Whisper Large V3 (Thai): Combined V1 This model is a fine tuned version of openai/whisper medium on augmented versions of the mozilla foundation/common voice 13 0 th, google/fleurs, and curated datasets. It achieves the following results on the common voice 13 test set: WER: 6.59 (with Deepcut Tokenizer) Model description Use the model with huggingface's transformers as follows: 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: AdamW with betas=(0.9,0.999) and epsilon=1e 08 lr scheduler type: linear lr scheduler warmup steps: 500 training steps: 10000 mixed precision training: Native AMP Framework versions Transformers 4.37.2 Pytorch 2.1.0 Datasets 2.16.1 Tokenizers 0.15.1 Citation Cite using Bibtex:
We use cookies for essential functionality and analytics. You can accept or reject analytics cookies.Cookie policy