Code review

February 2, 2025

LLM use for Code review #

The fundamental property of Large Language Models is their ability to understand the context of provided content. A straightforward application of this ability is text summarization, correction and any advanced work with natural language. For software developers it opens possibilities for an independent check of their code from the contextual point of view. There are plenty of tools already for this kind of tasks, mostly based on online services and therefore using the powerfull and really large language models. The obvious problem is that in the background your (maybe proprietary) code is sent to the cloud for analysis. And, maybe, this is quite an overkill. Let us take look if you can use local LLM for this task, in this case we use LokAi and its Smarter model (currently in LokAi we use Meta’s Llama-3.1-8B model).