wav2vec2-base-finetuned-ic-slurp

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

• Loss: 2.1101
• Accuracy: 0.7393

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: 24
• eval_batch_size: 24
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 96
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
4.0345	1.0	527	3.9813	0.0673
3.5622	2.0	1055	3.4634	0.1867
2.7737	3.0	1582	2.7252	0.3638
2.1285	4.0	2110	2.1754	0.4827
1.6216	5.0	2637	1.8169	0.5701
1.1786	6.0	3165	1.5773	0.6347
0.8747	7.0	3692	1.5024	0.6568
0.7565	8.0	4220	1.5020	0.6694
0.5236	9.0	4747	1.5287	0.6799
0.4517	10.0	5275	1.5165	0.6879
0.364	11.0	5802	1.5159	0.6949
0.3221	12.0	6330	1.5217	0.6996
0.227	13.0	6857	1.5718	0.7075
0.1828	14.0	7385	1.6979	0.6901
0.1691	15.0	7912	1.6162	0.7093
0.1642	16.0	8440	1.6973	0.7048
0.1254	17.0	8967	1.7060	0.7100
0.1578	18.0	9495	1.7328	0.7063
0.1509	19.0	10022	1.7658	0.7073
0.1409	20.0	10550	1.7770	0.7052
0.1085	21.0	11077	1.8033	0.7074
0.106	22.0	11605	1.7000	0.7149
0.0764	23.0	12132	1.7943	0.7104
0.0671	24.0	12660	1.8323	0.7155
0.0768	25.0	13187	1.8486	0.7146
0.0741	26.0	13715	1.8227	0.7187
0.0731	27.0	14242	1.7824	0.7230
0.0935	28.0	14770	1.8987	0.7164
0.0829	29.0	15297	1.8774	0.7202
0.0588	30.0	15825	1.8820	0.7211
0.059	31.0	16352	1.9535	0.7246
0.0431	32.0	16880	1.9621	0.7237
0.0324	33.0	17407	2.0160	0.7256
0.0447	34.0	17935	1.9392	0.7262
0.025	35.0	18462	2.0095	0.7284
0.0522	36.0	18990	1.9994	0.7244
0.0482	37.0	19517	2.0566	0.7262
0.0203	38.0	20045	2.0287	0.7295
0.0221	39.0	20572	2.0634	0.7300
0.0444	40.0	21100	2.0593	0.7302
0.0348	41.0	21627	2.0712	0.7298
0.0154	42.0	22155	2.0429	0.7351
0.024	43.0	22682	2.0708	0.7352
0.0157	44.0	23210	2.0701	0.7368
0.0222	45.0	23737	2.0963	0.7338
0.0126	46.0	24265	2.1329	0.7340
0.0211	47.0	24792	2.1230	0.7370
0.0288	48.0	25320	2.1101	0.7393
0.0347	49.0	25847	2.1201	0.7375
0.0162	49.95	26350	2.1197	0.7381

Framework versions

• Transformers 4.37.2
• Pytorch 2.2.0+cu121
• Datasets 2.17.1
• Tokenizers 0.15.2