wav2vec2-base-central-vi

minhtien2405/wav2vec2-base-central-vi

wav2vec2-base-central-vi is a automatic speech recognition model from minhtien2405. Use it when you need speech turned into text. It is set up for transformers. The card lists the license as cc-by-nc-4.0.

should probably proofread and complete it, then remove this comment. --

Downloads · 30 days

24% of all-time downloads

All-time downloads

Public

Parameters

94.5M

52.9 GB on disk

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.safetensors378 MB · 100%

At a glance

Task: Automatic Speech Recognition
Library: transformers
License: cc-by-nc-4.0
Model type: wav2vec2
Access: Public
Created: Jun 2, 2025
Updated: Jul 4, 2025
SHA: 181f1640

Base models

nguyenvulebinh/wav2vec2-base-vietnamese-250h

Model card

From the Hugging Face model README

wav2vec2-base-central-vi

This model is a fine-tuned version of nguyenvulebinh/wav2vec2-base-vietnamese-250h on the nguyendv02/ViMD_Dataset dataset. It achieves the following results on the evaluation set:

Loss: 0.4434
Wer: 0.1814

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
train_batch_size: 4
eval_batch_size: 4
seed: 42
gradient_accumulation_steps: 8
total_train_batch_size: 32
optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 20
num_epochs: 30
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.711	0.2719	40	0.4829	0.2439
0.6828	0.5438	80	0.4806	0.2658
0.6459	0.8156	120	0.4724	0.2732
0.6083	1.0816	160	0.4688	0.2412
0.6076	1.3534	200	0.4590	0.2501
0.5846	1.6253	240	0.4713	0.2296
0.5897	1.8972	280	0.4749	0.2302
0.5597	2.1631	320	0.4582	0.2261
0.5417	2.4350	360	0.4773	0.2373
0.5721	2.7069	400	0.4586	0.2320
0.5932	2.9788	440	0.4566	0.2273
0.5097	3.2447	480	0.4738	0.2210
0.5393	3.5166	520	0.4537	0.2317
0.5392	3.7884	560	0.4700	0.2222
0.506	4.0544	600	0.4844	0.2216
0.5163	4.3263	640	0.4556	0.2206
0.5178	4.5981	680	0.4507	0.2241
0.5208	4.8700	720	0.4542	0.2341
0.4426	5.1359	760	0.4574	0.2317
0.4777	5.4078	800	0.4485	0.2209
0.4672	5.6797	840	0.4802	0.2236
0.4765	5.9516	880	0.4677	0.2186
0.4618	6.2175	920	0.4747	0.2074
0.4519	6.4894	960	0.4654	0.2252
0.4574	6.7613	1000	0.4416	0.2149
0.4352	7.0272	1040	0.4580	0.2253
0.4646	7.2991	1080	0.4543	0.2166
0.4483	7.5709	1120	0.4617	0.2144
0.4322	7.8428	1160	0.4406	0.2161
0.4238	8.1088	1200	0.4864	0.2136
0.4091	8.3806	1240	0.4605	0.2162
0.4353	8.6525	1280	0.4496	0.2084
0.4293	8.9244	1320	0.4533	0.2072
0.421	9.1903	1360	0.4566	0.2096
0.3862	9.4622	1400	0.4578	0.2099
0.4184	9.7341	1440	0.4651	0.2081
0.3899	10.0	1480	0.4675	0.2100
0.4179	10.2719	1520	0.4555	0.2187
0.4151	10.5438	1560	0.4436	0.2140
0.3936	10.8156	1600	0.4401	0.2106
