{ "cells": [ { "cell_type": "code", "execution_count": 1, "id": "7d4e1012-3ee7-4482-b536-a5740f73a074", "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "2025-08-02 09:46:05,919 - INFO - ๐ŸŽฏ TD Learning Implementation - Complete with Logging\n", "2025-08-02 09:46:05,920 - INFO - Environment: 5 states (0 to 4)\n", "2025-08-02 09:46:05,920 - INFO - Goal: Reach terminal state 4 for reward +10\n", "2025-08-02 09:46:05,920 - INFO - \n", "๐Ÿš€ Starting TD Learning Training for 100 episodes\n", "2025-08-02 09:46:05,920 - INFO - Parameters - Alpha: 0.1, Gamma: 0.9\n", "2025-08-02 09:46:05,920 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,920 - INFO - TD Update - State: 0, Reward: -1, Target: -1.000, TD Error: -1.000, New V(s): -0.100\n", "2025-08-02 09:46:05,921 - INFO - TD Update - State: 0, Reward: 0, Target: 0.000, TD Error: 0.100, New V(s): -0.090\n", "2025-08-02 09:46:05,921 - INFO - TD Update - State: 1, Reward: 0, Target: 0.000, TD Error: 0.000, New V(s): 0.000\n", "2025-08-02 09:46:05,921 - INFO - TD Update - State: 2, Reward: -1, Target: -1.000, TD Error: -1.000, New V(s): -0.100\n", "2025-08-02 09:46:05,921 - INFO - TD Update - State: 2, Reward: 0, Target: 0.000, TD Error: 0.100, New V(s): -0.090\n", "2025-08-02 09:46:05,921 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 10.000, New V(s): 1.000\n", "2025-08-02 09:46:05,921 - INFO - Episode Complete - Total Reward: 8, Avg TD Error: 2.033\n", "2025-08-02 09:46:05,922 - INFO - \n", "๐Ÿ“Š Episode 0 Summary:\n", "2025-08-02 09:46:05,922 - INFO - Value Function: [-0.09 0. -0.09 1. 0. ]\n", "2025-08-02 09:46:05,922 - INFO - Recent Avg Reward: 8.00\n", "2025-08-02 09:46:05,922 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,922 - INFO - TD Update - State: 0, Reward: 0, Target: 0.000, TD Error: 0.090, New V(s): -0.081\n", "2025-08-02 09:46:05,922 - INFO - TD Update - State: 1, Reward: -1, Target: -1.000, TD Error: -1.000, New V(s): -0.100\n", "2025-08-02 09:46:05,922 - INFO - TD Update - State: 1, Reward: 0, Target: -0.081, TD Error: 0.019, New V(s): -0.098\n", "2025-08-02 09:46:05,923 - INFO - TD Update - State: 2, Reward: 0, Target: 0.900, TD Error: 0.990, New V(s): 0.009\n", "2025-08-02 09:46:05,923 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 9.000, New V(s): 1.900\n", "2025-08-02 09:46:05,923 - INFO - Episode Complete - Total Reward: 9, Avg TD Error: 2.220\n", "2025-08-02 09:46:05,923 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,923 - INFO - TD Update - State: 0, Reward: 0, Target: -0.088, TD Error: -0.007, New V(s): -0.082\n", "2025-08-02 09:46:05,923 - INFO - TD Update - State: 1, Reward: -1, Target: -1.088, TD Error: -0.990, New V(s): -0.197\n", "2025-08-02 09:46:05,924 - INFO - TD Update - State: 1, Reward: 0, Target: 0.008, TD Error: 0.205, New V(s): -0.177\n", "2025-08-02 09:46:05,924 - INFO - TD Update - State: 2, Reward: -1, Target: -0.992, TD Error: -1.001, New V(s): -0.091\n", "2025-08-02 09:46:05,924 - INFO - TD Update - State: 2, Reward: -1, Target: -1.082, TD Error: -0.991, New V(s): -0.190\n", "2025-08-02 09:46:05,924 - INFO - TD Update - State: 2, Reward: 0, Target: 1.710, TD Error: 1.900, New V(s): -0.000\n", "2025-08-02 09:46:05,924 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 8.100, New V(s): 2.710\n", "2025-08-02 09:46:05,924 - INFO - Episode Complete - Total Reward: 7, Avg TD Error: 1.885\n", "2025-08-02 09:46:05,924 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,925 - INFO - TD Update - State: 0, Reward: -1, Target: -1.074, TD Error: -0.992, New V(s): -0.181\n", "2025-08-02 09:46:05,925 - INFO - TD Update - State: 0, Reward: -1, Target: -1.163, TD Error: -0.982, New V(s): -0.279\n", "2025-08-02 09:46:05,925 - INFO - TD Update - State: 0, Reward: 0, Target: -0.159, TD Error: 0.120, New V(s): -0.267\n", "2025-08-02 09:46:05,925 - INFO - TD Update - State: 1, Reward: -1, Target: -1.159, TD Error: -0.982, New V(s): -0.275\n", "2025-08-02 09:46:05,925 - INFO - TD Update - State: 1, Reward: -1, Target: -1.247, TD Error: -0.973, New V(s): -0.372\n", "2025-08-02 09:46:05,925 - INFO - TD Update - State: 1, Reward: -1, Target: -1.335, TD Error: -0.963, New V(s): -0.468\n", "2025-08-02 09:46:05,925 - INFO - TD Update - State: 1, Reward: -1, Target: -1.422, TD Error: -0.953, New V(s): -0.564\n", "2025-08-02 09:46:05,926 - INFO - TD Update - State: 1, Reward: 0, Target: -0.000, TD Error: 0.564, New V(s): -0.507\n", "2025-08-02 09:46:05,926 - INFO - TD Update - State: 2, Reward: 0, Target: 2.439, TD Error: 2.439, New V(s): 0.244\n", "2025-08-02 09:46:05,926 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 7.290, New V(s): 3.439\n", "2025-08-02 09:46:05,926 - INFO - Episode Complete - Total Reward: 4, Avg TD Error: 1.626\n", "2025-08-02 09:46:05,926 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,926 - INFO - TD Update - State: 0, Reward: -1, Target: -1.240, TD Error: -0.973, New V(s): -0.364\n", "2025-08-02 09:46:05,927 - INFO - TD Update - State: 0, Reward: 0, Target: -0.457, TD Error: -0.092, New V(s): -0.374\n", "2025-08-02 09:46:05,927 - INFO - TD Update - State: 1, Reward: 0, Target: 0.219, TD Error: 0.727, New V(s): -0.435\n", "2025-08-02 09:46:05,927 - INFO - TD Update - State: 2, Reward: 0, Target: 3.095, TD Error: 2.851, New V(s): 0.529\n", "2025-08-02 09:46:05,927 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 6.561, New V(s): 4.095\n", "2025-08-02 09:46:05,927 - INFO - Episode Complete - Total Reward: 9, Avg TD Error: 2.241\n", "2025-08-02 09:46:05,927 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,927 - INFO - TD Update - State: 0, Reward: -1, Target: -1.336, TD Error: -0.963, New V(s): -0.470\n", "2025-08-02 09:46:05,928 - INFO - TD Update - State: 0, Reward: 0, Target: -0.391, TD Error: 0.079, New V(s): -0.462\n", "2025-08-02 09:46:05,928 - INFO - TD Update - State: 1, Reward: -1, Target: -1.391, TD Error: -0.957, New V(s): -0.530\n", "2025-08-02 09:46:05,928 - INFO - TD Update - State: 1, Reward: 0, Target: 0.476, TD Error: 1.006, New V(s): -0.430\n", "2025-08-02 09:46:05,928 - INFO - TD Update - State: 2, Reward: -1, Target: -0.524, TD Error: -1.053, New V(s): 0.424\n", "2025-08-02 09:46:05,928 - INFO - TD Update - State: 2, Reward: -1, Target: -0.619, TD Error: -1.042, New V(s): 0.319\n", "2025-08-02 09:46:05,928 - INFO - TD Update - State: 2, Reward: 0, Target: 3.686, TD Error: 3.366, New V(s): 0.656\n", "2025-08-02 09:46:05,929 - INFO - TD Update - State: 3, Reward: -1, Target: 2.686, TD Error: -1.410, New V(s): 3.954\n", "2025-08-02 09:46:05,929 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 6.046, New V(s): 4.559\n", "2025-08-02 09:46:05,929 - INFO - Episode Complete - Total Reward: 5, Avg TD Error: 1.769\n", "2025-08-02 09:46:05,929 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,930 - INFO - TD Update - State: 0, Reward: 0, Target: -0.387, TD Error: 0.075, New V(s): -0.454\n", "2025-08-02 09:46:05,930 - INFO - TD Update - State: 1, Reward: 0, Target: 0.590, TD Error: 1.020, New V(s): -0.328\n", "2025-08-02 09:46:05,930 - INFO - TD Update - State: 2, Reward: 0, Target: 4.103, TD Error: 3.447, New V(s): 1.001\n", "2025-08-02 09:46:05,930 - INFO - TD Update - State: 3, Reward: -1, Target: 3.103, TD Error: -1.456, New V(s): 4.413\n", "2025-08-02 09:46:05,930 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 5.587, New V(s): 4.972\n", "2025-08-02 09:46:05,931 - INFO - Episode Complete - Total Reward: 9, Avg TD Error: 2.317\n", "2025-08-02 09:46:05,931 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,931 - INFO - TD Update - State: 0, Reward: -1, Target: -1.409, TD Error: -0.955, New V(s): -0.550\n", "2025-08-02 09:46:05,931 - INFO - TD Update - State: 0, Reward: 0, Target: -0.295, TD Error: 0.255, New V(s): -0.524\n", "2025-08-02 09:46:05,931 - INFO - TD Update - State: 1, Reward: 0, Target: 0.901, TD Error: 1.228, New V(s): -0.205\n", "2025-08-02 09:46:05,931 - INFO - TD Update - State: 2, Reward: 0, Target: 4.475, TD Error: 3.474, New V(s): 1.348\n", "2025-08-02 09:46:05,932 - INFO - TD Update - State: 3, Reward: -1, Target: 3.475, TD Error: -1.497, New V(s): 4.822\n", "2025-08-02 09:46:05,932 - INFO - TD Update - State: 3, Reward: -1, Target: 3.340, TD Error: -1.482, New V(s): 4.674\n", "2025-08-02 09:46:05,932 - INFO - TD Update - State: 3, Reward: -1, Target: 3.207, TD Error: -1.467, New V(s): 4.527\n", "2025-08-02 09:46:05,932 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 5.473, New V(s): 5.074\n", "2025-08-02 09:46:05,932 - INFO - Episode Complete - Total Reward: 6, Avg TD Error: 1.979\n", "2025-08-02 09:46:05,933 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,933 - INFO - TD Update - State: 0, Reward: -1, Target: -1.472, TD Error: -0.948, New V(s): -0.619\n", "2025-08-02 09:46:05,933 - INFO - TD Update - State: 0, Reward: 0, Target: -0.184, TD Error: 0.435, New V(s): -0.576\n", "2025-08-02 09:46:05,933 - INFO - TD Update - State: 1, Reward: -1, Target: -1.184, TD Error: -0.980, New V(s): -0.303\n", "2025-08-02 09:46:05,933 - INFO - TD Update - State: 1, Reward: -1, Target: -1.273, TD Error: -0.970, New V(s): -0.400\n", "2025-08-02 09:46:05,933 - INFO - TD Update - State: 1, Reward: -1, Target: -1.360, TD Error: -0.960, New V(s): -0.496\n", "2025-08-02 09:46:05,934 - INFO - TD Update - State: 1, Reward: -1, Target: -1.446, TD Error: -0.950, New V(s): -0.591\n", "2025-08-02 09:46:05,934 - INFO - TD Update - State: 1, Reward: 0, Target: 1.213, TD Error: 1.804, New V(s): -0.410\n", "2025-08-02 09:46:05,934 - INFO - TD Update - State: 2, Reward: 0, Target: 4.567, TD Error: 3.219, New V(s): 1.670\n", "2025-08-02 09:46:05,934 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 4.926, New V(s): 5.567\n", "2025-08-02 09:46:05,934 - INFO - Episode Complete - Total Reward: 5, Avg TD Error: 1.688\n", "2025-08-02 09:46:05,934 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,934 - INFO - TD Update - State: 0, Reward: 0, Target: -0.369, TD Error: 0.206, New V(s): -0.555\n", "2025-08-02 09:46:05,935 - INFO - TD Update - State: 1, Reward: -1, Target: -1.369, TD Error: -0.959, New V(s): -0.506\n", "2025-08-02 09:46:05,935 - INFO - TD Update - State: 1, Reward: 0, Target: 1.503, TD Error: 2.009, New V(s): -0.305\n", "2025-08-02 09:46:05,935 - INFO - TD Update - State: 2, Reward: 0, Target: 5.010, TD Error: 3.340, New V(s): 2.004\n", "2025-08-02 09:46:05,935 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 4.433, New V(s): 6.010\n", "2025-08-02 09:46:05,936 - INFO - Episode Complete - Total Reward: 9, Avg TD Error: 2.190\n", "2025-08-02 09:46:05,936 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,936 - INFO - TD Update - State: 0, Reward: -1, Target: -1.500, TD Error: -0.944, New V(s): -0.650\n", "2025-08-02 09:46:05,936 - INFO - TD Update - State: 0, Reward: -1, Target: -1.585, TD Error: -0.935, New V(s): -0.743\n", "2025-08-02 09:46:05,936 - INFO - TD Update - State: 0, Reward: -1, Target: -1.669, TD Error: -0.926, New V(s): -0.836\n", "2025-08-02 09:46:05,937 - INFO - TD Update - State: 0, Reward: -1, Target: -1.752, TD Error: -0.916, New V(s): -0.927\n", "2025-08-02 09:46:05,937 - INFO - TD Update - State: 0, Reward: 0, Target: -0.275, TD Error: 0.652, New V(s): -0.862\n", "2025-08-02 09:46:05,937 - INFO - TD Update - State: 1, Reward: -1, Target: -1.275, TD Error: -0.969, New V(s): -0.402\n", "2025-08-02 09:46:05,937 - INFO - TD Update - State: 1, Reward: -1, Target: -1.362, TD Error: -0.960, New V(s): -0.498\n", "2025-08-02 09:46:05,937 - INFO - TD Update - State: 1, Reward: 0, Target: 1.804, TD Error: 2.302, New V(s): -0.268\n", "2025-08-02 09:46:05,937 - INFO - TD Update - State: 2, Reward: 0, Target: 5.409, TD Error: 3.405, New V(s): 2.345\n", "2025-08-02 09:46:05,938 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 3.990, New V(s): 6.409\n", "2025-08-02 09:46:05,938 - INFO - Episode Complete - Total Reward: 4, Avg TD Error: 1.600\n", "2025-08-02 09:46:05,938 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,938 - INFO - TD Update - State: 0, Reward: 0, Target: -0.241, TD Error: 0.621, New V(s): -0.800\n", "2025-08-02 09:46:05,938 - INFO - TD Update - State: 1, Reward: -1, Target: -1.241, TD Error: -0.973, New V(s): -0.365\n", "2025-08-02 09:46:05,938 - INFO - TD Update - State: 1, Reward: -1, Target: -1.329, TD Error: -0.963, New V(s): -0.462\n", "2025-08-02 09:46:05,939 - INFO - TD Update - State: 1, Reward: 0, Target: 2.110, TD Error: 2.572, New V(s): -0.205\n", "2025-08-02 09:46:05,939 - INFO - TD Update - State: 2, Reward: 0, Target: 5.768, TD Error: 3.424, New V(s): 2.687\n", "2025-08-02 09:46:05,939 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 3.591, New V(s): 6.768\n", "2025-08-02 09:46:05,939 - INFO - Episode Complete - Total Reward: 8, Avg TD Error: 2.024\n", "2025-08-02 09:46:05,939 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,939 - INFO - TD Update - State: 0, Reward: -1, Target: -1.720, TD Error: -0.920, New V(s): -0.892\n", "2025-08-02 09:46:05,939 - INFO - TD Update - State: 0, Reward: 0, Target: -0.184, TD Error: 0.708, New V(s): -0.821\n", "2025-08-02 09:46:05,940 - INFO - TD Update - State: 1, Reward: -1, Target: -1.184, TD Error: -0.980, New V(s): -0.303\n", "2025-08-02 09:46:05,940 - INFO - TD Update - State: 1, Reward: -1, Target: -1.272, TD Error: -0.970, New V(s): -0.400\n", "2025-08-02 09:46:05,940 - INFO - TD Update - State: 1, Reward: -1, Target: -1.360, TD Error: -0.960, New V(s): -0.496\n", "2025-08-02 09:46:05,940 - INFO - TD Update - State: 1, Reward: -1, Target: -1.446, TD Error: -0.950, New V(s): -0.591\n", "2025-08-02 09:46:05,940 - INFO - TD Update - State: 1, Reward: -1, Target: -1.532, TD Error: -0.941, New V(s): -0.685\n", "2025-08-02 09:46:05,940 - INFO - TD Update - State: 1, Reward: -1, Target: -1.616, TD Error: -0.932, New V(s): -0.778\n", "2025-08-02 09:46:05,940 - INFO - TD Update - State: 1, Reward: 0, Target: 2.418, TD Error: 3.196, New V(s): -0.458\n", "2025-08-02 09:46:05,941 - INFO - TD Update - State: 2, Reward: -1, Target: 1.418, TD Error: -1.269, New V(s): 2.560\n", "2025-08-02 09:46:05,941 - INFO - TD Update - State: 2, Reward: -1, Target: 1.304, TD Error: -1.256, New V(s): 2.434\n", "2025-08-02 09:46:05,941 - INFO - TD Update - State: 2, Reward: 0, Target: 6.092, TD Error: 3.657, New V(s): 2.800\n", "2025-08-02 09:46:05,941 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 3.232, New V(s): 7.092\n", "2025-08-02 09:46:05,941 - INFO - Episode Complete - Total Reward: 1, Avg TD Error: 1.536\n", "2025-08-02 09:46:05,941 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,942 - INFO - TD Update - State: 0, Reward: 0, Target: -0.412, TD Error: 0.409, New V(s): -0.780\n", "2025-08-02 09:46:05,942 - INFO - TD Update - State: 1, Reward: 0, Target: 2.520, TD Error: 2.978, New V(s): -0.160\n", "2025-08-02 09:46:05,942 - INFO - TD Update - State: 2, Reward: 0, Target: 6.382, TD Error: 3.582, New V(s): 3.158\n", "2025-08-02 09:46:05,942 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 2.908, New V(s): 7.382\n", "2025-08-02 09:46:05,942 - INFO - Episode Complete - Total Reward: 10, Avg TD Error: 2.469\n", "2025-08-02 09:46:05,942 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,942 - INFO - TD Update - State: 0, Reward: -1, Target: -1.702, TD Error: -0.922, New V(s): -0.872\n", "2025-08-02 09:46:05,943 - INFO - TD Update - State: 0, Reward: -1, Target: -1.785, TD Error: -0.913, New V(s): -0.964\n", "2025-08-02 09:46:05,943 - INFO - TD Update - State: 0, Reward: 0, Target: -0.144, TD Error: 0.819, New V(s): -0.882\n", "2025-08-02 09:46:05,943 - INFO - TD Update - State: 1, Reward: -1, Target: -1.144, TD Error: -0.984, New V(s): -0.259\n", "2025-08-02 09:46:05,943 - INFO - TD Update - State: 1, Reward: -1, Target: -1.233, TD Error: -0.974, New V(s): -0.356\n", "2025-08-02 09:46:05,943 - INFO - TD Update - State: 1, Reward: 0, Target: 2.843, TD Error: 3.199, New V(s): -0.036\n", "2025-08-02 09:46:05,943 - INFO - TD Update - State: 2, Reward: 0, Target: 6.644, TD Error: 3.486, New V(s): 3.507\n", "2025-08-02 09:46:05,944 - INFO - TD Update - State: 3, Reward: -1, Target: 5.644, TD Error: -1.738, New V(s): 7.209\n", "2025-08-02 09:46:05,944 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 2.791, New V(s): 7.488\n", "2025-08-02 09:46:05,944 - INFO - Episode Complete - Total Reward: 5, Avg TD Error: 1.758\n", "2025-08-02 09:46:05,944 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,944 - INFO - TD Update - State: 0, Reward: -1, Target: -1.794, TD Error: -0.912, New V(s): -0.973\n", "2025-08-02 09:46:05,944 - INFO - TD Update - State: 0, Reward: 0, Target: -0.033, TD Error: 0.940, New V(s): -0.879\n", "2025-08-02 09:46:05,945 - INFO - TD Update - State: 1, Reward: -1, Target: -1.033, TD Error: -0.996, New V(s): -0.136\n", "2025-08-02 09:46:05,945 - INFO - TD Update - State: 1, Reward: -1, Target: -1.122, TD Error: -0.986, New V(s): -0.235\n", "2025-08-02 