jialicheng/unlearn_speech_commands_wav2vec2-base_bad_teaching_2_42

superb_ks_42

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

• Loss: 1051591.375
• Accuracy: 0.6209
• Test Accuracy: 0.6209
• Df Accuracy: 0.1352
• Unlearn Overall Accuracy: 0.7429
• Unlearn Time: 11585.2870

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: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Overall Accuracy	Unlearn Overall Accuracy	Time
0.0	0.5	1564	2.0251	0.5642	0.6213	0.6213	-1
0.0	1.0	3128	24.9043	0.1352	0.7429	0.7429	-1
0.0	1.5	4692	118.8172	0.1342	0.7431	0.7431	-1
0.0	2.0	6256	360.7070	0.1342	0.7404	0.7404	-1
0.0	2.5	7820	1797.0188	0.1352	0.7429	0.7429	-1
0.0	3.0	9384	1833.0411	0.1342	0.7407	0.7407	-1
0.0	3.5	10948	5951.4312	0.1352	0.7391	0.7391	-1
0.0	4.0	12512	23488.5137	0.1352	0.7429	0.7429	-1
0.0	4.5	14076	43524.4297	0.1352	0.7429	0.7429	-1
0.0	5.0	15640	61965.0234	0.1352	0.7429	0.7429	-1
0.0	5.5	17204	83421.5312	0.1352	0.7429	0.7429	-1
0.0	6.0	18768	108193.9766	0.1352	0.7429	0.7429	-1
0.0	6.5	20332	136248.25	0.1352	0.7429	0.7429	-1
0.0	7.0	21896	167441.5938	0.1352	0.7429	0.7429	-1
0.0	7.5	23460	201594.4688	0.1352	0.7429	0.7429	-1
0.0	7.99	25024	238488.5	0.1352	0.7429	0.7429	-1
0.0	8.49	26588	277883.0625	0.1352	0.7429	0.7429	-1
0.0	8.99	28152	319393.5625	0.1352	0.7429	0.7429	-1
0.0	9.49	29716	362743.3438	0.1352	0.7429	0.7429	-1
0.0	9.99	31280	407556.3438	0.1352	0.7429	0.7429	-1
0.0	10.49	32844	453439.1875	0.1352	0.7429	0.7429	-1
0.0	10.99	34408	500049.6562	0.1352	0.7429	0.7429	-1
0.0	11.49	35972	546989.6875	0.1352	0.7429	0.7429	-1
0.0	11.99	37536	593870.5625	0.1352	0.7429	0.7429	-1
0.0	12.49	39100	640257.0	0.1352	0.7429	0.7429	-1
0.0	12.99	40664	685795.3125	0.1352	0.7429	0.7429	-1
0.0	13.49	42228	730008.5	0.1352	0.7429	0.7429	-1
0.0	13.99	43792	772677.1875	0.1352	0.7429	0.7429	-1
0.0	14.49	45356	813358.625	0.1352	0.7429	0.7429	-1
0.0	14.99	46920	851736.9375	0.1352	0.7429	0.7429	-1
0.0	15.49	48484	887515.625	0.1352	0.7429	0.7429	-1
0.0	15.99	50048	920330.625	0.1352	0.7429	0.7429	-1
0.0	16.49	51612	949979.0	0.1352	0.7429	0.7429	-1
0.0	16.99	53176	976159.25	0.1352	0.7429	0.7429	-1
0.0	17.49	54740	998714.6875	0.1352	0.7429	0.7429	-1
0.0	17.99	56304	1017460.9375	0.1352	0.7429	0.7429	-1
0.0	18.49	57868	1032206.9375	0.1352	0.7429	0.7429	-1
0.0	18.99	59432	1042871.625	0.1352	0.7429	0.7429	-1
0.0	19.49	60996	1049350.375	0.1352	0.7429	0.7429	-1
0.0	19.99	62560	1051591.375	0.1352	0.7429	0.7429	-1

Framework versions

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