wav2vec2-finetuned

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: 0.8859
• Accuracy: 0.6180

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: 0.0001
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 5
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.0181	0.1667	30	0.8603	0.6180
0.8372	0.3333	60	0.8716	0.6180
0.8483	0.5	90	0.8367	0.6180
1.0056	0.6667	120	0.8510	0.6180
0.9038	0.8333	150	0.8823	0.6180
0.8581	1.0	180	0.8741	0.6180
0.9678	1.1667	210	0.8928	0.3371
0.8662	1.3333	240	0.8506	0.6180
0.8186	1.5	270	0.9553	0.3371
0.8974	1.6667	300	0.8572	0.6180
0.9076	1.8333	330	0.8729	0.6180
0.7815	2.0	360	0.9393	0.3371
0.9007	2.1667	390	0.8691	0.6180
0.9566	2.3333	420	0.8597	0.6180
0.7802	2.5	450	0.8787	0.6180
0.8184	2.6667	480	0.9128	0.3371
0.96	2.8333	510	0.8731	0.6180
0.8662	3.0	540	0.8704	0.6180
0.8833	3.1667	570	0.8991	0.3371
0.8922	3.3333	600	0.8710	0.6180
0.7778	3.5	630	0.8623	0.6180
0.8933	3.6667	660	0.8579	0.6180
0.8636	3.8333	690	0.8572	0.6180
0.8913	4.0	720	0.8664	0.6180
0.9244	4.1667	750	0.8756	0.6180
0.8742	4.3333	780	0.8712	0.6180
0.9105	4.5	810	0.8704	0.6180
0.782	4.6667	840	0.8835	0.6180
0.9068	4.8333	870	0.8882	0.6180
0.8155	5.0	900	0.8859	0.6180

Framework versions

• Transformers 4.49.0
• Pytorch 2.6.0+cu124
• Datasets 3.3.2
• Tokenizers 0.21.0