5965 0.8556 0.8715 0.8556 0.8536 0.8978
0.6277 4.45 1650 0.5801 0.8570 0.8749 0.8570 0.8554 0.8996
0.6277 4.58 1700 0.6019 0.8516 0.8679 0.8516 0.8502 0.8960
0.6277 4.72 1750 0.6178 0.8502 0.8643 0.8502 0.8488 0.8937
0.6277 4.85 1800 0.5726 0.8637 0.8790 0.8637 0.8627 0.9059
0.6277 4.99 1850 0.5581 0.8596 0.8751 0.8596 0.8577 0.9020
0.5166 5.12 1900 0.6064 0.8448 0.8649 0.8448 0.8424 0.8907
0.5166 5.26 1950 0.6728 0.8516 0.8728 0.8516 0.8502 0.8958
0.5166 5.39 2000 0.5952 0.8650 0.8818 0.8650 0.8652 0.9067
0.5166 5.53 2050 0.4922 0.8704 0.8900 0.8704 0.8700 0.9100
0.5166 5.66 2100 0.5558 0.8812 0.8954 0.8812 0.8817 0.9165
0.5166 5.8 2150 0.6257 0.8596 0.8778 0.8596 0.8602 0.9030
0.5166 5.93 2200 0.5901 0.8650 0.8839 0.8650 0.8654 0.9054
0.4409 6.06 2250 0.5639 0.8650 0.8765 0.8650 0.8642 0.9050
0.4409 6.2 2300 0.5967 0.8610 0.8793 0.8610 0.8582 0.9022
0.4409 6.33 2350 0.5664 0.8704 0.8856 0.8704 0.8703 0.9086
0.4409 6.47 2400 0.5706 0.8745 0.8885 0.8745 0.8742 0.9119
0.4409 6.6 2450 0.5945 0.8637 0.8768 0.8637 0.8623 0.9039
0.4409 6.74 2500 0.6792 0.8556 0.8722 0.8556 0.8526 0.8973
0.4409 6.87 2550 0.6265 0.8623 0.8788 0.8623 0.8612 0.9038
0.3941 7.01 2600 0.5768 0.8691 0.8845 0.8691 0.8682 0.9090
0.3941 7.14 2650 0.5951 0.8610 0.8797 0.8610 0.8588 0.9039
0.3941 7.28 2700 0.6621 0.8596 0.8728 0.8596 0.8570 0.9008
0.3941 7.41 2750 0.5764 0.8745 0.8876 0.8745 0.8745 0.9128
0.3941 7.55 2800 0.6080 0.8677 0.8830 0.8677 0.8669 0.9067
0.3941 7.68 2850 0.6498 0.8691 0.8831 0.8691 0.8675 0.9086
0.3941 7.82 2900 0.6737 0.8475 0.8641 0.8475 0.8446 0.8928
0.3941 7.95 2950 0.7467 0.8462 0.8669 0.8462 0.8434 0.8926
0.3567 8.09 3000 0.5592 0.8745 0.8897 0.8745 0.8744 0.9117
0.3567 8.22 3050 0.5933 0.8772 0.8913 0.8772 0.8764 0.9128
0.3567 8.36 3100 0.5294 0.8826 0.8931 0.8826 0.8797 0.9171
0.3567 8.49 3150 0.6415 0.8664 0.8803 0.8664 0.8647 0.9062
0.3567 8.63 3200 0.6076 0.8704 0.8862 0.8704 0.8694 0.9085
0.3567 8.76 3250 0.5787 0.8812 0.8963 0.8812 0.8802 0.9171
0.3567 8.89 3300 0.5419 0.8799 0.8909 0.8799 0.8789 0.9161
0.3207 9.03 3350 0.5635 0.8731 0.8837 0.8731 0.8702 0.9108
0.3207 9.16 3400 0.5488 0.8839 0.8959 0.8839 0.8826 0.9182
0.3207 9.3 3450 0.5245 0.8839 0.8974 0.8839 0.8834 0.9185
0.3207 9.43 3500 0.6777 0.8637 0.8780 0.8637 0.8615 0.9038
0.3207 9.57 3550 0.6236 0.8704 0.8888 0.8704 0.8678 0.9081