09:46:05,945 - INFO - TD Update - State: 1, Reward: -1, Target: -1.211, TD Error: -0.977, New V(s): -0.332\n", "2025-08-02 09:46:05,945 - INFO - TD Update - State: 1, Reward: 0, Target: 3.156, TD Error: 3.488, New V(s): 0.017\n", "2025-08-02 09:46:05,945 - INFO - TD Update - State: 2, Reward: -1, Target: 2.156, TD Error: -1.351, New V(s): 3.372\n", "2025-08-02 09:46:05,945 - INFO - TD Update - State: 2, Reward: 0, Target: 6.739, TD Error: 3.367, New V(s): 3.709\n", "2025-08-02 09:46:05,946 - INFO - TD Update - State: 3, Reward: -1, Target: 5.739, TD Error: -1.749, New V(s): 7.313\n", "2025-08-02 09:46:05,946 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 2.687, New V(s): 7.582\n", "2025-08-02 09:46:05,946 - INFO - Episode Complete - Total Reward: 4, Avg TD Error: 1.745\n", "2025-08-02 09:46:05,946 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,946 - INFO - TD Update - State: 0, Reward: 0, Target: 0.015, TD Error: 0.894, New V(s): -0.790\n", "2025-08-02 09:46:05,946 - INFO - TD Update - State: 1, Reward: 0, Target: 3.338, TD Error: 3.321, New V(s): 0.349\n", "2025-08-02 09:46:05,946 - INFO - TD Update - State: 2, Reward: -1, Target: 2.338, TD Error: -1.371, New V(s): 3.571\n", "2025-08-02 09:46:05,947 - INFO - TD Update - State: 2, Reward: -1, Target: 2.214, TD Error: -1.357, New V(s): 3.436\n", "2025-08-02 09:46:05,947 - INFO - TD Update - State: 2, Reward: -1, Target: 2.092, TD Error: -1.344, New V(s): 3.301\n", "2025-08-02 09:46:05,947 - INFO - TD Update - State: 2, Reward: -1, Target: 1.971, TD Error: -1.330, New V(s): 3.168\n", "2025-08-02 09:46:05,947 - INFO - TD Update - State: 2, Reward: -1, Target: 1.852, TD Error: -1.317, New V(s): 3.037\n", "2025-08-02 09:46:05,947 - INFO - TD Update - State: 2, Reward: 0, Target: 6.823, TD Error: 3.787, New V(s): 3.415\n", "2025-08-02 09:46:05,947 - INFO - TD Update - State: 3, Reward: -1, Target: 5.823, TD Error: -1.758, New V(s): 7.406\n", "2025-08-02 09:46:05,948 - INFO - TD Update - State: 3, Reward: -1, Target: 5.665, TD Error: -1.741, New V(s): 7.232\n", "2025-08-02 09:46:05,948 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 2.768, New V(s): 7.508\n", "2025-08-02 09:46:05,948 - INFO - Episode Complete - Total Reward: 3, Avg TD Error: 1.908\n", "2025-08-02 09:46:05,948 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,948 - INFO - TD Update - State: 0, Reward: -1, Target: -1.711, TD Error: -0.921, New V(s): -0.882\n", "2025-08-02 09:46:05,948 - INFO - TD Update - State: 0, Reward: 0, Target: 0.314, TD Error: 1.196, New V(s): -0.762\n", "2025-08-02 09:46:05,949 - INFO - TD Update - State: 1, Reward: 0, Target: 3.074, TD Error: 2.725, New V(s): 0.621\n", "2025-08-02 09:46:05,949 - INFO - TD Update - State: 2, Reward: -1, Target: 2.074, TD Error: -1.342, New V(s): 3.281\n", "2025-08-02 09:46:05,949 - INFO - TD Update - State: 2, Reward: 0, Target: 6.758, TD Error: 3.476, New V(s): 3.629\n", "2025-08-02 09:46:05,949 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 2.492, New V(s): 7.758\n", "2025-08-02 09:46:05,949 - INFO - Episode Complete - Total Reward: 8, Avg TD Error: 2.025\n", "2025-08-02 09:46:05,949 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,949 - INFO - TD Update - State: 0, Reward: 0, Target: 0.559, TD Error: 1.321, New V(s): -0.630\n", "2025-08-02 09:46:05,950 - INFO - TD Update - State: 1, Reward: -1, Target: -0.441, TD Error: -1.062, New V(s): 0.515\n", "2025-08-02 09:46:05,950 - INFO - TD Update - State: 1, Reward: 0, Target: 3.266, TD Error: 2.751, New V(s): 0.790\n", "2025-08-02 09:46:05,950 - INFO - TD Update - State: 2, Reward: 0, Target: 6.982, TD Error: 3.353, New V(s): 3.964\n", "2025-08-02 09:46:05,950 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 2.242, New V(s): 7.982\n", "2025-08-02 09:46:05,950 - INFO - Episode Complete - Total Reward: 9, Avg TD Error: 2.146\n", "2025-08-02 09:46:05,950 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,951 - INFO - TD Update - State: 0, Reward: -1, Target: -1.567, TD Error: -0.937, New V(s): -0.724\n", "2025-08-02 09:46:05,951 - INFO - TD Update - State: 0, Reward: -1, Target: -1.651, TD Error: -0.928, New V(s): -0.816\n", "2025-08-02 09:46:05,951 - INFO - TD Update - State: 0, Reward: -1, Target: -1.735, TD Error: -0.918, New V(s): -0.908\n", "2025-08-02 09:46:05,951 - INFO - TD Update - State: 0, Reward: -1, Target: -1.817, TD Error: -0.909, New V(s): -0.999\n", "2025-08-02 09:46:05,951 - INFO - TD Update - State: 0, Reward: -1, Target: -1.899, TD Error: -0.900, New V(s): -1.089\n", "2025-08-02 09:46:05,951 - INFO - TD Update - State: 0, Reward: -1, Target: -1.980, TD Error: -0.891, New V(s): -1.178\n", "2025-08-02 09:46:05,951 - INFO - TD Update - State: 0, Reward: 0, Target: 0.711, TD Error: 1.889, New V(s): -0.989\n", "2025-08-02 09:46:05,952 - INFO - TD Update - State: 1, Reward: 0, Target: 3.568, TD Error: 2.778, New V(s): 1.068\n", "2025-08-02 09:46:05,952 - INFO - TD Update - State: 2, Reward: 0, Target: 7.184, TD Error: 3.220, New V(s): 4.286\n", "2025-08-02 09:46:05,952 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 2.018, New V(s): 8.184\n", "2025-08-02 09:46:05,952 - INFO - Episode Complete - Total Reward: 4, Avg TD Error: 1.539\n", "2025-08-02 09:46:05,952 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,953 - INFO - TD Update - State: 0, Reward: -1, Target: -1.890, TD Error: -0.901, New V(s): -1.079\n", "2025-08-02 09:46:05,953 - INFO - TD Update - State: 0, Reward: 0, Target: 0.961, TD Error: 2.041, New V(s): -0.875\n", "2025-08-02 09:46:05,953 - INFO - TD Update - State: 1, Reward: -1, Target: -0.039, TD Error: -1.107, New V(s): 0.957\n", "2025-08-02 09:46:05,953 - INFO - TD Update - State: 1, Reward: 0, Target: 3.858, TD Error: 2.900, New V(s): 1.247\n", "2025-08-02 09:46:05,953 - INFO - TD Update - State: 2, Reward: 0, Target: 7.365, TD Error: 3.079, New V(s): 4.594\n", "2025-08-02 09:46:05,953 - INFO - TD Update - State: 3, Reward: -1, Target: 6.365, TD Error: -1.818, New V(s): 8.002\n", "2025-08-02 09:46:05,954 - INFO - TD Update - State: 3, Reward: -1, Target: 6.202, TD Error: -1.800, New V(s): 7.822\n", "2025-08-02 09:46:05,954 - INFO - TD Update - State: 3, Reward: -1, Target: 6.040, TD Error: -1.782, New V(s): 7.644\n", "2025-08-02 09:46:05,954 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 2.356, New V(s): 7.879\n", "2025-08-02 09:46:05,954 - INFO - Episode Complete - Total Reward: 5, Avg TD Error: 1.976\n", "2025-08-02 09:46:05,954 - INFO - \n", "๐Ÿ“Š Episode 20 Summary:\n", "2025-08-02 09:46:05,955 - INFO - Value Function: [-0.87543578 1.24722903 4.59404087 7.87925002 0. ]\n", "2025-08-02 09:46:05,955 - INFO - Recent Avg Reward: 5.70\n", "2025-08-02 09:46:05,955 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,955 - INFO - TD Update - State: 0, Reward: 0, Target: 1.123, TD Error: 1.998, New V(s): -0.676\n", "2025-08-02 09:46:05,955 - INFO - TD Update - State: 1, Reward: -1, Target: 0.123, TD Error: -1.125, New V(s): 1.135\n", "2025-08-02 09:46:05,955 - INFO - TD Update - State: 1, Reward: -1, Target: 0.021, TD Error: -1.113, New V(s): 1.023\n", "2025-08-02 09:46:05,955 - INFO - TD Update - State: 1, Reward: 0, Target: 4.135, TD Error: 3.111, New V(s): 1.335\n", "2025-08-02 09:46:05,956 - INFO - TD Update - State: 2, Reward: -1, Target: 3.135, TD Error: -1.459, New V(s): 4.448\n", "2025-08-02 09:46:05,956 - INFO - TD Update - State: 2, Reward: -1, Target: 3.003, TD Error: -1.445, New V(s): 4.304\n", "2025-08-02 09:46:05,956 - INFO - TD Update - State: 2, Reward: -1, Target: 2.873, TD Error: -1.430, New V(s): 4.161\n", "2025-08-02 09:46:05,956 - INFO - TD Update - State: 2, Reward: 0, Target: 7.091, TD Error: 2.931, New V(s): 4.454\n", "2025-08-02 09:46:05,956 - INFO - TD Update - State: 3, Reward: -1, Target: 6.091, TD Error: -1.788, New V(s): 7.700\n", "2025-08-02 09:46:05,956 - INFO - TD Update - State: 3, Reward: -1, Target: 5.930, TD Error: -1.770, New V(s): 7.523\n", "2025-08-02 09:46:05,957 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 2.477, New V(s): 7.771\n", "2025-08-02 09:46:05,957 - INFO - Episode Complete - Total Reward: 3, Avg TD Error: 1.877\n", "2025-08-02 09:46:05,957 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,957 - INFO - TD Update - State: 0, Reward: -1, Target: -1.608, TD Error: -0.932, New V(s): -0.769\n", "2025-08-02 09:46:05,957 - INFO - TD Update - State: 0, Reward: 0, Target: 1.201, TD Error: 1.970, New V(s): -0.572\n", "2025-08-02 09:46:05,957 - INFO - TD Update - State: 1, Reward: 0, Target: 4.008, TD Error: 2.674, New V(s): 1.602\n", "2025-08-02 09:46:05,957 - INFO - TD Update - State: 2, Reward: -1, Target: 3.008, TD Error: -1.445, New V(s): 4.309\n", "2025-08-02 09:46:05,958 - INFO - TD Update - State: 2, Reward: 0, Target: 6.994, TD Error: 2.685, New V(s): 4.578\n", "2025-08-02 09:46:05,958 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 2.229, New V(s): 7.994\n", "2025-08-02 09:46:05,958 - INFO - Episode Complete - Total Reward: 8, Avg TD Error: 1.989\n", "2025-08-02 09:46:05,958 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,958 - INFO - TD Update - State: 0, Reward: -1, Target: -1.515, TD Error: -0.943, New V(s): -0.666\n", "2025-08-02 09:46:05,958 - INFO - TD Update - State: 0, Reward: -1, Target: -1.600, TD Error: -0.933, New V(s): -0.760\n", "2025-08-02 09:46:05,959 - INFO - TD Update - State: 0, Reward: -1, Target: -1.684, TD Error: -0.924, New V(s): -0.852\n", "2025-08-02 09:46:05,959 - INFO - TD Update - State: 0, Reward: 0, Target: 1.442, TD Error: 2.294, New V(s): -0.623\n", "2025-08-02 09:46:05,959 - INFO - TD Update - State: 1, Reward: -1, Target: 0.442, TD Error: -1.160, New V(s): 1.486\n", "2025-08-02 09:46:05,959 - INFO - TD Update - State: 1, Reward: -1, Target: 0.337, TD Error: -1.149, New V(s): 1.371\n", "2025-08-02 09:46:05,959 - INFO - TD Update - State: 1, Reward: 0, Target: 4.120, TD Error: 2.749, New V(s): 1.646\n", "2025-08-02 09:46:05,959 - INFO - TD Update - State: 2, Reward: -1, Target: 3.120, TD Error: -1.458, New V(s): 4.432\n", "2025-08-02 09:46:05,959 - INFO - TD Update - State: 2, Reward: -1, Target: 2.989, TD Error: -1.443, New V(s): 4.288\n", "2025-08-02 09:46:05,960 - INFO - TD Update - State: 2, Reward: -1, Target: 2.859, TD Error: -1.429, New V(s): 4.145\n", "2025-08-02 09:46:05,960 - INFO - TD Update - State: 2, Reward: -1, Target: 2.730, TD Error: -1.414, New V(s): 4.003\n", "2025-08-02 09:46:05,960 - INFO - TD Update - State: 2, Reward: 0, Target: 7.195, TD Error: 3.191, New V(s): 4.322\n", "2025-08-02 09:46:05,960 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 2.006, New V(s): 8.195\n", "2025-08-02 09:46:05,960 - INFO - Episode Complete - Total Reward: 1, Avg TD Error: 1.623\n", "2025-08-02 09:46:05,960 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,961 - INFO - TD Update - State: 0, Reward: -1, Target: -1.560, TD Error: -0.938, New V(s): -0.716\n", "2025-08-02 09:46:05,961 - INFO - TD Update - State: 0, Reward: -1, Target: -1.645, TD Error: -0.928, New V(s): -0.809\n", "2025-08-02 09:46:05,961 - INFO - TD Update - State: 0, Reward: -1, Target: -1.728, TD Error: -0.919, New V(s): -0.901\n", "2025-08-02 09:46:05,961 - INFO - TD Update - State: 0, Reward: -1, Target: -1.811, TD Error: -0.910, New V(s): -0.992\n", "2025-08-02 09:46:05,961 - INFO - TD Update - State: 0, Reward: 0, Target: 1.481, TD Error: 2.473, New V(s): -0.745\n", "2025-08-02 09:46:05,961 - INFO - TD Update - State: 1, Reward: 0, Target: 3.890, TD Error: 2.244, New V(s): 1.870\n", "2025-08-02 09:46:05,961 - INFO - TD Update - State: 2, Reward: 0, Target: 7.375, TD Error: 3.053, New V(s): 4.628\n", "2025-08-02 09:46:05,962 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.805, New V(s): 8.375\n", "2025-08-02 09:46:05,962 - INFO - Episode Complete - Total Reward: 6, Avg TD Error: 1.659\n", "2025-08-02 09:46:05,962 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,962 - INFO - TD Update - State: 0, Reward: -1, Target: -1.670, TD Error: -0.926, New V(s): -0.837\n", "2025-08-02 09:46:05,962 - INFO - TD Update - State: 0, Reward: -1, Target: -1.754, TD Error: -0.916, New V(s): -0.929\n", "2025-08-02 09:46:05,962 - INFO - TD Update - State: 0, Reward: -1, Target: -1.836, TD Error: -0.907, New V(s): -1.020\n", "2025-08-02 09:46:05,963 - INFO - TD Update - State: 0, Reward: -1, Target: -1.918, TD Error: -0.898, New V(s): -1.109\n", "2025-08-02 09:46:05,963 - INFO - TD Update - State: 0, Reward: -1, Target: -1.998, TD Error: -0.889, New V(s): -1.198\n", "2025-08-02 09:46:05,963 - INFO - TD Update - State: 0, Reward: 0, Target: 1.683, TD Error: 2.882, New V(s): -0.910\n", "2025-08-02 09:46:05,963 - INFO - TD Update - State: 1, Reward: 0, Target: 4.165, TD Error: 2.295, New V(s): 2.100\n", "2025-08-02 09:46:05,963 - INFO - TD Update - State: 2, Reward: -1, Target: 3.165, TD Error: -1.463, New V(s): 4.481\n", "2025-08-02 09:46:05,964 - INFO - TD Update - State: 2, Reward: -1, Target: 3.033, TD Error: -1.448, New V(s): 4.337\n", "2025-08-02 09:46:05,964 - INFO - TD Update - State: 2, Reward: 0, Target: 7.538, TD Error: 3.201, New V(s): 4.657\n", "2025-08-02 09:46:05,964 - INFO - TD Update - State: 3, Reward: -1, Target: 6.538, TD Error: -1.838, New V(s): 8.191\n", "2025-08-02 09:46:05,964 - INFO - TD Update - State: 3, Reward: -1, Target: 6.372, TD Error: -1.819, New V(s): 8.009\n", "2025-08-02 09:46:05,964 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.991, New V(s): 8.209\n", "2025-08-02 09:46:05,964 - INFO - Episode Complete - Total Reward: 1, Avg TD Error: 1.652\n", "2025-08-02 09:46:05,965 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,965 - INFO - TD Update - State: 0, Reward: 0, Target: 1.890, TD Error: 2.800, New V(s): -0.630\n", "2025-08-02 09:46:05,965 - INFO - TD Update - State: 1, Reward: 0, Target: 4.191, TD Error: 2.091, New V(s): 2.309\n", "2025-08-02 09:46:05,965 - INFO - TD Update - State: 2, Reward: -1, Target: 3.191, TD Error: -1.466, New V(s): 4.510\n", "2025-08-02 09:46:05,965 - INFO - TD Update - State: 2, Reward: 0, Target: 7.388, TD Error: 2.878, New V(s): 4.798\n", "2025-08-02 09:46:05,966 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.791, New V(s): 8.388\n", "2025-08-02 09:46:05,966 - INFO - Episode Complete - Total Reward: 9, Avg TD Error: 2.205\n", "2025-08-02 09:46:05,966 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,966 - INFO - TD Update - State: 0, Reward: 0, Target: 2.078, TD Error: 2.708, New V(s): -0.359\n", "2025-08-02 09:46:05,966 - INFO - TD Update - State: 1, Reward: 0, Target: 4.318, TD Error: 2.009, New V(s): 2.510\n", "2025-08-02 09:46:05,966 - INFO - TD Update - State: 2, Reward: -1, Target: 3.318, TD Error: -1.480, New V(s): 4.650\n", "2025-08-02 09:46:05,966 - INFO - TD Update - State: 2, Reward: 0, Target: 7.549, TD Error: 2.899, New V(s): 4.940\n", "2025-08-02 09:46:05,967 - INFO - TD Update - State: 3, Reward: -1, Target: 6.549, TD Error: -1.839, New V(s): 8.204\n", "2025-08-02 09:46:05,967 - INFO - TD Update - State: 3, Reward: -1, Target: 6.383, TD Error: -1.820, New V(s): 8.022\n", "2025-08-02 09:46:05,967 - INFO - TD Update - State: 3, Reward: -1, Target: 6.220, TD Error: -1.802, New V(s): 7.842\n", "2025-08-02 09:46:05,967 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 2.158, New V(s): 8.057\n", "2025-08-02 09:46:05,967 - INFO - Episode Complete - Total Reward: 6, Avg TD Error: 2.089\n", "2025-08-02 09:46:05,967 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,968 - INFO - TD Update - State: 0, Reward: 0, Target: 2.259, TD Error: 2.618, New V(s): -0.098\n", "2025-08-02 09:46:05,968 - INFO - TD Update - State: 1, Reward: -1, Target: 1.259, TD Error: -1.251, New V(s): 2.385\n", "2025-08-02 09:46:05,968 - INFO - TD Update - State: 1, Reward: -1, Target: 1.146, TD Error: -1.238, New V(s): 2.261\n", "2025-08-02 09:46:05,968 - INFO - TD Update - State: 1, Reward: 0, Target: 4.446, TD Error: 2.185, New V(s): 2.479\n", "2025-08-02 09:46:05,968 - INFO - TD Update - State: 2, Reward: 0, Target: 7.252, TD Error: 2.312, New V(s): 5.171\n", "2025-08-02 09:46:05,968 - INFO - TD Update - State: 3, Reward: -1, Target: 6.252, TD Error: -1.806, New V(s): 7.877\n", "2025-08-02 09:46:05,968 - INFO - TD Update - State: 3, Reward: -1, Target: 6.089, TD Error: -1.788, New V(s): 7.698\n", "2025-08-02 09:46:05,969 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 2.302, New V(s): 7.928\n", "2025-08-02 09:46:05,969 - INFO - Episode Complete - Total Reward: 6, Avg TD Error: 1.937\n", "2025-08-02 09:46:05,969 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,969 - INFO - TD Update - State: 0, Reward: 0, Target: 2.231, TD Error: 2.329, New V(s): 0.135\n", "2025-08-02 09:46:05,969 - INFO - TD Update - State: 1, Reward: 0, Target: 4.654, TD Error: 2.174, New V(s): 2.697\n", "2025-08-02 09:46:05,969 - INFO - TD Update - State: 2, Reward: -1, Target: 3.654, TD Error: -1.517, New V(s): 