Nate Nethercott

> some additional info in case it's useful before the PR gets merged --> as per the development branch @szelok mentioned (#5038), applying the changes to the file modules/models.py on...

Should also add this I guess ``` `Accelerate` version: 0.27.2 - Platform: Linux-5.10.0-28-cloud-amd64-x86_64-with-glibc2.31 - Python version: 3.9.2 - Numpy version: 1.24.4 - PyTorch version (GPU?): 2.0.1+cu118 (True) - PyTorch XPU...

# some logic Brain dumping here so the PR has a bit of context :) ## why a new codec? The `BinaryQuantized` codec maps an element to +1 or -1...

> Hey @nnethercott, sorry for the huge delay, I was on holiday for the past weeks, and currently I'm a bit in a rush, I really need to finish #118...

**Update:** I'm keeping this PR as a draft until #132 is merged (hopefully) cause that PR would allow us to store the splitting index as a usize in the node...

@irevoire hamming's back on the menu 🍝 SplitPlaneNormals now store a random index in the node header and an empty vec. Turns out hamming is pretty good on recall too...

> Hey, the https://github.com/meilisearch/arroy/pull/130 is almost done Ayy this is massive ! I tried looking into this a few days back but hit a wall, I'm glad a better mind...

> That would be me if my dog were smaller see you space cowboy...

> Something else I noticed is that 12% of the time is spent on this function #132 reduced the number of times `update_mean` gets called by a factor of 20...

Yo @irevoire I revived this PR in [a fork](https://github.com/meilisearch/arroy/compare/main...nnethercott:arroy:add-header-to-normal) cause it's concerning mathematically that the bias still isn't being used in arroy. I'll drop my thoughts below (déso pour le...