jharshraj/whisper-indian-names-tensorrt

TensorRT-LLM optimized Whisper model

This repository contains a TensorRT-LLM optimized version of the Whisper model from jharshraj/whisper-indian-names.

Optimization details

• —Precision: float16
• —Weight quantization: int8
• —Max batch size: 8
• —Max beam width: 4

Usage

To use this model, you need TensorRT-LLM installed. Please refer to the TensorRT-LLM Whisper documentation for usage instructions.