5.019\n", "2025-08-02 09:46:05,969 - INFO - TD Update - State: 2, Reward: 0, Target: 7.135, TD Error: 2.116, New V(s): 5.231\n", "2025-08-02 09:46:05,970 - INFO - TD Update - State: 3, Reward: -1, Target: 6.135, TD Error: -1.793, New V(s): 7.749\n", "2025-08-02 09:46:05,970 - INFO - TD Update - State: 3, Reward: -1, Target: 5.974, TD Error: -1.775, New V(s): 7.571\n", "2025-08-02 09:46:05,970 - INFO - TD Update - State: 3, Reward: -1, Target: 5.814, TD Error: -1.757, New V(s): 7.396\n", "2025-08-02 09:46:05,970 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 2.604, New V(s): 7.656\n", "2025-08-02 09:46:05,970 - INFO - Episode Complete - Total Reward: 6, Avg TD Error: 2.008\n", "2025-08-02 09:46:05,970 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,970 - INFO - TD Update - State: 0, Reward: 0, Target: 2.427, TD Error: 2.292, New V(s): 0.365\n", "2025-08-02 09:46:05,971 - INFO - TD Update - State: 1, Reward: 0, Target: 4.708, TD Error: 2.011, New V(s): 2.898\n", "2025-08-02 09:46:05,971 - INFO - TD Update - State: 2, Reward: -1, Target: 3.708, TD Error: -1.523, New V(s): 5.079\n", "2025-08-02 09:46:05,971 - INFO - TD Update - State: 2, Reward: -1, Target: 3.571, TD Error: -1.508, New V(s): 4.928\n", "2025-08-02 09:46:05,971 - INFO - TD Update - State: 2, Reward: -1, Target: 3.435, TD Error: -1.493, New V(s): 4.778\n", "2025-08-02 09:46:05,971 - INFO - TD Update - State: 2, Reward: 0, Target: 6.891, TD Error: 2.112, New V(s): 4.990\n", "2025-08-02 09:46:05,971 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 2.344, New V(s): 7.891\n", "2025-08-02 09:46:05,971 - INFO - Episode Complete - Total Reward: 7, Avg TD Error: 1.897\n", "2025-08-02 09:46:05,972 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,972 - INFO - TD Update - State: 0, Reward: 0, Target: 2.608, TD Error: 2.244, New V(s): 0.589\n", "2025-08-02 09:46:05,972 - INFO - TD Update - State: 1, Reward: 0, Target: 4.491, TD Error: 1.593, New V(s): 3.057\n", "2025-08-02 09:46:05,972 - INFO - TD Update - State: 2, Reward: 0, Target: 7.102, TD Error: 2.112, New V(s): 5.201\n", "2025-08-02 09:46:05,973 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 2.109, New V(s): 8.102\n", "2025-08-02 09:46:05,973 - INFO - Episode Complete - Total Reward: 10, Avg TD Error: 2.014\n", "2025-08-02 09:46:05,973 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,973 - INFO - TD Update - State: 0, Reward: 0, Target: 2.751, TD Error: 2.163, New V(s): 0.805\n", "2025-08-02 09:46:05,973 - INFO - TD Update - State: 1, Reward: -1, Target: 1.751, TD Error: -1.306, New V(s): 2.927\n", "2025-08-02 09:46:05,973 - INFO - TD Update - State: 1, Reward: 0, Target: 4.681, TD Error: 1.754, New V(s): 3.102\n", "2025-08-02 09:46:05,974 - INFO - TD Update - State: 2, Reward: 0, Target: 7.291, TD Error: 2.090, New V(s): 5.410\n", "2025-08-02 09:46:05,974 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.898, New V(s): 8.291\n", "2025-08-02 09:46:05,974 - INFO - Episode Complete - Total Reward: 9, Avg TD Error: 1.842\n", "2025-08-02 09:46:05,974 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,974 - INFO - TD Update - State: 0, Reward: 0, Target: 2.792, TD Error: 1.987, New V(s): 1.004\n", "2025-08-02 09:46:05,974 - INFO - TD Update - State: 1, Reward: 0, Target: 4.869, TD Error: 1.767, New V(s): 3.279\n", "2025-08-02 09:46:05,974 - INFO - TD Update - State: 2, Reward: 0, Target: 7.462, TD Error: 2.052, New V(s): 5.615\n", "2025-08-02 09:46:05,975 - INFO - TD Update - State: 3, Reward: -1, Target: 6.462, TD Error: -1.829, New V(s): 8.108\n", "2025-08-02 09:46:05,975 - INFO - TD Update - State: 3, Reward: -1, Target: 6.298, TD Error: -1.811, New V(s): 7.927\n", "2025-08-02 09:46:05,975 - INFO - TD Update - State: 3, Reward: -1, Target: 6.135, TD Error: -1.793, New V(s): 7.748\n", "2025-08-02 09:46:05,975 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 2.252, New V(s): 7.973\n", "2025-08-02 09:46:05,975 - INFO - Episode Complete - Total Reward: 7, Avg TD Error: 1.927\n", "2025-08-02 09:46:05,975 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,976 - INFO - TD Update - State: 0, Reward: 0, Target: 2.951, TD Error: 1.947, New V(s): 1.199\n", "2025-08-02 09:46:05,976 - INFO - TD Update - State: 1, Reward: 0, Target: 5.054, TD Error: 1.775, New V(s): 3.456\n", "2025-08-02 09:46:05,976 - INFO - TD Update - State: 2, Reward: 0, Target: 7.176, TD Error: 1.561, New V(s): 5.771\n", "2025-08-02 09:46:05,976 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 2.027, New V(s): 8.176\n", "2025-08-02 09:46:05,976 - INFO - Episode Complete - Total Reward: 10, Avg TD Error: 1.827\n", "2025-08-02 09:46:05,976 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,976 - INFO - TD Update - State: 0, Reward: -1, Target: 0.079, TD Error: -1.120, New V(s): 1.087\n", "2025-08-02 09:46:05,977 - INFO - TD Update - State: 0, Reward: 0, Target: 3.111, TD Error: 2.024, New V(s): 1.289\n", "2025-08-02 09:46:05,977 - INFO - TD Update - State: 1, Reward: -1, Target: 2.111, TD Error: -1.346, New V(s): 3.322\n", "2025-08-02 09:46:05,977 - INFO - TD Update - State: 1, Reward: 0, Target: 5.194, TD Error: 1.872, New V(s): 3.509\n", "2025-08-02 09:46:05,977 - INFO - TD Update - State: 2, Reward: -1, Target: 4.194, TD Error: -1.577, New V(s): 5.614\n", "2025-08-02 09:46:05,977 - INFO - TD Update - State: 2, Reward: 0, Target: 7.358, TD Error: 1.745, New V(s): 5.788\n", "2025-08-02 09:46:05,977 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.824, New V(s): 8.358\n", "2025-08-02 09:46:05,977 - INFO - Episode Complete - Total Reward: 7, Avg TD Error: 1.644\n", "2025-08-02 09:46:05,978 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,978 - INFO - TD Update - State: 0, Reward: -1, Target: 0.160, TD Error: -1.129, New V(s): 1.176\n", "2025-08-02 09:46:05,978 - INFO - TD Update - State: 0, Reward: 0, Target: 3.158, TD Error: 1.982, New V(s): 1.374\n", "2025-08-02 09:46:05,978 - INFO - TD Update - State: 1, Reward: 0, Target: 5.209, TD Error: 1.700, New V(s): 3.679\n", "2025-08-02 09:46:05,978 - INFO - TD Update - State: 2, Reward: 0, Target: 7.523, TD Error: 1.735, New V(s): 5.961\n", "2025-08-02 09:46:05,978 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.642, New V(s): 8.523\n", "2025-08-02 09:46:05,979 - INFO - Episode Complete - Total Reward: 9, Avg TD Error: 1.637\n", "2025-08-02 09:46:05,979 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,979 - INFO - TD Update - State: 0, Reward: 0, Target: 3.311, TD Error: 1.937, New V(s): 1.568\n", "2025-08-02 09:46:05,979 - INFO - TD Update - State: 1, Reward: 0, Target: 5.365, TD Error: 1.686, New V(s): 3.848\n", "2025-08-02 09:46:05,979 - INFO - TD Update - State: 2, Reward: -1, Target: 4.365, TD Error: -1.596, New V(s): 5.802\n", "2025-08-02 09:46:05,979 - INFO - TD Update - State: 2, Reward: 0, Target: 7.670, TD Error: 1.868, New V(s): 5.989\n", "2025-08-02 09:46:05,979 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.477, New V(s): 8.670\n", "2025-08-02 09:46:05,980 - INFO - Episode Complete - Total Reward: 9, Avg TD Error: 1.713\n", "2025-08-02 09:46:05,980 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,980 - INFO - TD Update - State: 0, Reward: -1, Target: 0.411, TD Error: -1.157, New V(s): 1.452\n", "2025-08-02 09:46:05,980 - INFO - TD Update - State: 0, Reward: -1, Target: 0.307, TD Error: -1.145, New V(s): 1.338\n", "2025-08-02 09:46:05,980 - INFO - TD Update - State: 0, Reward: 0, Target: 3.463, TD Error: 2.125, New V(s): 1.550\n", "2025-08-02 09:46:05,980 - INFO - TD Update - State: 1, Reward: 0, Target: 5.390, TD Error: 1.542, New V(s): 4.002\n", "2025-08-02 09:46:05,981 - INFO - TD Update - State: 2, Reward: 0, Target: 7.803, TD Error: 1.815, New V(s): 6.170\n", "2025-08-02 09:46:05,981 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.330, New V(s): 8.803\n", "2025-08-02 09:46:05,981 - INFO - Episode Complete - Total Reward: 8, Avg TD Error: 1.519\n", "2025-08-02 09:46:05,981 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,982 - INFO - TD Update - State: 0, Reward: -1, Target: 0.395, TD Error: -1.155, New V(s): 1.435\n", "2025-08-02 09:46:05,982 - INFO - TD Update - State: 0, Reward: -1, Target: 0.291, TD Error: -1.143, New V(s): 1.320\n", "2025-08-02 09:46:05,982 - INFO - TD Update - State: 0, Reward: -1, Target: 0.188, TD Error: -1.132, New V(s): 1.207\n", "2025-08-02 09:46:05,982 - INFO - TD Update - State: 0, Reward: 0, Target: 3.602, TD Error: 2.394, New V(s): 1.447\n", "2025-08-02 09:46:05,982 - INFO - TD Update - State: 1, Reward: 0, Target: 5.553, TD Error: 1.551, New V(s): 4.157\n", "2025-08-02 09:46:05,982 - INFO - TD Update - State: 2, Reward: -1, Target: 4.553, TD Error: -1.617, New V(s): 6.008\n", "2025-08-02 09:46:05,982 - INFO - TD Update - State: 2, Reward: -1, Target: 4.408, TD Error: -1.601, New V(s): 5.848\n", "2025-08-02 09:46:05,983 - INFO - TD Update - State: 2, Reward: -1, Target: 4.264, TD Error: -1.585, New V(s): 5.690\n", "2025-08-02 09:46:05,983 - INFO - TD Update - State: 2, Reward: 0, Target: 7.923, TD Error: 2.233, New V(s): 5.913\n", "2025-08-02 09:46:05,983 - INFO - TD Update - State: 3, Reward: -1, Target: 6.923, TD Error: -1.880, New V(s): 8.615\n", "2025-08-02 09:46:05,983 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.385, New V(s): 8.754\n", "2025-08-02 09:46:05,983 - INFO - Episode Complete - Total Reward: 3, Avg TD Error: 1.607\n", "2025-08-02 09:46:05,983 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,984 - INFO - TD Update - State: 0, Reward: -1, Target: 0.302, TD Error: -1.145, New V(s): 1.332\n", "2025-08-02 09:46:05,984 - INFO - TD Update - State: 0, Reward: 0, Target: 3.741, TD Error: 2.409, New V(s): 1.573\n", "2025-08-02 09:46:05,984 - INFO - TD Update - State: 1, Reward: 0, Target: 5.322, TD Error: 1.165, New V(s): 4.273\n", "2025-08-02 09:46:05,984 - INFO - TD Update - State: 2, Reward: 0, Target: 7.878, TD Error: 1.965, New V(s): 6.110\n", "2025-08-02 09:46:05,984 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.246, New V(s): 8.878\n", "2025-08-02 09:46:05,984 - INFO - Episode Complete - Total Reward: 9, Avg TD Error: 1.586\n", "2025-08-02 09:46:05,984 - INFO - \n", "๐Ÿ“Š Episode 40 Summary:\n", "2025-08-02 09:46:05,985 - INFO - Value Function: [1.57306672 4.27341598 6.10968602 8.87831698 0. ]\n", "2025-08-02 09:46:05,985 - INFO - Recent Avg Reward: 8.10\n", "2025-08-02 09:46:05,985 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,985 - INFO - TD Update - State: 0, Reward: 0, Target: 3.846, TD Error: 2.273, New V(s): 1.800\n", "2025-08-02 09:46:05,986 - INFO - TD Update - State: 1, Reward: 0, Target: 5.499, TD Error: 1.225, New V(s): 4.396\n", "2025-08-02 09:46:05,986 - INFO - TD Update - State: 2, Reward: 0, Target: 7.990, TD Error: 1.881, New V(s): 6.298\n", "2025-08-02 09:46:05,986 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.122, New V(s): 8.990\n", "2025-08-02 09:46:05,986 - INFO - Episode Complete - Total Reward: 10, Avg TD Error: 1.625\n", "2025-08-02 09:46:05,986 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,986 - INFO - TD Update - State: 0, Reward: 0, Target: 3.956, TD Error: 2.156, New V(s): 2.016\n", "2025-08-02 09:46:05,987 - INFO - TD Update - State: 1, Reward: 0, Target: 5.668, TD Error: 1.272, New V(s): 4.523\n", "2025-08-02 09:46:05,987 - INFO - TD Update - State: 2, Reward: 0, Target: 8.091, TD Error: 1.794, New V(s): 6.477\n", "2025-08-02 09:46:05,987 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.010, New V(s): 9.091\n", "2025-08-02 09:46:05,987 - INFO - Episode Complete - Total Reward: 10, Avg TD Error: 1.558\n", "2025-08-02 09:46:05,987 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,987 - INFO - TD Update - State: 0, Reward: 0, Target: 4.071, TD Error: 2.055, New V(s): 2.221\n", "2025-08-02 09:46:05,988 - INFO - TD Update - State: 1, Reward: 0, Target: 5.829, TD Error: 1.306, New V(s): 4.654\n", "2025-08-02 09:46:05,988 - INFO - TD Update - State: 2, Reward: 0, Target: 8.182, TD Error: 1.705, New V(s): 6.648\n", "2025-08-02 09:46:05,988 - INFO - TD Update - State: 3, Reward: -1, Target: 7.182, TD Error: -1.909, New V(s): 8.901\n", "2025-08-02 09:46:05,988 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.099, New V(s): 9.010\n", "2025-08-02 09:46:05,988 - INFO - Episode Complete - Total Reward: 9, Avg TD Error: 1.615\n", "2025-08-02 09:46:05,988 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,988 - INFO - TD Update - State: 0, Reward: 0, Target: 4.188, TD Error: 1.967, New V(s): 2.418\n", "2025-08-02 09:46:05,989 - INFO - TD Update - State: 1, Reward: -1, Target: 3.188, TD Error: -1.465, New V(s): 4.507\n", "2025-08-02 09:46:05,989 - INFO - TD Update - State: 1, Reward: -1, Target: 3.057, TD Error: -1.451, New V(s): 4.362\n", "2025-08-02 09:46:05,989 - INFO - TD Update - State: 1, Reward: -1, Target: 2.926, TD Error: -1.436, New V(s): 4.219\n", "2025-08-02 09:46:05,989 - INFO - TD Update - State: 1, Reward: 0, Target: 5.983, TD Error: 1.764, New V(s): 4.395\n", "2025-08-02 09:46:05,989 - INFO - TD Update - State: 2, Reward: -1, Target: 4.983, TD Error: -1.665, New V(s): 6.481\n", "2025-08-02 09:46:05,989 - INFO - TD Update - State: 2, Reward: -1, Target: 4.833, TD Error: -1.648, New V(s): 6.316\n", "2025-08-02 09:46:05,990 - INFO - TD Update - State: 2, Reward: -1, Target: 4.685, TD Error: -1.632, New V(s): 6.153\n", "2025-08-02 09:46:05,990 - INFO - TD Update - State: 2, Reward: 0, Target: 8.109, TD Error: 1.956, New V(s): 6.349\n", "2025-08-02 09:46:05,990 - INFO - TD Update - State: 3, Reward: -1, Target: 7.109, TD Error: -1.901, New V(s): 8.820\n", "2025-08-02 09:46:05,990 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.180, New V(s): 8.938\n", "2025-08-02 09:46:05,990 - INFO - Episode Complete - Total Reward: 3, Avg TD Error: 1.642\n", "2025-08-02 09:46:05,990 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,991 - INFO - TD Update - State: 0, Reward: -1, Target: 1.176, TD Error: -1.242, New V(s): 2.294\n", "2025-08-02 09:46:05,991 - INFO - TD Update - State: 0, Reward: 0, Target: 3.955, TD Error: 1.662, New V(s): 2.460\n", "2025-08-02 09:46:05,991 - INFO - TD Update - State: 1, Reward: 0, Target: 5.714, TD Error: 1.319, New V(s): 4.527\n", "2025-08-02 09:46:05,991 - INFO - TD Update - State: 2, Reward: -1, Target: 4.714, TD Error: -1.635, New V(s): 6.185\n", "2025-08-02 09:46:05,991 - INFO - TD Update - State: 2, Reward: 0, Target: 8.044, TD Error: 1.859, New V(s): 6.371\n", "2025-08-02 09:46:05,991 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.062, New V(s): 9.044\n", "2025-08-02 09:46:05,992 - INFO - Episode Complete - Total Reward: 8, Avg TD Error: 1.463\n", "2025-08-02 09:46:05,992 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,992 - INFO - TD Update - State: 0, Reward: 0, Target: 4.074, TD Error: 1.614, New V(s): 2.622\n", "2025-08-02 09:46:05,992 - INFO - TD Update - State: 1, Reward: 0, Target: 5.734, TD Error: 1.207, New V(s): 4.648\n", "2025-08-02 09:46:05,992 - INFO - TD Update - State: 2, Reward: 0, Target: 8.140, TD Error: 1.769, New V(s): 6.548\n", "2025-08-02 09:46:05,992 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 0.956, New V(s): 9.140\n", "2025-08-02 09:46:05,992 - INFO - Episode Complete - Total Reward: 10, Avg TD Error: 1.386\n", "2025-08-02 09:46:05,993 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,993 - INFO - TD Update - State: 0, Reward: -1, Target: 1.359, TD Error: -1.262, New V(s): 2.495\n", "2025-08-02 09:46:05,993 - INFO - TD Update - State: 0, Reward: -1, Target: 1.246, TD Error: -1.250, New V(s): 2.370\n", "2025-08-02 09:46:05,993 - INFO - TD Update - State: 0, Reward: -1, Target: 1.133, TD Error: -1.237, New V(s): 2.247\n", "2025-08-02 09:46:05,993 - INFO - TD Update - State: 0, Reward: -1, Target: 1.022, TD Error: -1.225, New V(s): 2.124\n", "2025-08-02 09:46:05,993 - INFO - TD Update - State: 0, Reward: -1, Target: 0.912, TD Error: -1.212, New V(s): 2.003\n", "2025-08-02 09:46:05,993 - INFO - TD Update - State: 0, Reward: 0, Target: 4.183, TD Error: 2.180, New V(s): 2.221\n", "2025-08-02 09:46:05,994 - INFO - TD Update - State: 1, Reward: -1, Target: 3.183, TD Error: -1.465, New V(s): 4.501\n", "2025-08-02 09:46:05,994 - INFO - TD Update - State: 1, Reward: 0, Target: 5.893, TD Error: 1.392, New V(s): 4.640\n", "2025-08-02 09:46:05,994 - INFO - TD Update - State: 2, Reward: 0, Target: 8.226, TD Error: 1.678, New V(s): 6.716\n", "2025-08-02 09:46:05,994 - INFO - TD Update - State: 3, Reward: -1, Target: 7.226, TD Error: -1.914, New V(s): 8.949\n", "2025-08-02 09:46:05,994 - INFO - TD Update - State: 3, Reward: -1, Target: 7.054, TD Error: -1.895, New V(s): 8.759\n", "2025-08-02 09:46:05,994 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.241, New V(s): 8.883\n", "2025-08-02 09:46:05,994 - INFO - Episode Complete - Total Reward: 2, Avg TD Error: 1.496\n", "2025-08-02 09:46:05,995 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,995 - INFO - TD Update - State: 0, Reward: -1, Target: 0.999, TD Error: -1.222, New V(s): 2.099\n", "2025-08-02 09:46:05,995 - INFO - TD Update - State: 0, Reward: 0, Target: 4.176, TD Error: 2.078, New V(s): 2.306\n", "2025-08-02 09:46:05,995 - INFO - TD Update - State: 1, Reward: -1, Target: 3.176, TD Error: -1.464, New V(s): 4.494\n", "2025-08-02 09:46:05,995 - INFO - TD Update - State: 1, Reward: -1, Target: 3.045, TD Error: -1.449, New V(s): 4.349\n", "2025-08-02 09:46:05,995 - INFO - TD Update - State: 1, Reward: -1, Target: 2.914, TD Error: -1.435, New V(s): 4.206\n", "2025-08-02 09:46:05,995 - INFO - TD Update - State: 1, Reward: -1, Target: 2.785, TD Error: -1.421, New V(s): 4.063\n", "2025-08-02 09:46:05,995 - INFO - TD Update - State: 1, Reward: 0, Target: 6.044, TD Error: 1.981, New V(s): 4.262\n", "2025-08-02 09:46:05,996 - INFO - TD Update - State: 2, Reward: -1, Target: 5.044, TD Error: -1.672, New V(s): 6.549\n", "2025-08-02 09:46:05,996 - INFO - TD Update - State: 2, Reward: 0, Target: 7.995, TD Error: 1.446, New V(s): 6.693\n", "2025-08-02 09:46:05,996 - INFO - TD Update - State: 3, Reward: -1, Target: 6.995, TD Error: -1.888, New V(s): 8.694\n", "2025-08-02 09:46:05,996 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.306, New V(s): 8.825\n", "2025-08-02 09:46:05,996 - INFO - Episode Complete - Total Reward: 3, Avg TD Error: 1.578\n", "2025-08-02 09:46:05,996 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,997 - INFO - TD Update - State: 0, Reward: 0, Target: 3.835, TD Error: 1.529, New V(s): 2.459\n", "2025-08-02 09:46:05,997 - INFO - TD Update - State: 1, Reward: 0, Target: 6.024, TD Error: 1.762, New V(s): 4.438\n", "2025-08-02 09:46:05,997 - INFO - TD Update - State: 2, Reward: 0, Target: 7.942, TD Error: 1.249, New V(s): 6.818\n", "2025-08-02 09:46:05,997 - INFO - TD Update - State: 3, Reward: -1, Target: 6.942, TD Error: -1.882, New V(s): 8.637\n", "2025-08-02 09:46:05,997 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.363, New V(s): 8.773\n", "2025-08-02 09:46:05,997 - INFO - Episode Complete - Total Reward: 9, Avg TD Error: 1.557\n", "2025-08-02 09:46:05,998 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,998 - INFO - TD Update - State: 0, Reward: 0, Target: 3.994, TD Error: 1.535, New V(s): 2.613\n", "2025-08-02 09:46:05,998 - INFO - TD Update - State: 1, Reward: 0, Target: 6.136, TD Error: 1.699, New V(s): 4.608\n", "2025-08-02 09:46:05,998 - INFO - TD Update - State: 2, Reward: -1, Target: 5.136, TD Error: -1.682, New V(s): 6.650\n", "2025-08-02 09:46:05,998 - INFO - TD Update - State: 2, Reward: -1, Target: 4.985, TD Error: -1.665, New V(s): 6.484\n", "2025-08-02 09:46:05,998 - INFO - TD Update - State: 2, Reward: 0, Target: 7.896, TD Error: 1.412, New V(s): 6.625\n", "2025-08-02 09:46:05,998 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.227, New V(s): 8.896\n", "2025-08-02 09:46:05,999 - INFO - Episode Complete - Total Reward: 8, Avg TD Error: 1.537\n", "2025-08-02 09:46:05,999 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:05,999 - INFO - TD Update - State: 0, Reward: 0, Target: 4.147, TD Error: 1.534, New V(s): 2.766\n", "2025-08-02 09:46:05,999 - INFO - TD Update - State: 1, Reward: 0, Target: 5.962, TD Error: 1.355, New V(s): 4.743\n", "2025-08-02 09:46:05,999 - INFO - TD Update - State: 2, Reward: 0, Target: 8.006, TD Error: 1.381, New V(s): 6.763\n", "2025-08-02 09:46:05,999 - INFO - TD Update - State: 3, Reward: -1, Target: 7.006, TD Error: -1.890, New V(s): 8.707\n", "2025-08-02 09:46:06,000 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.293, New V(s): 8.836\n", "2025-08-02 09:46:06,000 - INFO - Episode Complete - Total Reward: 9, Avg TD Error: 1.491\n", "2025-08-02 09:46:06,000 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,000 - INFO - TD Update - State: 0, Reward: 0, Target: 4.269, TD Error: 1.503, New V(s): 2.917\n", "2025-08-02 09:46:06,000 - INFO - TD Update - State: 1, Reward: 0, Target: 6.087, TD Error: 1.344, New V(s): 4.877\n", "2025-08-02 09:46:06,000 - INFO - TD Update - State: 2, Reward: -1, Target: 5.087, TD Error: -1.676, New V(s): 6.595\n", "2025-08-02 09:46:06,000 - INFO - TD Update - State: 2, Reward: -1, Target: 4.936, TD Error: -1.660, New V(s): 6.429\n", "2025-08-02 09:46:06,001 - INFO - TD Update - State: 2, Reward: 0, Target: 7.952, TD Error: 1.523, New V(s): 6.582\n", "2025-08-02 09:46:06,001 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.164, New V(s): 8.952\n", "2025-08-02 09:46:06,001 - INFO - Episode Complete - Total Reward: 8, Avg TD Error: 1.478\n", "2025-08-02 09:46:06,001 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,001 - INFO - TD Update - State: 0, Reward: 0, Target: 4.390, TD Error: 1.473, New V(s): 3.064\n", "2025-08-02 09:46:06,001 - INFO - TD Update - State: 1, Reward: -1, Target: 3.390, TD Error: -1.488, New V(s): 4.729\n", "2025-08-02 09:46:06,001 - INFO - TD Update - State: 1, Reward: -1, Target: 3.256, TD Error: -1.473, New V(s): 4.581\n", "2025-08-02 09:46:06,002 - INFO - TD Update - State: 1, Reward: 0, Target: 5.924, TD Error: 1.342, New V(s): 4.716\n", "2025-08-02 09:46:06,002 - INFO - TD Update - State: 2, Reward: 0, Target: 8.057, TD Error: 1.476, New V(s): 6.729\n", "2025-08-02 09:46:06,002 - INFO - TD Update - State: 3, Reward: -1, Target: 7.057, TD Error: -1.895, New V(s): 8.763\n", "2025-08-02 09:46:06,002 - INFO - TD Update - State: 3, Reward: -1, Target: 6.887, TD Error: -1.876, New V(s): 8.575\n", "2025-08-02 09:46:06,002 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.425, New V(s): 8.718\n", "2025-08-02 09:46:06,002 - INFO - Episode Complete - Total Reward: 6, Avg TD Error: 1.556\n", "2025-08-02 09:46:06,002 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,003 - INFO - TD Update - State: 0, Reward: -1, Target: 1.757, TD Error: -1.306, New V(s): 2.933\n", "2025-08-02 09:46:06,003 - INFO - TD Update - State: 0, Reward: 0, Target: 4.244, TD Error: 1.311, New V(s): 3.064\n", "2025-08-02 09:46:06,003 - INFO - TD Update - State: 1, Reward: 0, Target: 6.056, TD Error: 1.341, New V(s): 4.850\n", "2025-08-02 09:46:06,004 - INFO - TD Update - State: 2, Reward: 0, Target: 7.846, TD Error: 1.117, New V(s): 6.841\n", "2025-08-02 09:46:06,005 - INFO - TD Update - State: 3, Reward: -1, Target: 6.846, TD Error: -1.872, New V(s): 8.531\n", "2025-08-02 09:46:06,005 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.469, New V(s): 8.678\n", "2025-08-02 09:46:06,005 - INFO - Episode Complete - Total Reward: 8, Avg TD Error: 1.403\n", "2025-08-02 09:46:06,006 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,007 - INFO - TD Update - State: 0, Reward: 0, Target: 4.365, TD Error: 1.300, New V(s): 3.194\n", "2025-08-02 09:46:06,007 - INFO - TD Update - State: 1, Reward: 0, Target: 6.157, TD Error: 1.307, New V(s): 4.980\n", "2025-08-02 09:46:06,007 - INFO - TD Update - State: 2, Reward: -1, Target: 5.157, TD Error: -1.684, New V(s): 6.673\n", "2025-08-02 09:46:06,007 - INFO - TD Update - State: 2, Reward: -1, Target: 5.005, TD Error: -1.667, New V(s): 6.506\n", "2025-08-02 09:46:06,008 - INFO - TD Update - State: 2, Reward: -1, Target: 4.855, TD Error: -1.651, New V(s): 6.341\n", "2025-08-02 09:46:06,008 - INFO - TD Update - State: 2, Reward: 0, Target: 7.810, TD Error: 1.469, New V(s): 6.488\n", "2025-08-02 09:46:06,008 - INFO - TD Update - State: 3, Reward: -1, Target: 6.810, TD Error: -1.868, New V(s): 8.491\n", "2025-08-02 09:46:06,008 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.509, New V(s): 8.642\n", "2025-08-02 09:46:06,009 - INFO - Episode Complete - Total Reward: 6, Avg TD Error: 1.557\n", "2025-08-02 09:46:06,009 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,009 - INFO - TD Update - State: 0, Reward: 0, Target: 4.482, TD Error: 1.288, New V(s): 3.323\n", "2025-08-02 09:46:06,010 - INFO - TD Update - State: 1, Reward: -1, Target: 3.482, TD Error: -1.498, New V(s): 4.831\n", "2025-08-02 09:46:06,010 - INFO - TD Update - State: 1, Reward: 0, Target: 5.839, TD Error: 1.008, New V(s): 4.931\n", "2025-08-02 09:46:06,010 - INFO - TD Update - State: 2, Reward: 0, Target: 7.778, TD Error: 1.290, New V(s): 6.617\n", "2025-08-02 09:46:06,011 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.358, New V(s): 8.778\n", "2025-08-02 09:46:06,011 - INFO - Episode Complete - Total Reward: 9, Avg TD Error: 1.289\n", "2025-08-02 09:46:06,011 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,011 - INFO - TD Update - State: 0, Reward: 0, Target: 4.438, TD Error: 1.115, New V(s): 3.435\n", "2025-08-02 09:46:06,011 - INFO - TD Update - State: 1, Reward: 0, Target: 5.955, TD Error: 1.024, New V(s): 5.034\n", "2025-08-02 09:46:06,011 - INFO - TD Update - State: 2, Reward: 0, Target: 7.900, TD Error: 1.283, New V(s): 6.745\n", "2025-08-02 09:46:06,011 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.222, New V(s): 8.900\n", "2025-08-02 09:46:06,012 - INFO - Episode Complete - Total Reward: 10, Avg TD Error: 1.161\n", "2025-08-02 09:46:06,012 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,012 - INFO - TD Update - State: 0, Reward: 0, Target: 4.530, TD Error: 1.096, New V(s): 3.544\n", "2025-08-02 09:46:06,012 - INFO - TD Update - State: 1, Reward: -1, Target: 3.530, TD Error: -1.503, New V(s): 4.883\n", "2025-08-02 09:46:06,012 - INFO - TD Update - State: 1, Reward: 0, Target: 6.070, TD Error: 1.187, New V(s): 5.002\n", "2025-08-02 09:46:06,013 - INFO - TD Update - State: 2, Reward: -1, Target: 5.070, TD Error: -1.674, New V(s): 6.577\n", "2025-08-02 09:46:06,013 - INFO - TD Update - State: 2, Reward: 0, Target: 8.010, TD Error: 1.432, New V(s): 6.721\n", "2025-08-02 09:46:06,013 - INFO - TD Update - State: 3, Reward: -1, Target: 7.010, TD Error: -1.890, New V(s): 8.711\n", "2025-08-02 09:46:06,013 - INFO - TD Update - State: 3, Reward: -1, Target: 6.840, TD Error: -1.871, New V(s): 8.524\n", "2025-08-02 09:46:06,013 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.476, New V(s): 8.671\n", "2025-08-02 09:46:06,013 - INFO - Episode Complete - Total Reward: 6, Avg TD Error: 1.516\n", "2025-08-02 09:46:06,013 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,014 - INFO - TD Update - State: 0, Reward: -1, Target: 2.190, TD Error: -1.354, New V(s): 3.409\n", "2025-08-02 09:46:06,014 - INFO - TD Update - State: 0, Reward: -1, Target: 2.068, TD Error: -1.341, New V(s): 3.275\n", "2025-08-02 09:46:06,014 - INFO - TD Update - State: 0, Reward: -1, Target: 1.947, TD Error: -1.327, New V(s): 3.142\n", "2025-08-02 09:46:06,014 - INFO - TD Update - State: 0, Reward: 0, Target: 4.502, TD Error: 1.360, New V(s): 3.278\n", "2025-08-02 09:46:06,015 - INFO - TD Update - State: 1, Reward: 0, Target: 6.049, TD Error: 1.047, New V(s): 5.107\n", "2025-08-02 09:46:06,015 - INFO - TD Update - State: 2, Reward: -1, Target: 5.049, TD Error: -1.672, New V(s): 6.554\n", "2025-08-02 09:46:06,015 - INFO - TD Update - State: 2, Reward: -1, Target: 4.898, TD Error: -1.655, New V(s): 6.388\n", "2025-08-02 09:46:06,015 - INFO - TD Update - State: 2, Reward: -1, Target: 4.749, TD Error: -1.639, New V(s): 6.224\n", "2025-08-02 09:46:06,015 - INFO - TD Update - State: 2, Reward: -1, Target: 4.602, TD Error: -1.622, New V(s): 6.062\n", "2025-08-02 09:46:06,015 - INFO - TD Update - State: 2, Reward: 0, Target: 7.804, TD Error: 1.742, New V(s): 6.236\n", "2025-08-02 09:46:06,015 - INFO - TD Update - State: 3, Reward: -1, Target: 6.804, TD Error: -1.867, New V(s): 8.485\n", "2025-08-02 09:46:06,016 - INFO - TD Update - State: 3, Reward: -1, Target: 6.636, TD Error: -1.848, New V(s): 8.300\n", "2025-08-02 09:46:06,016 - INFO - TD Update - State: 3, Reward: -1, Target: 6.470, TD Error: -1.830, New V(s): 8.117\n", "2025-08-02 09:46:06,016 - INFO - TD Update - State: 3, Reward: -1, Target: 6.305, TD Error: -1.812, New V(s): 7.936\n", "2025-08-02 09:46:06,016 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 2.064, New V(s): 8.142\n", "2025-08-02 09:46:06,016 - INFO - Episode Complete - Total Reward: -1, Avg TD Error: 1.612\n", "2025-08-02 09:46:06,016 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,016 - INFO - TD Update - State: 0, Reward: 0, Target: 4.596, TD Error: 1.318, New V(s): 3.410\n", "2025-08-02 09:46:06,017 - INFO - TD Update - State: 1, Reward: -1, Target: 3.596, TD Error: -1.511, New V(s): 4.956\n", "2025-08-02 09:46:06,017 - INFO - TD Update - State: 1, Reward: -1, Target: 3.460, TD Error: -1.496, New V(s): 4.806\n", "2025-08-02 09:46:06,017 - INFO - TD Update - State: 1, Reward: -1, Target: 3.326, TD Error: -1.481, New V(s): 4.658\n", "2025-08-02 09:46:06,017 - INFO - TD Update - State: 1, Reward: -1, Target: 3.192, TD Error: -1.466, New V(s): 4.512\n", "2025-08-02 09:46:06,018 - INFO - TD Update - State: 1, Reward: 0, Target: 5.612, TD Error: 1.101, New V(s): 4.622\n", "2025-08-02 09:46:06,018 - INFO - TD Update - State: 2, Reward: -1, Target: 4.612, TD Error: -1.624, New V(s): 6.074\n", "2025-08-02 09:46:06,018 - INFO - TD Update - State: 2, Reward: 0, Target: 7.328, TD Error: 1.254, New V(s): 6.199\n", "2025-08-02 09:46:06,018 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.858, New V(s): 8.328\n", "2025-08-02 09:46:06,018 - INFO - Episode Complete - Total Reward: 5, Avg TD Error: 1.456\n", "2025-08-02 09:46:06,018 - INFO - \n", "๐Ÿ“Š Episode 60 Summary:\n", "2025-08-02 09:46:06,019 - INFO - Value Function: [3.40975648 4.62162398 6.19914033 8.32782678 0. ]\n", "2025-08-02 09:46:06,019 - INFO - Recent Avg Reward: 6.60\n", "2025-08-02 09:46:06,019 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,019 - INFO - TD Update - State: 0, Reward: -1, Target: 2.069, TD Error: -1.341, New V(s): 3.276\n", "2025-08-02 09:46:06,019 - INFO - TD Update - State: 0, Reward: -1, Target: 1.948, TD Error: -1.328, New V(s): 3.143\n", "2025-08-02 09:46:06,020 - INFO - TD Update - State: 0, Reward: -1, Target: 1.829, TD Error: -1.314, New V(s): 3.011\n", "2025-08-02 09:46:06,020 - INFO - TD Update - State: 0, Reward: 0, Target: 4.159, TD Error: 1.148, New V(s): 3.126\n", "2025-08-02 09:46:06,020 - INFO - TD Update - State: 1, Reward: 0, Target: 5.579, TD Error: 0.958, New V(s): 4.717\n", "2025-08-02 09:46:06,020 - INFO - TD Update - State: 2, Reward: -1, Target: 4.579, TD Error: -1.620, New V(s): 6.037\n", "2025-08-02 09:46:06,020 - INFO - TD Update - State: 2, Reward: 0, Target: 7.495, TD Error: 1.458, New V(s): 6.183\n", "2025-08-02 09:46:06,020 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.672, New V(s): 8.495\n", "2025-08-02 09:46:06,020 - INFO - Episode Complete - Total Reward: 6, Avg TD Error: 1.355\n", "2025-08-02 09:46:06,021 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,021 - INFO - TD Update - State: 0, Reward: 0, Target: 4.246, TD Error: 1.119, New V(s): 3.238\n", "2025-08-02 09:46:06,021 - INFO - TD Update - State: 1, Reward: 0, Target: 5.565, TD Error: 0.847, New V(s): 4.802\n", "2025-08-02 09:46:06,021 - INFO - TD Update - State: 2, Reward: -1, Target: 4.565, TD Error: -1.618, New V(s): 6.021\n", "2025-08-02 09:46:06,021 - INFO - TD Update - State: 2, Reward: 0, Target: 7.646, TD Error: 1.624, New V(s): 6.184\n", "2025-08-02 09:46:06,021 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.505, New V(s): 8.646\n", "2025-08-02 09:46:06,022 - INFO - Episode Complete - Total Reward: 9, Avg TD Error: 1.343\n", "2025-08-02 09:46:06,022 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,022 - INFO - TD Update - State: 0, Reward: 0, Target: 4.322, TD Error: 1.084, New V(s): 3.347\n", "2025-08-02 09:46:06,022 - INFO - TD Update - State: 1, Reward: 0, Target: 5.565, TD Error: 0.763, New V(s): 4.878\n", "2025-08-02 09:46:06,022 - INFO - TD Update - State: 2, Reward: -1, Target: 4.565, TD Error: -1.618, New V(s): 6.022\n", "2025-08-02 09:46:06,023 - INFO - TD Update - State: 2, Reward: -1, Target: 4.420, TD Error: -1.602, New V(s): 5.861\n", "2025-08-02 09:46:06,023 - INFO - TD Update - State: 2, Reward: -1, Target: 4.275, TD Error: -1.586, New V(s): 5.703\n", "2025-08-02 09:46:06,023 - INFO - TD Update - State: 2, Reward: 0, Target: 7.781, TD Error: 2.078, New V(s): 5.911\n", "2025-08-02 09:46:06,023 - INFO - TD Update - State: 3, Reward: -1, Target: 6.781, TD Error: -1.865, New V(s): 8.459\n", "2025-08-02 09:46:06,023 - INFO - TD Update - State: 3, Reward: -1, Target: 6.613, TD Error: -1.846, New V(s): 8.274\n", "2025-08-02 09:46:06,024 - INFO - TD Update - State: 3, Reward: -1, Target: 6.447, TD Error: -1.827, New V(s): 8.092\n", "2025-08-02 09:46:06,024 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.908, New V(s): 8.283\n", "2025-08-02 09:46:06,024 - INFO - Episode Complete - Total Reward: 4, Avg TD Error: 1.618\n", "2025-08-02 09:46:06,024 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,024 - INFO - TD Update - State: 0, Reward: 0, Target: 4.391, TD Error: 1.044, New V(s): 3.451\n", "2025-08-02 09:46:06,024 - INFO - TD