diff --git a/notebooks/gymv2_jax.ipynb b/notebooks/gymv2_jax.ipynb new file mode 100644 index 0000000..2dddc79 --- /dev/null +++ b/notebooks/gymv2_jax.ipynb @@ -0,0 +1,110 @@ +{ + "cells": [ + { + "cell_type": "code", + "execution_count": 1, + "metadata": {}, + "outputs": [], + "source": [ + "\n", + "from gymnasium.utils.env_checker import check_env as gym_check_env\n", + "from stable_baselines3 import TD3\n", + "from stable_baselines3.common.env_checker import check_env\n", + "from gymnasium.wrappers.jax_to_numpy import JaxToNumpy\n", + "from gymnasium.wrappers.vector import JaxToNumpy as VJaxToNumpy\n", + "from gymnax.wrappers.gym import GymnaxToVectorGymWrapper, GymnaxToGymWrapper\n", + "import matplotlib.pyplot as plt\n", + "import jax.numpy as jnp\n", + "from solarcarsim.simv2 import Snax\n", + "\n" + ] + }, + { + "cell_type": "code", + "execution_count": 5, + "metadata": {}, + "outputs": [], + "source": [ + "\n", + "env = Snax()\n", + "wrapped_env = GymnaxToGymWrapper(env)\n", + "vector_gym_env = GymnaxToVectorGymWrapper(env)\n", + "np_wrapper = JaxToNumpy(wrapped_env)\n", + "np_vec_wrapper = VJaxToNumpy(vector_gym_env)" + ] + }, + { + "cell_type": "code", + "execution_count": 7, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Using cuda device\n", + "Wrapping the env with a `Monitor` wrapper\n", + "Wrapping the env in a DummyVecEnv.\n" + ] + }, + { + "ename": "ValueError", + "evalue": "Non-hashable static arguments are not supported. An error occurred while trying to hash an object of type , Tracedwith. The error was:\nTypeError: unhashable type: 'DynamicJaxprTracer'\n", + "output_type": "error", + "traceback": [ + "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m", + "\u001b[0;31mValueError\u001b[0m Traceback (most recent call last)", + "Cell \u001b[0;32mIn[7], line 2\u001b[0m\n\u001b[1;32m 1\u001b[0m model \u001b[38;5;241m=\u001b[39m TD3(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mMlpPolicy\u001b[39m\u001b[38;5;124m\"\u001b[39m, np_wrapper, verbose\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m1\u001b[39m)\n\u001b[0;32m----> 2\u001b[0m \u001b[43mmodel\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mlearn\u001b[49m\u001b[43m(\u001b[49m\u001b[43mtotal_timesteps\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;241;43m1000\u001b[39;49m\u001b[43m)\u001b[49m\n", + "File \u001b[0;32m~/Documents/Code/solarcarsim/.venv/lib/python3.12/site-packages/stable_baselines3/td3/td3.py:222\u001b[0m, in \u001b[0;36mTD3.learn\u001b[0;34m(self, total_timesteps, callback, log_interval, tb_log_name, reset_num_timesteps, progress_bar)\u001b[0m\n\u001b[1;32m 213\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mlearn\u001b[39m(\n\u001b[1;32m 214\u001b[0m \u001b[38;5;28mself\u001b[39m: SelfTD3,\n\u001b[1;32m 215\u001b[0m total_timesteps: \u001b[38;5;28mint\u001b[39m,\n\u001b[0;32m (...)\u001b[0m\n\u001b[1;32m 220\u001b[0m progress_bar: \u001b[38;5;28mbool\u001b[39m \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;01mFalse\u001b[39;00m,\n\u001b[1;32m 221\u001b[0m ) \u001b[38;5;241m-\u001b[39m\u001b[38;5;241m>\u001b[39m SelfTD3:\n\u001b[0;32m--> 222\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43msuper\u001b[39;49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mlearn\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m 223\u001b[0m \u001b[43m \u001b[49m\u001b[43mtotal_timesteps\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mtotal_timesteps\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 224\u001b[0m \u001b[43m \u001b[49m\u001b[43mcallback\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mcallback\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 225\u001b[0m \u001b[43m \u001b[49m\u001b[43mlog_interval\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mlog_interval\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 226\u001b[0m \u001b[43m \u001b[49m\u001b[43mtb_log_name\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mtb_log_name\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 227\u001b[0m \u001b[43m \u001b[49m\u001b[43mreset_num_timesteps\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mreset_num_timesteps\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 228\u001b[0m \u001b[43m \u001b[49m\u001b[43mprogress_bar\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mprogress_bar\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 229\u001b[0m \u001b[43m \u001b[49m\u001b[43m)\u001b[49m\n", + "File \u001b[0;32m~/Documents/Code/solarcarsim/.venv/lib/python3.12/site-packages/stable_baselines3/common/off_policy_algorithm.py:328\u001b[0m, in \u001b[0;36mOffPolicyAlgorithm.learn\u001b[0;34m(self, total_timesteps, callback, log_interval, tb_log_name, reset_num_timesteps, progress_bar)\u001b[0m\n\u001b[1;32m 325\u001b[0m \u001b[38;5;28;01massert\u001b[39;00m \u001b[38;5;28misinstance\u001b[39m(\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mtrain_freq, TrainFreq) \u001b[38;5;66;03m# check done in _setup_learn()\u001b[39;00m\n\u001b[1;32m 327\u001b[0m \u001b[38;5;28;01mwhile\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mnum_timesteps \u001b[38;5;241m<\u001b[39m total_timesteps:\n\u001b[0;32m--> 328\u001b[0m rollout \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mcollect_rollouts\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m 329\u001b[0m \u001b[43m \u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43menv\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 330\u001b[0m \u001b[43m \u001b[49m\u001b[43mtrain_freq\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mtrain_freq\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 331\u001b[0m \u001b[43m \u001b[49m\u001b[43maction_noise\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43maction_noise\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 332\u001b[0m \u001b[43m \u001b[49m\u001b[43mcallback\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mcallback\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 333\u001b[0m \u001b[43m \u001b[49m\u001b[43mlearning_starts\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mlearning_starts\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 334\u001b[0m \u001b[43m \u001b[49m\u001b[43mreplay_buffer\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mreplay_buffer\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 335\u001b[0m \u001b[43m \u001b[49m\u001b[43mlog_interval\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mlog_interval\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 336\u001b[0m \u001b[43m \u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 338\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m rollout\u001b[38;5;241m.\u001b[39mcontinue_training:\n\u001b[1;32m 339\u001b[0m \u001b[38;5;28;01mbreak\u001b[39;00m\n", + "File \u001b[0;32m~/Documents/Code/solarcarsim/.venv/lib/python3.12/site-packages/stable_baselines3/common/off_policy_algorithm.py:560\u001b[0m, in \u001b[0;36mOffPolicyAlgorithm.collect_rollouts\u001b[0;34m(self, env, callback, train_freq, replay_buffer, action_noise, learning_starts, log_interval)\u001b[0m\n\u001b[1;32m 557\u001b[0m actions, buffer_actions \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_sample_action(learning_starts, action_noise, env\u001b[38;5;241m.