AI@SGMK - Revision history

Elrido: some corrections, linked Lina's blog post

2026-05-12T16:45:44Z

some corrections, linked Lina's blog post

← Older revision		Revision as of 18:45, 12 May 2026
Line 22:		Line 22:
	=== Projects Tackled ===		=== Projects Tackled ===

	* Ollama with Apertus and Gemma models got installed on a Macbook (using ~~homebrew~~) ~~and on a Raspberry Pi with AI-hat~~		* Ollama with Apertus and Gemma models got installed on a Macbook (using <code>brew install --cask ollama-app</code>)
	* We explored ComfyUI, as Zoe could show us her setup. Markus had a local install of it and you can read his short intro here: [[File:ComfyUI_VeryShortIntro.pdf]]		* We explored ComfyUI, as Zoe could show us her setup. Markus had a local install of it and you can read his short intro here: [[File:ComfyUI_VeryShortIntro.pdf]]
	* On the Radxa SBC presented at the last event we did manage to install Open WebUI and connected it to the RKllama service on the device, but failed selecting the installed models or to install Apertus - This setup will be worked on between sessions and hopefully we could set this up as a service at the Bitwäscherei for folks to toy with		* On the Radxa SBC presented at the last event we did manage to install Open WebUI and connected it to the RKllama service on the device, but failed selecting the installed models or to install Apertus - This setup will be worked on between sessions and hopefully we could set this up as a service at the Bitwäscherei for folks to toy with
	* We did experiment with text and images uploaded to Open WebUI (on an earlier installation) and ollama on the Macbook - Text works very well, Open WebUI lets you provide websites as context and some models like gemma on the Mac did provide decent image-to-text output		* We did experiment with text and images uploaded to Open WebUI (on an earlier installation) and ollama on the Macbook - Text works very well, Open WebUI lets you provide websites as context and some models like gemma on the Mac did provide decent image-to-text output
	* On the Raspberry Pi ~~AI hat~~, image recognition got setup (~~editors note: please add some more details on this~~)		* On the [https://blog.linalopes.info/raspberry-pi-5-local-ai-limits/ Raspberry Pi, image recognition got setup and local LLMs did run], even without the use of an AI hat (we mistook the M2 Hat for an AI one)
	* We didn't yet look into the [https://docs.espressif.com/projects/esp-dev-kits/en/latest/esp32p4/esp32-p4x-eye/user_guide.html ESP32-P4X-EYE camera & microcontroller]		* We didn't yet look into the [https://docs.espressif.com/projects/esp-dev-kits/en/latest/esp32p4/esp32-p4x-eye/user_guide.html ESP32-P4X-EYE camera & microcontroller]

Elrido: some of what we did today

2026-05-10T16:34:12Z

some of what we did today

← Older revision		Revision as of 18:34, 10 May 2026
Line 7:		Line 7:
	— Qwen 3B		— Qwen 3B
	</blockquote>		</blockquote>

			== Sunday, Date to be defined, 13:00–16:00, at Bitwäscherei in Zurich ==

			We have not yet decided on a new date and there was a suggestion for an alternative location, if that would give us access to other setups. Otherwise we'll fall back to the Bitwäscherei. Watch this space.

	== Sunday, 10 May 2026, 13:00–16:00, at Bitwäscherei in Zurich ==		== Sunday, 10 May 2026, 13:00–16:00, at Bitwäscherei in Zurich ==
Line 16:		Line 20:
	The idea is to bring your AI capable devices to the workshop (don't forget to also bring power cables) and we can help each other getting some AI installed on it. The Bitwäscherei offers fast internet, so slurping in those models from Hugging Face and Co. should take just a blink, but of course we will still be limited by the memory on your device.		The idea is to bring your AI capable devices to the workshop (don't forget to also bring power cables) and we can help each other getting some AI installed on it. The Bitwäscherei offers fast internet, so slurping in those models from Hugging Face and Co. should take just a blink, but of course we will still be limited by the memory on your device.

	~~To water your taste buds, lets collect our ideas here below, also so we can maybe prepare a little:~~		=== Projects Tackled ===

