]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/commit
graph : avoid huge warm-up graphs for MoE models (#14753)
authorGeorgi Gerganov <redacted>
Fri, 18 Jul 2025 11:31:15 +0000 (14:31 +0300)
committerGitHub <redacted>
Fri, 18 Jul 2025 11:31:15 +0000 (14:31 +0300)
commitd498af3d5a00f96bdd37b534860f03a6d9e98d39
tree5d38fc322fec3190c6bd695b393241d7e2c57825
parenteacdeb5bfcb6c6cd54461fd0e9f04cab78bf975b
graph : avoid huge warm-up graphs for MoE models (#14753)

* graph : avoid huge warm-up graphs for MoE models

ggml-ci

* cont : bump max nodes to 8x model tensors
src/llama-context.cpp
src/llama-graph.cpp