jacobgorm

Add support for float16 (half-precision floats) and related operations such as hgemm()

28

I am using BLIS for neural networks on embedded platforms (mostly ARMv8a), and I would like to reap the potential memory savings as well as possibly some speedups from running...

nvmpi-patched ffmpeg encode appears to hang for nonstandard resolutions

5

$ ffmpeg -loglevel debug -y -f rawvideo -s 480x272 -i /dev/zero -c:v h264_nvmpi foo.mp4 ffmpeg version 61853f7 Copyright (c) 2000-2019 the FFmpeg developers built with gcc 7 (Ubuntu/Linaro 7.4.0-1ubuntu1~18.04.1) configuration:...

Hashbrown crash due to bad malloc

1

hi, I am working on a Tauri project, and after updating to Tauri 1.5.2 I have been hitting this when using the auto-update feature: ```Thread 0 Crashed:: main Dispatch queue:...

### Describe the bug When attempting to cross-compile from Linux to Windows, I am encountering errors about lacking the GTK "appindicator". Looking at the code in tooling/cli/src/build.rs, it is clear...

bias defaults to True for nn.Conv2d

1

Thanks for publishing this quite interesting work. I believe it is customary to use either bias or BatchNorm, but not both at once. the nn.Conv2d layers not have bias=False added...

**Describe the bug** According the announcements from Microsoft, and according to AMD driver feature checks, HEVC video decoding and encoding is supported on AMD hardware under D3D12. However, any attempt...

Lost support for MacOS < 15.0 Sequoia

7

The commit 06387ae55 adds a call to compile_options.setMathMode() that is only supported from MacOS 15.0 and up. I am using Candle for an application that requires support for older MacOS...