\u001b[39mnum_envs)\n\u001b[1;32m 559\u001b[0m \u001b[38;5;66;03m# Rescale and perform action\u001b[39;00m\n\u001b[0;32m--> 560\u001b[0m new_obs, rewards, dones, infos \u001b[38;5;241m=\u001b[39m \u001b[43menv\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mstep\u001b[49m\u001b[43m(\u001b[49m\u001b[43mactions\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 562\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mnum_timesteps \u001b[38;5;241m+\u001b[39m\u001b[38;5;241m=\u001b[39m env\u001b[38;5;241m.\u001b[39mnum_envs\n\u001b[1;32m 563\u001b[0m num_collected_steps \u001b[38;5;241m+\u001b[39m\u001b[38;5;241m=\u001b[39m \u001b[38;5;241m1\u001b[39m\n", + "File \u001b[0;32m~/Documents/Code/solarcarsim/.venv/lib/python3.12/site-packages/stable_baselines3/common/vec_env/base_vec_env.py:206\u001b[0m, in \u001b[0;36mVecEnv.step\u001b[0;34m(self, actions)\u001b[0m\n\u001b[1;32m 199\u001b[0m \u001b[38;5;250m\u001b[39m\u001b[38;5;124;03m\"\"\"\u001b[39;00m\n\u001b[1;32m 200\u001b[0m \u001b[38;5;124;03mStep the environments with the given action\u001b[39;00m\n\u001b[1;32m 201\u001b[0m \n\u001b[1;32m 202\u001b[0m \u001b[38;5;124;03m:param actions: the action\u001b[39;00m\n\u001b[1;32m 203\u001b[0m \u001b[38;5;124;03m:return: observation, reward, done, information\u001b[39;00m\n\u001b[1;32m 204\u001b[0m \u001b[38;5;124;03m\"\"\"\u001b[39;00m\n\u001b[1;32m 205\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mstep_async(actions)\n\u001b[0;32m--> 206\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mstep_wait\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\n", + "File \u001b[0;32m~/Documents/Code/solarcarsim/.venv/lib/python3.12/site-packages/stable_baselines3/common/vec_env/dummy_vec_env.py:58\u001b[0m, in \u001b[0;36mDummyVecEnv.step_wait\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m 55\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mstep_wait\u001b[39m(\u001b[38;5;28mself\u001b[39m) \u001b[38;5;241m-\u001b[39m\u001b[38;5;241m>\u001b[39m VecEnvStepReturn:\n\u001b[1;32m 56\u001b[0m \u001b[38;5;66;03m# Avoid circular imports\u001b[39;00m\n\u001b[1;32m 57\u001b[0m \u001b[38;5;28;01mfor\u001b[39;00m env_idx \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28mrange\u001b[39m(\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mnum_envs):\n\u001b[0;32m---> 58\u001b[0m obs, \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mbuf_rews[env_idx], terminated, truncated, \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mbuf_infos[env_idx] \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43menvs\u001b[49m\u001b[43m[\u001b[49m\u001b[43menv_idx\u001b[49m\u001b[43m]\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mstep\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m 59\u001b[0m \u001b[43m \u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mactions\u001b[49m\u001b[43m[\u001b[49m\u001b[43menv_idx\u001b[49m\u001b[43m]\u001b[49m\n\u001b[1;32m 60\u001b[0m \u001b[43m \u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 61\u001b[0m \u001b[38;5;66;03m# convert to SB3 VecEnv api\u001b[39;00m\n\u001b[1;32m 62\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mbuf_dones[env_idx] \u001b[38;5;241m=\u001b[39m terminated \u001b[38;5;129;01mor\u001b[39;00m truncated\n", + "File \u001b[0;32m~/Documents/Code/solarcarsim/.venv/lib/python3.12/site-packages/stable_baselines3/common/monitor.py:94\u001b[0m, in \u001b[0;36mMonitor.step\u001b[0;34m(self, action)\u001b[0m\n\u001b[1;32m 92\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mneeds_reset:\n\u001b[1;32m 93\u001b[0m \u001b[38;5;28;01mraise\u001b[39;00m \u001b[38;5;167;01mRuntimeError\u001b[39;00m(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mTried to step environment that needs reset\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n\u001b[0;32m---> 94\u001b[0m observation, reward, terminated, truncated, info \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43menv\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mstep\u001b[49m\u001b[43m(\u001b[49m\u001b[43maction\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 95\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mrewards\u001b[38;5;241m.\u001b[39mappend(\u001b[38;5;28mfloat\u001b[39m(reward))\n\u001b[1;32m 96\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m terminated \u001b[38;5;129;01mor\u001b[39;00m truncated:\n", + "File \u001b[0;32m~/Documents/Code/solarcarsim/.venv/lib/python3.12/site-packages/gymnasium/wrappers/jax_to_numpy.py:166\u001b[0m, in \u001b[0;36mJaxToNumpy.step\u001b[0;34m(self, action)\u001b[0m\n\u001b[1;32m 157\u001b[0m \u001b[38;5;250m\u001b[39m\u001b[38;5;124;03m\"\"\"Transforms the action to a jax array .\u001b[39;00m\n\u001b[1;32m 158\u001b[0m \n\u001b[1;32m 159\u001b[0m \u001b[38;5;124;03mArgs:\u001b[39;00m\n\u001b[0;32m (...)\u001b[0m\n\u001b[1;32m 163\u001b[0m \u001b[38;5;124;03m A tuple containing numpy versions of the next observation, reward, termination, truncation, and extra info.\u001b[39;00m\n\u001b[1;32m 164\u001b[0m \u001b[38;5;124;03m\"\"\"\u001b[39;00m\n\u001b[1;32m 165\u001b[0m jax_action \u001b[38;5;241m=\u001b[39m numpy_to_jax(action)\n\u001b[0;32m--> 166\u001b[0m obs, reward, terminated, truncated, info \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43menv\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mstep\u001b[49m\u001b[43m(\u001b[49m\u001b[43mjax_action\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 168\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m (\n\u001b[1;32m 169\u001b[0m jax_to_numpy(obs),\n\u001b[1;32m 170\u001b[0m \u001b[38;5;28mfloat\u001b[39m(reward),\n\u001b[0;32m (...)\u001b[0m\n\u001b[1;32m 173\u001b[0m jax_to_numpy(info),\n\u001b[1;32m 174\u001b[0m )\n", + "File \u001b[0;32m~/Documents/Code/solarcarsim/.venv/lib/python3.12/site-packages/gymnax/wrappers/gym.py:70\u001b[0m, in \u001b[0;36mGymnaxToGymWrapper.step\u001b[0;34m(self, action)\u001b[0m\n\u001b[1;32m 68\u001b[0m \u001b[38;5;250m\u001b[39m\u001b[38;5;124;03m\"\"\"Step environment, follow new step API.\"\"\"\u001b[39;00m\n\u001b[1;32m 69\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mrng, step_key \u001b[38;5;241m=\u001b[39m jax\u001b[38;5;241m.\u001b[39mrandom\u001b[38;5;241m.\u001b[39msplit(\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mrng)\n\u001b[0;32m---> 70\u001b[0m o, \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39menv_state, r, d, info \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_env\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mstep\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m 71\u001b[0m \u001b[43m \u001b[49m\u001b[43mstep_key\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43menv_state\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43maction\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43menv_params\u001b[49m\n\u001b[1;32m 72\u001b[0m \u001b[43m\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 73\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m o, r, d, d, info\n", + " \u001b[0;31m[... skipping hidden 15 frame]\u001b[0m\n", + "File \u001b[0;32m~/Documents/Code/solarcarsim/.venv/lib/python3.12/site-packages/gymnax/environments/environment.py:45\u001b[0m, in \u001b[0;36mEnvironment.step\u001b[0;34m(self, key, state, action, params)\u001b[0m\n\u001b[1;32m 43\u001b[0m params \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mdefault_params\n\u001b[1;32m 44\u001b[0m key, key_reset \u001b[38;5;241m=\u001b[39m jax\u001b[38;5;241m.