My playground for experimenting with JEV and similar things (so far just Laya).
The two main things right now are jgrep.py and lgrep.py which are both
semantic grep implemented with jev and laya respectively.
For lgrep.py I use lserve.py to load the model once and keep it in memory
and then talk to it with lclient.py so it's fast.
lserve.py just exposes laya via zmq and then lclient.py accesses it.
I use uv for everything, so eg uv run lgrep.py ... or uv run jgrep.py ...
5 commits
Python
100.0%
My playground for experimenting with JEV and similar things (so far just Laya).
The two main things right now are jgrep.py and lgrep.py which are both
semantic grep implemented with jev and laya respectively.
For lgrep.py I use lserve.py to load the model once and keep it in memory
and then talk to it with lclient.py so it's fast.
lserve.py just exposes laya via zmq and then lclient.py accesses it.
I use uv for everything, so eg uv run lgrep.py ... or uv run jgrep.py ...
5 commits
Python
100.0%