tiny-audio

This model is a fine-tuned version of on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.2281

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: 0.0003
• train_batch_size: 6
• eval_batch_size: 6
• seed: 42
• gradient_accumulation_steps: 3
• total_train_batch_size: 18
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.95) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 500
• num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss
0.4431	0.0298	2000	0.3491
0.4701	0.0596	4000	0.3217
0.4086	0.0894	6000	0.3092
0.3937	0.1192	8000	0.2949
0.336	0.1490	10000	0.2896
0.3609	0.1788	12000	0.2827
0.342	0.3128	14000	0.2654
0.3576	0.3575	16000	0.2667
0.3266	0.4022	18000	0.2550
0.2951	0.3352	20000	0.2637
0.3089	0.3687	22000	0.2646
0.2892	0.4022	24000	0.2606
0.3752	0.4357	26000	0.2547
0.2865	0.4692	28000	0.2535
0.327	0.5027	30000	0.2494
0.3438	0.5363	32000	0.2453
0.2843	0.5698	34000	0.2405
0.3015	0.6033	36000	0.2374
0.2904	0.6368	38000	0.2364
0.2946	0.6703	40000	0.2340
0.3428	0.7038	42000	0.2323
0.3036	0.7374	44000	0.2299
0.3381	0.7709	46000	0.2293
0.2993	0.8044	48000	0.2291
0.302	0.8379	50000	0.2282
0.2779	0.8714	52000	0.2280
0.2856	0.9049	54000	0.2281
0.2904	0.9384	56000	0.2280
0.3048	0.9720	58000	0.2281

Framework versions

• Transformers 4.57.3
• Pytorch 2.8.0+cu128
• Datasets 3.6.0
• Tokenizers 0.22.1