rj45's learning exhaust
One thing I learned: weight streaming probably works well on GPUs

One thing I learned: weight streaming probably works well on GPUs

I wonder why it isn't a thing...

RJ Honicky
RJ Honicky
7 min read
One thing I learned: embeddings are task-specific

One thing I learned: embeddings are task-specific

RJ Honicky
RJ Honicky
5 min read
Is it the model or the data that's low rank?

Is it the model or the data that's low rank?

RJ Honicky
RJ Honicky
4 min read
RJ Honicky

RJ Honicky

8 followers

Ancient (e.g. 2009) data scientist learning in public about Gen AI and other advanced Machine Learning

Can we improve quantization by fine tuning?

Can we improve quantization by fine tuning?

RJ Honicky
8 min read
A toy problem and a bunch of models

A toy problem and a bunch of models

RJ Honicky
17 min read
Are All Large Language Models Really in 1.58 Bits?

Are All Large Language Models Really in 1.58 Bits?

RJ Honicky
21 min read