You can deploy this model and start earning money today!

Lguyogiro/
wav2vec2-large-mms-1b-NhiFromAzz-adapterft-ilv_fold1

Name: wav2vec2-large-mms-1b-NhiFromAzz-adapterft-ilv_fold1
Availability: InStock

safetensors

wav2vec2

generated_from_trainer

audiofolder

Lguyogiro/wav2vec2-large-mms-1b-nhi-adapterft-ilv_fold1

finetune

cc-by-nc-4.0

model-index

Views

No views yet

wav2vec2-large-mms-1b-NhiFromAzz-adapterft-ilv_fold1

This model is a fine-tuned version of Lguyogiro/wav2vec2-large-mms-1b-nhi-adapterft-ilv_fold1 on the audiofolder dataset. It achieves the following results on the evaluation set:

Loss: 0.6692
Wer: 0.3726
Cer: 0.1093

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: 0.001
train_batch_size: 20
eval_batch_size: 32
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 100
num_epochs: 100
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer	Cer
0.3928	1.6529	200	0.5688	0.4100	0.1202
0.3481	3.3058	400	0.5916	0.4081	0.1202
0.3464	4.9587	600	0.6038	0.4085	0.1196
0.3308	6.6116	800	0.5927	0.4180	0.1192
0.342	8.2645	1000	0.5926	0.4169	0.1208
0.336	9.9174	1200	0.6184	0.4016	0.1194
0.3286	11.5702	1400	0.6048	0.4104	0.1187
0.3291	13.2231	1600	0.6044	0.4081	0.1195
0.3103	14.8760	1800	0.5968	0.4104	0.1215
0.3065	16.5289	2000	0.5934	0.4177	0.1228
0.3032	18.1818	2200	0.5990	0.4096	0.1196
0.285	19.8347	2400	0.6101	0.4261	0.1217
0.2938	21.4876	2600	0.5793	0.3997	0.1154
0.2847	23.1405	2800	0.5926	0.4031	0.1191
0.283	24.7934	3000	0.5834	0.4104	0.1181
0.2692	26.4463	3200	0.6295	0.3997	0.1171
0.2678	28.0992	3400	0.5965	0.3993	0.1164
0.2724	29.7521	3600	0.5989	0.4001	0.1181
0.2603	31.4050	3800	0.6084	0.4070	0.1183
0.2686	33.0579	4000	0.6216	0.3978	0.1166
0.2428	34.7107	4200	0.5974	0.4005	0.1153
0.2555	36.3636	4400	0.5941	0.3982	0.1156
0.2519	38.0165	4600	0.6128	0.3982	0.1148
0.2344	39.6694	4800	0.6149	0.4008	0.1155
0.2299	41.3223	5000	0.6254	0.3974	0.1169
0.2279	42.9752	5200	0.6141	0.3989	0.1168
0.2306	44.6281	5400	0.6064	0.4031	0.1192
0.2128	46.2810	5600	0.6116	0.3821	0.1115
0.2176	47.9339	5800	0.6182	0.3997	0.1161
0.2185	49.5868	6000	0.6332	0.3791	0.1132
0.2061	51.2397	6200	0.6390	0.3848	0.1140
0.2201	52.8926	6400	0.6378	0.3905	0.1141
0.2029	54.5455	6600	0.6268	0.3932	0.1160
0.2028	56.1983	6800	0.6330	0.3878	0.1131
0.1922	57.8512	7000	0.6453	0.3856	0.1152
0.2042	59.5041	7200	0.6445	0.3848	0.1138
0.184	61.1570	7400	0.6315	0.3867	0.1131
0.1889	62.8099	7600	0.6593	0.3821	0.1113
0.19	64.4628	7800	0.6429	0.3905	0.1142
0.173	66.1157	8000	0.6404	0.3894	0.1130
0.1769	67.7686	8200	0.6519	0.3810	0.1124
0.1768	69.4215	8400	0.6278	0.3794	0.1113
0.1949	71.0744	8600	0.6536	0.3783	0.1118
0.1786	72.7273	8800	0.6473	0.3829	0.1138
0.1706	74.3802	9000	0.6658	0.3791	0.1111
0.168	76.0331	9200	0.6315	0.3840	0.1126
0.1617	77.6860	9400	0.6617	0.3840	0.1123
0.1544	79.3388	9600	0.6411	0.3779	0.1103
0.154	80.9917	9800	0.6612	0.3737	0.1096
0.1543	82.6446	10000	0.6523	0.3890	0.1126
0.1502	84.2975	10200	0.6636	0.3814	0.1110
0.1625	85.9504	10400	0.6518	0.3806	0.1102
0.1511	87.6033	10600	0.6600	0.3787	0.1095
0.146	89.2562	10800	0.6615	0.3802	0.1106
0.1579	90.9091	11000	0.6622	0.3779	0.1097
0.1536	92.5620	11200	0.6680	0.3791	0.1103
0.1579	94.2149	11400	0.6685	0.3794	0.1096
0.1487	95.8678	11600	0.6719	0.3745	0.1092
0.1482	97.5207	11800	0.6702	0.3756	0.1099
0.1513	99.1736	12000	0.6692	0.3726	0.1093

Framework versions

Transformers 4.41.2
Pytorch 2.4.0
Datasets 2.19.1
Tokenizers 0.19.1

Lguyogiro/wav2vec2-large-mms-1b-NhiFromAzz-adapterft-ilv_fold1

wav2vec2-large-mms-1b-NhiFromAzz-adapterft-ilv_fold1

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

Lguyogiro/
wav2vec2-large-mms-1b-NhiFromAzz-adapterft-ilv_fold1