jialicheng/unlearn_speech_commands_wav2vec2-large_scrub_4_42

superb_ks_42

This model is a fine-tuned version of facebook/wav2vec2-large on the superb dataset. It achieves the following results on the evaluation set:

• Loss: 393.8668
• Accuracy: 0.6209
• Test Accuracy: 0.6209
• Df Accuracy: 0.1334
• Unlearn Overall Accuracy: 0.7437
• Unlearn Time: 1625.0307

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: 2.5

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Overall Accuracy	Unlearn Overall Accuracy	Time
0.0	1.0	1534	393.8668	0.1334	0.7437	0.7437	-1
0.0	2.0	3068	1040.8676	0.1334	0.7437	0.7437	-1
0.0	2.5	3835	1152.0482	0.1334	0.7437	0.7437	-1

Framework versions

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