I would like to have the ability to create custom system prompts for Lumo. The background is that I want to adjust the behavior and also ensure that certain characters, such as the German sharp s (ß), are no longer used.
Like: Du verwendest ausschliesslich Schweizer Rechtschreibung („ss“ statt „ß“, z. B. „Strasse“, „dass“, „Fuss“).
Didn’t know they had an actual LLM
Anyone knows what it’s running in the back?
Copied from an earlier comment of mine. Prompt: Is Lumo training data its own, or is it based of other preestablished data?
Response: “Lumo is powered by several open-source large language models (LLMs) that have been optimized by Proton. The models used include Nemo (Mistral), OpenHands 32B, OLMO 2 32B (Allen Institute for AI), and Mistral Small 3 (Mistral). These models run exclusively on servers controlled by Proton, ensuring that all conversations remain private. Proton does not contribute any user data to the training of these models, maintaining a strong focus on user privacy.” This is what I got from Lumo.
Nice. I got nothing when I asked it
That was from the day it launched. so if they are tinkering with the backend, I at least got this out.
I also asked it on the day I saw the ennouncement posted. But it’s AI, the results are bound to be random.
Fair enough.
It’s a good thing to have an actual reputable company run an LLM model. Running performant and powerful LLMs on device just isn’t possible
It was just announced a week or two ago. As far as the models go, I don’t remember offhand, but I believe they had it listed in their press release or FAQ.
Mistral Large (I asked the model a few days ago and that’s what it said. Today it’s not saying it seems)
