my_14april_model

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

• Loss: 0.5186
• Accuracy: 0.8929

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: 2
• total_train_batch_size: 8
• optimizer: Use OptimizerNames.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: 35

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.6945	1.0	14	0.6906	0.75
0.6406	2.0	28	0.6579	0.75
0.5394	3.0	42	0.5893	0.7143
0.459	4.0	56	0.5432	0.75
0.3677	5.0	70	0.4503	0.7857
0.5435	6.0	84	0.3750	0.8929
0.4321	7.0	98	0.4493	0.7857
0.2022	8.0	112	0.3426	0.8929
0.2559	9.0	126	0.3465	0.8571
0.2626	10.0	140	0.3648	0.8571
0.288	11.0	154	0.4958	0.8571
0.2906	12.0	168	0.5387	0.8571
0.1413	13.0	182	0.4079	0.8929
0.395	14.0	196	0.4337	0.8929
0.2057	15.0	210	0.4214	0.8929
0.1814	16.0	224	0.5683	0.8571
0.1653	17.0	238	0.4052	0.8929
0.2064	18.0	252	0.4316	0.8929
0.1906	19.0	266	0.5650	0.8571
0.1084	20.0	280	0.6157	0.8571
0.0137	21.0	294	0.6055	0.8571
0.0668	22.0	308	0.4714	0.8929
0.0164	23.0	322	0.4272	0.8929
0.0139	24.0	336	0.4854	0.8929
0.0076	25.0	350	0.4991	0.8929
0.0081	26.0	364	0.5013	0.8929
0.0073	27.0	378	0.4924	0.8929
0.0071	28.0	392	0.4923	0.8929
0.0067	29.0	406	0.5015	0.8929
0.0059	30.0	420	0.5091	0.8929
0.0064	31.0	434	0.5164	0.8929
0.0063	32.0	448	0.5183	0.8929
0.0063	32.5185	455	0.5186	0.8929

Framework versions

• Transformers 4.50.3
• Pytorch 2.6.0+cu124
• Datasets 3.5.0
• Tokenizers 0.21.1