0.3876	11.0816	1640	0.4474	0.2103
0.3825	11.3534	1680	0.4729	0.2089
0.3973	11.6253	1720	0.4631	0.2084
0.3938	11.8972	1760	0.4598	0.2089
0.3654	12.1631	1800	0.4645	0.2105
0.3869	12.4350	1840	0.4812	0.2106
0.359	12.7069	1880	0.4711	0.2037
0.3916	12.9788	1920	0.4461	0.2110
0.3519	13.2447	1960	0.4900	0.2141
0.3872	13.5166	2000	0.4708	0.2159
0.3496	13.7884	2040	0.4672	0.2068
0.364	14.0544	2080	0.4944	0.2031
0.3418	14.3263	2120	0.4810	0.2067
0.368	14.5981	2160	0.4908	0.2007
0.3611	14.8700	2200	0.4825	0.2038
0.3442	15.1359	2240	0.4775	0.2098
0.3495	15.4078	2280	0.4957	0.2130
0.3514	15.6797	2320	0.4932	0.2066
0.3644	15.9516	2360	0.4733	0.2101
0.3197	16.2175	2400	0.4809	0.2019
0.3346	16.4894	2440	0.4915	0.2031
0.3716	16.7613	2480	0.5295	0.1991
0.3342	17.0272	2520	0.4961	0.2077
0.329	17.2991	2560	0.4672	0.2004
0.3169	17.5709	2600	0.5015	0.2066
0.3406	17.8428	2640	0.4845	0.2067
0.3039	18.1088	2680	0.4966	0.2008
0.3199	18.3806	2720	0.4911	0.2044
0.32	18.6525	2760	0.5061	0.2043
0.3153	18.9244	2800	0.5010	0.1972
0.2963	19.1903	2840	0.4992	0.2014
0.2979	19.4622	2880	0.4936	0.2039
0.2966	19.7341	2920	0.4999	0.2011
0.3099	20.0	2960	0.5097	0.1962
0.3187	20.2719	3000	0.5179	0.2055
0.308	20.5438	3040	0.4942	0.1942
0.3007	20.8156	3080	0.4960	0.1980
0.3001	21.0816	3120	0.5148	0.2003
0.3125	21.3534	3160	0.4803	0.2037
0.3183	21.6253	3200	0.5069	0.1929
0.2986	21.8972	3240	0.5012	0.1952
0.2656	22.1631	3280	0.5106	0.2005
0.2844	22.4350	3320	0.5098	0.1973
0.2887	22.7069	3360	0.5064	0.1951
0.2868	22.9788	3400	0.5193	0.1936
0.2935	23.2447	3440	0.5166	0.1961
0.2917	23.5166	3480	0.5163	0.1956
0.2822	23.7884	3520	0.5122	0.1951
0.2601	24.0544	3560	0.5235	0.1942
0.2709	24.3263	3600	0.5104	0.1931
0.2611	24.5981	3640	0.5203	0.1957
0.2681	24.8700	3680	0.5045	0.1926
0.2697	25.1359	3720	0.5234	0.1923
0.2597	25.4078	3760	0.5203	0.1939
0.2388	25.6797	3800	0.5169	0.1934
0.2798	25.9516	3840	0.5304	0.1921
0.2686	26.2175	3880	0.5208	0.1915
0.3073	26.4894	3920	0.5178	0.1901
0.2646	26.7613	3960	0.5350	0.1948
0.2358	27.0272	4000	0.5303	0.1932
0.2561	27.2991	4040	0.5240	0.1914
0.2355	27.5709	4080	0.5299	0.1912
0.252	27.8428	4120	0.5364	0.1903
0.225	28.1088	4160	0.5273	0.1914
0.2587	28.3806	4200	0.5330	0.1901
0.242	28.6525	4240	0.5343	0.1930
0.2619	28.9244	4280	0.5225	0.1902
0.2538	29.1903	4320	0.5319	0.1915
0.2295	29.4622	4360	0.5323	0.1898
0.2622	29.7341	4400	0.5281	0.1910
0.2672	30.0	4440	0.5303	0.1907

Framework versions

Transformers 4.53.0
Pytorch 2.7.1+cu126
Datasets 3.6.0
Tokenizers 0.21.2

Task: Automatic Speech Recognition
Library: transformers
Type: wav2vec2
License: cc-by-nc-4.0
Languages: vi
Created: Jun 2, 2025
Updated: Jul 4, 2025