Modelpublic

KasuleTrevor/runyankore_speech_to_intent_wav2vec_xlsr_4ctc

sourceHugging Faceupdated 2y agoView on Hugging Face

0likes3downloads

Model Card

runyankorespeechtointentwav2vecxlsr4ctc

This model is a fine-tuned version of KasuleTrevor/wav2vec2-xls-r-300m-nyn_filtered-yogera-v3 on an unknown dataset. It achieves the following results on the evaluation set:

—Loss: 0.1380
—Accuracy: 0.9836
—Precision: 0.9844
—Recall: 0.9836
—F1: 0.9837

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: 0.0001
—trainbatchsize: 32
—evalbatchsize: 8
—seed: 42
—gradientaccumulationsteps: 2
—totaltrainbatch_size: 64
—optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
—lrschedulertype: cosine
—lrschedulerwarmup_steps: 500
—num_epochs: 100
—mixedprecisiontraining: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
No log	1.0	88	2.8252	0.2648	0.1611	0.2648	0.1636
2.9339	2.0	176	1.2106	0.7984	0.7228	0.7984	0.7364
1.8912	3.0	264	0.1829	0.9756	0.9771	0.9756	0.9757
0.3454	4.0	352	0.1177	0.9776	0.9796	0.9776	0.9778
0.1317	5.0	440	0.1024	0.9796	0.9808	0.9796	0.9797
0.1059	6.0	528	0.0994	0.9796	0.9807	0.9796	0.9796
0.0926	7.0	616	0.0834	0.9776	0.9784	0.9776	0.9777
0.0775	8.0	704	0.1048	0.9776	0.9783	0.9776	0.9777
0.0775	9.0	792	0.0947	0.9756	0.9764	0.9756	0.9756
0.0522	10.0	880	0.1098	0.9796	0.9811	0.9796	0.9798
0.0487	11.0	968	0.1159	0.9796	0.9808	0.9796	0.9797
0.035	12.0	1056	0.1116	0.9817	0.9831	0.9817	0.9818
0.0335	13.0	1144	0.1134	0.9796	0.9804	0.9796	0.9797
0.0298	14.0	1232	0.1239	0.9776	0.9790	0.9776	0.9777
0.0248	15.0	1320	0.1186	0.9817	0.9823	0.9817	0.9817
0.0217	16.0	1408	0.1217	0.9776	0.9782	0.9776	0.9776
0.0217	17.0	1496	0.1101	0.9817	0.9823	0.9817	0.9817
0.0293	18.0	1584	0.1174	0.9837	0.9846	0.9837	0.9838
0.0222	19.0	1672	0.1326	0.9796	0.9802	0.9796	0.9797
0.0267	20.0	1760	0.1142	0.9837	0.9850	0.9837	0.9839
0.0217	21.0	1848	0.1166	0.9817	0.9824	0.9817	0.9817
0.0174	22.0	1936	0.1165	0.9796	0.9801	0.9796	0.9796
0.0198	23.0	2024	0.1274	0.9796	0.9806	0.9796	0.9797
0.0142	24.0	2112	0.1351	0.9796	0.9809	0.9796	0.9798
0.0204	25.0	2200	0.1377	0.9776	0.9782	0.9776	0.9776
0.0204	26.0	2288	0.1239	0.9776	0.9783	0.9776	0.9775
0.0173	27.0	2376	0.1565	0.9776	0.9794	0.9776	0.9778
0.0204	28.0	2464	0.1233	0.9817	0.9825	0.9817	0.9818

Framework versions

—Transformers 4.44.1
—Pytorch 2.1.0+cu118
—Datasets 2.21.0
—Tokenizers 0.19.1