You can deploy this model and start earning money today!

Faith-nchifor/
whisper-small-nep

Name: whisper-small-nep
Availability: InStock

automatic-speech-recognition

generated_from_trainer

Views

No views yet

whisper-small-nep

This model is a fine-tuned version of openai/whisper-small on the ASR Nepali dataset from openslr.org. It achieves the following results on the evaluation set:

Loss: 0.2274
Wer: 43.7277

Model description

Finetuned Whisper for Nepali dataset

Intended uses & limitations

This model is an initial finetuning model for ASR for Nepali language. More training is needed for better model performances

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: 1500
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.093	2.5	1000	0.2274	43.7277

Framework versions

Transformers 4.37.0.dev0
Pytorch 2.1.0+cu121
Datasets 2.16.0
Tokenizers 0.15.0

Faith-nchifor/whisper-small-nep