\u001b[39mrandom\u001b[38;5;241m.\u001b[39msplit(key)\n\u001b[0;32m---> 45\u001b[0m obs_st, state_st, reward, done, info \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mstep_env\u001b[49m\u001b[43m(\u001b[49m\u001b[43mkey\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mstate\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43maction\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mparams\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 46\u001b[0m obs_re, state_re \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mreset_env(key_reset, params)\n\u001b[1;32m 47\u001b[0m \u001b[38;5;66;03m# Auto-reset environment based on termination\u001b[39;00m\n", + "File \u001b[0;32m~/Documents/Code/solarcarsim/src/solarcarsim/simv2.py:138\u001b[0m, in \u001b[0;36mSnax.step_env\u001b[0;34m(self, key, state, action, params)\u001b[0m\n\u001b[1;32m 136\u001b[0m theta \u001b[38;5;241m=\u001b[39m state\u001b[38;5;241m.\u001b[39mslope[pos]\n\u001b[1;32m 137\u001b[0m velocity \u001b[38;5;241m=\u001b[39m jnp\u001b[38;5;241m.\u001b[39marray([action \u001b[38;5;241m*\u001b[39m params\u001b[38;5;241m.\u001b[39mcar\u001b[38;5;241m.\u001b[39mmax_speed])\u001b[38;5;241m.\u001b[39msqueeze()\n\u001b[0;32m--> 138\u001b[0m dragf \u001b[38;5;241m=\u001b[39m \u001b[43msim\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mdrag_force\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m 139\u001b[0m \u001b[43m \u001b[49m\u001b[43mvelocity\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mparams\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mcar\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mfrontal_area\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mparams\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mcar\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mdrag_coeff\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m1.184\u001b[39;49m\n\u001b[1;32m 140\u001b[0m \u001b[43m\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 141\u001b[0m rollf \u001b[38;5;241m=\u001b[39m sim\u001b[38;5;241m.\u001b[39mrolling_force(params\u001b[38;5;241m.\u001b[39mcar\u001b[38;5;241m.\u001b[39mmass, theta, params\u001b[38;5;241m.\u001b[39mcar\u001b[38;5;241m.\u001b[39mrolling_coeff)\n\u001b[1;32m 142\u001b[0m hillf \u001b[38;5;241m=\u001b[39m sim\u001b[38;5;241m.\u001b[39mdownslope_force(params\u001b[38;5;241m.\u001b[39mcar\u001b[38;5;241m.\u001b[39mmass, theta)\n", + " \u001b[0;31m[... skipping hidden 3 frame]\u001b[0m\n", + "File \u001b[0;32m~/Documents/Code/solarcarsim/.venv/lib/python3.12/site-packages/jax/_src/pjit.py:768\u001b[0m, in \u001b[0;36m_infer_params\u001b[0;34m(fun, ji, args, kwargs)\u001b[0m\n\u001b[1;32m 764\u001b[0m p, args_flat \u001b[38;5;241m=\u001b[39m _infer_params_impl(fun, ji, pjit_mesh, resource_env, args,\n\u001b[1;32m 765\u001b[0m kwargs, in_avals\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mNone\u001b[39;00m)\n\u001b[1;32m 766\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m p, p\u001b[38;5;241m.\u001b[39mconsts \u001b[38;5;241m+\u001b[39m args_flat\n\u001b[0;32m--> 768\u001b[0m entry \u001b[38;5;241m=\u001b[39m \u001b[43m_infer_params_cached\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m 769\u001b[0m \u001b[43m \u001b[49m\u001b[43mfun\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mji\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43msignature\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mavals\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mpjit_mesh\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mresource_env\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 770\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m entry\u001b[38;5;241m.\u001b[39mpjit_params \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n\u001b[1;32m 771\u001b[0m p, args_flat \u001b[38;5;241m=\u001b[39m _infer_params_impl(\n\u001b[1;32m 772\u001b[0m fun, ji, pjit_mesh, resource_env, args, kwargs, in_avals\u001b[38;5;241m=\u001b[39mavals)\n", + "\u001b[0;31mValueError\u001b[0m: Non-hashable static arguments are not supported. An error occurred while trying to hash an object of type , Tracedwith. The error was:\nTypeError: unhashable type: 'DynamicJaxprTracer'\n" + ] + } + ], + "source": [ + "\n", + "model = TD3(\"MlpPolicy\", np_wrapper, verbose=1)\n", + "model.learn(total_timesteps=1000)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [] + } + ], + "metadata": { + "kernelspec": { + "display_name": ".venv", + "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": "3.12.7" + } + }, + "nbformat": 4, + "nbformat_minor": 2 +} diff --git a/notebooks/testing.ipynb b/notebooks/testing.ipynb index ef441e7..7d6a093 100644 --- a/notebooks/testing.ipynb +++ b/notebooks/testing.ipynb @@ -2,7 +2,7 @@ "cells": [ { "cell_type": "code", - "execution_count": 2, + "execution_count": 1, "metadata": {}, "outputs": [ { @@ -12,7 +12,7 @@ " -1.1142221e-02, -1.1067827e-02, -1.1001030e-02], dtype=float32)" ] }, - "execution_count": 2, + "execution_count": 1, "metadata": {}, "output_type": "execute_result" } @@ -21,6 +21,8 @@ "import jax\n", "import jax.numpy as jnp\n", "from solarcarsim.physsim import CarParams, fractal_noise_1d\n", + "from solarcarsim.simv2 import Snax, SimParams\n", + "import chex\n", "\n", "\n", "key = jax.random.key(0)\n", @@ -32,13 +34,2877 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 2, "metadata": {}, "outputs": [], "source": [ - "# get an array of positions\n", - "positions = jnp.array([1.1,2.2,3.3,5,200.0], dtype=jnp.float32)" + "# make a simple triangle-type hill. Slope is 0.2 and then -0.2\n", + "dist = 1000 # it's 1 km long\n", + "xcoords = jnp.arange(dist)\n", + "slope = jnp.concat([jnp.full(int(dist/2), 0.2), jnp.full(int(dist/2), -0.2)])\n", + "def lerp(x):\n", + " return jnp.interp(x, xcoords, slope)" ] + }, + { + "cell_type": "code", + "execution_count": 3, + "metadata": {}, + "outputs": [], + "source": [ + "time_scale = 0.1\n" + ] + }, + { + "cell_type": "code", + "execution_count": 4, + "metadata": {}, + "outputs": [], + "source": [ + "import jax\n", + "import jax.numpy as jnp\n", + "import flax.linen as nn\n", + "import numpy as np\n", + "import optax\n", + "from flax.linen.initializers import constant, orthogonal\n", + "from typing import Sequence, NamedTuple, Any\n", + "from flax.training.train_state import TrainState\n", + "import distrax\n", + "from gymnax.wrappers.purerl import LogWrapper, FlattenObservationWrapper\n", + "\n", + "\n", + "class GymnaxWrapper(object):\n", + " \"\"\"Base class for Gymnax wrappers.\"\"\"\n", + "\n", + " def __init__(self, env):\n", + " self._env = env\n", + "\n", + " # provide proxy access to regular attributes of wrapped object\n", + " def __getattr__(self, name):\n", + " return getattr(self._env, name)\n", + "\n", + "\n", + "class VecEnv(GymnaxWrapper):\n", + " def __init__(self, env):\n", + " super().__init__(env)\n", + " self.reset = jax.vmap(self._env.reset, in_axes=(0, None))\n", + " self.step = jax.vmap(self._env.step, in_axes=(0, 0, 0, None))\n", + "\n", + "class ClipAction(GymnaxWrapper):\n", + " def __init__(self, env, low=-1.0, high=1.0):\n", + " super().__init__(env)\n", + " self.low = low\n", + " self.high = high\n", + "\n", + " def step(self, key, state, action, params=None):\n", + " \"\"\"TODO: In theory the below line should be the way to do this.