CoolFace
Modelpublic

simpragma/breeze-listen-w2v2-ml

sourceHugging Facecc-by-nc-4.0updated 3y agoView on Hugging Face
0likes24downloads
Model Card

<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->

breeze-listen-w2v2-ml

This model is a fine-tuned version of facebook/mms-1b-all on the MOZILLA-FOUNDATION/COMMONVOICE16_0 - ML dataset. It achieves the following results on the evaluation set:

  • —Loss: 0.2666
  • —Wer: 0.5349

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: 0.001
  • —trainbatchsize: 4
  • —evalbatchsize: 8
  • —seed: 42
  • —distributed_type: multi-GPU
  • —optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • —lrschedulertype: linear
  • —lrschedulerwarmup_steps: 100
  • —num_epochs: 4.0
  • —mixedprecisiontraining: Native AMP

Training results

Training LossEpochStepValidation LossWer
No log0.412005.47281.0757
No log0.814005.12741.0038
6.50371.226000.61670.8131
6.50371.638000.32840.5829
1.04822.0310000.31690.5667
1.04822.4412000.28760.5425
1.04822.8514000.28470.5522
0.43143.2516000.27460.5394
0.43143.6618000.26980.5346

Framework versions

  • —Transformers 4.38.0.dev0
  • —Pytorch 2.1.2+cu121
  • —Datasets 2.16.1
  • —Tokenizers 0.15.1