Hugging Face's logo

Gemmar
/
wav2vec2LugandaASR

Automatic Speech Recognition
Transformers
PyTorch
TensorBoard
wav2vec2
Generated from Trainer
Eval Results
Inference Endpoints
Model card Files Metrics Community
1
wav2vec2LugandaASR / README.md
Gemmar's picture
Gemmar
update model card README.md
0edebf9
preview code
|
raw history blame
5.41 kB
metadata
license: apache-2.0
tags:
  - generated_from_trainer
metrics:
  - wer
model-index:
  - name: wav2vec2LugandaASR
    results: []

wav2vec2LugandaASR

This model is a fine-tuned version of facebook/wav2vec2-large-xlsr-53 on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 0.6798
  • Wer: 0.5291

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: 0.0003
  • train_batch_size: 16
  • eval_batch_size: 8
  • seed: 42
  • gradient_accumulation_steps: 2
  • total_train_batch_size: 32
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_steps: 500
  • num_epochs: 130

Training results

Training Loss Epoch Step Validation Loss Wer
2.8406 1.94 100 2.8577 0.9993
2.7812 3.88 200 2.8315 0.9993
1.1352 5.83 300 1.0099 1.1813
0.5333 7.77 400 0.5782 0.7937
0.3341 9.71 500 0.5899 0.7265
0.2432 11.65 600 0.5352 0.7162
0.2146 13.59 700 0.5439 0.6466
0.1998 15.53 800 0.5865 0.6618
0.1576 17.48 900 0.5598 0.6309
0.1665 19.42 1000 0.5400 0.6135
0.1191 21.36 1100 0.5496 0.6004
0.1038 23.3 1200 0.6248 0.6084
0.104 25.24 1300 0.5517 0.5934
0.1025 27.18 1400 0.5933 0.6008
0.1024 29.13 1500 0.5693 0.5901
0.0935 31.07 1600 0.5842 0.5899
0.0851 33.01 1700 0.6291 0.6086
0.0773 34.95 1800 0.6138 0.5812
0.0873 36.89 1900 0.5944 0.5729
0.0634 38.83 2000 0.6180 0.5807
0.0631 40.78 2100 0.5904 0.5704
0.0709 42.72 2200 0.5855 0.5791
0.0576 44.66 2300 0.6096 0.5789
0.0605 46.6 2400 0.5749 0.5617
0.0795 48.54 2500 0.5974 0.5749
0.0543 50.49 2600 0.6386 0.5754
0.0531 52.43 2700 0.6469 0.5794
0.0554 54.37 2800 0.6340 0.5555
0.0515 56.31 2900 0.6500 0.5762
0.0439 58.25 3000 0.6376 0.5758
0.0461 60.19 3100 0.6265 0.5711
0.0479 62.14 3200 0.6230 0.5707
0.039 64.08 3300 0.6337 0.5584
0.0397 66.02 3400 0.6347 0.5736
0.0509 67.96 3500 0.5946 0.5483
0.0471 69.9 3600 0.6355 0.5584
0.0481 71.84 3700 0.6514 0.5559
0.0484 73.79 3800 0.6373 0.5566
0.041 75.73 3900 0.6736 0.5646
0.0349 77.67 4000 0.6375 0.5622
0.0349 79.61 4100 0.6158 0.5506
0.0273 81.55 4200 0.6914 0.5666
0.029 83.5 4300 0.6361 0.5399
0.0353 85.44 4400 0.6397 0.5584
0.0289 87.38 4500 0.6554 0.5499
0.0257 89.32 4600 0.6676 0.5557
0.0403 91.26 4700 0.6440 0.5584
0.0361 93.2 4800 0.6587 0.5521
0.0304 95.15 4900 0.6837 0.5454
0.0289 97.09 5000 0.6684 0.5370
0.0282 99.03 5100 0.6556 0.5296
0.0302 100.97 5200 0.6833 0.5394
0.0196 102.91 5300 0.6837 0.5291
0.0255 104.85 5400 0.6644 0.5374
0.0209 106.8 5500 0.6700 0.5289
0.0243 108.74 5600 0.6835 0.5338
0.0203 110.68 5700 0.6850 0.5410
0.0237 112.62 5800 0.6561 0.5349
0.0251 114.56 5900 0.6776 0.5298
0.0177 116.5 6000 0.6748 0.5282
0.0232 118.45 6100 0.6767 0.5296
0.0257 120.39 6200 0.6793 0.5320
0.0194 122.33 6300 0.6804 0.5303
0.0304 124.27 6400 0.6798 0.5287
0.0251 126.21 6500 0.6798 0.5291
0.0201 128.16 6600 0.6798 0.5291

Framework versions

  • Transformers 4.30.2
  • Pytorch 2.0.1+cu118
  • Datasets 2.13.0
  • Tokenizers 0.13.3