cc @ezyang @anjali411 @dylanbespalko

As documented in #5392 using the packages for tflite and tensorflow 2.1.0 the test as in the subject line segfaults. It has now been skipped in the testsuite but this needs to be fixed.

Support error function and fresnel integrals in https://docs.scipy.org/doc/scipy/reference/special.html#error-function-and-fresnel-integrals, those are not universal functions may not need to be supported.

Hey Contributor!

Thanks for checking out HyperLearn!! Super appreciate it.

Since the package is new (only started like August 27th)..., Issues are the best place to start helping out, and or check out the Projects tab. There's a whole list of stuff I envisioned to complete.

Also, if you have a NEW idea: please post an issue and label it new enhancement.

In terms of priorities, I wanted

Great work!

I discovered in https://github.com/arogozhnikov/einops/blob/master/einops/einops.py#L199 that you also support ellipsis. Its an important feature so you may want to add it to the documentation.

Since we're going to rewrite the parser dramatically, we need some tests to make sure that the functionality of the math output is preserved

We should have a lowering API that let's us specify which dimensions should be hard-coded in the emitted code. In the taco code generator tool (command-line/web) dimensions should not be hard-coded by default.

In the current C++ API, where an expression is with respect to specific tensors (with immutable dimensions) we should hard-code in all dimension sizes.

However, when we design an API t

The docs generated for #431 at https://mratsim.github.io/Arraymancer/pca.html have broken formatting:

• Upstream bug on the type:
  
• Random bolding:
  and
  ![image](https:/

In your RL examples in rollout.ml you create the initial observation by E.reset and put it into the frame stack but not into the s_states.

For A2C, IMHO, it leads to the very first calculation of the model outputs after creation of the rollout is made over all zeros.

I think that I’m still not using xtensor’s Python bindings as they were intended. I also think that this is because this part of the documentation (including that of the cookiecutter) is rather brief. It would be great to add a bunch of examples with different use cases and best practices.

Not sure if this is what we need. To be clear, here is how these things SHOULD work: There should be a CMake option (looks like DOCS) that turns on generation of documentation. If that option is not set, then no docs are generated. If the option IS set, then docs will be generated. If the option is set and the required programs for generating docs do not exist (pdflatex and texi2pdf), then

The Scaladoc should be similar to numpy's API reference

The fp_reduction_latency benchmarks were the very first benchmark, optimization and primitive code tested in Laser.

Unfortunately it is currently very confusing.

It should be reorganized:

• 1. Multiple accumulators: https://github.com/numforge/laser/blob/af191c086b4a98c49049ecf18f5519dc6856cc77/benchmarks/fp_reduction_latency/reduction_bench.nim and https://github.com/numforge/laser/blo

CI/CD on Google Cloud

Look over to Tensorflow Official Build Status.

We can have different badges for different mediums of deployment by using Google Cloud.

The *FlatIterator type implements Iterator, which is an interface for stateful iterators. It also supports reverse iteration (by means of the .SetReverse() method).

Currently, the *FlatIterator type supports the following operations

| - | Row Major Vector | Col Major Vector | Row Major N-DImensional | Col Major N-Dimensional |
|:-:|:-----------------------:|:----------------------