* Add --defer-experts flag to defer expert mmap residency on Linux * Disable warmup when defer-experts is enabled |
||
|---|---|---|
| .. | ||
| llama.h | ||
* Add --defer-experts flag to defer expert mmap residency on Linux * Disable warmup when defer-experts is enabled |
||
|---|---|---|
| .. | ||
| llama.h | ||
Powered by TurnKey Linux.