	=== ~~Your Ideas~~ ===

	* On the Radxa SBC presented at the last event:		* Ollama with Apertus and Gemma models got installed on a Macbook (using homebrew) and on a Raspberry Pi with AI-hat
	** Try out different LLM models, ~~such as [https://huggingface.co/swiss-ai/Apertus-70B-2509~~ the ~~small~~ Apertus ~~one]~~		* We explored ComfyUI, as Zoe could show us her setup. Markus had a local install of it and you can read his short intro here: [[File:ComfyUI_VeryShortIntro.pdf]]
	** Getting an [https://docs.openwebui.com/features/chat-~~conversations/rag/ RAG]~~ setup ~~going, to~~ be ~~able to get~~ the ~~LLM~~ to ~~incorporate specific documents into their answer~~		* On the Radxa SBC presented at the last event we did manage to install Open WebUI and connected it to the RKllama service on the device, but failed selecting the installed models or to install Apertus - This setup will be worked on between sessions and hopefully we could set this up as a service at the Bitwäscherei for folks to toy with
	** Get some image ~~generation model installed, get a small webservice deployed on it that lets us have Bitwäscherei~~-~~Image~~-~~Generator service~~		* We did experiment with text and images uploaded to Open WebUI (on an earlier installation) and ollama on the Macbook - Text works very well, Open WebUI lets you provide websites as context and some models like gemma on the Mac did provide decent image-to-text output
	** Get some ~~speech recognition connected to it, such as whisper - Can we build our own offline Alexa/Siri?~~		* On the Raspberry Pi AI hat, image recognition got setup (editors note: please add some more details on this)
	* ~~What AI can realistically run on~~ the [https://docs.espressif.com/projects/esp-dev-kits/en/latest/esp32p4/esp32-p4x-eye/user_guide.html ESP32-P4X-EYE camera & microcontroller~~]? Could we maybe get it to do face location and instead of boring boxes have it overlay the [https://ghostintheshell.fandom.com/wiki/Laughing_Man Laughing Man] smiley from the GitS SoC show?~~		* We didn't yet look into the [https://docs.espressif.com/projects/esp-dev-kits/en/latest/esp32p4/esp32-p4x-eye/user_guide.html ESP32-P4X-EYE camera & microcontroller]
	* ComfyUI very short intro: [[File:ComfyUI_VeryShortIntro.pdf]]

	== Sunday, 26 April 2026, 13:00–15:00, at Bitwäscherei in Zurich ==		== Sunday, 26 April 2026, 13:00–15:00, at Bitwäscherei in Zurich ==

Markus Sing: contribution for AI installation session Mai 10th, 2026

2026-05-09T20:02:21Z

contribution for AI installation session Mai 10th, 2026

← Older revision		Revision as of 22:02, 9 May 2026
Line 26:		Line 26:
	** Get some speech recognition connected to it, such as whisper - Can we build our own offline Alexa/Siri?		** Get some speech recognition connected to it, such as whisper - Can we build our own offline Alexa/Siri?
	* What AI can realistically run on the [https://docs.espressif.com/projects/esp-dev-kits/en/latest/esp32p4/esp32-p4x-eye/user_guide.html ESP32-P4X-EYE camera & microcontroller]? Could we maybe get it to do face location and instead of boring boxes have it overlay the [https://ghostintheshell.fandom.com/wiki/Laughing_Man Laughing Man] smiley from the GitS SoC show?		* What AI can realistically run on the [https://docs.espressif.com/projects/esp-dev-kits/en/latest/esp32p4/esp32-p4x-eye/user_guide.html ESP32-P4X-EYE camera & microcontroller]? Could we maybe get it to do face location and instead of boring boxes have it overlay the [https://ghostintheshell.fandom.com/wiki/Laughing_Man Laughing Man] smiley from the GitS SoC show?
			* ComfyUI very short intro: [[File:ComfyUI_VeryShortIntro.pdf]]

	== Sunday, 26 April 2026, 13:00–15:00, at Bitwäscherei in Zurich ==		== Sunday, 26 April 2026, 13:00–15:00, at Bitwäscherei in Zurich ==

Elrido at 05:24, 2 May 2026

2026-05-02T05:24:55Z

← Older revision		Revision as of 07:24, 2 May 2026
Line 22:		Line 22:
	* On the Radxa SBC presented at the last event:		* On the Radxa SBC presented at the last event:
	** Try out different LLM models, such as [https://huggingface.co/swiss-ai/Apertus-70B-2509 the small Apertus one]		** Try out different LLM models, such as [https://huggingface.co/swiss-ai/Apertus-70B-2509 the small Apertus one]
			** Getting an [https://docs.openwebui.com/features/chat-conversations/rag/ RAG] setup going, to be able to get the LLM to incorporate specific documents into their answer
	** Get some image generation model installed, get a small webservice deployed on it that lets us have Bitwäscherei-Image-Generator service		** Get some image generation model installed, get a small webservice deployed on it that lets us have Bitwäscherei-Image-Generator service
	** Get some speech recognition connected to it, such as whisper - Can we build our own offline Alexa/Siri?		** Get some speech recognition connected to it, such as whisper - Can we build our own offline Alexa/Siri?

Elrido: next workshop

2026-05-01T07:09:44Z

next workshop

@@ Line 7: / Line 7: @@
 — Qwen 3B
 </blockquote>
 == Sunday, 26 April 2026, 13:00–15:00, at Bitwäscherei in Zurich ==
@@ Line 47: / Line 66: @@
 Presentation: [[:File:Running local AI.zip]]
-=== Planning Future Sessions ===
+== Planning Future Sessions ==
 We'll collect what topics are of interest to the participants and will try to organize further sessions to dive deeper into the topics or work on concrete applications with the tools at hand. Let's have some fun!