\"\"\"\n", + " # action = jnp.clip(action, self.env.action_space.low, self.env.action_space.high)\n", + " action = jnp.clip(action, self.low, self.high)\n", + " return self._env.step(key, state, action, params)\n", + "\n", + "\n", + "\n", + "class ActorCritic(nn.Module):\n", + " action_dim: Sequence[int]\n", + " activation: str = \"tanh\"\n", + "\n", + " @nn.compact\n", + " def __call__(self, x):\n", + " if self.activation == \"relu\":\n", + " activation = nn.relu\n", + " else:\n", + " activation = nn.tanh\n", + " actor_mean = nn.Dense(\n", + " 256, kernel_init=orthogonal(np.sqrt(2)), bias_init=constant(0.0)\n", + " )(x)\n", + " actor_mean = activation(actor_mean)\n", + " actor_mean = nn.Dense(\n", + " 256, kernel_init=orthogonal(np.sqrt(2)), bias_init=constant(0.0)\n", + " )(actor_mean)\n", + " actor_mean = activation(actor_mean)\n", + " actor_mean = nn.Dense(\n", + " self.action_dim, kernel_init=orthogonal(0.01), bias_init=constant(0.0)\n", + " )(actor_mean)\n", + " actor_logtstd = self.param(\"log_std\", nn.initializers.zeros, (self.action_dim,))\n", + " pi = distrax.MultivariateNormalDiag(actor_mean, jnp.exp(actor_logtstd))\n", + "\n", + " critic = nn.Dense(\n", + " 256, kernel_init=orthogonal(np.sqrt(2)), bias_init=constant(0.0)\n", + " )(x)\n", + " critic = activation(critic)\n", + " critic = nn.Dense(\n", + " 256, kernel_init=orthogonal(np.sqrt(2)), bias_init=constant(0.0)\n", + " )(critic)\n", + " critic = activation(critic)\n", + " critic = nn.Dense(1, kernel_init=orthogonal(1.0), bias_init=constant(0.0))(\n", + " critic\n", + " )\n", + "\n", + " return pi, jnp.squeeze(critic, axis=-1)\n", + "\n", + "\n", + "class Transition(NamedTuple):\n", + " done: jnp.ndarray\n", + " action: jnp.ndarray\n", + " value: jnp.ndarray\n", + " reward: jnp.ndarray\n", + " log_prob: jnp.ndarray\n", + " obs: jnp.ndarray\n", + " info: jnp.ndarray\n", + "\n", + "\n", + "def make_train(config):\n", + " config[\"NUM_UPDATES\"] = (\n", + " config[\"TOTAL_TIMESTEPS\"] // config[\"NUM_STEPS\"] // config[\"NUM_ENVS\"]\n", + " )\n", + " config[\"MINIBATCH_SIZE\"] = (\n", + " config[\"NUM_ENVS\"] * config[\"NUM_STEPS\"] // config[\"NUM_MINIBATCHES\"]\n", + " )\n", + " env = Snax()\n", + " env_params = env.default_params\n", + " env = LogWrapper(env)\n", + " env = VecEnv(env)\n", + " #env = ClipAction(env)\n", + " # if config[\"NORMALIZE_ENV\"]:\n", + " # env = NormalizeVecObservation(env)\n", + " # env = NormalizeVecReward(env, config[\"GAMMA\"])\n", + "\n", + " def linear_schedule(count):\n", + " frac = (\n", + " 1.0\n", + " - (count // (config[\"NUM_MINIBATCHES\"] * config[\"UPDATE_EPOCHS\"]))\n", + " / config[\"NUM_UPDATES\"]\n", + " )\n", + " return config[\"LR\"] * frac\n", + "\n", + " def train(rng):\n", + " # INIT NETWORK\n", + " network = ActorCritic(\n", + " env.action_space(env_params).shape[0], activation=config[\"ACTIVATION\"]\n", + " )\n", + " rng, _rng = jax.random.split(rng)\n", + " init_x = jnp.zeros(env.observation_space(env_params).shape)\n", + " network_params = network.init(_rng, init_x)\n", + " if config[\"ANNEAL_LR\"]:\n", + " tx = optax.chain(\n", + " optax.clip_by_global_norm(config[\"MAX_GRAD_NORM\"]),\n", + " optax.adam(learning_rate=linear_schedule, eps=1e-5),\n", + " )\n", + " else:\n", + " tx = optax.chain(\n", + " optax.clip_by_global_norm(config[\"MAX_GRAD_NORM\"]),\n", + " optax.adam(config[\"LR\"], eps=1e-5),\n", + " )\n", + " train_state = TrainState.create(\n", + " apply_fn=network.apply,\n", + " params=network_params,\n", + " tx=tx,\n", + " )\n", + "\n", + " # INIT ENV\n", + " rng, _rng = jax.random.split(rng)\n", + " reset_rng = jax.random.split(_rng, config[\"NUM_ENVS\"])\n", + " obsv, env_state = env.reset(reset_rng, env_params)\n", + "\n", + " # TRAIN LOOP\n", + " def _update_step(runner_state, unused):\n", + " # COLLECT TRAJECTORIES\n", + " def _env_step(runner_state, unused):\n", + " train_state, env_state, last_obs, rng = runner_state\n", + "\n", + " # SELECT ACTION\n", + " rng, _rng = jax.random.split(rng)\n", + " pi, value = network.apply(train_state.params, last_obs)\n", + " action = pi.sample(seed=_rng)\n", + " log_prob = pi.log_prob(action)\n", + "\n", + " # STEP ENV\n", + " rng, _rng = jax.random.split(rng)\n", + " rng_step = jax.random.split(_rng, config[\"NUM_ENVS\"])\n", + " obsv, env_state, reward, done, info = env.step(\n", + " rng_step, env_state, action, env_params\n", + " )\n", + " transition = Transition(\n", + " done, action, value, reward, log_prob, last_obs, info\n", + " )\n", + " runner_state = (train_state, env_state, obsv, rng)\n", + " return runner_state, transition\n", + "\n", + " runner_state, traj_batch = jax.lax.scan(\n", + " _env_step, runner_state, None, config[\"NUM_STEPS\"]\n", + " )\n", + "\n", + " # CALCULATE ADVANTAGE\n", + " train_state, env_state, last_obs, rng = runner_state\n", + " _, last_val = network.apply(train_state.params, last_obs)\n", + "\n", + " def _calculate_gae(traj_batch, last_val):\n", + " def _get_advantages(gae_and_next_value, transition):\n", + " gae, next_value = gae_and_next_value\n", + " done, value, reward = (\n", + " transition.done,\n", + " transition.value,\n", + " transition.reward,\n", + " )\n", + " delta = reward + config[\"GAMMA\"] * next_value * (1 - done) - value\n", + " gae = (\n", + " delta\n", + " + config[\"GAMMA\"] * config[\"GAE_LAMBDA\"] * (1 - done) * gae\n", + " )\n", + " return (gae, value), gae\n", + "\n", + " _, advantages = jax.lax.scan(\n", + " _get_advantages,\n", + " (jnp.zeros_like(last_val), last_val),\n", + " traj_batch,\n", + " reverse=True,\n", + " unroll=16,\n", + " )\n", + " return advantages, advantages + traj_batch.value\n", + "\n", + " advantages, targets = _calculate_gae(traj_batch, last_val)\n", + "\n", + " # UPDATE NETWORK\n", + " def _update_epoch(update_state, unused):\n", + " def _update_minbatch(train_state, batch_info):\n", + " traj_batch, advantages, targets = batch_info\n", + "\n", + " def _loss_fn(params, traj_batch, gae, targets):\n", + " # RERUN NETWORK\n", + " pi, value = network.apply(params, traj_batch.obs)\n", + " log_prob = pi.log_prob(traj_batch.action)\n", + "\n", + " # CALCULATE VALUE LOSS\n", + " value_pred_clipped = traj_batch.value + (\n", + " value - traj_batch.value\n", + " ).clip(-config[\"CLIP_EPS\"], config[\"CLIP_EPS\"])\n", + " value_losses = jnp.square(value - targets)\n", + " value_losses_clipped = jnp.square(value_pred_clipped - targets)\n", + " value_loss = (\n", + " 0.5 * jnp.maximum(value_losses, value_losses_clipped).mean()\n", + " )\n", + "\n", + " # CALCULATE ACTOR LOSS\n", + " ratio = jnp.exp(log_prob - traj_batch.log_prob)\n", + " gae = (gae - gae.mean()) / (gae.std() + 1e-8)\n", + " loss_actor1 = ratio * gae\n", + " loss_actor2 = (\n", + " jnp.clip(\n", + " ratio,\n", + " 1.0 - config[\"CLIP_EPS\"],\n", + " 1.0 + config[\"CLIP_EPS\"],\n", + " )\n", + " * gae\n", + " )\n", + " loss_actor = -jnp.minimum(loss_actor1, loss_actor2)\n", + " loss_actor = loss_actor.mean()\n", + " entropy = pi.entropy().mean()\n", + "\n", + " total_loss = (\n", + " loss_actor\n", + " + config[\"VF_COEF\"] * value_loss\n", + " - config[\"ENT_COEF\"] * entropy\n", + " )\n", + " return total_loss, (value_loss, loss_actor, entropy)\n", + "\n", + " grad_fn = jax.value_and_grad(_loss_fn, has_aux=True)\n", + " total_loss, grads = grad_fn(\n", + " train_state.params, traj_batch, advantages, targets\n", + " )\n", + " train_state = train_state.apply_gradients(grads=grads)\n", + " return train_state, total_loss\n", + "\n", + " train_state, traj_batch, advantages, targets, rng = update_state\n", + " rng, _rng = jax.random.split(rng)\n", + " batch_size = config[\"MINIBATCH_SIZE\"] * config[\"NUM_MINIBATCHES\"]\n", + " assert (\n", + " batch_size == config[\"NUM_STEPS\"] * config[\"NUM_ENVS\"]\n", + " ), \"batch size must be equal to number of steps * number of envs\"\n", + " permutation = jax.random.permutation(_rng, batch_size)\n", + " batch = (traj_batch, advantages, targets)\n", + " batch = jax.tree_util.tree_map(\n", + " lambda x: x.reshape((batch_size,) + x.shape[2:]), batch\n", + " )\n", + " shuffled_batch = jax.tree_util.tree_map(\n", + " lambda x: jnp.take(x, permutation, axis=0), batch\n", + " )\n", + " minibatches = jax.tree_util.tree_map(\n", + " lambda x: jnp.reshape(\n", + " x, [config[\"NUM_MINIBATCHES\"], -1] + list(x.shape[1:])\n", + " ),\n", + " shuffled_batch,\n", + " )\n", + " train_state, total_loss = jax.lax.scan(\n", + " _update_minbatch, train_state, minibatches\n", + " )\n", + " update_state = (train_state, traj_batch, advantages, targets, rng)\n", + " return update_state, total_loss\n", + "\n", + " update_state = (train_state, traj_batch, advantages, targets, rng)\n", + " update_state, loss_info = jax.lax.scan(\n", + " _update_epoch, update_state, None, config[\"UPDATE_EPOCHS\"]\n", + " )\n", + " train_state = update_state[0]\n", + " metric = traj_batch.info\n", + " rng = update_state[-1]\n", + " if config.get(\"DEBUG\"):\n", + "\n", + " def callback(info):\n", + " return_values = info[\"returned_episode_returns\"][\n", + " info[\"returned_episode\"]\n", + " ]\n", + " timesteps = (\n", + " info[\"timestep\"][info[\"returned_episode\"]] * config[\"NUM_ENVS\"]\n", + " )\n", + " for t in range(len(timesteps)):\n", + " print(\n", + " f\"global step={timesteps[t]}, episodic return={return_values[t]}\"\n", + " )\n", + "\n", + " jax.debug.callback(callback, metric)\n", + "\n", + " runner_state = (train_state, env_state, last_obs, rng)\n", + " return runner_state, metric\n", + "\n", + " rng, _rng = jax.random.split(rng)\n", + " runner_state = (train_state, env_state, obsv, _rng)\n", + " metrics = []\n", + " for i in range(config[\"NUM_MAINLOOPS\"]):\n", + " runner_state, metric = jax.lax.scan(\n", + " _update_step, runner_state, None, config[\"NUM_UPDATES\"]\n", + " )\n", + " metrics.append(metric)\n", + " return {\"runner_state\": runner_state, \"metrics\": metrics}\n", + "\n", + " return train\n" + ] + }, + { + "cell_type": "code", + "execution_count": 5, + "metadata": {}, + "outputs": [], + "source": [ + "config = {\n", + " \"LR\": 3e-4,\n", + " \"NUM_ENVS\": 2048,\n", + " \"NUM_STEPS\": 10,\n", + " \"TOTAL_TIMESTEPS\": 5e7,\n", + " \"UPDATE_EPOCHS\": 4,\n", + " \"NUM_MINIBATCHES\": 64,\n", + " \"GAMMA\": 0.99,\n", + " \"GAE_LAMBDA\": 0.95,\n", + " \"CLIP_EPS\": 0.2,\n", + " \"ENT_COEF\": 0.0,\n", + " \"VF_COEF\": 0.5,\n", + " \"MAX_GRAD_NORM\": 0.5,\n", + " \"ACTIVATION\": \"tanh\",\n", + " \"ENV_NAME\": \"hopper\",\n", + " \"ANNEAL_LR\": False,\n", + " \"NORMALIZE_ENV\": True,\n", + " \"DEBUG\": False,\n", + " \"NUM_MAINLOOPS\": 3,\n", + "}\n", + "config_light = {\n", + " \"LR\": 3e-4,\n", + " \"NUM_ENVS\": 128,\n", + " \"NUM_STEPS\": 10,\n", + " \"TOTAL_TIMESTEPS\": 5e3,\n", + " \"UPDATE_EPOCHS\": 2,\n", + " \"NUM_MINIBATCHES\": 16,\n", + " \"GAMMA\": 0.99,\n", + " \"GAE_LAMBDA\": 0.95,\n", + " \"CLIP_EPS\": 0.2,\n", + " \"ENT_COEF\": 0.0,\n", + " \"VF_COEF\": 0.5,\n", + " \"MAX_GRAD_NORM\": 0.5,\n", + " \"ACTIVATION\": \"tanh\",\n", + " \"ENV_NAME\": \"hopper\",\n", + " \"ANNEAL_LR\": False,\n", + " \"NORMALIZE_ENV\": True,\n", + " \"DEBUG\": False,\n", + "}\n", + "\n", + "rng = jax.random.key(42)\n", + "train_jit = jax.jit(make_train(config))\n", + "out = train_jit(rng)" + ] + }, + { + "cell_type": "code", + "execution_count": 15, + "metadata": {}, + "outputs": [ + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "import matplotlib.pyplot as plt\n", + "data = jnp.concatenate([out[\"metrics\"][x][\"returned_episode_returns\"].mean(-1).reshape(-1) for x in range(3)])\n", + "plt.plot(data)\n", + "ax = plt.gca()\n", + "\n", + "plt.xlabel(\"Update Step\")\n", + "plt.ylabel(\"Return\")\n", + "plt.title(\"PPO Agent Returns\")\n", + "plt.savefig(\"PPO_results.pdf\")\n", + "plt.show()" + ] + }, + { + "cell_type": "code", + "execution_count": 19, + "metadata": {}, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "2024-12-17 20:40:06.756483: W external/xla/xla/tsl/framework/bfc_allocator.cc:497] Allocator (GPU_0_bfc) ran out of memory trying to allocate 7.63GiB (rounded to 8192000000)requested by op \n", + "2024-12-17 20:40:06.756558: W external/xla/xla/tsl/framework/bfc_allocator.cc:508] ***************************************************_________________________________________________\n", + "E1217 20:40:06.756588 512196 pjrt_stream_executor_client.cc:3086] Execution of replica 0 failed: RESOURCE_EXHAUSTED: Out of memory while trying to allocate 8192000000 bytes.\n" + ] + }, + { + "ename": "XlaRuntimeError", + "evalue": "RESOURCE_EXHAUSTED: Out of memory while trying to allocate 8192000000 bytes.", + "output_type": "error", + "traceback": [ + "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m", + "\u001b[0;31mXlaRuntimeError\u001b[0m Traceback (most recent call last)", + "Cell \u001b[0;32mIn[19], line 33\u001b[0m\n\u001b[1;32m 28\u001b[0m runner_state \u001b[38;5;241m=\u001b[39m (train_state, env_state, obsv, rng)\n\u001b[1;32m 29\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m runner_state, env_state\n\u001b[0;32m---> 33\u001b[0m runner_state, env_logs \u001b[38;5;241m=\u001b[39m \u001b[43mjax\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mlax\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mscan\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m 34\u001b[0m \u001b[43m \u001b[49m\u001b[43m_env_step\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mout\u001b[49m\u001b[43m[\u001b[49m\u001b[38;5;124;43m'\u001b[39;49m\u001b[38;5;124;43mrunner_state\u001b[39;49m\u001b[38;5;124;43m'\u001b[39;49m\u001b[43m]\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;28;43;01mNone\u001b[39;49;00m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m100\u001b[39;49m\n\u001b[1;32m 35\u001b[0m \u001b[43m)\u001b[49m\n", + " \u001b[0;31m[... skipping hidden 11 frame]\u001b[0m\n", + "File \u001b[0;32m~/Documents/Code/solarcarsim/.venv/lib/python3.12/site-packages/jax/_src/interpreters/pxla.py:1298\u001b[0m, in \u001b[0;36mExecuteReplicated.__call__\u001b[0;34m(self, *args)\u001b[0m\n\u001b[1;32m 1296\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_handle_token_bufs(result_token_bufs, sharded_runtime_token)\n\u001b[1;32m 1297\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[0;32m-> 1298\u001b[0m results \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mxla_executable\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mexecute_sharded\u001b[49m\u001b[43m(\u001b[49m\u001b[43minput_bufs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 1300\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m dispatch\u001b[38;5;241m.