Lee Miller
Lee Miller
``` public static async Task RunAsync() { var kernel = new KernelBuilder() .WithLogger(ConsoleLogger.Log) .WithAzureTextCompletionService( Env.Var("AZURE_OPENAI_DEPLOYMENT_NAME"), Env.Var("AZURE_OPENAI_ENDPOINT"), Env.Var("AZURE_OPENAI_KEY")) .Build(); var skill = kernel.ImportSkill(new TestSkill()); var plan = new Plan("Foobar plan for...
> See also #523 which wants to remove the planner altogether PlannerSkill - not Planner though :) But yeah, we will still want to do *something* to address this issue....
#2809 and #2147 made it easier to bring any tokenizer for use. #2840 demonstrates these capabilities specifically with MicrosoftML and DeepDev token counters along existing demonstration with SharpToken.
The planner has changed quite a bit. Please open a new issue if you are still seeing issues with 0.15 or greater.
Thanks for the PR @Sy0xFF -- we're in the process of refactoring things here, see https://github.com/microsoft/semantic-kernel/pull/523. I'll incorporate this fix with those changes.
We've transitioned away from the old PlannerSkill to a set of planners like ActionPlanner and SequentialPlanner (this was closest to previous implementation). Others on the way. Please open a new...
@stephentoub looks like there are some conflicts that need to be resolved before merging.
Currently holding for #598
Closing stale pr. We still cannot resolve these conflicts, so please feel free to re-open when the issues are resolved.
Sample Code: OpenAI service failed to complete the prompt, APIConnectionError issue with Python 3.10
We understand the issue is related to the virtual environment configuration. No changes to be made in the python SDK at present.