Comments (4)
I think onnx-parse
is simple and clear to the point. onnx-unify
could also work but I think it's not as clear.
from burn.
I think we can still keep the crate in the burn repo, since it's gonna be easier to update it, but we can give it a name without the burn prefix, something like onnx-loader
from burn.
This is a great idea. From the beginning I was intending Burn's ONNX IR to be its own crate. For now, however, we should keep it within the burn repo with a non-burn name. We are still making changing to the IR. In addition to missing ONNX node types, we need to add three big features: Subgraphs (#724), functions (#723) and scalar graph inputs (#1688).
Once we separate ONNX IR, we should put forward a design document that clearly states design goals that works for other projects as well.
Here are the names, I would like to suggest. It would be great if we somehow convey an idea that we are unifying ONNX representation from various OpSets into one.
- onnx-unify (I prefer this).
- onnx-north
- onnx-irgen
- onnx-nexus
- onnx-simplify
- onnx-unite
- onnx-parse
- onnx-clarity
- onnx-unified
- onnx-ir
- onnx-easy
- onnx-bridge
- onnx-harmony
- onnx-meridian
- onnx-fusion
@skewballfox @laggui @louisfd @nathanielsimard what do you think about the names?
from burn.
Same as @laggui, but I also like onnx-ir
!
from burn.
Related Issues (20)
- The trait `std::clone::Clone` is not implemented for `BenchmarkModuleRecord<B>
- [Book] Add custom dataset, loader and batcher detailed example HOT 1
- Crate zip 1.3.0 is yanked
- Alternative design for node, argument handling in burn import HOT 2
- Crate zip 1.3.0 is yanked
- Remove the Copy restriction from `impl<const N: usize, T, B> Module<B> for [T; N]`
- Batch matrix multiply leads to vulkan error on WGPU HOT 2
- CUDA backend does not work with rust nightly HOT 2
- Distributions from space bounds
- Backward pass has mismatched dimensions HOT 15
- burn-import: Add optional or configurable logging to generated burn graph HOT 1
- Crate zip 1.3.0 is yanked
- burn-import initializer tensor not added to scope HOT 16
- to_device not working HOT 2
- Loading record on initialized model set `Option<Module>` to None HOT 6
- The `model.load_record()` method turns off the activation function during the forward pass HOT 2
- Trouble importing FaceONNX detector model: `Only tensor indices is valid` HOT 2
- [Cube] Polishing
- "condvar wait not supported" in wasm-bindgen tests in browser with WebGPU backend
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from burn.