mcemilg

Joined 11 years ago

Karma
139
Submissions
86
Member since
Jun 17
https://mcemilg.dev

https://github.com/mcemilg

https://twitter.com/mcemilg

View on Hacker News

Recent submissions

Comment · 2 months ago
Started from an ugly repository and ended up making history in less than 10 years, from `pytorch-pretrained-bert` to huggingface, well deserved, its fascinating to follow.
Comment · 1 year ago
From my (just a user) perspective, GPUs are expensive, they shouldn't be left standing if they're not being used.
Comment · 1 year ago
Highly delusional.
Comment · 1 year ago
I'm building a local medical AI app for Mac, recently published on the App Store. https://apple.co/4mlYANu

It uses medgemma 4B for analyzing medical images and generating diagnostic insights and reports, ofc must be used by caution, its not for real diagnostics, can be something to have another view maybe.

Currently, it supports chat and report generation, but I'm stuck on what other features to add beyond these. Also experimenting with integrating the 27B model, even with 4bit quantization, looks better than 4b.

Comment · 1 year ago
Sorry, but it's garbage for now, at least in my case. From an asynchronous agent, I would expect it to get more done in 10 minutes than a regular agent like claude code. Instead, you give it a task, wait 10 minutes, and get garbage code. Then you provide feedback, wait another 10 minutes, and still get something that doesn't compile. Meanwhile, Claude Code does this in 10 seconds and usually produces runnable code.
Comment · 2 years ago
It’s great news that they provide it for free. It’s hard to subscribe to all the LLM providers. Even with a pro subscription, you need to buy credits to be able to use with the editors, which gets very expensive if you use them a lot.

On another side, I really like the experience of coding with GitHub Copilot. It suggests code directly in your editor without needing to switch tabs or ask separately. It feels much more natural and faster than having to switch tabs and request changes from an AI, which can slow down the coding process.

Comment · 2 years ago
I believe that if Mistral is serious about advancing in open source, they should consider sharing the corpus used for training their models, at least the base models pretraining data.
Comment · 2 years ago
I am curious about how they gained access to a huge cluster to train a 7B model from scratch. It seems like they used the EuroHPC cluster, Leonardo, but I am not sure if it was provided by EuroHPC or if they paid for it. If it was a giveaway, what privileges did they have to access such a large cluster?
Comment · 2 years ago
Looks wonderful. But I would like to add this, I hate einops, it doesn't make it simple to read unfortunately.
Comment · 3 years ago
Seems like, Ilya Sutskever ate Sam Altman.
Comment · 3 years ago
"For me"

> Be kind first, be right later.

Being kind is good, we should not forget to stand up for what you believe.

> People care more about stories than facts.

It doesn't feel virtuous, in the end, what's true matters more than stories.

> Think of yourself as the person you want to be, not the person you are.

It's good to know who you really are, not just who you want to be.

> Surround yourself with people who are passionate about their work.

Being passionate doesn't always mean it's good. Even a passionate slave is still a slave.

Comment · 3 years ago
Beside this, I also hate navigating through modules. I would like to see a part of code in a single file if it will not used anywhere else or the thing is very abstract.
Comment · 3 years ago
I think it's just by a chance, Ruby could win too if the data people first gather around it.
Comment · 3 years ago

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton