I've been using Pocketbase for several projects. While it's a delightful experience for solo-dev to quickly finish a project from backend to complicated frontend, I wish it has bulk-insert out of the box.
Zotero is an offline software with syncing capability. They charge for extra storage beyond 200MB, which is their only streams of revenue, you can setup a WebDAV server and point Zotero there
Because MongoDB's license change disallowed cloud service provider to host MongoDB. This reduces the competition and *probably* making MongoDB more expensive than says PostgreSQL. Which I can find a hosted solution anywhere I host my VPS.
I have access provided by the university but I didn’t dare to upload it elsewhere because my university said the downloaded file contains information that could link to me. If only we have a system for this, oh wait that thing is called Scihub
GGML is the framework for running deep neural network, mostly for interference. It's the same level as Pytorch or Tensorflow. So I would say GGML is the browser in your Javascript/React analogy.
llama.cpp is a project that uses GGML the framework under the hood, same authors. Some features were even developed in llama.cpp before being ported to GGML.
Ollama provides a user-friendly way to uses llama models. No ideas what it uses under the hood.
Isn't that the point of posting/ commenting on HN? So that we can discuss about the post? I see something on HN and after checking it, I would definitely comment on HN first.
I wouldn't count those comments as asking for more free stuff. I would think of them as feature requests, ideas, suggestions. Whether or not to implement them is totally up to the author, and the author could charge for those if they decide to implement it.
Wow. Thank you for making this. I'm frequently have to zip and unzip ~100GB of zip archive and I have to waste 10 minutes of waiting on a fast NVMe and 32 cores workstation. I know about ZSTD or pigz but the format must be zip.
I think Node.js is unsuitable for normal routing tasks like this because of single-threaded runtime. This could be achieved really easily with popular web servers like nginx or caddy. If I remember correctly, there was a lot of tutorial about serving static files using nginx in front of express.js/koa.js for better performance and that serving static files doesn't compete with the dynamic part of the website.
I was skeptical when decided to go for an MBA M2 but right now I’m absolutely love it. Being able to just pick the laptop up, walk elsewhere because of it’s extremely lightweight. Charge once every night at home and don’t have to worry about battery at all.
Note: I have a beefy server for compiling and AI training so the laptop is just a remote controller.
Correct. You need CUDA, or ROCm, MPS (native to macOS) backends for running deep learning. I found it relatively easy to train some Pytorch model on beefy server with CUDA and running interference on my Macbook Air.
As a non-native English speakers myself, ChatGPT has been a valuable tool. It's not about the small typos and errors (which tools like Grammarly could help), it's about the tone and everything.