zcbenz

5 か月前議論

It is a bug in MLX that has been fixed a few days ago: https://github.com/ml-explore/mlx/pull/3083

In the absence of hardware unified memory, CUDA will automatically copy data between CPU/GPU when there are page faults.

zcbenz

昨年議論

Most features are already supported on x86 CPUs, you can pip install mlx on Linux , and you can even use it on Windows (no official binary release yet but it is building and tests are passing).

zcbenz

昨年議論

For learning internals of ML frameworks I recommend reading the source code of MLX: https://github.com/ml-explore/mlx .

It is a modern and clean codebase without legacies, and I could understand most things without seeking external articles.

zcbenz

2 年前議論

You can use any serialization that supports numbers, in the repo I used BSER (a binary format of JSON), you can even use raw JSON but it would be very inefficient for float numbers though.

zcbenz

2 年前議論

No it is not normal, I only tested x64/arm64 macs, I will try on linux.

zcbenz

2 年前議論

That is sad, the name sisi comes from the sisi empress: https://en.m.wikipedia.org/wiki/Empress_Elisabeth_of_Austria

zcbenz

2 年前議論

A few months ago I ported ~15k lines of python code (10k are tests) to typescript, using GPT4. It cost me ~$70.

The python project is https://github.com/ml-explore/mlx and the converted project is https://github.com/frost-beta/node-mlx

I wrote a long prompt: https://github.com/frost-beta/node-mlx/blob/main/tests/promp...

The first result was almost always bad, but after manually modifying the assistant's answer, following generation usually went much better.

zcbenz

2 年前議論

I wrote a Node.js module that implemented similar feature years ago. https://github.com/atom-archive/node-runas (An updated fork can be found at https://www.npmjs.com/package/runas-redux)

zcbenz

3 年前議論

Shameless to insert my app here, also open source and provides quick access from menu bar.

https://chie.app/posts/minimalist-approach-to-using-chatgpt/

zcbenz

3 年前議論

I would argue it is not a large part a webview. The webview related code is 900 lines vs the whole project 10k lines.

zcbenz

3 年前議論

I actually asked ChatGPT what chie might mean in other languages, for multiple times, but it did not tell me this :(

投稿

Show HN: You need a few lines of JS, not a vector database

Show HN: Chrohime – cross-platform C library wrapping Chromium's native UI code

Show HN: Sisi – Semantic Image Search CLI tool, locally without third party APIs

V8 is damn fast, faster than C++ when it comes to numerics

V8 is damn fast, faster than C++ when it comes to numerics

Show HN: Train a language model from scratch with JavaScript

Show HN: node-mlx – Machine learning framework for Node.js

JavaScript's garbage collector is not very ready for machine learning

Show HN: Inference Llama2 with High-Level C++

Show HN: Chie – a cross-platform, native, and extensible desktop client for LLMs

コメント