\u001b[39mneeds_check_special():\n\u001b[1;32m 1301\u001b[0m out_arrays \u001b[38;5;241m=\u001b[39m results\u001b[38;5;241m.\u001b[39mdisassemble_into_single_device_arrays()\n", + "\u001b[0;31mXlaRuntimeError\u001b[0m: RESOURCE_EXHAUSTED: Out of memory while trying to allocate 8192000000 bytes." + ] + } + ], + "source": [ + "tstate = out['runner_state'][0]\n", + "lastobs = out['runner_state'][2]\n", + "# tstate.apply_fn(tstate.params, lastobs)\n", + "\n", + "env = Snax()\n", + "env_params = env.default_params\n", + "env = LogWrapper(env)\n", + "env = VecEnv(env)\n", + "\n", + "def _env_step(runner_state, unused):\n", + " train_state, env_state, last_obs, rng = runner_state\n", + "\n", + " # SELECT ACTION\n", + " rng, _rng = jax.random.split(rng)\n", + " pi, value = train_state.apply_fn(train_state.params, last_obs)\n", + " action = pi.sample(seed=_rng)\n", + " log_prob = pi.log_prob(action)\n", + "\n", + " # STEP ENV\n", + " rng, _rng = jax.random.split(rng)\n", + " rng_step = jax.random.split(_rng, config[\"NUM_ENVS\"])\n", + " obsv, env_state, reward, done, info = env.step(\n", + " rng_step, env_state, action, env_params\n", + " )\n", + " transition = Transition(\n", + " done, action, value, reward, log_prob, last_obs, info\n", + " )\n", + " runner_state = (train_state, env_state, obsv, rng)\n", + " return runner_state, env_state\n", + "\n", + "\n", + "\n", + "runner_state, env_logs = jax.lax.scan(\n", + " _env_step, out['runner_state'], None, 100\n", + ")" + ] + }, + { + "cell_type": "code", + "execution_count": 8, + "metadata": {}, + "outputs": [ + { + "data": { + "text/plain": [ + "TrainState(step=Array(1874688, dtype=int32, weak_type=True), apply_fn=, params={'params': {'Dense_0': {'bias': Array([-5.3029938e-04, 0.0000000e+00, 0.0000000e+00, 0.0000000e+00,\n", + " 1.9191748e-03, 0.0000000e+00, -1.7144637e-04, 2.5281330e-05,\n", + " 1.4641370e-03, 0.0000000e+00, -1.3245111e-03, 0.0000000e+00,\n", + " 0.0000000e+00, 8.3523680e-04, 0.0000000e+00, 0.0000000e+00,\n", + " 6.0847404e-05, -4.7602013e-04, 0.0000000e+00, 1.3794134e-03,\n", + " 0.0000000e+00, -7.6553306e-06, 0.0000000e+00, 2.1930558e-04,\n", + " 0.0000000e+00, -2.2188693e-03, 1.4511290e-03, 1.9425271e-03,\n", + " 0.0000000e+00, 0.0000000e+00, 0.0000000e+00, 0.0000000e+00,\n", + " 0.0000000e+00, 0.0000000e+00, 0.0000000e+00, 1.1256059e-03,\n", + " -5.8398215e-04, 0.0000000e+00, -1.8540439e-03, 0.0000000e+00,\n", + " 0.0000000e+00, 0.0000000e+00, 0.0000000e+00, 0.0000000e+00,\n", + " 0.0000000e+00, -6.9769150e-05, 0.0000000e+00, 0.0000000e+00,\n", + " 0.0000000e+00, 6.8441045e-04, 0.0000000e+00, 0.0000000e+00,\n", + " 0.0000000e+00, 0.0000000e+00, -8.8643702e-04, 1.3925527e-03,\n", + " -1.4111768e-04, 0.0000000e+00, 0.0000000e+00, -2.0299070e-03,\n", + " 6.8957923e-04, -1.2412324e-03, 5.5068691e-04, -5.2116567e-04,\n", + " 0.0000000e+00, 0.0000000e+00, 0.0000000e+00, 6.2125333e-04,\n", + " 9.1447402e-04, 0.0000000e+00, 0.0000000e+00, 0.0000000e+00,\n", + " 5.2126928e-04, 0.0000000e+00, 5.8222190e-04, 0.0000000e+00,\n", + " 0.0000000e+00, 2.2282777e-03, 1.3834524e-03, -1.0245681e-03,\n", + " 1.7563089e-04, -2.4510939e-03, 0.0000000e+00, -1.6243160e-03,\n", + " -9.4084017e-04, 0.0000000e+00, 0.0000000e+00, 0.0000000e+00,\n", + " -9.0897578e-04, 0.0000000e+00, -1.0096998e-03, 0.0000000e+00,\n", + " 0.0000000e+00, 0.0000000e+00, 0.0000000e+00, 3.1322434e-03,\n", + " 0.0000000e+00, 0.0000000e+00, 2.2586577e-03, -2.4337969e-03,\n", + " 5.6456536e-04, 3.0193112e-03, 0.0000000e+00, 0.0000000e+00,\n", + " -2.9874803e-03, 0.0000000e+00, 0.0000000e+00, 0.0000000e+00,\n", + " -3.7863411e-03, 0.0000000e+00, 0.0000000e+00, 0.0000000e+00,\n", + " 0.0000000e+00, 0.0000000e+00, 0.0000000e+00, -3.2294096e-04,\n", + " 2.7185181e-04, 1.4245387e-08, -1.9866133e-03, -2.2128341e-03,\n", + " 0.0000000e+00, -1.7956822e-04, 1.4105116e-03, 0.0000000e+00,\n", + " -6.7686941e-04, 0.0000000e+00, 0.0000000e+00, 4.0784094e-04,\n", + " 0.0000000e+00, 0.0000000e+00, -1.0925150e-03, -9.4625680e-04,\n", + " -1.0735401e-03, 1.7537770e-03, -1.4924963e-03, -1.0218532e-03,\n", + " 0.0000000e+00, -3.4914708e-03, -4.7835559e-04, 1.9216866e-03,\n", + " 0.0000000e+00, 7.1237684e-04, -3.2128301e-03, -1.7769258e-03,\n", + " 2.2146765e-03, 0.0000000e+00, 0.0000000e+00, 0.0000000e+00,\n", + " 6.1193650e-04, 0.0000000e+00, 0.0000000e+00, 1.7334922e-04,\n", + " -2.6329695e-03, -1.3060468e-03, 3.4790873e-03, 0.0000000e+00,\n", + " 1.2544972e-03, 0.0000000e+00, 0.0000000e+00, 0.0000000e+00,\n", + " 0.0000000e+00, 0.0000000e+00, 0.0000000e+00, 0.0000000e+00,\n", + " 0.0000000e+00, -1.2978822e-03, -3.8260728e-04, 0.0000000e+00,\n", + " 0.0000000e+00, 0.0000000e+00, -4.5437412e-04, 0.0000000e+00,\n", + " 0.0000000e+00, 7.1664042e-05, 0.0000000e+00, 0.0000000e+00,\n", + " 0.0000000e+00, 0.0000000e+00, 0.0000000e+00, 7.0464790e-05,\n", + " -8.9069502e-04, 0.0000000e+00, 6.6835230e-04, 0.0000000e+00,\n", + " 6.8251284e-05, 0.0000000e+00, 0.0000000e+00, 0.0000000e+00,\n", + " -1.0040104e-03, -2.3794189e-07, 0.0000000e+00, 0.0000000e+00,\n", + " 0.0000000e+00, 2.3229128e-04, 1.5453034e-03, -5.0179427e-05,\n", + " 2.2772038e-03, -2.0145031e-03, 9.4736563e-03, 7.9449776e-05,\n", + " -7.4491683e-05, 0.0000000e+00, 2.5824676e-04, 0.0000000e+00,\n", + " 1.2736652e-03, 0.0000000e+00, 3.0285574e-04, 2.4916211e-04,\n", + " 0.0000000e+00, 0.0000000e+00, 7.7232631e-04, 5.4959516e-04,\n", + " -1.9705489e-03, 0.0000000e+00, 1.1921050e-03, -3.9399034e-04,\n", + " 0.0000000e+00, 3.7107369e-04, -4.8485940e-04, 0.0000000e+00,\n", + " 1.5376735e-04, 1.8611597e-03, -6.0213718e-04, -2.2827655e-03,\n", + " 0.0000000e+00, 0.0000000e+00, -2.4243153e-04, 0.0000000e+00,\n", + " -2.2830942e-03, 4.1451986e-04, -2.6722133e-04, 2.5362780e-04,\n", + " 0.0000000e+00, 0.0000000e+00, 0.0000000e+00, 0.0000000e+00,\n", + " 0.0000000e+00, 0.0000000e+00, 0.0000000e+00, 0.0000000e+00,\n", + " 0.0000000e+00, 0.0000000e+00, 0.0000000e+00, 0.0000000e+00,\n", + " 0.0000000e+00, 0.0000000e+00, 0.0000000e+00, 0.0000000e+00,\n", + " 0.0000000e+00, 0.0000000e+00, -6.0310279e-04, 0.0000000e+00,\n", + " 0.0000000e+00, -4.7796275e-04, 0.0000000e+00, 0.0000000e+00], dtype=float32), 'kernel': Array([[-0.04059862, -0.05616716, 0.05436023, ..., -0.01495443,\n", + " -0.09417699, -0.03651878],\n", + " [-0.04345348, 0.0237919 , 0.13925986, ..., -0.03692472,\n", + " 0.03002992, -0.11007421],\n", + " [ 0.09764732, -0.03158703, -0.02001419, ..., -0.14909488,\n", + " 0.02449407, 0.01721021],\n", + " ...,\n", + " [ 0.11338337, -0.04206258, 0.11930869, ..., -0.01594234,\n", + " 0.0031549 , 0.07585711],\n", + " [-0.09437636, -0.10784481, -0.13662401, ..., 0.0185915 ,\n", + " 0.04616122, -0.02840761],\n", + " [-0.03904481, -0.04892467, 0.05431201, ..., 0.1215561 ,\n", + " -0.19458926, 0.04553339]], dtype=float32)}, 'Dense_1': {'bias': Array([ 0.00027806, 0.04620876, -0.00658554, -0.04454401, -0.00277375,\n", + " 0.03381715, 0.03254296, 0.00331175, -0.04037352, 0.04094049,\n", + " 0.00339541, -0.0373533 , -0.03950972, 0.00459826, 0.00176061,\n", + " -0.04285378, 0.03781348, 0.04358079, 0.03816335, 0.03898 ,\n", + " 0.03793908, -0.0067359 , -0.0411862 , 0.00432372, -0.03645715,\n", + " 0.03726813, -0.00039956, 0.03472266, -0.00593952, -0.04233544,\n", + " -0.00230428, -0.00433152, 0.03425088, -0.02485432, -0.0286654 ,\n", + " -0.00356171, 0.03057408, 0.03985466, 0.00293833, 