-
Explanation
I'm trying to change 1d_systolic_convolution's shard grid.
What I expected is activation's size is small enough to run in GS, that should be run on WH with same shard_grid.
But I fo…
-
Just building the tools under mingw64:
```
$ gcc --version
gcc.exe (Rev2, Built by MSYS2 project) 13.2.0
Copyright (C) 2023 Free Software Foundation, Inc.
This is free software; see the source …
-
### Search before asking
- [X] I have searched the Ultralytics YOLO [issues](https://github.com/ultralytics/ultralytics/issues) and [discussions](https://github.com/ultralytics/ultralytics/discussion…
-
I printed the value of -1 in triton and the result was 4294967295. I guess there is no concept of negative numbers in cuda.
But I printed the value of the expression 1>=0 in the triton, and the pri…
-
This 'issue' describes why certain MW changes were made today.
As described in https://github.com/funderburkjim/MWvlex/blob/master/step0/readme.txt,
step0 aims to categorize records of MW which are r…
-
Some users prefer and need the applications they use to be on f-droid.
### User who prefers to use F-Droid:
https://free2z.com/converse/e503693d-0dcc-4c87-8ef2-68962b9f044e
### What is F-Droid?
F-D…
-
logpt = F.log_softmax(input) # N*H*W,C
UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.
-
I wanted to implement the non local module in the mask rcnn pipeline. I've already written it as a standalone module but I'm not sure how to include it in the detectron codebase so that I can jointly …
-
I found a bug in llvm11 and it still exists in llvm18. The problem can be replicated by compiling with command: `clang++ 1.cpp -O2`, when use `-O1`, the result is expected. For a brief look, the probl…
-
Focus on some of the "before/after" questions in the surveys, that demonstrate that participants believe that they learned something.
@arokem and @aaarendt should look back at surveys from GHW and…