Modelpublic

foxxy-hm/wav2vec2-base-finetune-vi-v6

sourceHugging Facecc-by-nc-4.0updated 3y agoView on Hugging Face

0likes24downloads

Model Card

wav2vec2-base-finetune-vi-v6

This model is a fine-tuned version of foxxy-hm/wav2vec2-base-finetune-vi-v6 on the None dataset. It achieves the following results on the evaluation set:

—Loss: 0.2240
—Wer: 0.1296

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: 6e-05
—trainbatchsize: 16
—evalbatchsize: 8
—seed: 42
—optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
—lrschedulertype: linear
—lrschedulerwarmup_steps: 1000
—num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.0848	1.18	500	0.1792	0.1299
0.0881	2.37	1000	0.1978	0.1322
0.0789	3.55	1500	0.2050	0.1330
0.0729	4.74	2000	0.2013	0.1320
0.0611	5.92	2500	0.2034	0.1334
0.0597	7.11	3000	0.2116	0.1326
0.0581	8.29	3500	0.2123	0.1323
0.0481	9.48	4000	0.2134	0.1310
0.0507	10.66	4500	0.2222	0.1314
0.0441	11.85	5000	0.2142	0.1312
0.0427	13.03	5500	0.2270	0.1302
0.0402	14.22	6000	0.2212	0.1298
0.0402	15.4	6500	0.2150	0.1288
0.0381	16.59	7000	0.2240	0.1294
0.039	17.77	7500	0.2224	0.1286
0.0383	18.96	8000	0.2240	0.1296

Framework versions

—Transformers 4.30.2
—Pytorch 2.0.0
—Datasets 2.8.0
—Tokenizers 0.13.3