Distilled SPTQ and QAT Latency-Accuracy Tradeoffs in CodeLlama-33B Edge Deployment

This report synthesises findings from 3 peer-reviewed papers addressing the following research question: How does the latency-accuracy tradeoff of Distilled SPTQ on CodeLlama-33B compare to QAT when deployed on edge devices for code generation tasks, as measured by BLEU score and inference speed. Large Language Models (LLMs) have garnered remarkable advancements across diverse code-related tasks, known as Code LLMs, particularly in code generation that generates source code with LLM from natural language descriptions. This burgeoning field has captured significant. 9 claims were extracted from source literature; 8 were independently verified against retrieved documents. An automated multi-reviewer quality assessment produced a score of 7.5/10. This report is a machine-generated literature synthesis and does not constitute original research.

Research goal: How does the latency-accuracy tradeoff of Distilled SPTQ on CodeLlama-33B compare to QAT when deployed on edge devices for code generation tasks, as measured by BLEU score and inference speed?

Autonomous literature synthesis. Automated review score: 7.5/10. Full text and citation available at Assignee Research.