Yep, you figured out how it works! That was the easiest solution I could come up with. I'm sure theres additional context on other screens but this was a good 90/10 solution.
Thanks, yeah I do need to flesh out the debugging options. In the menu bar you can click the Dayflow icon which should allow you to view the recordings folder. The sqlite db is in that folder too if you want to poke around there as well.
Thank you! I can see that now. Checked, I have recordings. The SQLite seems to have very little data, despite recordings being present, and no cards show up in app. Possibly an LM Studio issue. I'll fiddle around with it and send an email if I can't get it working. Will test with Ollama in case there's some LMStudio error the API tester isn't catching.
This is the error I got: reason: No valid observations generated from frame fallback
Gemini 2.5 Pro is pretty expensive, mostly because videos take up a lot of tokens. It's roughly 1 million input tokens/hr, with a relatively insignificant amount of output tokens. Fortunately, Gemini has a very generous free tier, which is more than enough to cover daily usage. If you set up one paid project (and just don't consume any tokens), you can still use the free tier on a different project, and they can't train on your data.
If I had to put a grade on my own experience and evals, Gemini 2.5 pro produces A- results and qwen2.5vl is maybe like B-/C+. Obviously everything's nondetermistic, so it's hard to guarantee a level of quality.
I'm reading through papers that suggest it should be possible to get SOTA performance on local models via distillation, and that's what I'll experiment with next.
Yea, honestly I would hate if people used this to track _other_ people, especially bosses. I wanted to build something that gave people more agency to do more with their precious time, but there definitely is a fine line here.
Probably someone long ago said the same about hope for pointy sticks not being used just for hunting animals. Yet someone will likely make a pointy stick whether you do it or not.
Yep, helping people understand their distraction patterns would be an amazing feature. I find myself doing the same thing, funnily enough I also have that same Youtube extension.
Yep! Have tested it out on Qwen 2.5VL 3B and it works reasonably well on my 16GB Macbook Air. The only thing I will say is that I don't think it's a great idea to run local models on laptop battery, since it's quite compute intensive and drains kinda quickly. Have tested with Ollama and LMStudio, but you should be able to use any OpenAI compatible local server.
Would it be possible to check for the power adapter and run processing then? These are the types of things I've been thinking about for my own app: https://stardateapp.com
> Gemini leverages native video understanding for direct analysis, while Local models reconstruct understanding from individual frame descriptions - resulting in dramatically different processing complexity.
For people like me who haven't dabbled much with AI video processing and have no intuition for it, could you clarify the drawbacks of such a local-only approach vs what Gemini offers? I don't mean the performance or power/battery impact (that part is clear), just in terms of end-result and quality what the practical differences are.
I'm in the only-100%-offline camp here but would like to know what I'm missing out on since I won't even try Gemini here.
