metadata

language:
  - ko
license: apache-2.0
base_model: openai/whisper-small
tags:
  - whisper-event
  - generated_from_trainer
datasets:
  - GGarri/customdataset
metrics:
  - wer
model-index:
  - name: Whisper Small ko
    results:
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: customdata
          type: GGarri/customdataset
        metrics:
          - name: Wer
            type: wer
            value: 2.590564448188711

Whisper Small ko

This model is a fine-tuned version of openai/whisper-small on the customdata dataset. It achieves the following results on the evaluation set:

Loss: 0.0041
Cer: 2.4925
Wer: 2.5906

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: 1e-05
train_batch_size: 32
eval_batch_size: 8
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 500
training_steps: 500
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Cer	Wer
3.4996	0.89	25	3.1447	75.5887	19.7136
2.655	1.79	50	2.1647	74.1483	18.1761
1.7168	2.68	75	1.2822	71.8061	17.2283
0.9261	3.57	100	0.6754	63.5396	51.5586
0.4707	4.46	125	0.3511	40.5686	37.3842
0.2485	5.36	150	0.2027	27.9309	25.6950
0.1463	6.25	175	0.1315	24.7119	23.9890
0.1022	7.14	200	0.0881	21.1924	19.9242
0.0642	8.04	225	0.0501	18.7625	17.6917
0.0249	8.93	250	0.0144	27.2044	26.3479
0.0056	9.82	275	0.0082	12.4749	11.9208
0.0036	10.71	300	0.0067	8.5922	8.7616
0.0037	11.61	325	0.0119	6.4003	6.1500
0.0021	12.5	350	0.0054	3.7450	3.6015
0.0013	13.39	375	0.0052	2.8557	3.0329
0.0017	14.29	400	0.0062	9.0681	8.3825
0.0016	15.18	425	0.0081	4.9098	5.3917
0.0012	16.07	450	0.0108	14.5541	13.3530
0.0014	16.96	475	0.0033	3.4068	3.4120
0.0005	17.86	500	0.0041	2.4925	2.5906

Framework versions

Transformers 4.39.2
Pytorch 2.0.1
Datasets 2.18.0
Tokenizers 0.15.2