Commit 0640577
committed
fix(llama-cpp): drop upstreamed MiniMax M3 patch
The updated llama.cpp pin already includes its specialized MiniMax M3 parser, so the vendored patch now rejects during backend preparation. Remove the temporary patch and use the upstream implementation.
Assisted-by: Codex:gpt-51 parent 4db18b9 commit 0640577
1 file changed
Lines changed: 0 additions & 225 deletions
Lines changed: 0 additions & 225 deletions
This file was deleted.
0 commit comments