Cdial/Hausa_xlsr

This model is a fine-tuned version of facebook/wav2vec2-xls-r-300m It achieves the following results on the evaluation set (which is 10 percent of train data set merged with invalidated data, reported, other, and dev datasets):

Loss: 0.275118
Wer: 0.329955

Model description

"facebook/wav2vec2-xls-r-300m" was finetuned.

Intended uses & limitations

More information needed

Training and evaluation data

Training data - Common voice Hausa train.tsv, dev.tsv, invalidated.tsv, reported.tsv and other.tsv Only those points were considered where upvotes were greater than downvotes and duplicates were removed after concatenation of all the datasets given in common voice 7.0

Training procedure

For creating the training dataset, all possible datasets were appended and 90-10 split was used.

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 0.000096
train_batch_size: 16
eval_batch_size: 16
seed: 13
gradient_accumulation_steps: 2
lr_scheduler_type: cosine_with_restarts
lr_scheduler_warmup_steps: 500
num_epochs: 50
mixed_precision_training: Native AMP

Training results

Step	Training Loss	Validation Loss	Wer
500	5.175900	2.750914	1.000000
1000	1.028700	0.338649	0.497999
1500	0.332200	0.246896	0.402241
2000	0.227300	0.239640	0.395839
2500	0.175000	0.239577	0.373966
3000	0.140400	0.243272	0.356095
3500	0.119200	0.263761	0.365164
4000	0.099300	0.265954	0.353428
4500	0.084400	0.276367	0.349693
5000	0.073700	0.282631	0.343825
5500	0.068000	0.282344	0.341158
6000	0.064500	0.281591	0.342491

Framework versions

Transformers 4.16.0.dev0
Pytorch 1.10.0+cu102
Datasets 1.18.3
Tokenizers 0.10.3

Evaluation Commands

To evaluate on mozilla-foundation/common_voice_8_0 with split test

python eval.py --model_id Akashpb13/Hausa_xlsr --dataset mozilla-foundation/common_voice_8_0 --config ha --split test

Cdial
/

hausa-asr