hubert-base-superb-er-3kfoldfull30finetuned-bmd-20250824_095232-LOSO-section-out1

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

• Loss: 0.7662
• Accuracy: 0.625
• F1: 0.5017

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: 8
• eval_batch_size: 8
• seed: 1968
• gradient_accumulation_steps: 4
• total_train_batch_size: 32
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED 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: 30
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
No log	1.0	3	1.1018	0.3333	0.1720
No log	2.0	6	1.0994	0.3333	0.1667
No log	3.0	9	1.0954	0.3333	0.1667
1.0992	4.0	12	1.0900	0.4583	0.3913
1.0992	5.0	15	1.0839	0.5417	0.4938
1.0992	6.0	18	1.0747	0.5833	0.5352
1.0849	7.0	21	1.0583	0.625	0.6029
1.0849	8.0	24	1.0298	0.5417	0.5222
1.0849	9.0	27	0.9999	0.5417	0.4444
1.0315	10.0	30	0.9593	0.625	0.5944
1.0315	11.0	33	0.9270	0.6667	0.6291
1.0315	12.0	36	0.9139	0.625	0.5560
1.0315	13.0	39	0.9164	0.5833	0.4722
0.9493	14.0	42	0.8959	0.5833	0.4722
0.9493	15.0	45	0.8589	0.625	0.5705
0.9493	16.0	48	0.8279	0.625	0.5017
0.8856	17.0	51	0.8905	0.5833	0.4722
0.8856	18.0	54	0.9607	0.5	0.3956
0.8856	19.0	57	0.8622	0.5833	0.4722
0.8357	20.0	60	0.7882	0.625	0.5017
0.8357	21.0	63	0.7648	0.7917	0.7778
0.8357	22.0	66	0.7725	0.625	0.5017
0.8357	23.0	69	0.7825	0.625	0.5017
0.7695	24.0	72	0.7875	0.6667	0.5910
0.7695	25.0	75	0.7869	0.6667	0.5910
0.7695	26.0	78	0.7837	0.6667	0.5910
0.7523	27.0	81	0.7673	0.625	0.5017
0.7523	28.0	84	0.7661	0.625	0.5017
0.7523	29.0	87	0.7668	0.625	0.5017
0.7329	30.0	90	0.7662	0.625	0.5017

Framework versions

• Transformers 4.55.2
• Pytorch 2.8.0+cu126
• Datasets 3.6.0
• Tokenizers 0.21.4