Auralis/NatHACKS_Auralis

wav2vec2-large-xls-r-300m-dm32

This model is a fine-tuned version of facebook/wav2vec2-xls-r-300m on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.4880
• Accuracy: 0.7917

Model description

More information needed

Intended uses & limitations

Used for detecting Alzheimer's disease given voice samples

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: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 16
• 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: 22
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	2.3448	34	0.6847	0.5833
No log	4.6897	68	0.6828	0.5833
No log	7.0345	102	0.6775	0.5833
0.3495	9.3793	136	0.6757	0.5833
0.3495	11.7241	170	0.6739	0.5833
0.3495	14.0690	204	0.6081	0.6875
0.3335	16.4138	238	0.5084	0.7917
0.3335	18.7586	272	0.4868	0.8125
0.3335	21.1034	306	0.4880	0.7917

Framework versions

• Transformers 4.46.2
• Pytorch 2.5.1+cu121
• Datasets 3.1.0
• Tokenizers 0.20.3