o0dimplz0o/Fine-Tuned-Whisper-Large-v2-Zeroth-STT-KO

Whisper Large v2 Fine-Tuned for Korean ASR

Model Description

This model is a version of openai/whisper-large-v2, currently still being fine-tuned, specifically optimized for Korean automatic speech recognition (ASR) tasks. The fine-tuning process aims to deliver high accuracy and timestamped transcriptions for Korean speech.

Dataset Details

• Dataset Source: Custom dataset (https://huggingface.co/datasets/o0dimplz0o/Zeroth-STT-Korean)
• Number of Samples: 102,263
• Split: 90% train, 10% test
• Data Augmentation: 50% random, applied only to the training set

Training Details

• Hardware: L40S GPU
• Learning Rate Scheduler: Cosine
• Epochs: [pending completion]
• Optimizer: AdamW (with Bitsandbytes 8-bit optimization)