Studying quantization trade-offs for efficient inference deployment in machine translation

arXiv cs.CL Papers

Summary

This paper studies quantization trade-offs for deploying translation models like EuroLLM and Hy-MT2 on a single GPU, showing that combining document chunking with W4A8/W8A8 quantization improves latency-throughput trade-offs while introducing a document-level evaluation to assess long-context dynamics.

arXiv:2607.29397v1 Announce Type: new Abstract: Deploying large language models in realistic server environments poses challenges, as the system needs to provide high-quality responses with low latency. Quantization is a common approach to reduce the memory footprint and improve inference efficiency, yet its impact on latency and throughput is rarely evaluated under controlled, orchestration-level workloads. In this work we study the quantization trade-offs of two translation model families, EuroLLM \citep{martins2025eurollm} and Hy-MT2 \citep{zheng2026hy} across five models ranging from 1.7B to 22B for efficient deployment on a single A100 or H100 GPU. We demonstrate that combining a document-chunking strategy with W4A8 or W8A8 quantization improves the latency-throughput Pareto-curve under a wide range of workloads. Furthermore, since standard machine translation (MT) benchmarks rely on isolated sentences and fail to capture long-context dynamics, we introduce a document-level evaluation from WMT24++ to assess how text chunking strategies affect translation quality under quantization. Our results reveal that standard segment-level evaluation can fail to predict the interaction between quantization and long-context document translation. While Hy-MT2 remains robust under quantization, EuroLLM shows strong sensitivity and translation quality collapses rapidly for all considered quantization formats. Overall, our experiments show that the trade-off between inference efficiency and translation quality depends not only on the quantization format, but also on the choice of text chunking strategy.
Original Article
View Cached Full Text

Cached at: 08/03/26, 07:36 AM

# Studying quantization trade-offs for efficient inference deployment in machine translation
Source: [https://arxiv.org/abs/2607.29397](https://arxiv.org/abs/2607.29397)
[View PDF](https://arxiv.org/pdf/2607.29397)

> Abstract:Deploying large language models in realistic server environments poses challenges, as the system needs to provide high\-quality responses with low latency\. Quantization is a common approach to reduce the memory footprint and improve inference efficiency, yet its impact on latency and throughput is rarely evaluated under controlled, orchestration\-level workloads\. In this work we study the quantization trade\-offs of two translation model families, EuroLLM \\citep\{martins2025eurollm\} and Hy\-MT2 \\citep\{zheng2026hy\} across five models ranging from 1\.7B to 22B for efficient deployment on a single A100 or H100 GPU\. We demonstrate that combining a document\-chunking strategy with W4A8 or W8A8 quantization improves the latency\-throughput Pareto\-curve under a wide range of workloads\. Furthermore, since standard machine translation \(MT\) benchmarks rely on isolated sentences and fail to capture long\-context dynamics, we introduce a document\-level evaluation from WMT24\+\+ to assess how text chunking strategies affect translation quality under quantization\. Our results reveal that standard segment\-level evaluation can fail to predict the interaction between quantization and long\-context document translation\. While Hy\-MT2 remains robust under quantization, EuroLLM shows strong sensitivity and translation quality collapses rapidly for all considered quantization formats\. Overall, our experiments show that the trade\-off between inference efficiency and translation quality depends not only on the quantization format, but also on the choice of text chunking strategy\.

## Submission history

From: Jim Zhao \[[view email](https://arxiv.org/show-email/b19a3c83/2607.29397)\] **\[v1\]**Fri, 31 Jul 2026 13:15:11 UTC \(263 KB\)

Similar Articles

LLM Compression with Jointly Optimizing Architectural and Quantization choices

arXiv cs.LG

Researchers from UiT and University of Oslo propose a differentiable NAS framework that jointly optimizes architectural configurations and mixed-precision quantization for LLM compression, achieving up to 1.4× faster inference or 6% higher accuracy across seven reasoning tasks compared to sequential NAS-then-quantization baselines.

Llama-Mobile: Efficient 2.7-Bit Quantization of VLMs

Hugging Face Daily Papers

This paper presents a framework for quantizing vision-language models to 2.7 bits per parameter, enabling efficient mobile deployment by compressing the Llama 3.2 11B Vision Instruct model to 3.7 GB while preserving performance on visual QA tasks.