Hugging Face's logo

AhBotNLP
/
wav2vec2-xls-r-164m-id

Automatic Speech Recognition
Transformers
PyTorch
wav2vec2
Generated from Trainer
Eval Results
Inference Endpoints
Model card Files Community
1
wav2vec2-xls-r-164m-id / README.md
evanarlian's picture
evanarlian
update model card README.md
fdafed2
preview code
|
raw
history blame
4.37 kB
metadata
tags:
  - generated_from_trainer
datasets:
  - evanarlian/common_voice_11_0_id_filtered
metrics:
  - wer
model-index:
  - name: wav2vec2-xls-r-164m-id
    results:
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: evanarlian/common_voice_11_0_id_filtered
          type: evanarlian/common_voice_11_0_id_filtered
        metrics:
          - name: Wer
            type: wer
            value: 0.3199428097039019

wav2vec2-xls-r-164m-id

This model is a fine-tuned version of evanarlian/distil-wav2vec2-xls-r-164m-id on the evanarlian/common_voice_11_0_id_filtered dataset. It achieves the following results on the evaluation set:

  • Loss: 0.3215
  • Wer: 0.3199

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: 0.0002
  • train_batch_size: 24
  • eval_batch_size: 24
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_ratio: 0.3
  • num_epochs: 40.0
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Wer
3.5445 0.92 1000 3.0106 1.0000
2.5067 1.84 2000 1.6134 0.9905
1.0279 2.75 3000 0.7667 0.8217
0.7823 3.67 4000 0.6141 0.7224
0.6504 4.59 5000 0.5228 0.6503
0.5687 5.51 6000 0.4666 0.5963
0.5026 6.43 7000 0.4288 0.5612
0.4584 7.35 8000 0.4048 0.5267
0.4193 8.26 9000 0.4057 0.5218
0.3931 9.18 10000 0.3820 0.4813
0.3651 10.1 11000 0.3686 0.4709
0.3526 11.02 12000 0.3665 0.4655
0.3333 11.94 13000 0.3440 0.4485
0.3095 12.86 14000 0.3314 0.4331
0.2802 13.77 15000 0.3360 0.4157
0.2724 14.69 16000 0.3331 0.4107
0.2488 15.61 17000 0.3255 0.4037
0.231 16.53 18000 0.3089 0.3950
0.2146 17.45 19000 0.3398 0.3990
0.2103 18.37 20000 0.3080 0.3805
0.2035 19.28 21000 0.3158 0.3828
0.1933 20.2 22000 0.3118 0.3728
0.1839 21.12 23000 0.3076 0.3690
0.1791 22.04 24000 0.3041 0.3658
0.1696 22.96 25000 0.3092 0.3603
0.1608 23.88 26000 0.2936 0.3555
0.1568 24.79 27000 0.2936 0.3560
0.1456 25.71 28000 0.3257 0.3543
0.1399 26.63 29000 0.3100 0.3424
0.1345 27.55 30000 0.3172 0.3472
0.1264 28.47 31000 0.3276 0.3412
0.1289 29.38 32000 0.3104 0.3401
0.1246 30.3 33000 0.3204 0.3352
0.1156 31.22 34000 0.3013 0.3353
0.1143 32.14 35000 0.3102 0.3322
0.1152 33.06 36000 0.3240 0.3323
0.1093 33.98 37000 0.3105 0.3295
0.101 34.89 38000 0.3112 0.3263
0.1017 35.81 39000 0.3263 0.3239
0.0915 36.73 40000 0.3176 0.3226
0.0943 37.65 41000 0.3141 0.3210
0.0898 38.57 42000 0.3177 0.3183
0.0923 39.49 43000 0.3215 0.3199

Framework versions

  • Transformers 4.27.0.dev0
  • Pytorch 1.13.1+cu117
  • Datasets 2.9.1.dev0
  • Tokenizers 0.13.2