Alex Morisse

Here is the issue from the code above: `gg = graphistry.graph(gdn)` gg produces `{'bindings': {'edges': , 'nodes': None, 'source': None, 'destination': None, 'node': None,...`

I've added better pruning of implicit edges and a `filter_edges` function that allows pruning after the initial `g.umap(..)` call. The new logic means that `scale=0` returns the highest weighted edges,...

Appreciate this thread -- it seems that the UMAP algorithm is fundamentally non-commutative under `fit` and `transform` -- fitting produces a simplex (via the `umap.graph_`) AND n-dim (usually 2dim) coordinates....

Old stale PR

As per https://github.com/DIGITALCRIMINAL/OnlyFans/issues/1211 , modifying `langchain/cache.py` on line 7 and adding next line, solved the issue ``` from sqlalchemy.orm import Session #, declarative_base # comment out here from sqlalchemy.ext.declarative import...

This gives them a base class that doesn't use DGL, is lightweight, and stubs the net while I am away.

Thanks @molynyk Future roadmap has Mlflow/Optuna decorators on all models -- all metrics would be exposed in mlflow UI. a quick fix could be `g._metrics` with keys `accuracy`, `loss`, etc...

Looked green from Master, checking it from `cleanup` to check and docs are still failing. Please merge if you feel its worth it to master.

Yeah I just wanted to see if I moved Head to Cleanup (since thats on our docker for the preso). I moved it back to Master. Close it if you...

I can see a gpu version -- but thinking about scale -- these types of graphs are likely O(6) and less (ie they divide big graphs into lobes of smaller...