bug: Fix OpenAiCompatible's use and configuration of base address and…

… api key (#92)

README.md

@@ -166,6 +166,10 @@ These use cases are starting points. Experiment with different instructions to f
 
 ## Models
 
+Cellm supports both hosted and local models. These are configured via appsettings files. In general, you should leave `appsettings.json` alone and add your own configuration to `appsettings.Local.json` only. Any settings in this file will override the default settings in `appsettings.json`. 
+
+You can use `appsettings.Local.OpenAiCompatible.json` as a starting point for configuring any model provider that is compatible with OpenAI's API. Just rename it to `appsettings.Local.json` and edit the values. The following sections shows you how to configure `appsettings.Local.json` for commonly used hosted and local models.
+
 ### Hosted LLMs
 
 Cellm supports hosted models from Anthropic, Google, OpenAI, and any OpenAI-compatible provider. To use e.g. Claude 3.5 Sonnet from Anthropic:
@@ -221,7 +225,7 @@ Llamafile is a stand-alone executable that is very easy to setup. To get started
 3. Rename `appsettings.Llamafile.json` to `appsettings.Local.json`.
 4. Build and install Cellm.
 
-### Dockerized Ollama and vLLM
+#### Dockerized Ollama and vLLM
 
 Both Ollama and vLLM are packaged up with docker compose files in the `docker/` folder. Ollama is designed for easy of use and vLLM is designed to run many requests in parallel which is particularly useful for this Add-In. Open WebUI in included in both Ollama and vLLM docker compose files so you can test the local model outside of Cellm. It is available at `http://localhost:3000`.
 

src/Cellm/Models/Client.cs

@@ -11,7 +11,7 @@
 
 namespace Cellm.Models;
 
-public class Client(ISender sender)
+internal class Client(ISender sender)
 {
     public async Task<Prompt> Send(Prompt prompt, Provider? provider, Uri? baseAddress, CancellationToken cancellationToken)
     {
@@ -23,7 +23,7 @@ public async Task<Prompt> Send(Prompt prompt, Provider? provider, Uri? baseAddre
                 Provider.Llamafile => await sender.Send(new LlamafileRequest(prompt), cancellationToken),
                 Provider.Ollama => await sender.Send(new OllamaRequest(prompt), cancellationToken),
                 Provider.OpenAi => await sender.Send(new OpenAiRequest(prompt), cancellationToken),
-                Provider.OpenAiCompatible => await sender.Send(new OpenAiCompatibleRequest(prompt, baseAddress ?? throw new NullReferenceException($"{nameof(Provider.OpenAiCompatible)} requires BaseAddress")), cancellationToken),
+                Provider.OpenAiCompatible => await sender.Send(new OpenAiCompatibleRequest(prompt, baseAddress), cancellationToken),
                 _ => throw new NotSupportedException($"Provider {provider} is not supported")
             };
 

src/Cellm/Models/Providers/OpenAiCompatible/OpenAiCompatibleRequest.cs

@@ -4,5 +4,5 @@ namespace Cellm.Models.Providers.OpenAiCompatible;
 
 internal record OpenAiCompatibleRequest(
     Prompt Prompt,
-    Uri BaseAddress,
+    Uri? BaseAddress,
     string? ApiKey = null) : IModelRequest<OpenAiCompatibleResponse>;

src/Cellm/Models/Providers/OpenAiCompatible/OpenAiCompatibleRequestHandler.cs

@@ -1,18 +1,22 @@
 using Cellm.Models.Prompts;
+using Microsoft.Extensions.Options;
 
 namespace Cellm.Models.Providers.OpenAiCompatible;
 
 internal class OpenAiCompatibleRequestHandler(
-    OpenAiCompatibleChatClientFactory openAiCompatibleChatClientFactory)
+    OpenAiCompatibleChatClientFactory openAiCompatibleChatClientFactory,
+    IOptions<OpenAiCompatibleConfiguration> openAiCompatibleConfiguration)
     : IModelRequestHandler<OpenAiCompatibleRequest, OpenAiCompatibleResponse>
 {
 
     public async Task<OpenAiCompatibleResponse> Handle(OpenAiCompatibleRequest request, CancellationToken cancellationToken)
     {
+        var api_key = string.IsNullOrEmpty(request.ApiKey) ? "API_KEY" : request.ApiKey;
+
         var chatClient = openAiCompatibleChatClientFactory.Create(
-            request.BaseAddress,
+            request.BaseAddress ?? openAiCompatibleConfiguration.Value.BaseAddress,
             request.Prompt.Options.ModelId ?? string.Empty,
-            request.ApiKey ?? "API_KEY");
+            api_key);
 
         var chatCompletion = await chatClient.CompleteAsync(request.Prompt.Messages, request.Prompt.Options, cancellationToken);
 

src/Cellm/appsettings.Local.OpenAiCompatible.json

@@ -0,0 +1,10 @@
+{
+    "OpenAiCompatibleConfiguration": {
+        "BaseAddress": "YOUR_API_ENDPOINT",
+        "DefaultModel": "YOUR_DEFAULT_MODEL",
+        "ApiKey": "YOUR_API_KEY_OPTIONAL"
+    },
+    "ProviderConfiguration": {
+        "DefaultProvider": "OpenAiCompatible"
+    }
+}

src/Cellm/appsettings.json

@@ -2,19 +2,28 @@
     "AnthropicConfiguration": {
         "BaseAddress": "https://api.anthropic.com",
         "DefaultModel": "claude-3-5-sonnet-latest",
-        "Version": "2023-06-01"
+        "Version": "2023-06-01",
+        "ApiKey": "YOUR_ANTHROPIC_API_KEY"
     },
     "LlamafileConfiguration": {
-        "BaseAddress": "http://127.0.0.1/v1",
-        "DefaultModel": "gemma-2-2b"
+        "BaseAddress": "http://127.0.0.1:8080/v1",
+        "DefaultModel": "gemma-2-2b",
+        "ApiKey": "YOUR_LLAMAFILE_API_KEY_OPTIONAL"
     },
     "OllamaConfiguration": {
         "BaseAddress": "http://127.0.0.1:11434",
-        "DefaultModel": "gemma2:2b"
+        "DefaultModel": "gemma2:2b",
+        "ApiKey": "YOUR_OLLAMA_API_KEY_OPTIONAL"
     },
     "OpenAiConfiguration": {
         "BaseAddress": "https://api.openai.com/v1",
-        "DefaultModel": "gpt-4o-mini"
+        "DefaultModel": "gpt-4o-mini",
+        "ApiKey": "YOUR_OPENAI_API_KEY"
+    },
+    "OpenAiCompatibleConfiguration": {
+        "BaseAddress": "https://api.openai.com/v1",
+        "DefaultModel": "gpt-4o-mini",
+        "ApiKey": "YOUR_API_KEY"
     },
     "CellmConfiguration": {
         "Debug": false