Create basic experiment for semantic kernel handlebar planners.

Author: alkampfergit

99_CSharp/Program.cs

@@ -3,19 +3,13 @@
 using Microsoft.SemanticKernel;
 using Microsoft.SemanticKernel.ChatCompletion;
 using Microsoft.SemanticKernel.Connectors.OpenAI;
+using Microsoft.SemanticKernel.Planning.Handlebars;
 using Microsoft.SemanticKernel.PromptTemplates.Handlebars;
 using SemanticKernelExperiments.Helper;
 using System;
 using System.IO;
 using System.Linq;
 using System.Threading.Tasks;
-using System;
-using System.Net;
-using System.Threading.Tasks;
-using Microsoft.Extensions.DependencyInjection;
-using Microsoft.Extensions.Http.Resilience;
-using Microsoft.Extensions.Logging;
-using Microsoft.SemanticKernel;
 
 namespace SemanticKernelExperiments;
 
@@ -32,7 +26,8 @@ static async Task Main(string[] args)
         // await Ex03_DirectSequentialCallToExtractVideo();
         // await Ex03_b_DirectSequentialCallToExtractVideo();
 
-        await Ex04_Load_function_in_builder();
+        //await Ex04_Load_function_in_builder();
+        await Ex05_basic_planner();
         Console.ReadLine();
     }
 
@@ -208,6 +203,45 @@ private static async Task Ex04_Load_function_in_builder()
         }
     }
 
+    private static async Task Ex05_basic_planner()
+    {
+        var pluginsDirectory = Path.Combine(System.IO.Directory.GetCurrentDirectory(), "plugins", "PublishingPlugin");
+        var kernelBuilder = CreateBasicKernelBuilder();
+        kernelBuilder
+            .Plugins
+                .AddFromType<AudioVideoPlugin.AudioVideoPlugin>("AudioVideoPlugin")
+                .AddFromPromptDirectory(pluginsDirectory);
+        var kernel = kernelBuilder.Build();
+
+#pragma warning disable SKEXP0060 // Type is for evaluation purposes only and is subject to change or removal in future updates. Suppress this diagnostic to proceed.
+        var planner = new HandlebarsPlanner();
+
+        Console.WriteLine("\n\nProcess followed to answer the question:\n");
+        var question = "I want to extract summarized timeline from video file C:\\temp\\ssh.mp4";
+
+        var plan = await planner.CreatePlanAsync(kernel, question);
+
+        var textPlan = plan.ToString();
+        Console.WriteLine(textPlan);
+
+        //now we can invoke the plan
+        var result = await plan.InvokeAsync(kernel);
+
+#pragma warning restore SKEXP0060 // Type is for evaluation purposes only and is subject to change or removal in future updates. Suppress this diagnostic to proceed.
+        DumpTextSection("RAW FUNCTION CALLS");
+        var llmCalls = _loggingProvider.GetLLMCalls();
+        foreach (var llmCall in llmCalls)
+        {
+            Console.WriteLine($"Function {llmCall.ResponseFunctionCall} with arguments {llmCall.ResponseFunctionCallParameters}");
+        }
+
+        DumpTextSection("FULL INFORMATION DUMP");
+        foreach (var llmCall in llmCalls)
+        {
+            Console.WriteLine(llmCall.Dump());
+        }
+    }
+
     private static void DumpTextSection(string text)
     {
         int totalWidth = 80;
@@ -232,7 +266,7 @@ private static IKernelBuilder CreateBasicKernelBuilder()
             .AddLogger(s => _loggingProvider.CreateHttpRequestBodyLogger(s.GetRequiredService<ILogger<DumpLoggingProvider>>())));
 
         kernelBuilder.Services.AddAzureOpenAIChatCompletion(
-            "GPT35_2",//"GPT42",
+            "GPT42", //"GPT35_2",//"GPT42",
             Dotenv.Get("OPENAI_API_BASE"),
             Dotenv.Get("OPENAI_API_KEY"));
 

99_CSharp/SemanticKernelExperiments.csproj

@@ -31,6 +31,7 @@
       <PackageReference Include="Microsoft.Extensions.Logging.Console" Version="8.0.0" />
       <PackageReference Include="Microsoft.Extensions.Logging.Debug" Version="8.0.0" />
       <PackageReference Include="Microsoft.SemanticKernel" Version="1.1.0" />
+      <PackageReference Include="Microsoft.SemanticKernel.Planners.Handlebars" Version="1.1.0-preview" />
       <PackageReference Include="Microsoft.SemanticKernel.PromptTemplates.Handlebars" Version="1.1.0" />
       <PackageReference Include="Microsoft.SemanticKernel.Yaml" Version="1.1.0" />
       <PackageReference Include="Microsoft.SemanticKernel.Abstractions" Version="1.1.0" />

99_CSharp/python/transcript_timeline.py

@@ -2,7 +2,7 @@
 import sys
 
 def transcript_timeline(audiofile: str) -> str:
-    model = whisper.load_model("tiny.en")
+    model = whisper.load_model("large-v3")
 
     transcription_options = {
         "verbose" : False,