Data used for training is the initial 25% of train and validation of Italian Common Voice 11.0 from Mozilla Foundation. The dataset used for evaluation is the initial 10% of test of Italian Common Voice.

Training procedure

After loading the pre trained model, it has been trained on the dataset.

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: 4000
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.584600	0.95	1000	0.801204	48.980865
0.496100	1.91	2000	0.713927	46.283971
0.406000	2.86	3000	0.680141	43.268164
0.402000	3.82	4000	0.670211	42.276761

Framework versions

Transformers 4.26.0.dev0
Pytorch 1.12.1+cu113
Datasets 2.7.1
Tokenizers 0.13.2

FCameCode/whisper-tiny-it-11

Whisper Tiny it 11

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

FCameCode/
whisper-tiny-it-11