Update - State: 1, Reward: 0, Target: 5.320, TD Error: 0.441, New V(s): 4.923\n", "2025-08-02 09:46:06,025 - INFO - TD Update - State: 2, Reward: -1, Target: 4.320, TD Error: -1.591, New V(s): 5.752\n", "2025-08-02 09:46:06,025 - INFO - TD Update - State: 2, Reward: -1, Target: 4.176, TD Error: -1.575, New V(s): 5.594\n", "2025-08-02 09:46:06,025 - INFO - TD Update - State: 2, Reward: 0, Target: 7.454, TD Error: 1.860, New V(s): 5.780\n", "2025-08-02 09:46:06,025 - INFO - TD Update - State: 3, Reward: -1, Target: 6.454, TD Error: -1.828, New V(s): 8.100\n", "2025-08-02 09:46:06,025 - INFO - TD Update - State: 3, Reward: -1, Target: 6.290, TD Error: -1.810, New V(s): 7.919\n", "2025-08-02 09:46:06,025 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 2.081, New V(s): 8.127\n", "2025-08-02 09:46:06,026 - INFO - Episode Complete - Total Reward: 6, Avg TD Error: 1.529\n", "2025-08-02 09:46:06,026 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,026 - INFO - TD Update - State: 0, Reward: -1, Target: 2.106, TD Error: -1.345, New V(s): 3.316\n", "2025-08-02 09:46:06,026 - INFO - TD Update - State: 0, Reward: 0, Target: 4.430, TD Error: 1.114, New V(s): 3.428\n", "2025-08-02 09:46:06,026 - INFO - TD Update - State: 1, Reward: -1, Target: 3.430, TD Error: -1.492, New V(s): 4.773\n", "2025-08-02 09:46:06,026 - INFO - TD Update - State: 1, Reward: -1, Target: 3.296, TD Error: -1.477, New V(s): 4.626\n", "2025-08-02 09:46:06,026 - INFO - TD Update - State: 1, Reward: 0, Target: 5.202, TD Error: 0.577, New V(s): 4.683\n", "2025-08-02 09:46:06,027 - INFO - TD Update - State: 2, Reward: 0, Target: 7.314, TD Error: 1.534, New V(s): 5.934\n", "2025-08-02 09:46:06,027 - INFO - TD Update - State: 3, Reward: -1, Target: 6.314, TD Error: -1.813, New V(s): 7.946\n", "2025-08-02 09:46:06,027 - INFO - TD Update - State: 3, Reward: -1, Target: 6.151, TD Error: -1.795, New V(s): 7.766\n", "2025-08-02 09:46:06,027 - INFO - TD Update - State: 3, Reward: -1, Target: 5.990, TD Error: -1.777, New V(s): 7.588\n", "2025-08-02 09:46:06,027 - INFO - TD Update - State: 3, Reward: -1, Target: 5.830, TD Error: -1.759, New V(s): 7.413\n", "2025-08-02 09:46:06,027 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 2.587, New V(s): 7.671\n", "2025-08-02 09:46:06,027 - INFO - Episode Complete - Total Reward: 3, Avg TD Error: 1.570\n", "2025-08-02 09:46:06,028 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,028 - INFO - TD Update - State: 0, Reward: 0, Target: 4.215, TD Error: 0.787, New V(s): 3.507\n", "2025-08-02 09:46:06,028 - INFO - TD Update - State: 1, Reward: 0, Target: 5.340, TD Error: 0.657, New V(s): 4.749\n", "2025-08-02 09:46:06,028 - INFO - TD Update - State: 2, Reward: 0, Target: 6.904, TD Error: 0.971, New V(s): 6.031\n", "2025-08-02 09:46:06,028 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 2.329, New V(s): 7.904\n", "2025-08-02 09:46:06,028 - INFO - Episode Complete - Total Reward: 10, Avg TD Error: 1.186\n", "2025-08-02 09:46:06,074 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,075 - INFO - TD Update - State: 0, Reward: -1, Target: 2.156, TD Error: -1.351, New V(s): 3.371\n", "2025-08-02 09:46:06,075 - INFO - TD Update - State: 0, Reward: 0, Target: 4.274, TD Error: 0.903, New V(s): 3.462\n", "2025-08-02 09:46:06,075 - INFO - TD Update - State: 1, Reward: 0, Target: 5.428, TD Error: 0.679, New V(s): 4.817\n", "2025-08-02 09:46:06,075 - INFO - TD Update - State: 2, Reward: 0, Target: 7.114, TD Error: 1.083, New V(s): 6.139\n", "2025-08-02 09:46:06,075 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 2.096, New V(s): 8.114\n", "2025-08-02 09:46:06,076 - INFO - Episode Complete - Total Reward: 9, Avg TD Error: 1.222\n", "2025-08-02 09:46:06,076 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,076 - INFO - TD Update - State: 0, Reward: -1, Target: 2.116, TD Error: -1.346, New V(s): 3.327\n", "2025-08-02 09:46:06,076 - INFO - TD Update - State: 0, Reward: 0, Target: 4.335, TD Error: 1.008, New V(s): 3.428\n", "2025-08-02 09:46:06,076 - INFO - TD Update - State: 1, Reward: 0, Target: 5.525, TD Error: 0.708, New V(s): 4.888\n", "2025-08-02 09:46:06,077 - INFO - TD Update - State: 2, Reward: 0, Target: 7.302, TD Error: 1.164, New V(s): 6.255\n", "2025-08-02 09:46:06,077 - INFO - TD Update - State: 3, Reward: -1, Target: 6.302, TD Error: -1.811, New V(s): 7.933\n", "2025-08-02 09:46:06,077 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 2.067, New V(s): 8.139\n", "2025-08-02 09:46:06,077 - INFO - Episode Complete - Total Reward: 8, Avg TD Error: 1.351\n", "2025-08-02 09:46:06,077 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,078 - INFO - TD Update - State: 0, Reward: 0, Target: 4.399, TD Error: 0.971, New V(s): 3.525\n", "2025-08-02 09:46:06,078 - INFO - TD Update - State: 1, Reward: 0, Target: 5.630, TD Error: 0.742, New V(s): 4.962\n", "2025-08-02 09:46:06,078 - INFO - TD Update - State: 2, Reward: 0, Target: 7.325, TD Error: 1.070, New V(s): 6.362\n", "2025-08-02 09:46:06,078 - INFO - TD Update - State: 3, Reward: -1, Target: 6.325, TD Error: -1.814, New V(s): 7.958\n", "2025-08-02 09:46:06,078 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 2.042, New V(s): 8.162\n", "2025-08-02 09:46:06,079 - INFO - Episode Complete - Total Reward: 9, Avg TD Error: 1.328\n", "2025-08-02 09:46:06,079 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,079 - INFO - TD Update - State: 0, Reward: 0, Target: 4.466, TD Error: 0.941, New V(s): 3.619\n", "2025-08-02 09:46:06,079 - INFO - TD Update - State: 1, Reward: -1, Target: 3.466, TD Error: -1.496, New V(s): 4.812\n", "2025-08-02 09:46:06,079 - INFO - TD Update - State: 1, Reward: -1, Target: 3.331, TD Error: -1.481, New V(s): 4.664\n", "2025-08-02 09:46:06,080 - INFO - TD Update - State: 1, Reward: -1, Target: 3.198, TD Error: -1.466, New V(s): 4.517\n", "2025-08-02 09:46:06,080 - INFO - TD Update - State: 1, Reward: 0, Target: 5.726, TD Error: 1.209, New V(s): 4.638\n", "2025-08-02 09:46:06,080 - INFO - TD Update - State: 2, Reward: 0, Target: 7.346, TD Error: 0.984, New V(s): 6.461\n", "2025-08-02 09:46:06,080 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.838, New V(s): 8.346\n", "2025-08-02 09:46:06,080 - INFO - Episode Complete - Total Reward: 7, Avg TD Error: 1.345\n", "2025-08-02 09:46:06,080 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,081 - INFO - TD Update - State: 0, Reward: 0, Target: 4.174, TD Error: 0.555, New V(s): 3.675\n", "2025-08-02 09:46:06,081 - INFO - TD Update - State: 1, Reward: 0, Target: 5.815, TD Error: 1.176, New V(s): 4.756\n", "2025-08-02 09:46:06,081 - INFO - TD Update - State: 2, Reward: -1, Target: 4.815, TD Error: -1.646, New V(s): 6.296\n", "2025-08-02 09:46:06,081 - INFO - TD Update - State: 2, Reward: -1, Target: 4.666, TD Error: -1.630, New V(s): 6.133\n", "2025-08-02 09:46:06,082 - INFO - TD Update - State: 2, Reward: -1, Target: 4.520, TD Error: -1.613, New V(s): 5.972\n", "2025-08-02 09:46:06,082 - INFO - TD Update - State: 2, Reward: -1, Target: 4.375, TD Error: -1.597, New V(s): 5.812\n", "2025-08-02 09:46:06,082 - INFO - TD Update - State: 2, Reward: -1, Target: 4.231, TD Error: -1.581, New V(s): 5.654\n", "2025-08-02 09:46:06,082 - INFO - TD Update - State: 2, Reward: -1, Target: 4.089, TD Error: -1.565, New V(s): 5.497\n", "2025-08-02 09:46:06,082 - INFO - TD Update - State: 2, Reward: 0, Target: 7.511, TD Error: 2.014, New V(s): 5.699\n", "2025-08-02 09:46:06,082 - INFO - TD Update - State: 3, Reward: -1, Target: 6.511, TD Error: -1.835, New V(s): 8.163\n", "2025-08-02 09:46:06,083 - INFO - TD Update - State: 3, Reward: -1, Target: 6.346, TD Error: -1.816, New V(s): 7.981\n", "2025-08-02 09:46:06,083 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 2.019, New V(s): 8.183\n", "2025-08-02 09:46:06,083 - INFO - Episode Complete - Total Reward: 2, Avg TD Error: 1.587\n", "2025-08-02 09:46:06,083 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,083 - INFO - TD Update - State: 0, Reward: -1, Target: 2.307, TD Error: -1.367, New V(s): 3.538\n", "2025-08-02 09:46:06,083 - INFO - TD Update - State: 0, Reward: -1, Target: 2.184, TD Error: -1.354, New V(s): 3.402\n", "2025-08-02 09:46:06,084 - INFO - TD Update - State: 0, Reward: -1, Target: 2.062, TD Error: -1.340, New V(s): 3.268\n", "2025-08-02 09:46:06,084 - INFO - TD Update - State: 0, Reward: 0, Target: 4.280, TD Error: 1.012, New V(s): 3.370\n", "2025-08-02 09:46:06,084 - INFO - TD Update - State: 1, Reward: -1, Target: 3.280, TD Error: -1.476, New V(s): 4.608\n", "2025-08-02 09:46:06,084 - INFO - TD Update - State: 1, Reward: 0, Target: 5.129, TD Error: 0.521, New V(s): 4.660\n", "2025-08-02 09:46:06,084 - INFO - TD Update - State: 2, Reward: -1, Target: 4.129, TD Error: -1.570, New V(s): 5.542\n", "2025-08-02 09:46:06,085 - INFO - TD Update - State: 2, Reward: 0, Target: 7.365, TD Error: 1.823, New V(s): 5.724\n", "2025-08-02 09:46:06,085 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.817, New V(s): 8.365\n", "2025-08-02 09:46:06,085 - INFO - Episode Complete - Total Reward: 5, Avg TD Error: 1.364\n", "2025-08-02 09:46:06,085 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,085 - INFO - TD Update - State: 0, Reward: 0, Target: 4.194, TD Error: 0.825, New V(s): 3.452\n", "2025-08-02 09:46:06,086 - INFO - TD Update - State: 1, Reward: 0, Target: 5.152, TD Error: 0.491, New V(s): 4.710\n", "2025-08-02 09:46:06,086 - INFO - TD Update - State: 2, Reward: 0, Target: 7.528, TD Error: 1.804, New V(s): 5.904\n", "2025-08-02 09:46:06,086 - INFO - TD Update - State: 3, Reward: -1, Target: 6.528, TD Error: -1.836, New V(s): 8.181\n", "2025-08-02 09:46:06,086 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.819, New V(s): 8.363\n", "2025-08-02 09:46:06,086 - INFO - Episode Complete - Total Reward: 9, Avg TD Error: 1.355\n", "2025-08-02 09:46:06,087 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,087 - INFO - TD Update - State: 0, Reward: 0, Target: 4.239, TD Error: 0.786, New V(s): 3.531\n", "2025-08-02 09:46:06,087 - INFO - TD Update - State: 1, Reward: -1, Target: 3.239, TD Error: -1.471, New V(s): 4.562\n", "2025-08-02 09:46:06,087 - INFO - TD Update - State: 1, Reward: -1, Target: 3.106, TD Error: -1.456, New V(s): 4.417\n", "2025-08-02 09:46:06,087 - INFO - TD Update - State: 1, Reward: 0, Target: 5.314, TD Error: 0.897, New V(s): 4.507\n", "2025-08-02 09:46:06,088 - INFO - TD Update - State: 2, Reward: -1, Target: 4.314, TD Error: -1.590, New V(s): 5.745\n", "2025-08-02 09:46:06,088 - INFO - TD Update - State: 2, Reward: -1, Target: 4.171, TD Error: -1.575, New V(s): 5.588\n", "2025-08-02 09:46:06,088 - INFO - TD Update - State: 2, Reward: 0, Target: 7.527, TD Error: 1.939, New V(s): 5.782\n", "2025-08-02 09:46:06,088 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.637, New V(s): 8.527\n", "2025-08-02 09:46:06,088 - INFO - Episode Complete - Total Reward: 6, Avg TD Error: 1.419\n", "2025-08-02 09:46:06,088 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,089 - INFO - TD Update - State: 0, Reward: -1, Target: 2.178, TD Error: -1.353, New V(s): 3.395\n", "2025-08-02 09:46:06,089 - INFO - TD Update - State: 0, Reward: -1, Target: 2.056, TD Error: -1.340, New V(s): 3.262\n", "2025-08-02 09:46:06,089 - INFO - TD Update - State: 0, Reward: -1, Target: 1.935, TD Error: -1.326, New V(s): 3.129\n", "2025-08-02 09:46:06,089 - INFO - TD Update - State: 0, Reward: 0, Target: 4.056, TD Error: 0.927, New V(s): 3.222\n", "2025-08-02 09:46:06,089 - INFO - TD Update - State: 1, Reward: 0, Target: 5.204, TD Error: 0.697, New V(s): 4.576\n", "2025-08-02 09:46:06,089 - INFO - TD Update - State: 2, Reward: 0, Target: 7.674, TD Error: 1.892, New V(s): 5.971\n", "2025-08-02 09:46:06,090 - INFO - TD Update - State: 3, Reward: -1, Target: 6.674, TD Error: -1.853, New V(s): 8.341\n", "2025-08-02 09:46:06,090 - INFO - TD Update - State: 3, Reward: -1, Target: 6.507, TD Error: -1.834, New V(s): 8.158\n", "2025-08-02 09:46:06,090 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.842, New V(s): 8.342\n", "2025-08-02 09:46:06,090 - INFO - Episode Complete - Total Reward: 5, Avg TD Error: 1.452\n", "2025-08-02 09:46:06,090 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,091 - INFO - TD Update - State: 0, Reward: 0, Target: 4.119, TD Error: 0.897, New V(s): 3.311\n", "2025-08-02 09:46:06,091 - INFO - TD Update - State: 1, Reward: -1, Target: 3.119, TD Error: -1.458, New V(s): 4.430\n", "2025-08-02 09:46:06,091 - INFO - TD Update - State: 1, Reward: -1, Target: 2.987, TD Error: -1.443, New V(s): 4.286\n", "2025-08-02 09:46:06,091 - INFO - TD Update - State: 1, Reward: -1, Target: 2.858, TD Error: -1.429, New V(s): 4.143\n", "2025-08-02 09:46:06,091 - INFO - TD Update - State: 1, Reward: -1, Target: 2.729, TD Error: -1.414, New V(s): 4.002\n", "2025-08-02 09:46:06,091 - INFO - TD Update - State: 1, Reward: -1, Target: 2.602, TD Error: -1.400, New V(s): 3.862\n", "2025-08-02 09:46:06,092 - INFO - TD Update - State: 1, Reward: -1, Target: 2.476, TD Error: -1.386, New V(s): 3.723\n", "2025-08-02 09:46:06,092 - INFO - TD Update - State: 1, Reward: -1, Target: 2.351, TD Error: -1.372, New V(s): 3.586\n", "2025-08-02 09:46:06,092 - INFO - TD Update - State: 1, Reward: -1, Target: 2.227, TD Error: -1.359, New V(s): 3.450\n", "2025-08-02 09:46:06,092 - INFO - TD Update - State: 1, Reward: 0, Target: 5.374, TD Error: 1.924, New V(s): 3.643\n", "2025-08-02 09:46:06,092 - INFO - TD Update - State: 2, Reward: 0, Target: 7.508, TD Error: 1.537, New V(s): 6.125\n", "2025-08-02 09:46:06,093 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.658, New V(s): 8.508\n", "2025-08-02 09:46:06,093 - INFO - Episode Complete - Total Reward: 2, Avg TD Error: 1.440\n", "2025-08-02 09:46:06,093 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,093 - INFO - TD Update - State: 0, Reward: 0, Target: 3.278, TD Error: -0.033, New V(s): 3.308\n", "2025-08-02 09:46:06,093 - INFO - TD Update - State: 1, Reward: 0, Target: 5.512, TD Error: 1.870, New V(s): 3.830\n", "2025-08-02 09:46:06,093 - INFO - TD Update - State: 2, Reward: 0, Target: 7.657, TD Error: 1.532, New V(s): 6.278\n", "2025-08-02 09:46:06,093 - INFO - TD Update - State: 3, Reward: -1, Target: 6.657, TD Error: -1.851, New V(s): 8.323\n", "2025-08-02 09:46:06,094 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.677, New V(s): 8.490\n", "2025-08-02 09:46:06,094 - INFO - Episode Complete - Total Reward: 9, Avg TD Error: 1.393\n", "2025-08-02 09:46:06,094 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,094 - INFO - TD Update - State: 0, Reward: -1, Target: 1.977, TD Error: -1.331, New V(s): 3.175\n", "2025-08-02 09:46:06,094 - INFO - TD Update - State: 0, Reward: 0, Target: 3.447, TD Error: 0.272, New V(s): 3.202\n", "2025-08-02 09:46:06,094 - INFO - TD Update - State: 1, Reward: -1, Target: 2.447, TD Error: -1.383, New V(s): 3.691\n", "2025-08-02 09:46:06,095 - INFO - TD Update - State: 1, Reward: -1, Target: 2.322, TD Error: -1.369, New V(s): 3.554\n", "2025-08-02 09:46:06,095 - INFO - TD Update - State: 1, Reward: 0, Target: 5.650, TD Error: 2.096, New V(s): 3.764\n", "2025-08-02 09:46:06,095 - INFO - TD Update - State: 2, Reward: 0, Target: 7.641, TD Error: 1.364, New V(s): 6.414\n", "2025-08-02 09:46:06,095 - INFO - TD Update - State: 3, Reward: -1, Target: 6.641, TD Error: -1.849, New V(s): 8.306\n", "2025-08-02 09:46:06,095 - INFO - TD Update - State: 3, Reward: -1, Target: 6.475, TD Error: -1.831, New V(s): 8.123\n", "2025-08-02 09:46:06,095 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.877, New V(s): 8.310\n", "2025-08-02 09:46:06,096 - INFO - Episode Complete - Total Reward: 5, Avg TD Error: 1.486\n", "2025-08-02 09:46:06,096 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,096 - INFO - TD Update - State: 0, Reward: -1, Target: 1.882, TD Error: -1.320, New V(s): 3.070\n", "2025-08-02 09:46:06,096 - INFO - TD Update - State: 0, Reward: 0, Target: 3.387, TD Error: 0.317, New V(s): 3.102\n", "2025-08-02 09:46:06,096 - INFO - TD Update - State: 1, Reward: 0, Target: 5.773, TD Error: 2.009, New V(s): 3.965\n", "2025-08-02 09:46:06,096 - INFO - TD Update - State: 2, Reward: -1, Target: 4.773, TD Error: -1.641, New V(s): 6.250\n", "2025-08-02 09:46:06,097 - INFO - TD Update - State: 2, Reward: 0, Target: 7.479, TD Error: 1.229, New V(s): 6.373\n", "2025-08-02 09:46:06,097 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.690, New V(s): 8.479\n", "2025-08-02 09:46:06,097 - INFO - Episode Complete - Total Reward: 8, Avg TD Error: 1.368\n", "2025-08-02 09:46:06,097 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,097 - INFO - TD Update - State: 0, Reward: 0, Target: 3.568, TD Error: 0.466, New V(s): 3.148\n", "2025-08-02 09:46:06,097 - INFO - TD Update - State: 1, Reward: 0, Target: 5.736, TD Error: 1.771, New V(s): 4.142\n", "2025-08-02 09:46:06,098 - INFO - TD Update - State: 2, Reward: 