Update README.md

Author: intitni

CustomSuggestionService/ContentView.swift

@@ -178,8 +178,10 @@ struct RequestStrategyPicker: View {
                     case .continue:
                         Text("Continue").tag(option.rawValue)
                     case .codeLlamaFillInTheMiddle:
-                        Text("CodeLlama Fill-in-the-Middle (Good for Codellama:xb-code)")
-                            .tag(option.rawValue)
+                        Text(
+                            "CodeLlama Fill-in-the-Middle (Good for Codellama:xb-code and other models with Fill-in-the-Middle support)"
+                        )
+                        .tag(option.rawValue)
                     case .codeLlamaFillInTheMiddleWithSystemPrompt:
                         Text("CodeLlama Fill-in-the-Middle with System Prompt")
                             .tag(option.rawValue)

README.md

@@ -1,4 +1,4 @@
-#  Custom Suggestion Service for Copilot for Xcode
+# Custom Suggestion Service for Copilot for Xcode
 
 This extension offers a custom suggestion service for [Copilot for Xcode](https://github.com/intitni/CopilotForXcode), allowing you to leverage a chat model to enhance the suggestions provided as you write code.
 
@@ -27,31 +27,39 @@ The app supports three types of suggestion services:
 - Models with completions API
 - [Tabby](https://tabby.tabbyml.com)
 
-It is recommended to use Tabby since they have extensive experience in crafting prompts.
+If you are new to running a model locally, you can try [LM Studio](https://lmstudio.ai).
+
+### Recommended Settings
+
+- Use Tabby since they have extensive experience in code completion.
+- Use models with completions API with Fill-in-the-Middle support (for example, codellama:7b-code), and use the "Codellama Fill-in-the-Middle" strategy.
+
+### Others
 
-If you choose not to use Tabby, it is advisable to use a custom model with the completions API and employ the default request strategy.
+In other situations, it is advisable to use a custom model with the completions API over a chat completions API, and employ the default request strategy.
 
 Ensure that the prompt format remains as simple as the following:
 
-``` 
+```
 {System}
 {User}
 {Assistant}
 ```
 
-If you are new to running a model locally, you can try [LM Studio](https://lmstudio.ai).
-
 ## Strategies
 
 - Default: This strategy meticulously explains the context to the model, prompting it to generate a suggestion.
 - Naive: This strategy rearranges the code in a naive way to trick the model into believing it's appending code at the end of a file.
 - Continue: This strategy employs the "Please Continue" technique to persuade the model that it has started a suggestion and must continue to complete it. (Only effective with the chat completion API).
+- CodeLlama Fill-in-the-Middle: It uses special tokens to guide the models to generate suggestions. The models need to support FIM to use it (codellama:xb-code, startcoder, etc.). This strategy uses the special tokens documented by CodeLlama.
+- CodeLlama Fill-in-the-Middle with System Prompt: The previous one doesn't have a system prompt telling it what to do. You can try to use it in models that don't support FIM.
 
 ## Contribution
 
-Prompt engineering is a challenging task, and your assistance is invaluable. 
+Prompt engineering is a challenging task, and your assistance is invaluable.
 
-The most complex things are located within the `Core` package. 
+The most complex things are located within the `Core` package.
 
-- To add a new service, please refer to the `CodeCompletionService` folder. 
+- To add a new service, please refer to the `CodeCompletionService` folder.
 - To add new request strategies, check out the `SuggestionService` folder.
+