You can deploy this model and start earning money today!

manifestasi/
whisper-tiny-manifestasi-indo

Name: whisper-tiny-manifestasi-indo
Availability: InStock

Views

No views yet

whisper-tiny-manifestasi-indo-v4

This model is a fine-tuned version of manifestasi/whisper-tiny-manifestasi-indo-v2 on an unknown dataset. It achieves the following results on the evaluation set:

More information needed

More information needed

More information needed

The following hyperparameters were used during training:

learning_rate: 1e-05
train_batch_size: 16
eval_batch_size: 8
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: 100
training_steps: 4000
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss	Wer
0.3604	0.48	300	0.3503	15.2016
0.322	0.96	600	0.3194	14.0440
0.2093	1.44	900	0.3105	15.5197
0.2232	1.92	1200	0.3031	13.3698
0.1335	2.4	1500	0.3041	13.0645
0.121	2.88	1800	0.3046	13.0645
0.0734	3.36	2100	0.3077	12.9882
0.0706	3.84	2400	0.3081	13.0263
0.0476	4.32	2700	0.3111	13.6242
0.0571	4.8	3000	0.3120	14.6419
0.0455	5.28	3300	0.3138	13.0899
0.0293	5.76	3600	0.3161	12.8864
0.0265	6.24	3900	0.3161	14.2984