0, Target: 7.631, TD Error: 1.258, New V(s): 6.499\n", "2025-08-02 09:46:06,098 - INFO - TD Update - State: 3, Reward: -1, Target: 6.631, TD Error: -1.848, New V(s): 8.294\n", "2025-08-02 09:46:06,098 - INFO - TD Update - State: 3, Reward: -1, Target: 6.465, TD Error: -1.829, New V(s): 8.112\n", "2025-08-02 09:46:06,098 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.888, New V(s): 8.300\n", "2025-08-02 09:46:06,098 - INFO - Episode Complete - Total Reward: 8, Avg TD Error: 1.510\n", "2025-08-02 09:46:06,098 - INFO - \n", "๐Ÿ“Š Episode 80 Summary:\n", "2025-08-02 09:46:06,099 - INFO - Value Function: [3.14845117 4.14187375 6.49887922 8.30035909 0. ]\n", "2025-08-02 09:46:06,099 - INFO - Recent Avg Reward: 5.90\n", "2025-08-02 09:46:06,099 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,099 - INFO - TD Update - State: 0, Reward: 0, Target: 3.728, TD Error: 0.579, New V(s): 3.206\n", "2025-08-02 09:46:06,099 - INFO - TD Update - State: 1, Reward: -1, Target: 2.728, TD Error: -1.414, New V(s): 4.000\n", "2025-08-02 09:46:06,099 - INFO - TD Update - State: 1, Reward: -1, Target: 2.600, TD Error: -1.400, New V(s): 3.860\n", "2025-08-02 09:46:06,100 - INFO - TD Update - State: 1, Reward: -1, Target: 2.474, TD Error: -1.386, New V(s): 3.722\n", "2025-08-02 09:46:06,100 - INFO - TD Update - State: 1, Reward: -1, Target: 2.350, TD Error: -1.372, New V(s): 3.585\n", "2025-08-02 09:46:06,100 - INFO - TD Update - State: 1, Reward: -1, Target: 2.226, TD Error: -1.358, New V(s): 3.449\n", "2025-08-02 09:46:06,100 - INFO - TD Update - State: 1, Reward: 0, Target: 5.849, TD Error: 2.400, New V(s): 3.689\n", "2025-08-02 09:46:06,100 - INFO - TD Update - State: 2, Reward: 0, Target: 7.470, TD Error: 0.971, New V(s): 6.596\n", "2025-08-02 09:46:06,100 - INFO - TD Update - State: 3, Reward: -1, Target: 6.470, TD Error: -1.830, New V(s): 8.117\n", "2025-08-02 09:46:06,101 - INFO - TD Update - State: 3, Reward: -1, Target: 6.306, TD Error: -1.812, New V(s): 7.936\n", "2025-08-02 09:46:06,101 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 2.064, New V(s): 8.143\n", "2025-08-02 09:46:06,101 - INFO - Episode Complete - Total Reward: 3, Avg TD Error: 1.508\n", "2025-08-02 09:46:06,101 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,101 - INFO - TD Update - State: 0, Reward: -1, Target: 1.886, TD Error: -1.321, New V(s): 3.074\n", "2025-08-02 09:46:06,101 - INFO - TD Update - State: 0, Reward: -1, Target: 1.767, TD Error: -1.307, New V(s): 2.944\n", "2025-08-02 09:46:06,101 - INFO - TD Update - State: 0, Reward: 0, Target: 3.320, TD Error: 0.376, New V(s): 2.981\n", "2025-08-02 09:46:06,102 - INFO - TD Update - State: 1, Reward: -1, Target: 2.320, TD Error: -1.369, New V(s): 3.552\n", "2025-08-02 09:46:06,102 - INFO - TD Update - State: 1, Reward: 0, Target: 5.936, TD Error: 2.385, New V(s): 3.790\n", "2025-08-02 09:46:06,102 - INFO - TD Update - State: 2, Reward: -1, Target: 4.936, TD Error: -1.660, New V(s): 6.430\n", "2025-08-02 09:46:06,104 - INFO - TD Update - State: 2, Reward: 0, Target: 7.328, TD Error: 0.898, New V(s): 6.520\n", "2025-08-02 09:46:06,104 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.857, New V(s): 8.328\n", "2025-08-02 09:46:06,104 - INFO - Episode Complete - Total Reward: 6, Avg TD Error: 1.397\n", "2025-08-02 09:46:06,105 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,105 - INFO - TD Update - State: 0, Reward: 0, Target: 3.411, TD Error: 0.430, New V(s): 3.024\n", "2025-08-02 09:46:06,105 - INFO - TD Update - State: 1, Reward: 0, Target: 5.868, TD Error: 2.078, New V(s): 3.998\n", "2025-08-02 09:46:06,105 - INFO - TD Update - State: 2, Reward: 0, Target: 7.495, TD Error: 0.976, New V(s): 6.617\n", "2025-08-02 09:46:06,105 - INFO - TD Update - State: 3, Reward: -1, Target: 6.495, TD Error: -1.833, New V(s): 8.145\n", "2025-08-02 09:46:06,106 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.855, New V(s): 8.331\n", "2025-08-02 09:46:06,106 - INFO - Episode Complete - Total Reward: 9, Avg TD Error: 1.434\n", "2025-08-02 09:46:06,106 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,106 - INFO - TD Update - State: 0, Reward: -1, Target: 1.722, TD Error: -1.302, New V(s): 2.894\n", "2025-08-02 09:46:06,107 - INFO - TD Update - State: 0, Reward: 0, Target: 3.598, TD Error: 0.704, New V(s): 2.964\n", "2025-08-02 09:46:06,107 - INFO - TD Update - State: 1, Reward: -1, Target: 2.598, TD Error: -1.400, New V(s): 3.858\n", "2025-08-02 09:46:06,107 - INFO - TD Update - State: 1, Reward: 0, Target: 5.956, TD Error: 2.098, New V(s): 4.068\n", "2025-08-02 09:46:06,107 - INFO - TD Update - State: 2, Reward: -1, Target: 4.956, TD Error: -1.662, New V(s): 6.451\n", "2025-08-02 09:46:06,108 - INFO - TD Update - State: 2, Reward: 0, Target: 7.497, TD Error: 1.046, New V(s): 6.556\n", "2025-08-02 09:46:06,108 - INFO - TD Update - State: 3, Reward: -1, Target: 6.497, TD Error: -1.833, New V(s): 8.147\n", "2025-08-02 09:46:06,108 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.853, New V(s): 8.332\n", "2025-08-02 09:46:06,108 - INFO - Episode Complete - Total Reward: 6, Avg TD Error: 1.487\n", "2025-08-02 09:46:06,109 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,109 - INFO - TD Update - State: 0, Reward: -1, Target: 1.668, TD Error: -1.296, New V(s): 2.835\n", "2025-08-02 09:46:06,109 - INFO - TD Update - State: 0, Reward: -1, Target: 1.551, TD Error: -1.283, New V(s): 2.706\n", "2025-08-02 09:46:06,110 - INFO - TD Update - State: 0, Reward: 0, Target: 3.661, TD Error: 0.955, New V(s): 2.802\n", "2025-08-02 09:46:06,110 - INFO - TD Update - State: 1, Reward: -1, Target: 2.661, TD Error: -1.407, New V(s): 3.927\n", "2025-08-02 09:46:06,110 - INFO - TD Update - State: 1, Reward: -1, Target: 2.534, TD Error: -1.393, New V(s): 3.788\n", "2025-08-02 09:46:06,110 - INFO - TD Update - State: 1, Reward: 0, Target: 5.900, TD Error: 2.112, New V(s): 3.999\n", "2025-08-02 09:46:06,111 - INFO - TD Update - State: 2, Reward: -1, Target: 4.900, TD Error: -1.656, New V(s): 6.390\n", "2025-08-02 09:46:06,111 - INFO - TD Update - State: 2, Reward: 0, Target: 7.499, TD Error: 1.109, New V(s): 6.501\n", "2025-08-02 09:46:06,111 - INFO - TD Update - State: 3, Reward: -1, Target: 6.499, TD Error: -1.833, New V(s): 8.149\n", "2025-08-02 09:46:06,111 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.851, New V(s): 8.334\n", "2025-08-02 09:46:06,111 - INFO - Episode Complete - Total Reward: 4, Avg TD Error: 1.490\n", "2025-08-02 09:46:06,112 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,112 - INFO - TD Update - State: 0, Reward: -1, Target: 1.522, TD Error: -1.280, New V(s): 2.674\n", "2025-08-02 09:46:06,112 - INFO - TD Update - State: 0, Reward: -1, Target: 1.406, TD Error: -1.267, New V(s): 2.547\n", "2025-08-02 09:46:06,112 - INFO - TD Update - State: 0, Reward: -1, Target: 1.292, TD Error: -1.255, New V(s): 2.422\n", "2025-08-02 09:46:06,112 - INFO - TD Update - State: 0, Reward: -1, Target: 1.179, TD Error: -1.242, New V(s): 2.297\n", "2025-08-02 09:46:06,112 - INFO - TD Update - State: 0, Reward: -1, Target: 1.068, TD Error: -1.230, New V(s): 2.174\n", "2025-08-02 09:46:06,113 - INFO - TD Update - State: 0, Reward: -1, Target: 0.957, TD Error: -1.217, New V(s): 2.053\n", "2025-08-02 09:46:06,113 - INFO - TD Update - State: 0, Reward: -1, Target: 0.847, TD Error: -1.205, New V(s): 1.932\n", "2025-08-02 09:46:06,113 - INFO - TD Update - State: 0, Reward: -1, Target: 0.739, TD Error: -1.193, New V(s): 1.813\n", "2025-08-02 09:46:06,113 - INFO - TD Update - State: 0, Reward: -1, Target: 0.632, TD Error: -1.181, New V(s): 1.695\n", "2025-08-02 09:46:06,113 - INFO - TD Update - State: 0, Reward: 0, Target: 3.599, TD Error: 1.905, New V(s): 1.885\n", "2025-08-02 09:46:06,114 - INFO - TD Update - State: 1, Reward: 0, Target: 5.851, TD Error: 1.852, New V(s): 4.184\n", "2025-08-02 09:46:06,114 - INFO - TD Update - State: 2, Reward: -1, Target: 4.851, TD Error: -1.650, New V(s): 6.336\n", "2025-08-02 09:46:06,114 - INFO - TD Update - State: 2, Reward: -1, Target: 4.703, TD Error: -1.634, New V(s): 6.173\n", "2025-08-02 09:46:06,114 - INFO - TD Update - State: 2, Reward: -1, Target: 4.556, TD Error: -1.617, New V(s): 6.011\n", "2025-08-02 09:46:06,114 - INFO - TD Update - State: 2, Reward: 0, Target: 7.501, TD Error: 1.490, New V(s): 6.160\n", "2025-08-02 09:46:06,115 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.666, New V(s): 8.501\n", "2025-08-02 09:46:06,115 - INFO - Episode Complete - Total Reward: -2, Avg TD Error: 1.430\n", "2025-08-02 09:46:06,115 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,115 - INFO - TD Update - State: 0, Reward: 0, Target: 3.766, TD Error: 1.881, New V(s): 2.073\n", "2025-08-02 09:46:06,116 - INFO - TD Update - State: 1, Reward: 0, Target: 5.544, TD Error: 1.360, New V(s): 4.320\n", "2025-08-02 09:46:06,116 - INFO - TD Update - State: 2, Reward: -1, Target: 4.544, TD Error: -1.616, New V(s): 5.998\n", "2025-08-02 09:46:06,116 - INFO - TD Update - State: 2, Reward: -1, Target: 4.399, TD Error: -1.600, New V(s): 5.839\n", "2025-08-02 09:46:06,116 - INFO - TD Update - State: 2, Reward: -1, Target: 4.255, TD Error: -1.584, New V(s): 5.680\n", "2025-08-02 09:46:06,116 - INFO - TD Update - State: 2, Reward: 0, Target: 7.651, TD Error: 1.971, New V(s): 5.877\n", "2025-08-02 09:46:06,116 - INFO - TD Update - State: 3, Reward: -1, Target: 6.651, TD Error: -1.850, New V(s): 8.316\n", "2025-08-02 09:46:06,117 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.684, New V(s): 8.484\n", "2025-08-02 09:46:06,117 - INFO - Episode Complete - Total Reward: 6, Avg TD Error: 1.693\n", "2025-08-02 09:46:06,117 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,117 - INFO - TD Update - State: 0, Reward: -1, Target: 0.866, TD Error: -1.207, New V(s): 1.953\n", "2025-08-02 09:46:06,117 - INFO - TD Update - State: 0, Reward: -1, Target: 0.757, TD Error: -1.195, New V(s): 1.833\n", "2025-08-02 09:46:06,117 - INFO - TD Update - State: 0, Reward: 0, Target: 3.888, TD Error: 2.055, New V(s): 2.039\n", "2025-08-02 09:46:06,117 - INFO - TD Update - State: 1, Reward: -1, Target: 2.888, TD Error: -1.432, New V(s): 4.177\n", "2025-08-02 09:46:06,118 - INFO - TD Update - State: 1, Reward: 0, Target: 5.289, TD Error: 1.112, New V(s): 4.288\n", "2025-08-02 09:46:06,118 - INFO - TD Update - State: 2, Reward: 0, Target: 7.636, TD Error: 1.759, New V(s): 6.053\n", "2025-08-02 09:46:06,118 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.516, New V(s): 8.636\n", "2025-08-02 09:46:06,118 - INFO - Episode Complete - Total Reward: 7, Avg TD Error: 1.468\n", "2025-08-02 09:46:06,118 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,118 - INFO - TD Update - State: 0, Reward: 0, Target: 3.860, TD Error: 1.821, New V(s): 2.221\n", "2025-08-02 09:46:06,118 - INFO - TD Update - State: 1, Reward: 0, Target: 5.448, TD Error: 1.159, New V(s): 4.404\n", "2025-08-02 09:46:06,119 - INFO - TD Update - State: 2, Reward: 0, Target: 7.772, TD Error: 1.719, New V(s): 6.225\n", "2025-08-02 09:46:06,119 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.364, New V(s): 8.772\n", "2025-08-02 09:46:06,119 - INFO - Episode Complete - Total Reward: 10, Avg TD Error: 1.516\n", "2025-08-02 09:46:06,119 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,119 - INFO - TD Update - State: 0, Reward: -1, Target: 0.999, TD Error: -1.222, New V(s): 2.098\n", "2025-08-02 09:46:06,119 - INFO - TD Update - State: 0, Reward: -1, Target: 0.889, TD Error: -1.210, New V(s): 1.977\n", "2025-08-02 09:46:06,119 - INFO - TD Update - State: 0, Reward: -1, Target: 0.780, TD Error: -1.198, New V(s): 1.858\n", "2025-08-02 09:46:06,120 - INFO - TD Update - State: 0, Reward: 0, Target: 3.964, TD Error: 2.106, New V(s): 2.068\n", "2025-08-02 09:46:06,120 - INFO - TD Update - State: 1, Reward: -1, Target: 2.964, TD Error: -1.440, New V(s): 4.260\n", "2025-08-02 09:46:06,120 - INFO - TD Update - State: 1, Reward: 0, Target: 5.602, TD Error: 1.342, New V(s): 4.394\n", "2025-08-02 09:46:06,120 - INFO - TD Update - State: 2, Reward: -1, Target: 4.602, TD Error: -1.622, New V(s): 6.063\n", "2025-08-02 09:46:06,120 - INFO - TD Update - State: 2, Reward: 0, Target: 7.895, TD Error: 1.832, New V(s): 6.246\n", "2025-08-02 09:46:06,120 - INFO - TD Update - State: 3, Reward: -1, Target: 6.895, TD Error: -1.877, New V(s): 8.585\n", "2025-08-02 09:46:06,120 - INFO - TD Update - State: 3, Reward: -1, Target: 6.726, TD Error: -1.858, New V(s): 8.399\n", "2025-08-02 09:46:06,120 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.601, New V(s): 8.559\n", "2025-08-02 09:46:06,121 - INFO - Episode Complete - Total Reward: 3, Avg TD Error: 1.574\n", "2025-08-02 09:46:06,121 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,121 - INFO - TD Update - State: 0, Reward: -1, Target: 0.861, TD Error: -1.207, New V(s): 1.948\n", "2025-08-02 09:46:06,121 - INFO - TD Update - State: 0, Reward: 0, Target: 3.955, TD Error: 2.007, New V(s): 2.148\n", "2025-08-02 09:46:06,121 - INFO - TD Update - State: 1, Reward: -1, Target: 2.955, TD Error: -1.439, New V(s): 4.251\n", "2025-08-02 09:46:06,121 - INFO - TD Update - State: 1, Reward: 0, Target: 5.621, TD Error: 1.371, New V(s): 4.388\n", "2025-08-02 09:46:06,122 - INFO - TD Update - State: 2, Reward: -1, Target: 4.621, TD Error: -1.625, New V(s): 6.083\n", "2025-08-02 09:46:06,122 - INFO - TD Update - State: 2, Reward: -1, Target: 4.475, TD Error: -1.608, New V(s): 5.923\n", "2025-08-02 09:46:06,122 - INFO - TD Update - State: 2, Reward: 0, Target: 7.703, TD Error: 1.780, New V(s): 6.101\n", "2025-08-02 09:46:06,122 - INFO - TD Update - State: 3, Reward: -1, Target: 6.703, TD Error: -1.856, New V(s): 8.373\n", "2025-08-02 09:46:06,122 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.627, New V(s): 8.536\n", "2025-08-02 09:46:06,122 - INFO - Episode Complete - Total Reward: 5, Avg TD Error: 1.613\n", "2025-08-02 09:46:06,122 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,123 - INFO - TD Update - State: 0, Reward: 0, Target: 3.949, TD Error: 1.801, New V(s): 2.328\n", "2025-08-02 09:46:06,123 - INFO - TD Update - State: 1, Reward: 0, Target: 5.491, TD Error: 1.103, New V(s): 4.498\n", "2025-08-02 09:46:06,123 - INFO - TD Update - State: 2, Reward: -1, Target: 4.491, TD Error: -1.610, New V(s): 5.940\n", "2025-08-02 09:46:06,123 - INFO - TD Update - State: 2, Reward: 0, Target: 7.682, TD Error: 1.743, New V(s): 6.114\n", "2025-08-02 09:46:06,123 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.464, New V(s): 8.682\n", "2025-08-02 09:46:06,123 - INFO - Episode Complete - Total Reward: 9, Avg TD Error: 1.544\n", "2025-08-02 09:46:06,123 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,124 - INFO - TD Update - State: 0, Reward: -1, Target: 1.096, TD Error: -1.233, New V(s): 2.205\n", "2025-08-02 09:46:06,124 - INFO - TD Update - State: 0, Reward: 0, Target: 4.048, TD Error: 1.843, New V(s): 2.389\n", "2025-08-02 09:46:06,124 - INFO - TD Update - State: 1, Reward: 0, Target: 5.503, TD Error: 1.005, New V(s): 4.598\n", "2025-08-02 09:46:06,124 - INFO - TD Update - State: 2, Reward: -1, Target: 4.503, TD Error: -1.611, New V(s): 5.953\n", "2025-08-02 09:46:06,124 - INFO - TD Update - State: 2, Reward: 0, Target: 7.814, TD Error: 1.861, New V(s): 6.139\n", "2025-08-02 09:46:06,124 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.318, New V(s): 8.814\n", "2025-08-02 09:46:06,125 - INFO - Episode Complete - Total Reward: 8, Avg TD Error: 1.478\n", "2025-08-02 09:46:06,125 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,125 - INFO - TD Update - State: 0, Reward: 0, Target: 4.139, TD Error: 1.749, New V(s): 2.564\n", "2025-08-02 09:46:06,125 - INFO - TD Update - State: 1, Reward: 0, Target: 5.525, TD Error: 0.927, New V(s): 4.691\n", "2025-08-02 09:46:06,125 - INFO - TD Update - State: 2, Reward: -1, Target: 4.525, TD Error: -1.614, New V(s): 5.978\n", "2025-08-02 09:46:06,125 - INFO - TD Update - State: 2, Reward: 0, Target: 7.933, TD Error: 1.955, New V(s): 6.173\n", "2025-08-02 09:46:06,125 - INFO - TD Update - State: 3, Reward: -1, Target: 6.933, TD Error: -1.881, New V(s): 8.626\n", "2025-08-02 09:46:06,125 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.374, New V(s): 8.763\n", "2025-08-02 09:46:06,126 - INFO - Episode Complete - Total Reward: 8, Avg TD Error: 1.583\n", "2025-08-02 09:46:06,126 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,126 - INFO - TD Update - State: 0, Reward: 0, Target: 4.222, TD Error: 1.658, New V(s): 2.730\n", "2025-08-02 09:46:06,126 - INFO - TD Update - State: 1, Reward: 0, Target: 5.556, TD Error: 0.865, New V(s): 4.778\n", "2025-08-02 09:46:06,126 - INFO - TD Update - State: 2, Reward: 0, Target: 7.887, TD Error: 1.714, New V(s): 6.344\n", "2025-08-02 09:46:06,126 - INFO - TD Update - State: 3, Reward: -1, Target: 6.887, TD Error: -1.876, New V(s): 