From 04e28294d29fd7e43e622770091ab6961f3b0f10 Mon Sep 17 00:00:00 2001
From: dhb <1084714805@qq.com>
Date: Mon, 23 Sep 2024 23:53:17 +0800
Subject: [PATCH] merge parameters of spsa tune

---
 Rapfi/search/mcts/parameter.h | 31 +++++++++++++++----------------
 Rapfi/search/mcts/search.cpp  |  2 +-
 2 files changed, 16 insertions(+), 17 deletions(-)

diff --git a/Rapfi/search/mcts/parameter.h b/Rapfi/search/mcts/parameter.h
index 0e29d01..4296f67 100644
--- a/Rapfi/search/mcts/parameter.h
+++ b/Rapfi/search/mcts/parameter.h
@@ -22,29 +22,28 @@
 
 namespace Search::MCTS {
 
-constexpr float MaxNewVisitsProp = 0.2f;
+constexpr float MaxNewVisitsProp = 0.275f;
 
-constexpr float CpuctExploration     = 1.0f;
-constexpr float CpuctExplorationLog  = 0.4f;
-constexpr float CpuctExplorationBase = 500;
+constexpr float CpuctExploration     = 0.40f;
+constexpr float CpuctExplorationLog  = 0.75f;
+constexpr float CpuctExplorationBase = 336;
 constexpr float CpuctParentVisitBias = 0.1f;
 
-constexpr float CpuctUtilityStdevScale     = 0.0f;
-constexpr float CpuctUtilityVarPrior       = 0.16f;
-constexpr float CpuctUtilityVarPriorWeight = 2.0f;
+constexpr float CpuctUtilityStdevScale     = 0.035f;
+constexpr float CpuctUtilityVarPrior       = 0.18f;
+constexpr float CpuctUtilityVarPriorWeight = 2.14f;
 
-constexpr float FpuReductionMax     = 0.1f;
-constexpr float FpuLossProp         = 0.0f;
-constexpr float RootFpuReductionMax = 0.05f;
-constexpr float RootFpuLossProp     = 0.0f;
-constexpr float FpuUtilityBlendPow  = 2.0f;
+constexpr float FpuReductionMax     = 0.075f;
+constexpr float FpuLossProp         = 0.001f;
+constexpr float RootFpuReductionMax = 0.075f;
+constexpr float RootFpuLossProp     = 0.0036f;
+constexpr float FpuUtilityBlendPow  = 1.73f;
 
-constexpr uint32_t MinTranspositionSkipVisits = 10;
+constexpr uint32_t MinTranspositionSkipVisits = 12;
 
 constexpr bool  UseLCBForBestmoveSelection = true;
-constexpr float LCBStdevs =
-    4.265f;  // quantile of student's t dist(n) at confidence level 1-1e-5 as n -> inf
-constexpr float LCBMinVisitProp = 0.15f;
+constexpr float LCBStdevs                  = 5.0f;
+constexpr float LCBMinVisitProp            = 0.12f;
 
 constexpr float PolicyTemperature = 1.0f;
 
diff --git a/Rapfi/search/mcts/search.cpp b/Rapfi/search/mcts/search.cpp
index 77301f6..dac1e6a 100644
--- a/Rapfi/search/mcts/search.cpp
+++ b/Rapfi/search/mcts/search.cpp
@@ -38,7 +38,7 @@ inline float cpuctExplorationFactor(uint32_t parentVisits)
     float cpuct = CpuctExploration;
     if (CpuctExplorationLog != 0)
         cpuct += CpuctExplorationLog * std::log(1.0f + parentVisits / CpuctExplorationBase);
-    return cpuct * std::sqrt(parentVisits + CpuctParentVisitBias);
+    return cpuct * std::sqrt(parentVisits + 1e-2f);
 }
 
 /// Compute the initial utility value for unexplored children, considering first play urgency.