Thanks for well explained intro to Semantic Kernel. Subscribed! You asked about intrest on more videos about Semantic Kernel. Yes, please. If you would explain in simple samples like in this video how to use vector databases with Semantic Kernel that would be very intresting.
Thank you for the well-done video. You clearly explained how to add custom code execution plugins. I’m eagerly awaiting your video demonstrating how I can connect to more than one service and intelligently divert requests based on context.
Awesome video, thank you! Would love to see how you would use semantic kernel with a custom program to control a conversation (I.e. - follow a script).
awesome video!! Thank you very much. Wonderful if you have have a video talk about how to chain multiple AI models together? and call different API for database query and so on.
Hi. Good video , kindly add more content on the possibilites we have out there for SK. can you do it also with VS Code and Python instead of .NET? Also, how would a Multi-Agent scenario could be implemented together with Microsoft AutoGen is highly recommended to explore as well as RAG options . this will be appreciated.
Thanks! These are good ideas. I could look into a Python version of this, I do have a bit of Python experience, though its not really my area of expertise.
@@alexthecodewolf please more dotnet )) not python. Or you can show how to fine tune models with python and use them with sem kernel. I think this will collect a tone of reviews.
first of all tanks for the video, i have a question i build the project like this but with phi3 mini model and ollama the plugin functions not working ... is it about PromptExecutionSettings functionCnoisceBehavier?
Hey thanks, I'm not sure if I fully understand the issue here, but Ollama with Semantic Kernel recently switched to a dedicated connector package, you can see how that works here: learn.microsoft.com/en-us/dotnet/ai/quickstarts/quickstart-local-ai You'll definitely need to configure your app to allow functions to execute as shown in the video, but I"m not sure if that setting is different somehow for ollama.
at th-cam.com/video/kCGZPhnTGHM/w-d-xo.html the tech.txt file popped up before the thing even finished... hows that possible. Its like it created the file first, then searched the web, then populated the file contents.
Interesting observations, my understanding is that the orchestrator aspect of SK doesn't necessarily do things in a specific order, it decides what order is best as it goes. So yeah it probably creates the file while it went out and searched the endpoint and then populated it.
Yes, please! More Semantic Kernel! What a fun technology. I like the idea of the new AgentGroupChat framework.
What a clear and concise demo of semantic kernel👍👍, can't wait to see more about semantic kernel.
Thanks for the feedback!
Thanks!
Best Semantic Kernel showcase video I've seen
Thanks! I'll try to create a follow up with more concepts
Thank you so much... This was very clear explanation.
Thanks, I appreciate the feedback.
Thanks for well explained intro to Semantic Kernel. Subscribed! You asked about intrest on more videos about Semantic Kernel. Yes, please. If you would explain in simple samples like in this video how to use vector databases with Semantic Kernel that would be very intresting.
That's a good topic!
Superb presentation - subscribed to channel!
Thank you!
Thank you for the well-done video. You clearly explained how to add custom code execution plugins. I’m eagerly awaiting your video demonstrating how I can connect to more than one service and intelligently divert requests based on context.
Thanks for watching
Great instructor! thanks brother send you a cup of coffee! appreciated
Thanks so much!!
great demo
Thanks for watching
Awesome video, thank you! Would love to see how you would use semantic kernel with a custom program to control a conversation (I.e. - follow a script).
awesome video!! Thank you very much. Wonderful if you have have a video talk about how to chain multiple AI models together? and call different API for database query and so on.
Glad you enjoyed it, thanks for the recommendations here!
Please create more of these videos
Thanks for the feedback! The channel is always growing and I'm hoping to do more.
I really enjoy this video. Is there a way you can create a video or provide some sample code to create a new plugin to read data from our own pdf?
Can u please also cover semantic Kernel Connectors.
Hi. Good video , kindly add more content on the possibilites we have out there for SK. can you do it also with VS Code and Python instead of .NET? Also, how would a Multi-Agent scenario could be implemented together with Microsoft AutoGen is highly recommended to explore as well as RAG options . this will be appreciated.
Thanks! These are good ideas. I could look into a Python version of this, I do have a bit of Python experience, though its not really my area of expertise.
@@alexthecodewolf please more dotnet )) not python. Or you can show how to fine tune models with python and use them with sem kernel. I think this will collect a tone of reviews.
@@alexthecodewolf let’s give it a shot
first of all tanks for the video, i have a question i build the project like this but with phi3 mini model and ollama the plugin functions not working ... is it about PromptExecutionSettings functionCnoisceBehavier?
Hey thanks, I'm not sure if I fully understand the issue here, but Ollama with Semantic Kernel recently switched to a dedicated connector package, you can see how that works here: learn.microsoft.com/en-us/dotnet/ai/quickstarts/quickstart-local-ai
You'll definitely need to configure your app to allow functions to execute as shown in the video, but I"m not sure if that setting is different somehow for ollama.
great, can you share the repository?
Does functions works fine with local LLM, would be great to make a video for Sem Kernel limitations for local setup.
I believe recent versions of Ollama support function calls, though I haven't tested this extensively myself.
at th-cam.com/video/kCGZPhnTGHM/w-d-xo.html the tech.txt file popped up before the thing even finished... hows that possible. Its like it created the file first, then searched the web, then populated the file contents.
Interesting observations, my understanding is that the orchestrator aspect of SK doesn't necessarily do things in a specific order, it decides what order is best as it goes. So yeah it probably creates the file while it went out and searched the endpoint and then populated it.