From 635c45e1a0b2dd642112e48380b5bd0bbcffab4e Mon Sep 17 00:00:00 2001
From: Concedo <39025047+LostRuins@users.noreply.github.com>
Date: Thu, 18 Jun 2026 21:09:35 +0800
Subject: [PATCH] fix incorrect mtp layers setting (+1 squashed commits)

Squashed commits:

[8dad1a5c0] fix incorrect mtp layers setting
---
 gpttype_adapter.cpp | 2 +-
 koboldcpp.py        | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/gpttype_adapter.cpp b/gpttype_adapter.cpp
index 97dab95e0..0c0dc7436 100644
--- a/gpttype_adapter.cpp
+++ b/gpttype_adapter.cpp
@@ -893,7 +893,7 @@ static void mtp_decoding_setup(llama_model * main_model, llama_context * main_ct
     }
 
     draft_is_mtp = true;
-    speculative_state_setup(main_ctx, mtp_ctx_params, 0, COMMON_SPECULATIVE_TYPE_DRAFT_MTP);
+    speculative_state_setup(main_ctx, mtp_ctx_params, static_cast<int32_t>(main_model->n_gpu_layers()), COMMON_SPECULATIVE_TYPE_DRAFT_MTP);
 }
 
 //loads a model for speculative decoding.
diff --git a/koboldcpp.py b/koboldcpp.py
index 0459b5192..f6efb5741 100644
--- a/koboldcpp.py
+++ b/koboldcpp.py
@@ -78,7 +78,7 @@ dry_seq_break_max = 128
 extra_images_max = 4 # for kontext/qwen img
 
 # global vars
-KcppVersion = "1.115.2"
+KcppVersion = "1.116"
 showdebug = True
 kcpp_instance = None #global running instance
 global_memory = {"tunnel_url": "", "restart_target":"", "input_to_exit":False, "load_complete":False, "restart_override_base_config":"", "last_active_timestamp":datetime.now(), "triggered_sleeping":False, "current_model":"initial_model", "base_config":"", "swapReqType": None, "autoswapmode": False}