0.00458675,\n", + " 0.02603373, -0.03051696, 0.03516515, 0.00130115, 0.0465891 ,\n", + " -0.04041 , -0.04491264, 0.01050323, -0.03533997, 0.03221814,\n", + " -0.00215372, 0.04099191, 0.03301746, 0.04153331, -0.00470519,\n", + " -0.00428926, -0.05097796, -0.03040344, -0.03735339, -0.03491513,\n", + " 0.01251048, 0.03140776, 0.00783039, 0.00627123, -0.02943583,\n", + " 0.03217795, -0.03619361, 0.04503313, 0.00422618, 0.03048726,\n", + " -0.04530523, -0.05115746, 0.00748269, 0.04205422, 0.04298782,\n", + " 0.03045189, -0.03948338, 0.00261836, 0.03464689, 0.00538703,\n", + " -0.00227268, -0.03553307, 0.00145856, 0.02827597, -0.03024151,\n", + " -0.04150978, -0.03633608, -0.00803869, -0.00474264, -0.03425602,\n", + " 0.03940321, 0.03186259, -0.00434997, 0.00952879, 0.04427051,\n", + " 0.00112212, -0.01438359, 0.02742064, 0.00732329, -0.03644897,\n", + " 0.02770537, -0.00053647, -0.03326581, 0.03436317, -0.03182915,\n", + " 0.00353186, -0.0258979 , -0.00346143, -0.00030614, 0.00350122,\n", + " 0.04173567, -0.04064614, -0.01026691, -0.04184798, -0.0351134 ,\n", + " 0.02159838, 0.03210129, 0.02979252, -0.00021149, -0.03641422,\n", + " -0.00728356, -0.00786742, 0.03055024, -0.0375047 , 0.03636042,\n", + " -0.03306672, 0.03809343, -0.00785372, 0.03908067, -0.02755055,\n", + " -0.0311247 , -0.03837723, 0.03915109, 0.00293649, -0.02566425,\n", + " -0.03642197, 0.00442762, -0.0285236 , -0.03219024, -0.03771985,\n", + " 0.03348036, 0.04192547, 0.04804754, -0.00390566, 0.03801316,\n", + " -0.0048022 , -0.02534208, 0.00532888, -0.00960828, 0.0399802 ,\n", + " 0.00194347, 0.04182222, -0.02789724, -0.03810298, -0.0374799 ,\n", + " -0.00092357, -0.03994033, -0.03276777, 0.0271517 , -0.03100638,\n", + " -0.04451067, -0.00705037, -0.02453071, 0.03519889, 0.03063235,\n", + " -0.03020622, -0.03580233, -0.03554969, 0.03846539, -0.00717684,\n", + " 0.0024833 , 0.0319246 , 0.02931642, 0.03770342, -0.03255963,\n", + " 0.03416643, 0.01054269, -0.03945149, -0.00351726, 0.04458899,\n", + " -0.00589401, -0.00648886, 0.0339095 , 0.02671813, 0.04369878,\n", + " 0.03415771, 0.03542192, 0.00600839, -0.0109551 , -0.037458 ,\n", + " -0.03981134, 0.03337134, 0.00489916, 0.00273095, -0.00618493,\n", + " 0.04161482, 0.03454573, -0.01003036, 0.00425193, 0.01817769,\n", + " 0.03138407, -0.03242443, 0.00188284, -0.00314224, 0.03295113,\n", + " 0.04222574, 0.03434506, 0.04464654, -0.03982402, -0.04468929,\n", + " 0.04249624, 0.00770373, -0.03094419, 0.00289513, 0.04945203,\n", + " 0.03187467, -0.04758247, 0.00177818, 0.00270677, -0.04519045,\n", + " -0.0402201 , -0.0426771 , 0.00278877, -0.03111694, -0.03696601,\n", + " -0.03909003, -0.03003668, 0.00376004, 0.03456468, 0.01219636,\n", + " 0.00517881, -0.0396665 , 0.04550315, -0.04317153, -0.00225738,\n", + " -0.02796482, -0.00212918, 0.0041019 , -0.03426032, -0.04060804,\n", + " 0.00500863, 0.0303803 , -0.03806275, -0.0432169 , 0.03569881,\n", + " 0.04669258, -0.00873193, -0.00346405, 0.02921363, 0.03867232,\n", + " 0.04459479, -0.03213866, -0.03638731, -0.03427472, 0.04575037,\n", + " -0.00120791], dtype=float32), 'kernel': Array([[ 0.0554822 , -0.07003432, 0.09695947, ..., -0.17977887,\n", + " 0.0366156 , 0.02077846],\n", + " [ 0.17207822, 0.09883893, 0.01466036, ..., -0.02480283,\n", + " 0.08038498, -0.09478654],\n", + " [ 0.00384633, 0.1082558 , 0.02827624, ..., -0.02226412,\n", + " 0.05913283, 0.08880645],\n", + " ...,\n", + " [ 0.0052855 , 0.08274142, -0.02554986, ..., 0.08520137,\n", + " 0.01621614, -0.10647655],\n", + " [ 0.08706143, 0.09899756, -0.05577984, ..., -0.05286903,\n", + " -0.06213011, -0.07769346],\n", + " [ 0.11625612, 0.03126499, -0.13369204, ..., -0.05385393,\n", + " -0.092626 , -0.06382608]], dtype=float32)}, 'Dense_2': {'bias': Array([0.47612667], dtype=float32), 'kernel': Array([[-5.04823923e-01],\n", + " [ 4.65460658e-01],\n", + " [-5.04793406e-01],\n", + " [-4.72569168e-01],\n", + " [ 5.04825413e-01],\n", + " [ 4.79185104e-01],\n", + " [ 4.80066299e-01],\n", + " [-5.04783750e-01],\n", + " [-4.73151356e-01],\n", + " [ 4.51244444e-01],\n", + " [ 5.04835188e-01],\n", + " [-4.73466516e-01],\n", + " [-4.83210683e-01],\n", + " [ 5.04850090e-01],\n", + " [ 5.04784822e-01],\n", + " [-4.91237015e-01],\n", + " [ 4.74481404e-01],\n", + " [ 4.75008428e-01],\n", + " [ 4.78206635e-01],\n", + " [ 4.73737508e-01],\n", + " [ 4.78371084e-01],\n", + " [ 5.04845023e-01],\n", + " [-4.80819106e-01],\n", + " [-4.72271591e-01],\n", + " [-4.73856926e-01],\n", + " [ 4.75976706e-01],\n", + " [ 9.55915993e-07],\n", + " [ 4.79922265e-01],\n", + " [ 5.04790187e-01],\n", + " [-4.81541991e-01],\n", + " [ 5.04812837e-01],\n", + " [ 5.04815102e-01],\n", + " [ 4.71088231e-01],\n", + " [-4.76686895e-01],\n", + " [-4.73729491e-01],\n", + " [-5.04812241e-01],\n", + " [ 4.79393840e-01],\n", + " [ 4.72649992e-01],\n", + " [-5.04822433e-01],\n", + " [-5.04823506e-01],\n", + " [ 4.76308912e-01],\n", + " [-4.75529820e-01],\n", + " [ 4.74255502e-01],\n", + " [-5.04849136e-01],\n", + " [ 4.87328082e-01],\n", + " [-4.72647727e-01],\n", + " [-4.74320143e-01],\n", + " [ 5.04813969e-01],\n", + " [-4.75550056e-01],\n", + " [ 4.75366712e-01],\n", + " [-5.04834890e-01],\n", + " [ 4.62294102e-01],\n", + " [ 4.76273477e-01],\n", + " [ 4.81204808e-01],\n", + " [ 5.04829705e-01],\n", + " [-4.72716808e-01],\n", + " [-4.76078093e-01],\n", + " [-4.71584499e-01],\n", + " [-4.79490250e-01],\n", + " [-4.75963056e-01],\n", + " [ 5.04846215e-01],\n", + " [ 4.72023100e-01],\n", + " [ 5.04857779e-01],\n", + " [ 5.04864097e-01],\n", + " [-4.75473702e-01],\n", + " [ 4.74616170e-01],\n", + " [-4.75657642e-01],\n", + " [ 4.74489599e-01],\n", + " [-5.04848897e-01],\n", + " [ 4.76847380e-01],\n", + " [-4.80367184e-01],\n", + " [-4.77631360e-01],\n", + " [ 5.04855454e-01],\n", + " [ 4.69335794e-01],\n", + " [ 4.70332831e-01],\n", + " [ 4.77089763e-01],\n", + " [-4.76223826e-01],\n", + " [-5.04711330e-01],\n", + " [ 4.76124525e-01],\n", + " [ 4.71102148e-01],\n", + " [-5.04813373e-01],\n", + " [-4.76169169e-01],\n", + " [ 5.04840791e-01],\n", + " [ 4.74855334e-01],\n", + " [-4.73940402e-01],\n", + " [-4.77510095e-01],\n", + " [-4.69742119e-01],\n", + " [ 5.04834712e-01],\n", + " [ 4.07728885e-06],\n", + " [-4.77828860e-01],\n", + " [ 4.78402615e-01],\n", + " [ 4.75762784e-01],\n", + " [ 5.04799962e-01],\n", + " [ 5.04824638e-01],\n", + " [ 4.77130145e-01],\n", + " [-5.04835606e-01],\n", + " [-5.04765272e-01],\n", + " [ 4.85664934e-01],\n", + " [ 5.04874110e-01],\n", + " [-4.76336956e-01],\n", + " [ 4.73611504e-01],\n", + " [ 5.04758954e-01],\n", + " [-4.73996580e-01],\n", + " [ 4.81847137e-01],\n", + " [-4.73559260e-01],\n", + " [ 5.04823327e-01],\n", + " [-4.75162297e-01],\n", + " [ 5.04778862e-01],\n", + " [ 5.04798412e-01],\n", + " [-5.04776955e-01],\n", + " [ 4.74137306e-01],\n", + " [-4.81388301e-01],\n", + " [-5.04833639e-01],\n", + " [-4.77248192e-01],\n", + " [-4.74423558e-01],\n", + " [ 4.75161046e-01],\n", + " [ 4.78681266e-01],\n", + " [ 4.74670380e-01],\n", + " [ 5.04786909e-01],\n", + " [-4.53696519e-01],\n", + " [-5.04794300e-01],\n", + " [ 5.04820824e-01],\n", + " [ 4.74956840e-01],\n", + " [-4.66621697e-01],\n", + " [ 4.78065014e-01],\n", + " [-4.74870712e-01],\n", + " [ 4.75875765e-01],\n", + " [-5.04830480e-01],\n", + " [ 4.70007926e-01],\n", + " [-4.74391580e-01],\n", + " [ 1.86074972e-02],\n", + " [-4.84576911e-01],\n", + " [ 