whispertiny_kokoro_styletts_25sil_v1_2_15k – AI Model by albertuspekerti | AlphaNeural AI

You can deploy this model and start earning money today!

albertuspekerti/
whispertiny_kokoro_styletts_25sil_v1_2_15k

like

0

generated_from_trainer

openai/whisper-tiny

Views

No views yet

whispertiny_kokoro_styletts_25sil_v1_2_15k

This model is a fine-tuned version of openai/whisper-tiny on the None dataset. It achieves the following results on the evaluation set:

Loss: 0.1072
Wer: 7.1101

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: 16
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: 15000
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.0903	0.2286	1000	0.1369	14.5474
0.0656	0.4571	2000	0.1046	10.9415
0.0543	0.6857	3000	0.1151	11.6249
0.0426	0.9143	4000	0.1558	14.0531
0.0272	1.1429	5000	0.1032	9.0004
0.0269	1.3714	6000	0.1073	8.2879
0.0193	1.6	7000	0.0996	8.0407
0.0218	1.8286	8000	0.0993	7.6409
0.0119	2.0571	9000	0.0973	7.2410
0.011	2.2857	10000	0.1014	7.4591
0.0118	2.5143	11000	0.1007	7.1320
0.0116	2.7429	12000	0.1027	7.1174
0.0111	2.9714	13000	0.1069	7.5100
0.0095	3.2	14000	0.1060	7.2192
0.0042	3.4286	15000	0.1072	7.1101

Framework versions

Transformers 4.41.2
Pytorch 2.3.0+cu121
Datasets 4.0.0
Tokenizers 0.19.1