Scrya
/

whisper-medium-ms-augmented

Automatic Speech Recognition

Generated from Trainer

Inference Endpoints

Model card Files Files and versions Metrics Training metrics Community

Edit model card

Whisper Medium MS - Augmented

This model is a fine-tuned version of openai/whisper-medium on the google/fleurs dataset. It achieves the following results on the evaluation set:

Loss: 0.2066
Wer: 9.5784
Cer: 2.8109

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

Training:

google/fleurs (train+validation)

Evaluation:

google/fleurs (test)

Training procedure

Datasets were augmented on-the-fly using audiomentations via PitchShift and TimeStretch transformations at p=0.3.

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 1e-05
train_batch_size: 2
eval_batch_size: 2
seed: 42
gradient_accumulation_steps: 16
total_train_batch_size: 32
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 100
training_steps: 1000
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer	Cer
0.0876	2.15	200	0.1949	10.3105	3.0685
0.0064	4.3	400	0.1974	9.7004	2.9596
0.0014	6.45	600	0.2031	9.6190	2.8955
0.001	8.6	800	0.2058	9.6055	2.8440
0.0009	10.75	1000	0.2066	9.5784	2.8109

Framework versions

Transformers 4.26.0.dev0
Pytorch 1.13.0+cu117
Datasets 2.7.1.dev0
Tokenizers 0.13.2

Downloads last month: 14

Inference Examples

Automatic Speech Recognition

This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Dataset used to train Scrya/whisper-medium-ms-augmented

Spaces using Scrya/whisper-medium-ms-augmented 2

Evaluation results

WER on google/fleurs
test set self-reported

9.578
CER on google/fleurs
test set self-reported

2.811

View on Papers With Code