wav2vec2-base-Malawi-small

This model is a fine-tuned version of facebook/wav2vec2-base-960h on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.8095
• Wer: 0.4636
• Cer: 0.1534

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: 6
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 24
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• training_steps: 900
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer	Cer
7.3684	14.6341	300	3.2704	1.0	1.0
1.9223	29.2683	600	0.8459	0.5406	0.1849
0.2949	43.9024	900	0.8095	0.4636	0.1534

Framework versions

• Transformers 4.44.2
• Pytorch 2.5.0+cu121
• Datasets 1.18.3
• Tokenizers 0.19.1