HN user

t0mk

121 karma
Posts11
Comments87
View on HN

I think there's a lot of sideloading. Buying non-{english,mandarin,spanish} books happens often in other book eshops, usually language- or -country specific. If you supply your kindle ID, they are able to send it straight to the reader, but the book collection is then stored in the local eshop (or more eshops if you buy from multiple). That's why Calibre exists.

Yesterday I realized that I can just copy an epub to a directory on my Kobo reader to the `root/<auhtor's name>` of the mountpoint in /media/t0mk/koboreader (on Ubunutu 22.04). It will show up in the list of books on the device.

For some reason I thought that uploading books to the Kobo reader was somehow complicated, and I only used Caliber to get books from laptop to Kobo.

I know Calibre has lots of other functions, but I didn't like the interface and I'm happy I don't have to use it, or any other application like that, anymore.

At first I thought that in effect, this project only removes a couple of Ctrl+Enter keystrokes in Jupyter-notebook workflow. But after trying out the intro I think it looks good, I really like the simple convert to a webapp.

I wonder if the state/data in the generated app are stored server-side of sent to browser.

I went through the slider example in the intro and I noticed that when I change the icon, the slider position goes back to 1. I tried to fix it so that the slider-selected value is preserved over icon changes, but didn't manage, it doesn't seem straightforward.

Gemini AI 3 years ago

Google has the possibility to roll and integrate small LLM(!) to the Pixel phones, that's something OpenAI can't do easily. Too bad MSFT dropped the Windows phone.

I agree so much! I've always shivered when they asked me what tech I'm passionate about on job interviews.

When I hear/read people who say that they're passionate about a javascript lib or a programming language, it sounds really disingenuous to me (or like marketing bs, depending on context). Although they probably mean well, in most cases.

Talk-Llama 3 years ago

I was struggling with the same error on PopOS 22.04, and this helped: https://github.com/ggerganov/whisper.cpp/issues/352#issuecom...

I'm not sure what changed, but basically I purged ffmpeg and libsdl2-dev and the `make` in the root of the repo. Then I installed libsdl2 and ffmpeg and `make talk-llama`.

It's quite slow on 4 core i7-8550U and 16 GB of RAM.

basically, in the root of the repo:

$ sudo apt purge ffmpeg $ make clean $ git pull $ make $ sudo apt install libsdl2-dev $ make talk-llama $ ./talk-llama -mw ./models/ggml-small.en.bin -ml ../llama.cpp/models/llama-2-13b.Q4_0.gguf -p "t0mk" -t 8\n\n

HTH

In Finland, you can get credit data of a person from official source (or service resellers) for ~9 EUR. You need to know their social security number though. It's used by landlords (private and corporate) to vet potential tenants.

Not sure if there's a telegram bot for that yet :D

Was the data of Orben & Przybylski (2019) public? It seems that if it would be public, and it was properly labeled (social media vs digital media, boys vs girls), the obscured relationship between social media and the quality of girls' mental health could be discovered by anybody looking at a simple correlation matrix of the data in Jupyter and Pandas.

I had some negative feedback about the fact that the cells are changing color. It because the price range of visible area is mapped on the whole color pallette on every zoom or pan.

I thought that it will be helpful in order to see differences in prices on higher zoom on an area that are quite similarly priced from global point of view. But some people apparently find it annoying.