wav2vec2-base-music_genre_classifier-g3b

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: 1.3709
• Accuracy: 0.7380
• F1: 0.7356
• Recall: 0.7395
• Precision: 0.7400

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: 3e-05
• train_batch_size: 12
• eval_batch_size: 12
• 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
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 15
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Recall	Precision
2.3444	1.0	276	2.2888	0.3618	0.2663	0.3495	0.2702
1.946	2.0	552	1.7679	0.4880	0.4277	0.4778	0.5072
1.6394	3.0	828	1.4655	0.5565	0.4966	0.5463	0.5089
1.2346	4.0	1104	1.3279	0.5974	0.5654	0.5937	0.6372
0.8945	5.0	1380	1.2718	0.6226	0.6021	0.6178	0.6240
0.7872	6.0	1656	1.1310	0.6671	0.6594	0.6691	0.6826
0.5562	7.0	1932	1.1743	0.6743	0.6677	0.6730	0.6857
0.65	8.0	2208	1.0722	0.7163	0.7178	0.7179	0.7394
0.3239	9.0	2484	1.1846	0.6899	0.6863	0.6909	0.6997
0.3885	10.0	2760	1.2243	0.7031	0.6994	0.7072	0.7126
0.1529	11.0	3036	1.2539	0.7175	0.7193	0.7195	0.7245
0.4527	12.0	3312	1.3231	0.7188	0.7116	0.7182	0.7220
0.324	13.0	3588	1.3190	0.7344	0.7360	0.7368	0.7409
0.0277	14.0	3864	1.3623	0.7356	0.7340	0.7370	0.7407
0.0276	15.0	4140	1.3709	0.7380	0.7356	0.7395	0.7400

Framework versions

• Transformers 4.46.2
• Pytorch 2.5.0+cu121
• Datasets 3.1.0
• Tokenizers 0.20.3