Seyfelislem/afrispeech_large_A100
16
<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
afrispeechlargeA100
This model is a fine-tuned version of openai/whisper-large-v2 on the afrispeech-200 dataset.
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 1e-05
- trainbatchsize: 32
- evalbatchsize: 8
- seed: 42
- gradientaccumulationsteps: 2
- totaltrainbatch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lrschedulertype: linear
- lrschedulerwarmup_steps: 500
- training_steps: 2000
- mixedprecisiontraining: Native AMP
Training results
https://huggingface.co/Seyfelislem/afrispeechlargeA100/tensorboard
Framework versions
- Transformers 4.29.1
- Pytorch 1.13.1
- Datasets 2.12.0
- Tokenizers 0.13.3