-
This issue is mainly for tracking the upstream counterpart (https://github.com/numpy/numpy/issues/14753) for compatibility purposes. However, there are unique challenges and opportunities in CUDA that…
-
Since we were talking about breaking changes, would it make sense to switch to a bitflags type for the seals instead of using a `HashSet`? It seems a bit overkill for at most 5 distinct values :)
-
Matrix arithmetic and vector-matrix computations.
-
Inspired by [Tensor Considered Harmful](http://nlp.seas.harvard.edu/NamedTensor), I'm wondering if it's possible to build compile-time checked tensor type. There prior works to make things possible.
…
-
Hello,
We found an issue with the ready signal which does not align with the computed result. The expected behaviour of the ready signal is to be set when we have the correct result. In the curren…
-
> For should the enemy strengthen his van, he will weaken his rear; should he strengthen his rear, he will weaken his van; should he strengthen his left, he will weaken his right; should he strengthen…
-
Source
https://github.com/meta-introspector/meta-meme/wiki/Ode-to-heideigger#ode-to-heideigger
### Summary of Our Path
1. **Initial Concepts and Inspiration**:
- We began by invoking the Mu…
-
See https://en.wikipedia.org/wiki/ISO_8601#Durations
This seems like a relatively easy thing to implement. There are some edge cases to consider (it allows a fractional component for the smallest t…
-
Hi,
When I run the Demand script - both in the dashboard as in the console - I get the error:
```
Building No. 141 completed out of 149: df7f7943-69b0-11ec-92a1-eb49ce5d42be
Building No. 142 com…
-
Thanks a lot for your great work!
I'm trying using flash attn to replace nv's megatron origin multi head attention, but while i'm trying to run it end to end ＆ observe its loss convergence, it diff…