8.576\n", "2025-08-02 09:46:06,126 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.424, New V(s): 8.718\n", "2025-08-02 09:46:06,127 - INFO - Episode Complete - Total Reward: 9, Avg TD Error: 1.507\n", "2025-08-02 09:46:06,127 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,127 - INFO - TD Update - State: 0, Reward: -1, Target: 1.457, TD Error: -1.273, New V(s): 2.603\n", "2025-08-02 09:46:06,127 - INFO - TD Update - State: 0, Reward: -1, Target: 1.343, TD Error: -1.260, New V(s): 2.477\n", "2025-08-02 09:46:06,127 - INFO - TD Update - State: 0, Reward: -1, Target: 1.229, TD Error: -1.248, New V(s): 2.352\n", "2025-08-02 09:46:06,127 - INFO - TD Update - State: 0, Reward: -1, Target: 1.117, TD Error: -1.235, New V(s): 2.228\n", "2025-08-02 09:46:06,127 - INFO - TD Update - State: 0, Reward: 0, Target: 4.300, TD Error: 2.071, New V(s): 2.436\n", "2025-08-02 09:46:06,128 - INFO - TD Update - State: 1, Reward: -1, Target: 3.300, TD Error: -1.478, New V(s): 4.630\n", "2025-08-02 09:46:06,128 - INFO - TD Update - State: 1, Reward: 0, Target: 5.710, TD Error: 1.080, New V(s): 4.738\n", "2025-08-02 09:46:06,128 - INFO - TD Update - State: 2, Reward: 0, Target: 7.846, TD Error: 1.502, New V(s): 6.495\n", "2025-08-02 09:46:06,128 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.282, New V(s): 8.846\n", "2025-08-02 09:46:06,128 - INFO - Episode Complete - Total Reward: 5, Avg TD Error: 1.381\n", "2025-08-02 09:46:06,128 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,129 - INFO - TD Update - State: 0, Reward: -1, Target: 1.192, TD Error: -1.244, New V(s): 2.311\n", "2025-08-02 09:46:06,129 - INFO - TD Update - State: 0, Reward: -1, Target: 1.080, TD Error: -1.231, New V(s): 2.188\n", "2025-08-02 09:46:06,129 - INFO - TD Update - State: 0, Reward: 0, Target: 4.264, TD Error: 2.076, New V(s): 2.396\n", "2025-08-02 09:46:06,129 - INFO - TD Update - State: 1, Reward: -1, Target: 3.264, TD Error: -1.474, New V(s): 4.590\n", "2025-08-02 09:46:06,129 - INFO - TD Update - State: 1, Reward: 0, Target: 5.845, TD Error: 1.255, New V(s): 4.716\n", "2025-08-02 09:46:06,129 - INFO - TD Update - State: 2, Reward: 0, Target: 7.962, TD Error: 1.467, New V(s): 6.641\n", "2025-08-02 09:46:06,129 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.154, New V(s): 8.962\n", "2025-08-02 09:46:06,130 - INFO - Episode Complete - Total Reward: 7, Avg TD Error: 1.414\n", "2025-08-02 09:46:06,130 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,130 - INFO - TD Update - State: 0, Reward: -1, Target: 1.156, TD Error: -1.240, New V(s): 2.272\n", "2025-08-02 09:46:06,130 - INFO - TD Update - State: 0, Reward: 0, Target: 4.244, TD Error: 1.973, New V(s): 2.469\n", "2025-08-02 09:46:06,130 - INFO - TD Update - State: 1, Reward: -1, Target: 3.244, TD Error: -1.472, New V(s): 4.569\n", "2025-08-02 09:46:06,130 - INFO - TD Update - State: 1, Reward: 0, Target: 5.977, TD Error: 1.408, New V(s): 4.710\n", "2025-08-02 09:46:06,130 - INFO - TD Update - State: 2, Reward: 0, Target: 8.066, TD Error: 1.424, New V(s): 6.784\n", "2025-08-02 09:46:06,131 - INFO - TD Update - State: 3, Reward: -1, Target: 7.066, TD Error: -1.896, New V(s): 8.772\n", "2025-08-02 09:46:06,131 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.228, New V(s): 8.895\n", "2025-08-02 09:46:06,131 - INFO - Episode Complete - Total Reward: 7, Avg TD Error: 1.520\n", "2025-08-02 09:46:06,131 - INFO - \n", "=== Starting New Episode ===\n", "2025-08-02 09:46:06,131 - INFO - TD Update - State: 0, Reward: -1, Target: 1.222, TD Error: -1.247, New V(s): 2.344\n", "2025-08-02 09:46:06,131 - INFO - TD Update - State: 0, Reward: -1, Target: 1.110, TD Error: -1.234, New V(s): 2.221\n", "2025-08-02 09:46:06,131 - INFO - TD Update - State: 0, Reward: 0, Target: 4.239, TD Error: 2.018, New V(s): 2.423\n", "2025-08-02 09:46:06,132 - INFO - TD Update - State: 1, Reward: 0, Target: 6.105, TD Error: 1.396, New V(s): 4.849\n", "2025-08-02 09:46:06,132 - INFO - TD Update - State: 2, Reward: 0, Target: 8.005, TD Error: 1.222, New V(s): 6.906\n", "2025-08-02 09:46:06,132 - INFO - TD Update - State: 3, Reward: -1, Target: 7.005, TD Error: -1.889, New V(s): 8.706\n", "2025-08-02 09:46:06,132 - INFO - TD Update - State: 3, Reward: -1, Target: 6.835, TD Error: -1.871, New V(s): 8.519\n", "2025-08-02 09:46:06,132 - INFO - TD Update - State: 3, Reward: 10, Target: 10.000, TD Error: 1.481, New V(s): 8.667\n", "2025-08-02 09:46:06,132 - INFO - Episode Complete - Total Reward: 6, Avg TD Error: 1.545\n", "2025-08-02 09:46:06,132 - INFO - \n", "โœ… Training Complete!\n", "2025-08-02 09:46:06,133 - INFO - Final Value Function: [2.42265579 4.84913975 6.90591591 8.66697041 0. ]\n", "2025-08-02 09:46:06,134 - INFO - ๐Ÿ’พ Results saved to td_learning_20250802_094606.json\n", "2025-08-02 09:46:06,559 - INFO - ๐Ÿ“ˆ Plots saved to td_learning_plots_20250802_094606.png\n" ] }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "name": "stderr", "output_type": "stream", "text": [ "2025-08-02 09:46:06,655 - INFO - \n", "๐ŸŽ‰ Experiment Complete!\n", "2025-08-02 09:46:06,655 - INFO - ๐Ÿ“ Files saved: td_learning_20250802_094606.json, td_learning_plots_20250802_094606.png\n" ] } ], "source": [ "import numpy as np\n", "import matplotlib.pyplot as plt\n", "import random\n", "import logging\n", "from collections import defaultdict\n", "import json\n", "import os\n", "from datetime import datetime\n", "\n", "logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s')\n", "logger = logging.getLogger(__name__)\n", "\n", "class TDLearningEnvironment:\n", " def __init__(self, num_states=5):\n", " self.num_states = num_states\n", " self.current_state = 0\n", " self.terminal_state = num_states - 1\n", " self.reset()\n", " \n", " def reset(self):\n", " self.current_state = 0\n", " return self.current_state\n", " \n", " def step(self, action):\n", " if self.current_state == self.terminal_state:\n", " return self.current_state, 0, True\n", " \n", " reward = 0\n", " if action == 1:\n", " self.current_state += 1\n", " if self.current_state == self.terminal_state:\n", " reward = 10\n", " else:\n", " reward = -1\n", " \n", " done = self.current_state == self.terminal_state\n", " return self.current_state, reward, done\n", "\n", "class TDLearningAgent:\n", " def __init__(self, num_states, alpha=0.1, gamma=0.9):\n", " self.num_states = num_states\n", " self.alpha = alpha\n", " self.gamma = gamma\n", " self.V = np.zeros(num_states)\n", " self.episode_rewards = []\n", " self.value_history = []\n", " self.td_errors = []\n", " self.training_metrics = {\n", " 'episodes': [],\n", " 'total_rewards': [],\n", " 'avg_td_error': [],\n", " 'convergence_rate': []\n", " }\n", " \n", " def get_action(self, state):\n", " return random.choice([0, 1])\n", " \n", " def td_update(self, state, reward, next_state, done):\n", " if done:\n", " target = reward\n", " else:\n", " target = reward + self.gamma * self.V[next_state]\n", " \n", " td_error = target - self.V[state]\n", " self.V[state] += self.alpha * td_error\n", " \n", " self.td_errors.append(abs(td_error))\n", " \n", " logger.info(f\"TD Update - State: {state}, Reward: {reward}, Target: {target:.3f}, \"\n", " f\"TD Error: {td_error:.3f}, New V(s): {self.V[state]:.3f}\")\n", " \n", " return td_error\n", " \n", " def train_episode(self, env):\n", " state = env.reset()\n", " total_reward = 0\n", " episode_td_errors = []\n", " episode_states = []\n", " \n", " logger.info(f\"\\n=== Starting New Episode ===\")\n", " \n", " while True:\n", " action = self.get_action(state)\n", " next_state, reward, done = env.step(action)\n", " \n", " td_error = self.td_update(state, reward, next_state, done)\n", " episode_td_errors.append(abs(td_error))\n", " episode_states.append(state)\n", " \n", " total_reward += reward\n", " \n", " if done:\n", " break\n", " \n", " state = next_state\n", " \n", " self.episode_rewards.append(total_reward)\n", " self.value_history.append(self.V.copy())\n", " \n", " avg_td_error = np.mean(episode_td_errors) if episode_td_errors else 0\n", " logger.info(f\"Episode Complete - Total Reward: {total_reward}, \"\n", " f\"Avg TD Error: {avg_td_error:.3f}\")\n", " \n", " return total_reward, avg_td_error, episode_states\n", " \n", " def train(self, env, num_episodes=100):\n", " logger.info(f\"\\n๐Ÿš€ Starting TD Learning Training for {num_episodes} episodes\")\n", " logger.info(f\"Parameters - Alpha: {self.alpha}, Gamma: {self.gamma}\")\n", " \n", " for episode in range(num_episodes):\n", " total_reward, avg_td_error, states = self.train_episode(env)\n", " \n", " self.training_metrics['episodes'].append(episode)\n", " self.training_metrics['total_rewards'].append(total_reward)\n", " self.training_metrics['avg_td_error'].append(avg_td_error)\n", " \n", " convergence_rate = np.std(self.V) if len(self.value_history) > 1 else 0\n", " self.training_metrics['convergence_rate'].append(convergence_rate)\n", " \n", " if episode % 20 == 0:\n", " logger.info(f\"\\n๐Ÿ“Š Episode {episode} Summary:\")\n", " logger.info(f\"Value Function: {self.V}\")\n", " logger.info(f\"Recent Avg Reward: {np.mean(self.episode_rewards[-10:]):.2f}\")\n", " \n", " logger.info(f\"\\nโœ… Training Complete!\")\n", " logger.info(f\"Final Value Function: {self.V}\")\n", " \n", " def save_results(self, filename_prefix=\"td_learning\"):\n", " timestamp = datetime.now().strftime(\"%Y%m%d_%H%M%S\")\n", " \n", " results = {\n", " 'parameters': {\n", " 'alpha': self.alpha,\n", " 'gamma': self.gamma,\n", " 'num_states': self.num_states\n", " },\n", " 'final_values': self.V.tolist(),\n", " 'training_metrics': self.training_metrics,\n", " 'episode_rewards': self.episode_rewards\n", " }\n", " \n", " filename = f\"{filename_prefix}_{timestamp}.json\"\n", " with open(filename, 'w') as f:\n", " json.dump(results, f, indent=2)\n", " \n", " logger.info(f\"๐Ÿ’พ Results saved to {filename}\")\n", " return filename\n", " \n", " def visualize_training(self):\n", " fig, ((ax1, ax2), (ax3, ax4)) = plt.subplots(2, 2, figsize=(15, 10))\n", " \n", " ax1.plot(self.training_metrics['episodes'], self.training_metrics['total_rewards'])\n", " ax1.set_title('Episode Rewards Over Time')\n", " ax1.set_xlabel('Episode')\n", " ax1.set_ylabel('Total Reward')\n", " ax1.grid(True)\n", " \n", " if len(self.value_history) > 0:\n", " value_array = np.array(self.value_history)\n", " for state in range(self.num_states):\n", " ax2.plot(range(len(self.value_history)), value_array[:, state], \n", " label=f'State {state}', linewidth=2)\n", " ax2.set_title('Value Function Evolution')\n", " ax2.set_xlabel('Episode')\n", " ax2.set_ylabel('Value V(s)')\n", " ax2.legend()\n", " ax2.grid(True)\n", " \n", " ax3.plot(self.training_metrics['episodes'], self.training_metrics['avg_td_error'])\n", " ax3.set_title('Average TD Error Over Time')\n", " ax3.set_xlabel('Episode')\n", " ax3.set_ylabel('TD Error')\n", " ax3.grid(True)\n", " \n", " final_values = self.V\n", " bars = ax4.bar(range(len(final_values)), final_values, \n", " color=['lightblue' if i != len(final_values)-1 else 'gold' \n", " for i in range(len(final_values))])\n", " ax4.set_title('Final Value Function')\n", " ax4.set_xlabel('State')\n", " ax4.set_ylabel('Value V(s)')\n", " ax4.grid(True, alpha=0.3)\n", " \n", " for i, v in enumerate(final_values):\n", " ax4.text(i, v + 0.1, f'{v:.2f}', ha='center', va='bottom', fontweight='bold')\n", " \n", " plt.tight_layout()\n", " \n", " timestamp = datetime.now().strftime(\"%Y%m%d_%H%M%S\")\n", " plot_filename = f\"td_learning_plots_{timestamp}.png\"\n", " plt.savefig(plot_filename, dpi=300, bbox_inches='tight')\n", " logger.info(f\"๐Ÿ“ˆ Plots saved to {plot_filename}\")\n", " \n", " plt.show()\n", " return plot_filename\n", "\n", "def run_td_learning_experiment():\n", " logger.info(\"๐ŸŽฏ TD Learning Implementation - Complete with Logging\")\n", " \n", " env = TDLearningEnvironment(num_states=5)\n", " agent = TDLearningAgent(num_states=5, alpha=0.1, gamma=0.9)\n", " \n", " logger.info(f\"Environment: {env.num_states} states (0 to {env.num_states-1})\")\n", " logger.info(f\"Goal: Reach terminal state {env.terminal_state} for reward +10\")\n", " \n", " agent.train(env, num_episodes=100)\n", " \n", " results_file = agent.save_results()\n", " plot_file = agent.visualize_training()\n", " \n", " logger.info(f\"\\n๐ŸŽ‰ Experiment Complete!\")\n", " logger.info(f\"๐Ÿ“ Files saved: {results_file}, {plot_file}\")\n", " \n", " return agent, env, results_file, plot_file\n", "\n", "if __name__ == \"__main__\":\n", " agent, env, results_file, plot_file = run_td_learning_experiment()" ] }, { "cell_type": "code", "execution_count": null, "id": "6f90a7a5-2341-4df8-bc4f-a28c3f8296a7", "metadata": {}, "outputs": [], "source": [ "import numpy as np\n", "import matplotlib.pyplot as plt\n", "import random\n", "import logging\n", "from collections import defaultdict\n", "import json\n", "import os\n", "from datetime import datetime\n", "\n", "# Configure logging to show detailed information about the training process\n", "# This helps us track exactly what the algorithm is learning at each step\n", "logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s')\n", "logger = logging.getLogger(__name__)\n", "\n", "class TDLearningEnvironment:\n", " \"\"\"\n", " Simple environment for demonstrating TD Learning\n", " - States: 0, 1, 2, 3, 4 (where 4 is terminal)\n", " - Actions: 0 (stay/wrong), 1 (move forward/right)\n", " - Rewards: +10 for reaching terminal, -1 for wrong moves, 0 otherwise\n", " \"\"\"\n", " def __init__(self, num_states=5):\n", " self.num_states = num_states # Total number of states in our environment\n", " self.current_state = 0 # Agent always starts at state 0\n", " self.terminal_state = num_states - 1 # Last state (4) is the goal\n", " self.reset()\n", " \n", " def reset(self):\n", " \"\"\"Reset environment to starting state - called at beginning of each episode\"\"\"\n", " self.current_state = 0\n", " return self.current_state\n", " \n", " def step(self, action):\n", " \"\"\"\n", " Execute one step in the environment\n", " Args:\n", " action: 0 (stay/wrong move) or 1 (move forward)\n", " Returns:\n", " next_state, reward, done\n", " \"\"\"\n", " # If already at terminal state, no more moves possible\n", " if self.current_state == self.terminal_state:\n", " return self.current_state, 0, True\n", " \n", " reward = 0\n", " \n", " # Action 1 = move forward toward goal\n", " if action == 1:\n", " self.current_state += 1\n", " # Give big reward (+10) when reaching the terminal state\n", " if self.current_state == self.terminal_state:\n", " reward = 10\n", " else:\n", " # Action 0 = wrong move, give negative reward\n", " reward = -1\n", " \n", " # Episode is done when we reach the terminal state\n", " done = self.current_state == self.terminal_state\n", " return self.current_state, reward, done\n", "\n", "class TDLearningAgent:\n", " \"\"\"\n", " TD Learning Agent that learns state values V(s) using the TD(0) algorithm\n", " Key insight: Updates value estimates using bootstrapping - learning from partial experience\n", " \"\"\"\n", " def __init__(self, num_states, alpha=0.1, gamma=0.9):\n", " self.num_states = num_states # Number of states in the environment\n", " self.alpha = alpha # Learning rate: how fast we update our estimates\n", " self.gamma = gamma # Discount factor: how much we value future rewards\n", " \n", " # Initialize all state values to zero - these are our V(s) estimates\n", " self.V = np.zeros(num_states)\n", " \n", " # Track training progress for analysis and visualization\n", " self.episode_rewards = [] # Total reward per episode\n", " self.value_history = [] # V(s) values after each episode\n", " self.td_errors = [] # All TD errors during training\n", " \n", " # Metrics for tracking learning progress\n", " self.training_metrics = {\n", " 'episodes': [], # Episode numbers\n", " 'total_rewards': [], # Cumulative reward per episode\n", " 'avg_td_error': [], # Average TD error per episode\n", " 'convergence_rate': [] # How much V(s) values are changing\n", " }\n", " \n", " def get_action(self, state):\n", " \"\"\"\n", " Simple random policy for action selection\n", " In a real scenario, this could be epsilon-greedy or policy-based\n", " \"\"\"\n", " return random.choice([0, 1])\n", " \n", " def td_update(self, state, reward, next_state, done):\n", " \"\"\"\n", " The core TD(0) update rule: V(s) โ† V(s) + ฮฑ[r + ฮณV(s') - V(s)]\n", " This is the heart of temporal difference learning!