s2t-small-uit-vimd_20250514_144603

QuanHoangNgoc/s2t-small-uit-vimd_20250514_144603

s2t-small-uit-vimd_20250514_144603 is a automatic speech recognition model from QuanHoangNgoc. Use it when you need speech turned into text. It is set up for transformers.

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

Downloads · 30 days

14% of all-time downloads

All-time downloads

Public

Parameters

32.6M

5.6 GB on disk

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.safetensors131 MB · 99%

At a glance

Task: Automatic Speech Recognition
Library: transformers
Model type: speech_to_text
Access: Public
Created: May 14, 2025
Updated: May 14, 2025
SHA: 6cff3637

Base models

s2t-small-librispeech-asr

Datasets

uit-vimd

Model card

From the Hugging Face model README

s2t-small-uit-vimd-finetuned

This model is a fine-tuned version of s2t-small-librispeech-asr on the UIT-ViMD dataset. It achieves the following results on the evaluation set:

Loss: 4.1774
Wer: 156.0641

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: 3e-05
train_batch_size: 16
eval_batch_size: 16
seed: 42
optimizer: Use OptimizerNames.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: 500
training_steps: 4000
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
8.5697	0.0990	93	6.4030	100.9153
6.4339	0.1981	186	6.2123	267.7346
6.2304	0.2971	279	5.9593	183.6766
5.9361	0.3962	372	5.4394	119.9085
5.611	0.4952	465	5.0782	217.6201
5.4075	0.5942	558	5.0541	183.9054
5.2848	0.6933	651	4.9875	102.9748
5.1533	0.7923	744	4.8326	175.1335
5.0337	0.8914	837	4.9110	210.6026
4.9704	0.9904	930	4.7458	243.0206
4.8977	1.0895	1023	5.0089	153.5469
4.8621	1.1885	1116	4.4668	159.4966
4.7784	1.2875	1209	4.7943	234.4775
4.7841	1.3866	1302	4.4658	249.8856
4.7533	1.4856	1395	4.5277	249.9619
4.7271	1.5847	1488	4.7147	167.1243
4.7165	1.6837	1581	4.4671	181.6934
4.7034	1.7827	1674	4.6343	174.5995
4.6697	1.8818	1767	4.6031	181.0831
4.6511	1.9808	1860	4.5789	210.7551
4.582	2.0799	1953	4.4299	202.5934
4.5487	2.1789	2046	4.5597	158.7338
4.5324	2.2780	2139	4.1907	184.3631
4.5213	2.3770	2232	4.3382	187.4905
4.5139	2.4760	2325	4.1682	203.5088
4.4873	2.5751	2418	4.8307	153.6995
4.4675	2.6741	2511	4.3182	180.5492
4.5047	2.7732	2604	4.1314	186.3463
4.4808	2.8722	2697	4.0971	184.8207
4.4626	2.9712	2790	4.2918	187.9481
4.4058	3.0703	2883	4.2169	190.5416
4.3554	3.1693	2976	4.2198	163.7681
4.3842	3.2684	3069	4.1875	187.5667
4.3664	3.3674	3162	4.4539	177.8032
4.3472	3.4665	3255	4.3428	173.6079
4.35	3.5655	3348	4.2917	176.7353
4.3769	3.6645	3441	3.8971	179.4813
4.3513	3.7636	3534	3.9561	185.4310
4.3595	3.8626	3627	4.3058	176.8116
4.3246	3.9617	3720	4.4962	160.0305
4.2921	4.0607	3813	4.5301	164.6834
4.3408	4.1597	3906	4.0017	172.6163
4.2625	4.2588	3999	4.1774	156.0641

Framework versions

Transformers 4.49.0
Pytorch 2.6.0+cu124
Datasets 3.6.0
Tokenizers 0.21.0

Task: Automatic Speech Recognition
Library: transformers
Type: speech_to_text
Languages: vi
Created: May 14, 2025
Updated: May 14, 2025