jialicheng/unlearn_speech_commands_hubert-base_scrub_2_42

superb_ks_42

This model is a fine-tuned version of facebook/hubert-base-ls960 on the superb dataset. It achieves the following results on the evaluation set:

• Loss: 84.4856
• Accuracy: 0.6214
• Test Accuracy: 0.6214
• Df Accuracy: 0.2380
• Unlearn Overall Accuracy: 0.6917
• Unlearn Time: 1183.4350

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: 5e-05
• train_batch_size: 32
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Overall Accuracy	Unlearn Overall Accuracy	Time
0.0	1.0	1565	84.4856	0.2380	0.6917	0.6917	-1
0.0	2.0	3130	363.7596	0.3800	0.6257	0.6257	-1
0.0	3.0	4695	714.0270	0.4006	0.6171	0.6171	-1
0.0	4.0	6260	971.4188	0.2635	0.6789	0.6789	-1
0.0	5.0	7825	1100.3213	0.2449	0.6880	0.6880	-1

Framework versions

• Transformers 4.39.3
• Pytorch 2.2.2+cu118
• Datasets 2.18.0
• Tokenizers 0.15.2