Bhasha-Rupantarika: Algorithm-Hardware Co-design approach for Multilingual Neural Machine Translation

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Lokhande, Mukul, Dewangan, Tanushree, Mansoori, Mohd Sharik, Chaudhari, Tejas, J., Akarsh, Lokhande, Damayanti, Teman, Adam, Vishvakarma, Santosh Kumar
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866913177268125696
author Lokhande, Mukul
Dewangan, Tanushree
Mansoori, Mohd Sharik
Chaudhari, Tejas
J., Akarsh
Lokhande, Damayanti
Teman, Adam
Vishvakarma, Santosh Kumar
author_facet Lokhande, Mukul
Dewangan, Tanushree
Mansoori, Mohd Sharik
Chaudhari, Tejas
J., Akarsh
Lokhande, Damayanti
Teman, Adam
Vishvakarma, Santosh Kumar
contents This paper introduces Bhasha-Rupantarika, a light and efficient multilingual translation system tailored through algorithm-hardware codesign for resource-limited settings. The method investigates model deployment at sub-octet precision levels (FP8, INT8, INT4, and FP4), with experimental results indicating a 4.1x reduction in model size (FP4) and a 4.2x speedup in inference speed, which correlates with an increased throughput of 66 tokens/s (improvement by 4.8x). This underscores the importance of ultra-low precision quantization for real-time deployment in IoT devices using FPGA accelerators, achieving performance on par with expectations. Our evaluation covers bidirectional translation between Indian and international languages, showcasing its adaptability in low-resource linguistic contexts. The FPGA deployment demonstrated a 1.96x reduction in LUTs and a 1.65x decrease in FFs, resulting in a 2.2x enhancement in throughput compared to OPU and a 4.6x enhancement compared to HPTA. Overall, the evaluation provides a viable solution based on quantisation-aware translation along with hardware efficiency suitable for deployable multilingual AI systems. The entire codes [https://github.com/mukullokhande99/Bhasha-Rupantarika/] and dataset for reproducibility are publicly available, facilitating rapid integration and further development by researchers.
format Preprint
id arxiv_https___arxiv_org_abs_2510_10676
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Bhasha-Rupantarika: Algorithm-Hardware Co-design approach for Multilingual Neural Machine Translation
Lokhande, Mukul
Dewangan, Tanushree
Mansoori, Mohd Sharik
Chaudhari, Tejas
J., Akarsh
Lokhande, Damayanti
Teman, Adam
Vishvakarma, Santosh Kumar
Hardware Architecture
Computation and Language
Robotics
Audio and Speech Processing
This paper introduces Bhasha-Rupantarika, a light and efficient multilingual translation system tailored through algorithm-hardware codesign for resource-limited settings. The method investigates model deployment at sub-octet precision levels (FP8, INT8, INT4, and FP4), with experimental results indicating a 4.1x reduction in model size (FP4) and a 4.2x speedup in inference speed, which correlates with an increased throughput of 66 tokens/s (improvement by 4.8x). This underscores the importance of ultra-low precision quantization for real-time deployment in IoT devices using FPGA accelerators, achieving performance on par with expectations. Our evaluation covers bidirectional translation between Indian and international languages, showcasing its adaptability in low-resource linguistic contexts. The FPGA deployment demonstrated a 1.96x reduction in LUTs and a 1.65x decrease in FFs, resulting in a 2.2x enhancement in throughput compared to OPU and a 4.6x enhancement compared to HPTA. Overall, the evaluation provides a viable solution based on quantisation-aware translation along with hardware efficiency suitable for deployable multilingual AI systems. The entire codes [https://github.com/mukullokhande99/Bhasha-Rupantarika/] and dataset for reproducibility are publicly available, facilitating rapid integration and further development by researchers.
title Bhasha-Rupantarika: Algorithm-Hardware Co-design approach for Multilingual Neural Machine Translation
topic Hardware Architecture
Computation and Language
Robotics
Audio and Speech Processing
url https://arxiv.org/abs/2510.10676