Coolcoder360

I went with quantized Gemma

Well, was it quantized in a way that iphone 16 supports?

Often it’s the quantization where things break down, and the hardware needs to support the quantization, can’t run FP16 on int8 hardware… And sometimes the act of quantization can cause problems too.

And yeah, LLMs are likely going to be very hit or miss anyway.

I’ve not found them useful yet for more than basic things. I tried Ollama, it let’s you run locally, has simple setup, stays out of the way.

Blogging, Newsletter & Co.: Well, as you can see, I’m writing on Substack. There are no alternatives except to host it entirely yourself, but that doesn’t make sense to me right now.

Let me introduce you to bear blog: https://bearblog.dev/

I don’t yet use it, but I read a few of the blogs. Very RSS feed reader friendly and it’s simple without excess crap.

I think when I set up vault warden with the docker compose it had scripts to generate it’s own self-signed certificate. So it was already set up to use https.

I have a CA I created with easyrsa so I went and found the csr from vault warden and signed it with my own CA, so I didn’t have to juggle two certs.

But otherwise yeah, running it on my local LAN, no let’s encrypt.

The problem is the AI tool you’re using is making your comments longer than necessary while still saying little to nothing.

Shorter content is faster to both write and read. Using AI to fluff up your comments just makes it seem like you’re out of touch with how using it makes you sound, and trying to feel like you didn’t waste all your money paying for an AI model.

It’s also not concise at all, extra words with no purpose make it less clear and don’t make it any more helpful.

Just say what you prompted it with.

I never set up the actual bank sync, I think what I did was basically a CSV export and then import, rather than the bank sync.

But it’s been about a year so not remembering exactly how the import was.

I use actual as well.

The docker compose works really well, basically set it up once and then it works, even with running updates by pulling new container versions.

I used the account importing to start but now input everything manually and don’t do live sync.

Never heard of the other options so didn’t know about them to compare before setting up actual. I do like the methodology of actual, where it has you only budget money that actually exists in your account, that feels very sane to me.