whisper-tiny-kor_eng_tiny_ps_ev

CianKim/whisper-tiny-kor_eng_tiny_ps_ev

whisper-tiny-kor_eng_tiny_ps_ev is a automatic speech recognition model from CianKim. 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

31% of all-time downloads

All-time downloads

Public

Parameters

37.8M

2.4 GB on disk

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.safetensors151 MB · 99%

At a glance

Task: Automatic Speech Recognition
Library: transformers
Model type: whisper
Access: Public
Created: May 8, 2025
Updated: May 15, 2025
SHA: 8aaebdcd

Model card

From the Hugging Face model README

whisper-tiny-kor_eng_tiny_ps_ev

This model was trained from scratch on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 0.6602
Cer: 12.7506

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: 12
eval_batch_size: 6
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	Cer
0.8641	0.7194	100	0.7351	26.3331
0.5593	1.4388	200	0.5476	20.0705
0.3777	2.1583	300	0.4698	17.0663
0.2209	2.8777	400	0.4626	25.7207
0.1185	3.5971	500	0.4861	25.2239
0.0852	4.3165	600	0.5005	12.3635
0.0494	5.0360	700	0.5416	20.0127
0.03	5.7554	800	0.5177	18.4817
0.0236	6.4748	900	0.5654	13.2706
0.0178	7.1942	1000	0.5850	12.3404
0.0154	7.9137	1100	0.5761	14.1256
0.0126	8.6331	1200	0.5862	15.1135
0.0067	9.3525	1300	0.6111	17.1645
0.0054	10.0719	1400	0.6148	12.7217
0.0048	10.7914	1500	0.6131	11.6356
0.0054	11.5108	1600	0.6126	12.0226
0.0028	12.2302	1700	0.6276	14.6744
0.0041	12.9496	1800	0.6194	13.3514
0.0016	13.6691	1900	0.6282	12.0284
0.0028	14.3885	2000	0.6314	17.6093
0.0013	15.1079	2100	0.6539	11.8204
0.002	15.8273	2200	0.6369	14.2585
0.0022	16.5468	2300	0.6491	14.5762
0.0004	17.2662	2400	0.6476	13.0048
0.0012	17.9856	2500	0.6457	12.7217
0.0007	18.7050	2600	0.6482	12.9644
0.0008	19.4245	2700	0.6517	12.3173
0.0004	20.1439	2800	0.6521	12.2075
0.0003	20.8633	2900	0.6556	12.4559
0.0005	21.5827	3000	0.6553	12.5773
0.0003	22.3022	3100	0.6560	12.5079
0.0003	23.0216	3200	0.6571	12.5599
0.0003	23.7410	3300	0.6570	12.6177
0.0002	24.4604	3400	0.6580	12.6697
0.0003	25.1799	3500	0.6590	12.6582
0.0003	25.8993	3600	0.6591	12.5426
0.0003	26.6187	3700	0.6595	12.6755
0.0002	27.3381	3800	0.6601	12.6755
0.0002	28.0576	3900	0.6601	12.7506
0.0002	28.7770	4000	0.6602	12.7506

Framework versions

Transformers 4.51.3
Pytorch 2.6.0+cu126
Datasets 3.5.0
Tokenizers 0.21.1

Task: Automatic Speech Recognition
Library: transformers
Type: whisper
Created: May 8, 2025
Updated: May 15, 2025