C#: TrainingCollector新增q_value列, IsmctsBot传入winRate CSV: reward→q_value, ISMCTS状态带真实评估值, Neural填0 Python: 优先取ISMCTS状态+评估值训练, 无q_value的Neural状态跳过 旧CSV格式(reward)自动fallback到Neural-only过滤 信号从'这局谁赢'改为'ISMCTS评估这步有多好'
C#: TrainingCollector新增q_value列, IsmctsBot传入winRate CSV: reward→q_value, ISMCTS状态带真实评估值, Neural填0 Python: 优先取ISMCTS状态+评估值训练, 无q_value的Neural状态跳过 旧CSV格式(reward)自动fallback到Neural-only过滤 信号从'这局谁赢'改为'ISMCTS评估这步有多好'