OpenSource/ollama

Autor	SHA1 Nachricht	Datum
Jeffrey Morgan	6164f378f2 revert cuda overhead to 20%	vor 1 Jahr
Jeffrey Morgan	6566387ae3 add `TODO` for cuda overhead	vor 1 Jahr
Jeffrey Morgan	37708931fb update cuda overhead to 20% to fix crashes when switching between models and large context sizes	vor 1 Jahr
Jeffrey Morgan	f6cb0a553c update cuda overhead to 15% or 400MiB	vor 1 Jahr
Jeffrey Morgan	2680078c13 fix build on linux	vor 1 Jahr
Jeffrey Morgan	f1b7e5f560 update overhead to 15%	vor 1 Jahr
Jeffrey Morgan	cb534e6ac2 use 10% vram overhead for cuda	vor 1 Jahr
Jeffrey Morgan	08f1e18965 Offload layers to GPU based on new model size estimates (#1850)	vor 1 Jahr
Daniel Hiltgen	d74ce6bd4f Detect very old CUDA GPUs and fall back to CPU	vor 1 Jahr
Daniel Hiltgen	a2ad952440 Fix windows system memory lookup	vor 1 Jahr
Daniel Hiltgen	d966b730ac Switch windows build to fully dynamic	vor 1 Jahr
Daniel Hiltgen	7555ea44f8 Revamp the dynamic library shim	vor 1 Jahr
Daniel Hiltgen	1b991d0ba9 Refine build to support CPU only	vor 1 Jahr
Daniel Hiltgen	35934b2e05 Adapted rocm support to cgo based llama.cpp	vor 1 Jahr