README.md

metadata

license: apache-2.0
tags:
  - generated_from_trainer
model-index:
  - name: wav2vec2-base-timit-demo-google-colab
    results: []

wav2vec2-base-timit-demo-google-colab

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: 0.5182
• Wer: 0.3329

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
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 1000
• num_epochs: 30
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
3.5177	1.0	500	1.8932	0.9837
0.854	2.01	1000	0.5295	0.5266
0.4205	3.01	1500	0.4299	0.4453
0.2934	4.02	2000	0.3940	0.4180
0.2272	5.02	2500	0.4269	0.4149
0.1856	6.02	3000	0.4277	0.4335
0.1668	7.03	3500	0.4214	0.3852
0.1388	8.03	4000	0.4410	0.3805
0.1254	9.04	4500	0.4152	0.3716
0.1073	10.04	5000	0.4257	0.3726
0.1	11.04	5500	0.4405	0.3642
0.0928	12.05	6000	0.4823	0.3708
0.0829	13.05	6500	0.4636	0.3548
0.0682	14.06	7000	0.4718	0.3599
0.0643	15.06	7500	0.4965	0.3583
0.0609	16.06	8000	0.5279	0.3576
0.0586	17.07	8500	0.4869	0.3528
0.055	18.07	9000	0.4671	0.3567
0.0465	19.08	9500	0.5090	0.3508
0.0432	20.08	10000	0.5024	0.3543
0.0427	21.08	10500	0.4658	0.3417
0.033	22.09	11000	0.5276	0.3418
0.0297	23.09	11500	0.5095	0.3415
0.0317	24.1	12000	0.5061	0.3364
0.0262	25.1	12500	0.4910	0.3367
0.0257	26.1	13000	0.4869	0.3331
0.0237	27.11	13500	0.5023	0.3333
0.0228	28.11	14000	0.5131	0.3333
0.021	29.12	14500	0.5182	0.3329

Framework versions

• Transformers 4.17.0
• Pytorch 1.11.0+cu113
• Datasets 1.18.3
• Tokenizers 0.12.1