Appease linter

Signed-off-by: Anastas Stoyanovsky <[email protected]>

Author: anastasds

docs/adr-chat-module-architecture.md

@@ -2,11 +2,11 @@
 
 ## Context
 
-InstructLab recently gained the ability of [retrieval augmented generation (RAG) for chat](https://github.com/instructlab/instructlab/pull/2886). The initial implementation is the simplest possible, inserting retrieved document chunks in the chat session history before submitting the user query to the target model. This is done by adding a retriever instance into the chat class which is called by submitting the user query as-is to the vector store as a search query to retrieve those chunks. 
+InstructLab recently gained the ability of [retrieval augmented generation (RAG) for chat](https://github.com/instructlab/instructlab/pull/2886). The initial implementation is the simplest possible, inserting retrieved document chunks in the chat session history before submitting the user query to the target model. This is done by adding a retriever instance into the chat class which is called by submitting the user query as-is to the vector store as a search query to retrieve those chunks.
 
-Tuning retrieval systems requires significant expertise and testing, and there are many techniques that may be needed specifically and only for that use case. For example, query reformulation or agentic approaches may be necessary to achieve reliable retrieval quality while being redundant and expensive in a non-retrieval scenario. 
+Tuning retrieval systems requires significant expertise and testing, and there are many techniques that may be needed specifically and only for that use case. For example, query reformulation or agentic approaches may be necessary to achieve reliable retrieval quality while being redundant and expensive in a non-retrieval scenario.
 
-Having a single code path that conditionally executes various strategies based on a mixture of configuration, application state, and user input can quickly become difficult to maintain, debug, and even understand. That chat module in InstructLab would benefit from a principled approach; that is, to be purposefully architected before becoming a [big ball of mud](http://www.laputan.org/mud/?ref=blog.codinghorror.com). 
+Having a single code path that conditionally executes various strategies based on a mixture of configuration, application state, and user input can quickly become difficult to maintain, debug, and even understand. That chat module in InstructLab would benefit from a principled approach; that is, to be purposefully architected before becoming a [big ball of mud](http://www.laputan.org/mud/?ref=blog.codinghorror.com).
 
 Prompt template management, model interaction pattern, chat session history, and any interactions with third party systems can vary by use case and so should be logically grouped for purposes of at least maintainability, understandability, modifiability, and testability.