Closed alexracape closed 3 months ago
Ideally there would be some way to load the model so that it could be queried client side. Could look for a rust library to load a pytorch model. May need to look into hosting the model.
It looks like easiest way might be to deploy model as separate microservice. Could use the same model server code which would be nice.
Issue right now is using tonic / async grpc requests and compiling to wasm
Ideally there would be some way to load the model so that it could be queried client side. Could look for a rust library to load a pytorch model. May need to look into hosting the model.