\n", " \n", " Args:\n", " state: Current state s_t\n", " reward: Reward r_{t+1} received after taking action\n", " next_state: Next state s_{t+1}\n", " done: Whether episode is finished\n", " \n", " Returns:\n", " td_error: The temporal difference error ฮด\n", " \"\"\"\n", " # Calculate the TD target\n", " if done:\n", " # If episode is done, there's no next state value to consider\n", " target = reward\n", " else:\n", " # TD target = immediate reward + discounted future value\n", " # This is the \"bootstrapping\" - we use our current estimate of V(s')\n", " target = reward + self.gamma * self.V[next_state]\n", " \n", " # Calculate TD error: ฮด = target - current_estimate\n", " # This tells us how wrong our current value estimate was\n", " td_error = target - self.V[state]\n", " \n", " # Update the value function using the TD error\n", " # ฮฑ controls how much we trust this new experience vs our old estimate\n", " self.V[state] += self.alpha * td_error\n", " \n", " # Store TD error for analysis\n", " self.td_errors.append(abs(td_error))\n", " \n", " # Log the update for detailed tracking\n", " logger.info(f\"TD Update - State: {state}, Reward: {reward}, Target: {target:.3f}, \"\n", " f\"TD Error: {td_error:.3f}, New V(s): {self.V[state]:.3f}\")\n", " \n", " return td_error\n", " \n", " def train_episode(self, env):\n", " \"\"\"\n", " Run one complete episode of TD learning\n", " An episode goes from start state to terminal state (or max steps)\n", " \"\"\"\n", " state = env.reset() # Start at initial state\n", " total_reward = 0 # Track cumulative reward for this episode\n", " episode_td_errors = [] # Track TD errors for this episode\n", " episode_states = [] # Track which states we visited\n", " \n", " logger.info(f\"\\n=== Starting New Episode ===\")\n", " \n", " # Continue until episode is done\n", " while True:\n", " # Choose an action (random policy in this simple example)\n", " action = self.get_action(state)\n", " \n", " # Take action in environment and observe results\n", " next_state, reward, done = env.step(action)\n", " \n", " # *** THIS IS THE KEY STEP: TD UPDATE ***\n", " # Update our value function using the TD(0) rule\n", " td_error = self.td_update(state, reward, next_state, done)\n", " \n", " # Track statistics for analysis\n", " episode_td_errors.append(abs(td_error))\n", " episode_states.append(state)\n", " total_reward += reward\n", " \n", " # If episode is finished, break out of loop\n", " if done:\n", " break\n", " \n", " # Move to next state for next iteration\n", " state = next_state\n", " \n", " # Store episode results for tracking progress\n", " self.episode_rewards.append(total_reward)\n", " self.value_history.append(self.V.copy()) # Save snapshot of current V(s)\n", " \n", " # Calculate average TD error for this episode\n", " avg_td_error = np.mean(episode_td_errors) if episode_td_errors else 0\n", " \n", " logger.info(f\"Episode Complete - Total Reward: {total_reward}, \"\n", " f\"Avg TD Error: {avg_td_error:.3f}\")\n", " \n", " return total_reward, avg_td_error, episode_states\n", " \n", " def train(self, env, num_episodes=100):\n", " \"\"\"\n", " Train the TD learning agent for multiple episodes\n", " Each episode provides more experience to improve our value estimates\n", " \"\"\"\n", " logger.info(f\"\\n๐Ÿš€ Starting TD Learning Training for {num_episodes} episodes\")\n", " logger.info(f\"Parameters - Alpha: {self.alpha}, Gamma: {self.gamma}\")\n", " \n", " # Run the specified number of training episodes\n", " for episode in range(num_episodes):\n", " # Train for one episode and get results\n", " total_reward, avg_td_error, states = self.train_episode(env)\n", " \n", " # Store metrics for later analysis and plotting\n", " self.training_metrics['episodes'].append(episode)\n", " self.training_metrics['total_rewards'].append(total_reward)\n", " self.training_metrics['avg_td_error'].append(avg_td_error)\n", " \n", " # Calculate convergence rate (how much V(s) is still changing)\n", " convergence_rate = np.std(self.V) if len(self.value_history) > 1 else 0\n", " self.training_metrics['convergence_rate'].append(convergence_rate)\n", " \n", " # Print progress every 20 episodes\n", " if episode % 20 == 0:\n", " logger.info(f\"\\n๐Ÿ“Š Episode {episode} Summary:\")\n", " logger.info(f\"Value Function: {self.V}\")\n", " logger.info(f\"Recent Avg Reward: {np.mean(self.episode_rewards[-10:]):.2f}\")\n", " \n", " logger.info(f\"\\nโœ… Training Complete!\")\n", " logger.info(f\"Final Value Function: {self.V}\")\n", " \n", " def save_results(self, filename_prefix=\"td_learning\"):\n", " \"\"\"\n", " Save all training results to JSON file for later analysis\n", " This includes parameters, final values, and all training metrics\n", " \"\"\"\n", " timestamp = datetime.now().strftime(\"%Y%m%d_%H%M%S\")\n", " \n", " # Package all results into a dictionary\n", " results = {\n", " 'parameters': {\n", " 'alpha': self.alpha,\n", " 'gamma': self.gamma,\n", " 'num_states': self.num_states\n", " },\n", " 'final_values': self.V.tolist(), # Final learned V(s) values\n", " 'training_metrics': self.training_metrics, # All episode data\n", " 'episode_rewards': self.episode_rewards # Reward progression\n", " }\n", " \n", " # Save to timestamped JSON file\n", " filename = f\"{filename_prefix}_{timestamp}.json\"\n", " with open(filename, 'w') as f:\n", " json.dump(results, f, indent=2)\n", " \n", " logger.info(f\"๐Ÿ’พ Results saved to {filename}\")\n", " return filename\n", " \n", " def visualize_training(self):\n", " \"\"\"\n", " Create comprehensive visualizations of the TD learning process\n", " Shows how the algorithm learned over time\n", " \"\"\"\n", " # Create 2x2 subplot layout for multiple visualizations\n", " fig, ((ax1, ax2), (ax3, ax4)) = plt.subplots(2, 2, figsize=(15, 10))\n", " \n", " # Plot 1: Episode rewards over time - shows learning progress\n", " ax1.plot(self.training_metrics['episodes'], self.training_metrics['total_rewards'])\n", " ax1.set_title('Episode Rewards Over Time')\n", " ax1.set_xlabel('Episode')\n", " ax1.set_ylabel('Total Reward')\n", " ax1.grid(True)\n", " \n", " # Plot 2: Value function evolution - shows how V(s) changed during learning\n", " if len(self.value_history) > 0:\n", " value_array = np.array(self.value_history)\n", " # Plot each state's value over time\n", " for state in range(self.num_states):\n", " ax2.plot(range(len(self.value_history)), value_array[:, state], \n", " label=f'State {state}', linewidth=2)\n", " ax2.set_title('Value Function Evolution')\n", " ax2.set_xlabel('Episode')\n", " ax2.set_ylabel('Value V(s)')\n", " ax2.legend()\n", " ax2.grid(True)\n", " \n", " # Plot 3: TD error over time - shows convergence\n", " ax3.plot(self.training_metrics['episodes'], self.training_metrics['avg_td_error'])\n", " ax3.set_title('Average TD Error Over Time')\n", " ax3.set_xlabel('Episode')\n", " ax3.set_ylabel('TD Error')\n", " ax3.grid(True)\n", " \n", " # Plot 4: Final value function - shows what the agent learned\n", " final_values = self.V\n", " bars = ax4.bar(range(len(final_values)), final_values, \n", " color=['lightblue' if i != len(final_values)-1 else 'gold' \n", " for i in range(len(final_values))])\n", " ax4.set_title('Final Value Function')\n", " ax4.set_xlabel('State')\n", " ax4.set_ylabel('Value V(s)')\n", " ax4.grid(True, alpha=0.3)\n", " \n", " # Add value labels on top of bars\n", " for i, v in enumerate(final_values):\n", " ax4.text(i, v + 0.1, f'{v:.2f}', ha='center', va='bottom', fontweight='bold')\n", " \n", " plt.tight_layout()\n", " \n", " # Save plot with timestamp\n", " timestamp = datetime.now().strftime(\"%Y%m%d_%H%M%S\")\n", " plot_filename = f\"td_learning_plots_{timestamp}.png\"\n", " plt.savefig(plot_filename, dpi=300, bbox_inches='tight')\n", " logger.info(f\"๐Ÿ“ˆ Plots saved to {plot_filename}\")\n", " \n", " plt.show()\n", " return plot_filename\n", "\n", "def run_td_learning_experiment():\n", " \"\"\"\n", " Main function to run the complete TD Learning experiment\n", " This ties together the environment and agent for a full demonstration\n", " \"\"\"\n", " logger.info(\"๐ŸŽฏ TD Learning Implementation - Complete with Logging\")\n", " \n", " # Create environment and agent\n", " env = TDLearningEnvironment(num_states=5) # 5-state environment\n", " agent = TDLearningAgent(num_states=5, alpha=0.1, gamma=0.9) # TD learning agent\n", " \n", " logger.info(f\"Environment: {env.num_states} states (0 to {env.num_states-1})\")\n", " logger.info(f\"Goal: Reach terminal state {env.terminal_state} for reward +10\")\n", " \n", " # Run the training process\n", " agent.train(env, num_episodes=100)\n", " \n", " # Save results and create visualizations\n", " results_file = agent.save_results()\n", " plot_file = agent.visualize_training()\n", " \n", " logger.info(f\"\\n๐ŸŽ‰ Experiment Complete!\")\n", " logger.info(f\"๐Ÿ“ Files saved: {results_file}, {plot_file}\")\n", " \n", " return agent, env, results_file, plot_file\n", "\n", "# Run the experiment when script is executed\n", "if __name__ == \"__main__\":\n", " agent, env, results_file, plot_file = run_td_learning_experiment()" ] }, { "cell_type": "markdown", "id": "c8566b53-0ebd-46de-b501-9d2a4ce87ceb", "metadata": {}, "source": [ "# PHASE 4: TRAINING FLOW\n", "\n", "Let me walk you through **exactly how TD Learning learns step-by-step** using simple analogies and the actual numbers from your run!\n", "\n", "--------------------------------------------\n", "## ๐ŸŽฏ **The Big Picture: Learning Restaurant Quality**\n", "--------------------------------------------\n", "\n", "Think of TD Learning like **rating restaurants on a street**:\n", "- **State 0**: Far from the best restaurant (your starting point)\n", "- **State 1-3**: Getting closer to the amazing restaurant\n", "- **State 4**: The amazing restaurant (+10 reward)\n", "\n", "The agent learns: *\"How good is each location, assuming I'll walk optimally from there?\"*\n", "\n", "--------------------------------------------\n", "## ๐Ÿ”„ **Step-by-Step Training Flow**\n", "--------------------------------------------\n", "\n", "### **STEP 1: Initialize Everything to Zero**\n", "```\n", "V(0) = 0.0 V(1) = 0.0 V(2) = 0.0 V(3) = 0.0 V(4) = 0.0\n", "```\n", "**Analogy**: *\"I have no idea how good any restaurant location is yet\"*\n", "\n", "---\n", "\n", "### **STEP 2: First Experience (Episode 1)**\n", "\n", "**๐Ÿšถ Agent's Journey:**\n", "```\n", "State 0 โ†’ (wrong action) โ†’ State 0, reward = -1\n", "State 0 โ†’ (right action) โ†’ State 1, reward = 0 \n", "State 1 โ†’ (right action) โ†’ State 2, reward = 0\n", "State 2 โ†’ (wrong action) โ†’ State 2, reward = -1\n", "State 2 โ†’ (right action) โ†’ State 3, reward = 0\n", "State 3 โ†’ (right action) โ†’ State 4, reward = +10 ๐ŸŽ‰\n", "```\n", "\n", "**๐Ÿง  TD Updates (The Learning Magic):**\n", "\n", "**Update 1**: State 0 gets wrong action\n", "```\n", "TD Target = reward + ฮณ ร— V(next_state) = -1 + 0.9 ร— 0.0 = -1.0\n", "TD Error = target - current = -1.0 - 0.0 = -1.0\n", "V(0) = 0.0 + 0.1 ร— (-1.0) = -0.1\n", "```\n", "*\"Being at State 0 and making wrong moves is bad!\"*\n", "\n", "**Update 2**: State 0 โ†’ State 1 (right move)\n", "```\n", "TD Target = 0 + 0.9 ร— 0.0 = 0.0\n", "TD Error = 0.0 - (-0.1) = 0.1\n", "V(0) = -0.1 + 0.1 ร— 0.1 = -0.09\n", "```\n", "*\"State 0 is slightly better when I move right\"*\n", "\n", "**Update 6**: State 3 โ†’ State 4 (BIG REWARD!)\n", "```\n", "TD Target = 10 + 0.9 ร— 0.0 = 10.0\n", "TD Error = 10.0 - 0.0 = 10.0\n", "V(3) = 0.0 + 0.1 ร— 10.0 = 1.0\n", "```\n", "*\"WOW! State 3 is really valuable - it leads to the prize!\"*\n", "\n", "**After Episode 1:**\n", "```\n", "V(0) = -0.09 V(1) = 0.0 V(2) = -0.09 V(3) = 1.0 V(4) = 0.0\n", "```\n", "\n", "---\n", "\n", "### **STEP 3: Information Spreads Backwards (Episode 2)**\n", "\n", "**๐Ÿ”„ The Magic of Bootstrapping:**\n", "\n", "When agent reaches State 2 โ†’ State 3:\n", "```\n", "TD Target = 0 + 0.9 ร— V(3) = 0 + 0.9 ร— 1.0 = 0.9\n", "TD Error = 0.9 - (-0.09) = 0.99\n", "V(2) = -0.09 + 0.1 ร— 0.99 = 0.009\n", "```\n", "*\"State 2 is valuable because it can reach the valuable State 3!\"*\n", "\n", "**Analogy**: Like hearing *\"The restaurant at location 3 is amazing!\"* so you start valuing location 2 because it's close to location 3.\n", "\n", "---\n", "\n", "### **STEP 4: Values Propagate Through the Chain**\n", "\n", "**Episodes 3-10: Watch the pattern!**\n", "```\n", "Episode 1: V = [-0.09, 0.00, -0.09, 1.00, 0.00]\n", "Episode 2: V = [-0.08, -0.10, 0.01, 1.90, 0.00]\n", "Episode 5: V = [-0.55, -0.41, 1.67, 4.97, 0.00]\n", "Episode 10: V = [-0.86, -0.27, 2.00, 5.57, 0.00]\n", "```\n", "\n", "**What's Happening:**\n", "- **State 3** learns fastest (closest to reward)\n", "- **State 2** learns it's good because it reaches State 3\n", "- **State 1** learns it's good because it reaches State 2\n", "- **State 0** slowly learns it's the starting point\n", "\n", "---\n", "\n", "### **STEP 5: Convergence (Episodes 50-100)**\n", "\n", "**Final Values from Your Run:**\n", "```\n", "Final: V = [2.42, 4.85, 6.91, 8.67, 0.00]\n", "```\n", "\n", "**Perfect Learning! ๐ŸŽฏ**\n", "- **Higher states = Higher values** (closer to reward)\n", "- **State 3**: 8.67 (almost as good as the +10 reward)\n", "- **State 2**: 6.91 (good because leads to State 3)\n", "- **State 1**: 4.85 (decent because 2 steps from reward)\n", "- **State 0**: 2.42 (starting point, but can reach goal)\n", "\n", "---\n", "\n", "### **STEP 6: How TD Error Shows Learning**\n", "\n", "**Early Episodes**: Large TD errors (2.0+)\n", "```\n", "\"I thought State 2 was worth 0, but I just learned it leads to State 3 worth 5!\"\n", "```\n", "\n", "**Later Episodes**: Small TD errors (1.4-)\n", "```\n", "\"I thought State 2 was worth 6.8, and I just learned it's worth 6.9 - close!\"\n", "```\n", "\n", "**Analogy**: Like a GPS that starts with wrong estimates but gets more accurate with each trip!\n", "\n", "--------------------------------------------\n", "## ๐Ÿ”ฅ **The \"Aha!\" Moments During Training**\n", "--------------------------------------------\n", "\n", "### **Moment 1: First Reward Discovery**\n", "```\n", "Episode 1: \"Holy cow! State 3 โ†’ +10 reward! State 3 is VALUABLE!\"\n", "```\n", "\n", "### **Moment 2: Backward Propagation**\n", "```\n", "Episode 2-5: \"Wait, State 2 is valuable too because it leads to State 3!\"\n", "```\n", "\n", "### **Moment 3: Chain Reaction**\n", "```\n", "Episode 10-20: \"State 1 is good because it leads to State 2, which leads to State 3!\"\n", "```\n", "\n", "### **Moment 4: Convergence**\n", "```\n", "Episode 80+: \"I've got it! Each state's value reflects how close it is to the goal!\"\n", "```\n", "\n", "--------------------------------------------\n", "## ๐Ÿ“Š **Numbers Changing Through Iterations**\n", "--------------------------------------------\n", "\n", "**Watch V(3) learn (closest to reward):**\n", "```\n", "Episode 1: V(3) = 1.00 (first discovery)\n", "Episode 5: V(3) = 4.97 (learning it's really good)\n", "Episode 20: V(3) = 7.77 (almost perfect)\n", "Episode 100: V(3) = 8.67 (converged!)\n", "```\n", "\n", "**Watch V(0) learn (furthest from reward):**\n", "```\n", "Episode 1: V(0) = -0.09 (seems bad at first)\n", "Episode 20: V(0) = 1.57 (realizing it can reach goal)\n", "Episode 50: V(0) = 3.15 (getting more optimistic)\n", "Episode 100: V(0) = 2.42 (settled at realistic value)\n", "```\n", "\n", "--------------------------------------------\n", "## ๐ŸŽจ **Visual Walkthrough**\n", "--------------------------------------------\n", "\n", "```\n", "Episode 1: [0] โ†’ [0] โ†’ [1] โ†’ [2] โ†’ [3] โ†’ [4] ๐Ÿ’ฐ\n", "Values: -0.09 0.0 -0.09 1.0 0.0\n", "\n", "Episode 50: [0] โ†’ [1] โ†’ [2] โ†’ [3] โ†’ [4] ๐Ÿ’ฐ \n", "Values: 3.15 4.27 6.11 8.88 0.0\n", "\n", "Episode 100: [0] โ†’ [1] โ†’ [2] โ†’ [3] โ†’ [4] ๐Ÿ’ฐ\n", "Values: 2.42 4.85 6.91 8.67 0.0\n", " โ†‘ โ†‘ โ†‘ โ†‘ โ†‘\n", " Start Better Good Great Goal!\n", "```\n", "\n", "--------------------------------------------\n", "## ๐Ÿงช **Why This Works (The Magic Explained)**\n", "--------------------------------------------\n", "\n", "**Traditional Learning**: *\"Wait until I finish the whole journey, then update everything\"*\n", "\n", "**TD Learning**: *\"Update my beliefs immediately based on what I just experienced + what I currently believe about the future\"*\n", "\n", "**The Bootstrap Formula:**\n", "```\n", "New Belief = Old Belief + Learning_Rate ร— (Reality - Old Belief)\n", " = Old Belief + ฮฑ ร— (reward + ฮณ ร— future_estimate - Old Belief)\n", "```\n", "\n", "**Why It's Powerful:**\n", "1. **Learns online** - no waiting for episode to end\n", "2. **Uses current knowledge** - bootstraps from existing estimates \n", "3. **Balances old vs new** - learning rate controls the blend\n", "4. **Propagates value backwards** - good states make previous states good\n", "\n", "**The Result**: The agent learns *\"How good is it to be in each state?\"* which is exactly what we want for decision making! ๐ŸŽฏ\n" ] }, { "cell_type": "code", "execution_count": null, "id": "f6e9f145-5a14-45d6-8a33-0ec83d4b7cce", "metadata": {}, "outputs": [], "source": [] } ], "metadata": { "kernelspec": { "display_name": "Python 3 (ipykernel)", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.11.13" } }, "nbformat": 4, "nbformat_minor": 5 }