whisper-tiny-kor_eng_tiny_ps_ob

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

Model description

More information needed

More information needed

More information needed

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 Loss	Epoch	Step	Validation Loss	Cer
1.1215	0.4167	100	0.9467	26.4777
0.8211	0.8333	200	0.7182	22.5991
0.5379	1.25	300	0.6071	15.1115
0.4346	1.6667	400	0.5519	11.8711
0.3403	2.0833	500	0.5439	12.8501
0.1993	2.5	600	0.5605	14.4827
0.1992	2.9167	700	0.5525	12.3203
0.1056	3.3333	800	0.5637	11.2175
0.0831	3.75	900	0.5797	19.1976
0.0666	4.1667	1000	0.6070	11.7379
0.0465	4.5833	1100	0.6067	14.9133
0.0453	5.0	1200	0.6096	12.7974
0.0221	5.4167	1300	0.6307	14.0613
0.0262	5.8333	1400	0.6271	11.9052
0.017	6.25	1500	0.6371	11.3259
0.0124	6.6667	1600	0.6460	12.4938
0.0126	7.0833	1700	0.6564	10.8922
0.0091	7.5	1800	0.6503	10.9139
0.007	7.9167	1900	0.6666	12.0973
0.0077	8.3333	2000	0.6709	10.9696
0.0046	8.75	2100	0.6740	10.7404
0.0049	9.1667	2200	0.6936	11.4157
0.0035	9.5833	2300	0.6826	11.6109
0.0025	10.0	2400	0.7020	11.9300
0.0028	10.4167	2500	0.7051	15.3779
0.0028	10.8333	2600	0.6954	12.4938
0.0023	11.25	2700	0.6995	11.4870
0.0013	11.6667	2800	0.7008	12.8160
0.0011	12.0833	2900	0.7056	12.8625
0.0012	12.5	3000	0.7057	13.0235
0.0024	12.9167	3100	0.7057	12.8903
0.001	13.3333	3200	0.7100	12.9833
0.0008	13.75	3300	0.7135	11.3910
0.0014	14.1667	3400	0.7108	12.8222
0.0007	14.5833	3500	0.7127	12.7757
0.0006	15.0	3600	0.7142	12.8470
0.0006	15.4167	3700	0.7152	12.9244
0.0007	15.8333	3800	0.7158	12.6983
0.0006	16.25	3900	0.7167	12.5991
0.0006	16.6667	4000	0.7169	12.5898