Modelpublic

tadgeis/xls-r-1b-ckt-staged-bible-radio-to-chuklang-source-eval-stage2

sourceHugging Faceapache-2.0updated 4mo agoView on Hugging Face

0likes11downloads

Model Card

xls-r-1b-ckt-staged-bible-radio-to-chuklang-source-eval-stage2

This model is a fine-tuned version of facebook/wav2vec2-xls-r-1b.

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: 0.0003
—trainbatchsize: 4
—evalbatchsize: 4
—seed: 42
—optimizer: Use OptimizerNames.ADAMWTORCHFUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
—lrschedulertype: linear
—num_epochs: 8
—mixedprecisiontraining: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer	Cer
0.5914	0.1515	25	1.5204	0.9413	0.3508
1.1904	0.3030	50	1.3849	0.9556	0.3428
1.1642	0.4545	75	1.4699	0.9365	0.3512
1.5551	0.6061	100	1.3179	0.9429	0.3378
1.3518	0.7576	125	1.3624	0.9397	0.3658
1.2616	0.9091	150	1.2519	0.9508	0.3375
1.2567	1.0606	175	1.3471	0.9254	0.3060
0.9603	1.2121	200	1.2992	0.9460	0.3283
1.2213	1.3636	225	1.3064	0.9429	0.3130
1.1614	1.5152	250	1.3027	0.9365	0.3349
1.0144	1.6667	275	1.2166	0.9175	0.3020
1.2407	1.8182	300	1.1266	0.9143	0.2903
0.9871	1.9697	325	1.3958	0.9365	0.3311
0.8599	2.1212	350	1.2405	0.9095	0.2927
0.9353	2.2727	375	1.1928	0.9032	0.3049
0.7067	2.4242	400	1.2234	0.8841	0.2877
0.9908	2.5758	425	1.1693	0.8841	0.2755
0.814	2.7273	450	1.1659	0.9143	0.2928
0.7993	2.8788	475	1.1450	0.9	0.2846
0.9854	3.0303	500	1.0528	0.8841	0.2696
0.5801	3.1818	525	1.0923	0.8683	0.2605
0.7924	3.3333	550	1.0705	0.8571	0.2557
0.6309	3.4848	575	1.1535	0.8905	0.2841
0.6079	3.6364	600	1.1247	0.8683	0.2579
0.6872	3.7879	625	1.1277	0.8841	0.2687
0.5814	3.9394	650	1.1812	0.9032	0.2700
0.7309	4.0909	675	1.2485	0.8905	0.2603
0.5441	4.2424	700	1.2010	0.9302	0.2744
0.4768	4.3939	725	1.1432	0.8698	0.2612

Framework versions

—Transformers 4.57.1
—Pytorch 2.11.0+cu130
—Datasets 4.8.5
—Tokenizers 0.22.2