ast-finetuned-audioset-10-10-0.4593_ft_env_0-12

This model is a fine-tuned version of MIT/ast-finetuned-audioset-10-10-0.4593 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.3804
• Accuracy: 0.9643
• Precision: 0.9702
• Recall: 0.9643
• F1: 0.9643

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: 1.5e-06
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 8
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 56
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
2.0371	1.0	28	1.9267	0.1429	0.3214	0.1429	0.1482
1.7315	2.0	56	1.5823	0.3214	0.3667	0.3214	0.2973
1.3081	3.0	84	1.2250	0.75	0.8423	0.75	0.7499
0.9664	4.0	112	0.9526	0.8214	0.8616	0.8214	0.8078
0.6607	5.0	140	0.7525	0.8571	0.8795	0.8571	0.8520
0.5239	6.0	168	0.6080	0.8929	0.9152	0.8929	0.8866
0.453	7.0	196	0.5089	0.9286	0.9286	0.9286	0.9286
0.323	8.0	224	0.4353	0.9286	0.9286	0.9286	0.9286
0.296	9.0	252	0.3804	0.9643	0.9702	0.9643	0.9643
0.2167	10.0	280	0.3382	0.9643	0.9702	0.9643	0.9643
0.186	11.0	308	0.3157	0.9643	0.9702	0.9643	0.9643
0.1748	12.0	336	0.2931	0.9643	0.9702	0.9643	0.9643
0.1367	13.0	364	0.2781	0.9643	0.9702	0.9643	0.9643
0.1469	14.0	392	0.2705	0.9643	0.9702	0.9643	0.9643
0.1308	15.0	420	0.2679	0.9643	0.9702	0.9643	0.9643

Framework versions

• Transformers 4.27.4
• Pytorch 2.0.0
• Datasets 2.10.1
• Tokenizers 0.11.0