Edit model card

finetuning-wav2vec-large-swahili-asr-model_v10

This model is a fine-tuned version of Joshua-Abok/finetuned_wav2vec_asr on the common_voice_12_0 dataset. It achieves the following results on the evaluation set:

  • eval_loss: 0.3257
  • eval_wer: 0.1481
  • eval_runtime: 611.0298
  • eval_samples_per_second: 17.966
  • eval_steps_per_second: 2.247
  • epoch: 14.79
  • step: 20000

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: 15
  • mixed_precision_training: Native AMP

Framework versions

  • Transformers 4.36.0.dev0
  • Pytorch 2.0.1+cu118
  • Datasets 2.14.6
  • Tokenizers 0.14.1
Downloads last month
4
Safetensors
Model size
315M params
Tensor type
F32
·

Finetuned from