0.3207 9.7 3600 0.6140 0.8718 0.8865 0.8718 0.8714 0.9111
0.3207 9.84 3650 0.6249 0.8623 0.8718 0.8623 0.8593 0.9040
0.3207 9.97 3700 0.5656 0.8772 0.8874 0.8772 0.8757 0.9138
0.3047 10.11 3750 0.6042 0.8731 0.8821 0.8731 0.8709 0.9109
0.3047 10.24 3800 0.5685 0.8826 0.8921 0.8826 0.8823 0.9170
0.3047 10.38 3850 0.6586 0.8758 0.8885 0.8758 0.8742 0.9135
0.3047 10.51 3900 0.6546 0.8758 0.8877 0.8758 0.8743 0.9124
0.3047 10.65 3950 0.6802 0.8677 0.8796 0.8677 0.8652 0.9076
0.3047 10.78 4000 0.6282 0.8799 0.8937 0.8799 0.8785 0.9166
0.3047 10.92 4050 0.6671 0.8677 0.8830 0.8677 0.8663 0.9072
0.2817 11.05 4100 0.5854 0.8812 0.8957 0.8812 0.8797 0.9166
0.2817 11.19 4150 0.6261 0.8758 0.8887 0.8758 0.8740 0.9132
0.2817 11.32 4200 0.6103 0.8799 0.8949 0.8799 0.8790 0.9165
0.2817 11.46 4250 0.5799 0.8799 0.8893 0.8799 0.8781 0.9161
0.2817 11.59 4300 0.5591 0.8866 0.8985 0.8866 0.8865 0.9212
0.2817 11.73 4350 0.5359 0.8893 0.9010 0.8893 0.8892 0.9231
0.2817 11.86 4400 0.6664 0.8677 0.8811 0.8677 0.8674 0.9076
0.2817 11.99 4450 0.6034 0.8799 0.8923 0.8799 0.8799 0.9159
0.2736 12.13 4500 0.6436 0.8745 0.8873 0.8745 0.8722 0.9113
0.2736 12.26 4550 0.6724 0.8799 0.8963 0.8799 0.8792 0.9161
0.2736 12.4 4600 0.5840 0.8893 0.9005 0.8893 0.8886 0.9227
0.2736 12.53 4650 0.6570 0.8785 0.8918 0.8785 0.8779 0.9151
0.2736 12.67 4700 0.6322 0.8745 0.8877 0.8745 0.8737 0.9119
0.2736 12.8 4750 0.6748 0.8880 0.9002 0.8880 0.8878 0.9212
0.2736 12.94 4800 0.7166 0.8718 0.8864 0.8718 0.8695 0.9105
0.2541 13.07 4850 0.5717 0.8866 0.9001 0.8866 0.8858 0.9198
0.2541 13.21 4900 0.6211 0.8745 0.8910 0.8745 0.8735 0.9123
0.2541 13.34 4950 0.5923 0.8799 0.8975 0.8799 0.8805 0.9151
0.2541 13.48 5000 0.5885 0.8758 0.8891 0.8758 0.8759 0.9132
0.2541 13.61 5050 0.6245 0.8866 0.8998 0.8866 0.8862 0.9192
0.2541 13.75 5100 0.6897 0.8718 0.8866 0.8718 0.8706 0.9090
0.2541 13.88 5150 0.6919 0.8677 0.8807 0.8677 0.8677 0.9076
0.2384 14.02 5200 0.5996 0.8961 0.9079 0.8961 0.8951 0.9269
0.2384 14.15 5250 0.6649 0.8880 0.9013 0.8880 0.8861 0.9208
0.2384 14.29 5300 0.7136 0.8664 0.8854 0.8664 0.8625 0.9061
0.2384 14.42 5350 0.6670 0.8812 0.8970 0.8812 0.8812 0.9158
0.2384 14.56 5400 0.6286 0.8826 0.8952 0.8826 0.8820 0.9175
0.2384 14.69 5450 0.6297 0.8785 0.8877 0.8785 0.8755 0.9151
0.2384 14.82 5500 0.7010 0.8799 0.8953 0.8799 0.8795 0.9159
