Preprint
Jul 2026
Robust Assamese Speech Recognition through Controlled Fine-Tuning of Whisper Models
A controlled, fine-tuned Whisper-based Assamese ASR system trained on the Mozilla Common Voice 24.0-Assamese corpus is presented, employing mixed-precision training and gradient accumulation on Tesla 4 Graphics Processing Units (T4 GPUs).
Ganapati Das, Dwipen Laskar, Hasin Afzal Ahmed et al.
· 0 citations