muk-luganda-digits-classification-clean

This model is a fine-tuned version of facebook/wav2vec2-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 2.1182
• Accuracy: 0.4074

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: 1e-05
• train_batch_size: 4
• eval_batch_size: 1
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.3019	1.0	54	2.3080	0.0741
2.3011	2.0	108	2.3017	0.1852
2.2894	3.0	162	2.3069	0.0370
2.2761	4.0	216	2.3072	0.1481
2.2574	5.0	270	2.2921	0.1111
2.2345	6.0	324	2.2809	0.1481
2.2061	7.0	378	2.2674	0.2593
2.1754	8.0	432	2.2683	0.2593
2.1463	9.0	486	2.2374	0.1852
2.1	10.0	540	2.2400	0.2593
2.0551	11.0	594	2.2009	0.2963
2.0238	12.0	648	2.1784	0.2593
1.9852	13.0	702	2.1676	0.2963
1.9571	14.0	756	2.1834	0.2593
1.9168	15.0	810	2.1502	0.3333
1.8981	16.0	864	2.1212	0.3704
1.8546	17.0	918	2.1315	0.4074
1.8475	18.0	972	2.1182	0.4074
1.8346	19.0	1026	2.1224	0.4074
1.8144	20.0	1080	2.1182	0.4074

Framework versions

• Transformers 4.56.2
• Pytorch 2.8.0+cu128
• Datasets 3.6.0
• Tokenizers 0.22.1