0.2384 14.96 5550 0.6078 0.8853 0.8985 0.8853 0.8822 0.9197
0.2218 15.09 5600 0.6684 0.8758 0.8917 0.8758 0.8751 0.9127
0.2218 15.23 5650 0.6672 0.8799 0.8917 0.8799 0.8774 0.9157
0.2218 15.36 5700 0.6440 0.8839 0.8998 0.8839 0.8828 0.9189
0.2218 15.5 5750 0.6807 0.8866 0.9002 0.8866 0.8863 0.9204
0.2218 15.63 5800 0.6325 0.8839 0.8949 0.8839 0.8831 0.9184
0.2218 15.77 5850 0.6078 0.8934 0.9046 0.8934 0.8918 0.9250
0.2218 15.9 5900 0.6638 0.8866 0.9005 0.8866 0.8860 0.9202
0.2192 16.04 5950 0.5822 0.8920 0.9044 0.8920 0.8910 0.9240
0.2192 16.17 6000 0.6028 0.8785 0.8922 0.8785 0.8765 0.9138
0.2192 16.31 6050 0.6012 0.8893 0.9013 0.8893 0.8884 0.9227
0.2192 16.44 6100 0.5819 0.8853 0.8980 0.8853 0.8838 0.9193
0.2192 16.58 6150 0.6055 0.8826 0.8998 0.8826 0.8818 0.9170
0.2192 16.71 6200 0.6642 0.9001 0.9123 0.9001 0.8991 0.9297
0.2192 16.85 6250 0.6235 0.8880 0.8976 0.8880 0.8857 0.9219
0.2192 16.98 6300 0.5460 0.8920 0.9007 0.8920 0.8905 0.9240
0.2103 17.12 6350 0.5525 0.8920 0.9031 0.8920 0.8908 0.9242
0.2103 17.25 6400 0.5847 0.8974 0.9092 0.8974 0.8960 0.9279
0.2103 17.39 6450 0.5585 0.8961 0.9081 0.8961 0.8958 0.9260
0.2103 17.52 6500 0.5424 0.8920 0.9008 0.8920 0.8911 0.9247
0.2103 17.65 6550 0.5473 0.9042 0.9141 0.9042 0.9032 0.9331
0.2103 17.79 6600 0.5548 0.9001 0.9081 0.9001 0.8990 0.9308
0.2103 17.92 6650 0.6355 0.8866 0.8983 0.8866 0.8839 0.9194
0.1962 18.06 6700 0.5878 0.9015 0.9120 0.9015 0.8983 0.9306
0.1962 18.19 6750 0.6067 0.8907 0.9015 0.8907 0.8890 0.9231
0.1962 18.33 6800 0.5797 0.8880 0.8989 0.8880 0.8863 0.9212
0.1962 18.46 6850 0.5842 0.8907 0.9020 0.8907 0.8894 0.9236
0.1962 18.6 6900 0.5838 0.8961 0.9078 0.8961 0.8947 0.9260
0.1962 18.73 6950 0.5655 0.8880 0.9021 0.8880 0.8876 0.9212
0.1962 18.87 7000 0.5601 0.8988 0.9088 0.8988 0.8977 0.9287
0.1881 19.0 7050 0.5815 0.8988 0.9077 0.8988 0.8972 0.9283
0.1881 19.14 7100 0.6380 0.8947 0.9054 0.8947 0.8932 0.9259
0.1881 19.27 7150 0.6770 0.8907 0.9004 0.8907 0.8892 0.9227
0.1881 19.41 7200 0.6608 0.8907 0.8997 0.8907 0.8890 0.9227
0.1881 19.54 7250 0.7075 0.8826 0.8974 0.8826 0.8815 0.9174
0.1881 19.68 7300 0.6649 0.8853 0.8992 0.8853 0.8827 0.9193
0.1881 19.81 7350 0.6430 0.8880 0.8983 0.8880 0.8868 0.9212

Framework versions

  • Transformers 4.38.2
  • Pytorch 2.3.0
  • Datasets 2.19.1
  • Tokenizers 0.15.1