Edit model card

WavLM-large-CORAA-pt

This model is a fine-tuned version of microsoft/wavlm-large on CORAA dataset. It achieves the following results on the evaluation set:

  • Loss: 0.6144
  • Wer: 0.3840

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: 8
  • eval_batch_size: 8
  • seed: 42
  • gradient_accumulation_steps: 2
  • total_train_batch_size: 16
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_steps: 1000
  • training_steps: 40000
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Wer
No log 0.04 1000 1.9230 0.9960
5.153 0.08 2000 1.3733 0.8444
5.153 0.13 3000 1.1992 0.7362
1.367 0.17 4000 1.1289 0.6957
1.367 0.21 5000 1.0357 0.6470
1.1824 0.25 6000 1.0216 0.6201
1.1824 0.29 7000 0.9338 0.6036
1.097 0.33 8000 0.9149 0.5760
1.097 0.38 9000 0.8885 0.5541
1.0254 0.42 10000 0.8678 0.5366
1.0254 0.46 11000 0.8349 0.5323
0.9782 0.5 12000 0.8230 0.5155
0.9782 0.54 13000 0.8245 0.5049
0.9448 0.59 14000 0.7802 0.4990
0.9448 0.63 15000 0.7650 0.4900
0.9092 0.67 16000 0.7665 0.4796
0.9092 0.71 17000 0.7568 0.4795
0.8764 0.75 18000 0.7403 0.4615
0.8764 0.8 19000 0.7219 0.4644
0.8498 0.84 20000 0.7180 0.4502
0.8498 0.88 21000 0.7017 0.4436
0.8278 0.92 22000 0.6992 0.4395
0.8278 0.96 23000 0.7021 0.4329
0.8077 1.0 24000 0.6892 0.4265
0.8077 1.05 25000 0.6940 0.4248
0.7486 1.09 26000 0.6767 0.4202
0.7486 1.13 27000 0.6734 0.4150
0.7459 1.17 28000 0.6650 0.4152
0.7459 1.21 29000 0.6559 0.4078
0.7304 1.26 30000 0.6536 0.4088
0.7304 1.3 31000 0.6537 0.4025
0.7183 1.34 32000 0.6462 0.4008
0.7183 1.38 33000 0.6381 0.3973
0.7059 1.42 34000 0.6266 0.3930
0.7059 1.46 35000 0.6280 0.3921
0.6983 1.51 36000 0.6248 0.3897
0.6983 1.55 37000 0.6275 0.3872
0.6892 1.59 38000 0.6199 0.3852
0.6892 1.63 39000 0.6180 0.3842
0.691 1.67 40000 0.6144 0.3840

Framework versions

  • Transformers 4.16.0.dev0
  • Pytorch 1.10.1+cu102
  • Datasets 1.17.1.dev0
  • Tokenizers 0.11.0
Downloads last month
4