delijati

Adding the interpreter is fairly easy. I had in project to do it for windows. If someone is interested. https://github.com/delijati/docker-shiv?tab=readme-ov-file#a...

40+; python sever side rendering + sprinkels htmx where needed; all works is fast and yes you can post forms without JS :D edit: and is still use bootstrap css because it works

Even a directly attached eGPU via thunderbold 4 was after some time too slow for machine learning aka training. As i work now fully remote i just have a beefy midi tower. Some context about eGPU [1].

But hey i'm happy to be proofed wrong ;)

[1] https://news.ycombinator.com/item?id=38890182#38905888

Can i run this in a sandbox or dry ... Not that i'm not trusting my AI Overlord ;)

Nice aka perfect timing.I just restored "some" files (40GB) with [1] But the filetype detection of photorec set some wrong file types.

Edit: It would be super helpful if the "suffix" could be added as output so i can move the files to the right directory [2] ;)

[1] https://www.cgsecurity.org/wiki/PhotoRec [2] https://github.com/google/magika/issues/63

I can recommend droidian or ubuntu touch (uborts.com) and if need (for me it is) use waydroid to run android apps https://docs.waydro.id/usage/install-on-desktops#ubuntu-debi...

Got my Razer Core X (arch linux cuda was already installed as i have in my T14 a NVIDIA GeForce MX450) so far it works pretty well, thunderbold cable in set CUDA_VISIBLE_DEVICES=1 and have a lot more cuda cores ... yay. I use it with my spare 1070ti but will add my 3090 soon.

I fell for the TrackPoint. I really miss the X200 that had no Touchpad.

Currently i do exactly that. I have a ATX box (AMD 2700x + Nvidia 3090) and access it remote, but i hate that it is idle most of the day and to transfer the data that i'm working on to it. (My notebook is my daily driver with company access)

I just bought a eGPU case and will give it a shot, will report back :)

I'm also considering that P14s AMD Gen4. But to do some ML stuff i need the damn Nvidia cuda cores :/ It has a USB4 port and could in theory run a eGPU but i can't find any infos if it would work.

Also it seams that the [1] intel 1360p is a bit faster? and i'm on the safer side because of real thunderbold support.

[1] https://nanoreview.net/en/cpu-compare/intel-core-i7-1360p-vs...

I currently have a insane long Markdown file as brain dump. Will give it a shot. Something like an overview or minimap or a way to structure like mindmap would be nice :)

It is more lightweight than PySide, PyQT and was easier to bundle as one click pkg for ubuntu-touch

Pyotherside [1] == QML+Python is really a nice combo. Build a schedule app [2] some time ago for ubuntu touch.

[1] https://github.com/thp/pyotherside [2] https://github.com/delijati/fosdem-qml/tree/master

will go on my "to try" list where i already have cosmopolitan [2]. my last setup (windows) was shiv + wine + nsis (used that as pyinstaller had some issues)[2]

[1] https://github.com/jart/cosmopolitan/issues/141#issuecomment... [2] https://github.com/delijati/docker-shiv#add-python-interpret...

Funny as i'm waiting for the Z13 Gen 2. It was announced for June. I hope it will be available this year.

Woot AMD now supports APU? I sold my notebook as i hit a wall when trying rocm [1] Is there a list oft Wirkung apu's ?

[1] https://github.com/RadeonOpenCompute/ROCm/issues/1587

How big are the executabels ? I would like to move our interence code from tensorflow to a executable

ohh thats some really nice readable c-code

I went with a minimal approach ... I still use my `.vimrc` file, removed all code completion stuff and added LSP https://github.com/VonHeikemen/lsp-zero.nvim/blob/v2.x/doc/m... ... next step treesitter but i'm in no hurry :)

just looked into the docs [1] and indeed it has CBV ... I will definitely take a closer look.

[1] https://docs.litestar.dev/latest/#feature-comparison-with-si...

Has litestar CBV (class based views) fastapi refuses to implement them so we build [1]. Or proper lifetime event [2]?

[1] https://github.com/KiraPC/fastapi-router-controller [2] https://github.com/tiangolo/fastapi/issues/617

that's just a justification to force user to create an account ... he killed the api that already had an rate limit ... its so obvious that it hurts

Exactly. That's what i'm doing. I also rank the tech that i really do not like to work with lowest ;)

Now let's wait another day or two to get to 100% ;)

You mean this? https://i.stack.imgur.com/XriCj.jpg Anyway a DE independent power management tool sounds nice.

has anyone numbers on how many queries/second in inference time chatGPT can serve?

If they do it via "pyrolyse" (burning plastic without oxygen) there is this Quantafuel [1] doing this in "large scale".

[1] https://www.quantafuel.com/

What are the current requirements to run that model?

i would really love to give it a try but with an offline LLM :/

I work on mutlirepos and i really dislike gits subtree, subrepo i use https://pypi.org/project/zc.buildout/. Yes i know i can do [1]. But editing on multirepos at the same time i can only do with zc.buildout [2] Still not perfect but it does the job.

[1]

   $ cat req.txt
   requests
   git+ssh://git@gitlab.com/foo/bar/amber.git@0.4.0
   git+ssh://git@gitlab.com/foo/bar/rebam.git@0.4.0

[2]

   $ cat buildout.cfg
   [buildout]
   extends = versions.cfg
   extensions = mr.developer
   auto-checkout = \*
   develop = .
   show-picked-versions = true
   update-versions-file = versions.cfg
   sources-dir = git-sources

   parts = py

   eggs = 
       amber
       rebam
       hammer

   [sources]
   amber = git git@gitlab.com/foo/bar/amber.git@0.4.0
   rebma = git git@gitlab.com/foo/bar/rebma.git@0.4.0

   [py]
   recipe = zc.recipe.egg
   eggs =
       ${buildout:eggs}
   interpreter = py-backend
   dependent-scripts = true