IDAT_aug_red_696_Wav2Vec

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

• Loss: 0.5564
• Accuracy: 0.7667

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: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 16
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 40

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.6862	0.97	17	0.6526	0.675
0.609	2.0	35	0.6533	0.6917
0.5659	2.97	52	0.7129	0.6583
0.6223	4.0	70	0.5106	0.7833
0.5163	4.97	87	0.7914	0.7
0.6306	6.0	105	0.5960	0.75
0.5651	6.97	122	0.7364	0.5583
0.7081	8.0	140	0.7070	0.5
0.699	8.97	157	0.6951	0.5
0.6958	10.0	175	0.6934	0.5
0.6978	10.97	192	0.6916	0.5
0.6786	12.0	210	0.6940	0.5
0.6981	12.97	227	0.6932	0.5
0.6931	14.0	245	0.6933	0.5
0.6939	14.97	262	0.6931	0.5
0.6935	16.0	280	0.6931	0.5
0.6972	16.97	297	0.6931	0.5
0.6935	18.0	315	0.6931	0.5
0.6931	18.97	332	0.6951	0.5
0.6955	20.0	350	0.6921	0.7
0.6992	20.97	367	0.6843	0.5
0.6785	22.0	385	0.6919	0.5
0.6661	22.97	402	0.6317	0.6417
0.597	24.0	420	0.5620	0.75
0.5849	24.97	437	0.6103	0.75
0.5955	26.0	455	0.6245	0.725
0.4777	26.97	472	0.5215	0.7833
0.4726	28.0	490	0.5657	0.775
0.4487	28.97	507	0.5306	0.7833
0.4478	30.0	525	0.6591	0.7333
0.5039	30.97	542	0.5304	0.7833
0.5173	32.0	560	0.7111	0.7
0.5266	32.97	577	0.5587	0.7667
0.4677	34.0	595	0.5515	0.7667
0.4775	34.97	612	0.5116	0.7917
0.4651	36.0	630	0.5450	0.7667
0.4738	36.97	647	0.5371	0.7833
0.4862	38.0	665	0.5497	0.7667
0.4695	38.86	680	0.5564	0.7667

Framework versions

• Transformers 4.32.1
• Pytorch 2.1.2
• Datasets 2.16.1
• Tokenizers 0.13.3