whisper-turbo-fr-pho – AI Model by Ycharo | AlphaNeural AI

You can deploy this model and start earning money today!

Ycharo/
whisper-turbo-fr-pho

like

0

automatic-speech-recognition

generated_from_trainer

openai/whisper-large-v3-turbo

endpoints_compatible

Views

No views yet

whisper-turbo-fr-pho

This model is a fine-tuned version of openai/whisper-large-v3-turbo on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 2.5281
Wer Ortho: 380.0798
Wer: 395.3620

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: 16
seed: 42
optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: constant_with_warmup
lr_scheduler_warmup_steps: 50
num_epochs: 3.0
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer Ortho	Wer
3.028	0.4	500	3.0232	343.3288	357.0751
2.7971	0.8	1000	2.7895	352.7828	366.6841
2.6552	1.2	1500	2.6937	301.3842	313.4266
2.6097	1.6	2000	2.6318	342.1635	355.7224
2.5566	2.0	2500	2.5847	301.3496	313.2575
2.4767	2.4	3000	2.5525	667.3102	694.8316
2.4376	2.8	3500	2.5281	380.0798	395.3620

Framework versions

Transformers 4.51.3
Pytorch 2.6.0+cu124
Datasets 3.5.0
Tokenizers 0.21.1