Closed decahedron1 closed 4 months ago
@mush42 Are you able to upload the converted .ort
model?
@decahedron1
Here's the model. It is a quantized model using neural compressor
.
model.zip
Can you confirm that https://github.com/pykeio/ort/commit/80be2068296b04e8ae0a442d4b9c2d635362d07e fixed this issue by setting ort
as a git dependency:
[dependencies]
ort = { git = "https://github.com/pykeio/ort.git", rev = "80be2068296b04e8ae0a442d4b9c2d635362d07e" }
@decahedron1 I confirm that the commit fixes the issue.
@decahedron1 If possible, make a new rc release.
My model is now live at: https://mush42.github.io/libtashkeel/
Do you have any idea of what has been added to nightly to make ort work in wasm-unknown-unknown?
Just came by this: https://github.com/Liamolucko/find_wasm_import
No idea. My best guesses are https://github.com/rust-lang/rust/pull/124036 or the LLVM 18 upgrade (which should land in v1.78 stable, releasing tomorrow).
OK.
Whatever it is, it seams to add around 5-6 MB to the .wasm
module bundle.
Using stable, my build is around 6 MB when compiled with stable, jumped to 12 MB when compiled with nightly.
@decahedron1
Build and load of wasm module in the browser is successful.
But when inferencing the model, I face following error:
I'm using an
ort
model.Originally posted by @mush42 in https://github.com/pykeio/ort/issues/75#issuecomment-2081213832