mirror of
https://github.com/tinygrad/tinygrad.git
synced 2026-06-24 02:14:17 +00:00
* remove cpu and torch backends * don't copy to cpu * use clang instead of cpu * multitensor gathers on the first device * clang is cpu + use default * fixup * bugfix |
||
|---|---|---|
| .. | ||
| workflows | ||