Elrido: added my talk

2026-04-26T10:54:59Z

added my talk

← Older revision		Revision as of 12:54, 26 April 2026
Line 44:		Line 44:

	Simon / El RIDO will show some practical applications of small LLMs, that we can run locally on single board computers (SBC) with a small NPU. For practical trials, a Radxa Rock 5b, an SBC based on the Rockchip RK3588 chipset with 16 GB of RAM and a NPU with 6 TOPS will be available on site, letting us run local, offline inference using the [https://github.com/airockchip/rknn-llm RKNN-LLM framework] in a terminal based chat. For some more bells and whistles, such an SBC or a gaming GPU (300 TOPS and more) can be combined with the [https://github.com/open-webui/open-webui Open WebUI] dashboard through the [https://github.com/ollama/ollama ollama-API]. This lets us come close to the experiance of cloudbased commercial AI chat solutions.		Simon / El RIDO will show some practical applications of small LLMs, that we can run locally on single board computers (SBC) with a small NPU. For practical trials, a Radxa Rock 5b, an SBC based on the Rockchip RK3588 chipset with 16 GB of RAM and a NPU with 6 TOPS will be available on site, letting us run local, offline inference using the [https://github.com/airockchip/rknn-llm RKNN-LLM framework] in a terminal based chat. For some more bells and whistles, such an SBC or a gaming GPU (300 TOPS and more) can be combined with the [https://github.com/open-webui/open-webui Open WebUI] dashboard through the [https://github.com/ollama/ollama ollama-API]. This lets us come close to the experiance of cloudbased commercial AI chat solutions.

			Presentation: [[:File:Running local AI.zip]]

	=== Planning Future Sessions ===		=== Planning Future Sessions ===

	We'll collect what topics are of interest to the participants and will try to organize further sessions to dive deeper into the topics or work on concrete applications with the tools at hand. Let's have some fun!		We'll collect what topics are of interest to the participants and will try to organize further sessions to dive deeper into the topics or work on concrete applications with the tools at hand. Let's have some fun!

Markus Sing: update presentation file

2026-04-25T22:01:20Z

update presentation file

← Older revision		Revision as of 00:01, 26 April 2026
Line 39:		Line 39:
	Markus will provide us with the theoretical tools to grasp how neural network operate, so we can start to make sense of all the new acronyms and terms that can make your head spin when dealing with AI.		Markus will provide us with the theoretical tools to grasp how neural network operate, so we can start to make sense of all the new acronyms and terms that can make your head spin when dealing with AI.

	~~First draft, more details on unsupervised learning and convolution layers will be added~~:		Presentation 26.4.2026: [[File:AI_SomeTheoreticalBasics.pdf]]
	[[File:~~AI_TheoreticalBasics~~.pdf]]

	=== Running Local AI Inference ===		=== Running Local AI Inference ===

Markus Sing at 19:34, 22 April 2026

2026-04-22T19:34:10Z

← Older revision		Revision as of 21:34, 22 April 2026
Line 40:		Line 40:

	First draft, more details on unsupervised learning and convolution layers will be added:		First draft, more details on unsupervised learning and convolution layers will be added:
			[[File:AI_TheoreticalBasics.pdf]]

	=== Running Local AI Inference ===		=== Running Local AI Inference ===

Markus Sing: /* Basic Theory */

2026-04-22T19:31:33Z

Basic Theory

← Older revision		Revision as of 21:31, 22 April 2026
Line 35:		Line 35:
	* What hardware is required?		* What hardware is required?

	=== ~~Basic Theory~~ ===		=== Theoretical Basics ===

	Markus will provide us with the theoretical tools to grasp how neural network operate, so we can start to make sense of all the new acronyms and terms that can make your head spin when dealing with AI.		Markus will provide us with the theoretical tools to grasp how neural network operate, so we can start to make sense of all the new acronyms and terms that can make your head spin when dealing with AI.

			First draft, more details on unsupervised learning and convolution layers will be added:

	=== Running Local AI Inference ===		=== Running Local AI Inference ===

ChrisMicro at 05:14, 20 April 2026

2026-04-20T05:14:04Z

← Older revision		Revision as of 07:14, 20 April 2026
Line 21:		Line 21:
	Chrismicro will introduce this workshop series, and will give us an overview of what AI-related hardware is out there, from tiny micro controllers (MCUs), consumer graphics cards (GPU), up to server grade neural processing units (NPU).		Chrismicro will introduce this workshop series, and will give us an overview of what AI-related hardware is out there, from tiny micro controllers (MCUs), consumer graphics cards (GPU), up to server grade neural processing units (NPU).

	== What to expect: ==		==== What to expect: ====

	* What do artificial neural networks really have in common with the human brain?		* What do artificial neural networks really have in common with the human brain?
Line 29:		Line 29:
	* What hardware is worth investing in today—and what’s coming next?		* What hardware is worth investing in today—and what’s coming next?

	== Local AI – take control of your data:==		==== Local AI – take control of your data:====
	More and more people don’t want to share their data with the cloud. Local AI opens up powerful alternatives.		More and more people don’t want to share their data with the cloud. Local AI opens up powerful alternatives.