You can deploy this model and start earning money today!

VladS159/
whisper-medium-6000-steps_0_percent_synthetic_data_18_12_2025

Name: whisper-medium-6000-steps_0_percent_synthetic_data_18_12_2025
Availability: InStock

VladS159/common_voice_17_0_without_synthetic_data

openai/whisper-medium

apache-2.0

model-index

Views

No views yet

Whisper Medium Ro - PEFT

This model is a fine-tuned version of openai/whisper-medium on the Common Voice 17.0 0% Synthetic dataset. It achieves the following results on the evaluation set:

Loss: 0.1599
Wer: 8.2418

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: 8
eval_batch_size: 8
seed: 42
optimizer: Use adamw_torch_fused 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: 600
training_steps: 6000
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
1.4319	0.1133	500	1.2351	11.8956
0.2438	0.2267	1000	0.2124	10.4976
0.2407	0.3400	1500	0.1926	9.6856
0.2152	0.4533	2000	0.1813	9.3407
0.206	0.5666	2500	0.1750	8.9591
0.1986	0.6800	3000	0.1709	8.8370
0.2124	0.7933	3500	0.1670	8.6203
0.1881	0.9066	4000	0.1648	8.4035
0.1833	1.0199	4500	0.1619	8.3578
0.1926	1.1333	5000	0.1607	8.3394
0.177	1.2466	5500	0.1597	8.3669
0.1707	1.3599	6000	0.1599	8.2418

Framework versions

PEFT 0.18.1.dev0
Transformers 4.57.1
Pytorch 2.9.1+rocm6.4
Datasets 3.6.0
Tokenizers 0.22.1

VladS159/whisper-medium-6000-steps_0_percent_synthetic_data_18_12_2025

Whisper Medium Ro - PEFT

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

VladS159/
whisper-medium-6000-steps_0_percent_synthetic_data_18_12_2025