hubert-custom

This model is a fine-tuned version of facebook/hubert-base-ls960 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.6927
• Accuracy: 0.5156

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
• 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
• num_epochs: 30
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.6835	2.7778	100	0.6809	0.5625
0.6998	5.5556	200	0.6928	0.5156
0.711	8.3333	300	0.7050	0.4844
0.6965	11.1111	400	0.6932	0.4844
0.6994	13.8889	500	0.7030	0.4844
0.7002	16.6667	600	0.6927	0.5156
0.6982	19.4444	700	0.6935	0.4844
0.6981	22.2222	800	0.6928	0.5156
0.6942	25.0	900	0.6923	0.5156
0.6951	27.7778	1000	0.6927	0.5156

Framework versions

• Transformers 4.51.1
• Pytorch 2.6.0+cu124
• Datasets 3.5.0
• Tokenizers 0.21.1