wav2vec2-large-xlsr-hausa2-demo-colab

This model is a fine-tuned version of facebook/wav2vec2-large-xlsr-53 on the common_voice dataset. It achieves the following results on the evaluation set:

• Loss: 0.2993
• Wer: 0.4826

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: 9.6e-05
• train_batch_size: 12
• eval_batch_size: 8
• seed: 13
• gradient_accumulation_steps: 3
• total_train_batch_size: 36
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 400
• num_epochs: 50
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
6.1549	12.5	400	2.7289	1.0
2.0566	25.0	800	0.4582	0.6768
0.4423	37.5	1200	0.3037	0.5138
0.2991	50.0	1600	0.2993	0.4826

Framework versions

• Transformers 4.16.2
• Pytorch 1.10.0+cu111
• Datasets 1.18.3
• Tokenizers 0.11.0