Aditya Goel

I believe onnxoptimizer is currently not on conda-forge. Would the maintainer(s) of this package be open to me adding it to conda-forge so users may install onnxoptimizer via conda? Thanks!

win32 wheels

5

Currently it looks like wheels are not built for win32 ([skipped in CI](https://github.com/andreasvc/pyre2/blob/master/.github/workflows/release.yml#L43)). Was there a reason I should be aware of for this, and would it be possible to...

[Feature Request] Use read-write lock instead of global lock when sampling is non-mutating

2

## Motivation The use of a global threading.RLock in replay buffer prototypes limits the number of concurrent reads that can be performed on a replay buffer unnecessarily when contents are...

If node adaptation fails when node outside subgraph is referenced.

3

The described error can be reproduced with the following example: ```python import spox.opset.ai.onnx.v19 as op19 import spox.opset.ai.onnx.v18 as op18 from spox import Tensor, Var, argument, build import onnx import numpy...

### Describe the issue The OneHot operator CPU EP implementation features a division operation in calculating the output shape. When the ``indices`` input has a shape with a zero dimension,...

# Ask a Question In the specification for a label encoder it says the following: > For key look-up, bit-wise comparison is used so even a float NaN can be...

RFC: clarify `broadcast_to` semantics

7

I'm finding the ``broadcast_to`` [specification](https://data-apis.org/array-api/latest/API_specification/generated/array_api.broadcast_to.html#broadcast-to) a little underspecified. In the docs we see the following for the shape parameter: > shape (Tuple[int, ...]) – array shape. Must be compatible with...