4.70582545e-01],\n", + " [-5.04855335e-01],\n", + " [-4.74273682e-01],\n", + " [-4.81450081e-01],\n", + " [-5.04836082e-01],\n", + " [-4.72928852e-01],\n", + " [-4.76107746e-01],\n", + " [-4.86656040e-01],\n", + " [ 4.78675216e-01],\n", + " [ 4.63450849e-01],\n", + " [ 4.78642941e-01],\n", + " [-5.04807711e-01],\n", + " [ 4.72376823e-01],\n", + " [ 5.04851758e-01],\n", + " [-4.75505918e-01],\n", + " [ 5.04856586e-01],\n", + " [ 5.04849017e-01],\n", + " [ 4.73896414e-01],\n", + " [ 5.04803658e-01],\n", + " [ 4.65696543e-01],\n", + " [-4.75109488e-01],\n", + " [-4.84716296e-01],\n", + " [-4.82673854e-01],\n", + " [-5.04828036e-01],\n", + " [-4.82922703e-01],\n", + " [-4.74735767e-01],\n", + " [ 4.76470083e-01],\n", + " [-4.74129856e-01],\n", + " [-4.67900127e-01],\n", + " [-5.04822850e-01],\n", + " [-4.77026075e-01],\n", + " [ 4.84468341e-01],\n", + " [ 4.75914717e-01],\n", + " [-4.69322681e-01],\n", + " [-4.79369819e-01],\n", + " [-4.73021626e-01],\n", + " [ 4.77468073e-01],\n", + " [ 5.04799247e-01],\n", + " [ 4.73501444e-01],\n", + " [ 4.77092564e-01],\n", + " [ 4.74903047e-01],\n", + " [ 4.64455873e-01],\n", + " [-4.74970967e-01],\n", + " [ 4.76718366e-01],\n", + " [-5.04804611e-01],\n", + " [-4.72296417e-01],\n", + " [ 5.04768193e-01],\n", + " [ 4.85322595e-01],\n", + " [-5.04860520e-01],\n", + " [ 5.04836917e-01],\n", + " [ 4.70947295e-01],\n", + " [ 4.75462735e-01],\n", + " [ 4.79784906e-01],\n", + " [ 4.77948338e-01],\n", + " [ 4.75362688e-01],\n", + " [ 5.04798710e-01],\n", + " [-5.04820824e-01],\n", + " [-4.81628805e-01],\n", + " [-4.71553355e-01],\n", + " [ 4.70683396e-01],\n", + " [-5.04853904e-01],\n", + " [ 4.68065441e-01],\n", + " [ 5.04831374e-01],\n", + " [ 4.78038251e-01],\n", + " [ 4.75826532e-01],\n", + " [ 5.04855931e-01],\n", + " [ 5.03943205e-01],\n", + " [-2.45599379e-03],\n", + " [ 4.73539799e-01],\n", + " [-4.75195169e-01],\n", + " [-5.04856825e-01],\n", + " [ 5.04774928e-01],\n", + " [ 1.11883745e-01],\n", + " [ 4.77893203e-01],\n", + " [ 4.68741715e-01],\n", + " [ 4.79511768e-01],\n", + " [-4.78091866e-01],\n", + " [-4.70777363e-01],\n", + " [ 4.74244267e-01],\n", + " [ 5.04789829e-01],\n", + " [-4.83687103e-01],\n", + " [-5.04752696e-01],\n", + " [ 4.75559562e-01],\n", + " [ 4.74621952e-01],\n", + " [-4.87164825e-01],\n", + " [-5.04829168e-01],\n", + " [ 5.04815757e-01],\n", + " [-4.90391821e-01],\n", + " [-4.68300521e-01],\n", + " [-4.67967004e-01],\n", + " [ 5.04820704e-01],\n", + " [ 2.23571993e-02],\n", + " [-4.72705662e-01],\n", + " [-4.66774702e-01],\n", + " [-4.77931947e-01],\n", + " [ 5.04782379e-01],\n", + " [ 4.75416005e-01],\n", + " [ 5.04854858e-01],\n", + " [ 5.04825830e-01],\n", + " [-4.73938853e-01],\n", + " [ 4.71472412e-01],\n", + " [-4.69849199e-01],\n", + " [ 5.04805148e-01],\n", + " [-4.73442256e-01],\n", + " [-5.04803061e-01],\n", + " [-5.04784584e-01],\n", + " [-4.77962017e-01],\n", + " [-4.72930998e-01],\n", + " [ 5.04825056e-01],\n", + " [ 4.75001872e-01],\n", + " [-4.77844417e-01],\n", + " [-4.69303727e-01],\n", + " [ 4.82429415e-01],\n", + " [ 4.70820099e-01],\n", + " [ 5.04818201e-01],\n", + " [-5.04762232e-01],\n", + " [ 4.78444785e-01],\n", + " [ 4.84044403e-01],\n", + " [ 1.10305548e-01],\n", + " [-4.72621500e-01],\n", + " [-4.72007573e-01],\n", + " [-4.69124794e-01],\n", + " [ 4.82611150e-01],\n", + " [-5.04812717e-01]], dtype=float32)}, 'Dense_3': {'bias': Array([ 0.00000000e+00, 0.00000000e+00, 1.33985464e-04, 0.00000000e+00,\n", + " 1.72092405e-04, 0.00000000e+00, 1.10953239e-04, 0.00000000e+00,\n", + " -7.64621655e-04, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,\n", + " -2.41837231e-04, 0.00000000e+00, 1.13354428e-04, 3.29057017e-04,\n", + " 2.16200497e-04, 2.19336202e-04, -3.66769207e-04, -4.79893264e-04,\n", + " 0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,\n", + " -1.69424180e-04, -4.80763149e-04, -2.22473827e-05, 0.00000000e+00,\n", + " 0.00000000e+00, 0.00000000e+00, 2.85213537e-05, -2.57758162e-04,\n", + " 0.00000000e+00, -7.53246495e-05, 0.00000000e+00, 0.00000000e+00,\n", + " 7.56459194e-04, 0.00000000e+00, 0.00000000e+00, 7.03259939e-05,\n", + " -1.14754995e-03, 1.92435517e-04, 0.00000000e+00, 0.00000000e+00,\n", + " 0.00000000e+00, 0.00000000e+00, 0.00000000e+00, -2.27750847e-04,\n", + " -9.26916346e-06, 0.00000000e+00, -1.29773716e-05, -2.41517424e-04,\n", + " 0.00000000e+00, 1.77495138e-04, 6.55367854e-04, 0.00000000e+00,\n", + " 0.00000000e+00, 1.65232050e-04, 4.73026412e-05, 0.00000000e+00,\n", + " 0.00000000e+00, -1.73691460e-05, 0.00000000e+00, 0.00000000e+00,\n", + " 0.00000000e+00, 0.00000000e+00, -8.86134421e-06, 0.00000000e+00,\n", + " 0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 3.70853086e-04,\n", + " -5.84304449e-04, 1.24640821e-04, 0.00000000e+00, -1.71510528e-05,\n", + " -1.06271233e-04, 4.95114073e-04, 0.00000000e+00, 0.00000000e+00,\n", + " -9.10749368e-05, -6.45977125e-05, -2.22136936e-04, 0.00000000e+00,\n", + " 0.00000000e+00, 0.00000000e+00, -3.36646590e-05, 0.00000000e+00,\n", + " 0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 9.56642965e-04,\n", + " 0.00000000e+00, 0.00000000e+00, -1.40761127e-04, 0.00000000e+00,\n", + " 0.00000000e+00, 3.49420065e-04, 9.81494231e-05, 0.00000000e+00,\n", + " -3.01860314e-06, 4.25145874e-04, 0.00000000e+00, -2.97043298e-04,\n", + " 0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,\n", + " 0.00000000e+00, -4.22585785e-04, 3.35624827e-05, -1.63179357e-04,\n", + " -3.85784428e-04, 0.00000000e+00, 0.00000000e+00, -1.18570577e-04,\n", + " 0.00000000e+00, 4.95775254e-04, 0.00000000e+00, 0.00000000e+00,\n", + " 0.00000000e+00, 1.57581700e-04, 2.71000172e-05, 0.00000000e+00,\n", + " 0.00000000e+00, 0.00000000e+00, 0.00000000e+00, -3.49011447e-04,\n", + " 0.00000000e+00, 2.14735373e-05, 0.00000000e+00, 0.00000000e+00,\n", + " 0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,\n", + " 0.00000000e+00, 1.23570091e-03, 4.56354406e-04, -5.37516189e-06,\n", + " 0.00000000e+00, -4.59517993e-04, 0.00000000e+00, -1.05757033e-03,\n", + " -8.03241505e-07, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,\n", + " -8.58970961e-05, 0.00000000e+00, 0.00000000e+00, -8.31345751e-05,\n", + " 0.00000000e+00, 0.00000000e+00, 0.00000000e+00, -1.75031179e-04,\n", + " 0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 3.40558800e-05,\n", + " 0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,\n", + " 4.68007347e-04, 0.00000000e+00, -6.02232467e-05, 7.55003784e-05,\n", + " 0.00000000e+00, 1.04056271e-04, -4.94510285e-04, 8.19228473e-04,\n", + " 0.00000000e+00, -2.76178849e-04, -4.28266299e-04, 0.00000000e+00,\n", + " 0.00000000e+00, 0.00000000e+00, 0.00000000e+00, -1.01641235e-04,\n", + " 7.85850134e-05, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,\n", + " -1.81269148e-04, 0.00000000e+00, -1.07560321e-04, 3.96017676e-05,\n", + " 0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,\n", + " 5.02542243e-04, 0.00000000e+00, -1.22958256e-04, 0.00000000e+00,\n", + " 0.00000000e+00, -5.88585157e-04, 0.00000000e+00, -4.56086855e-04,\n", + " 3.73370094e-05, 0.00000000e+00, 0.00000000e+00, -1.42131088e-04,\n", + " 0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 5.83865003e-05,\n", + " 0.00000000e+00, -7.49289902e-05, 0.00000000e+00, -6.43308813e-05,\n", + " -5.08234507e-06, 0.00000000e+00, 0.00000000e+00, -7.32544358e-05,\n", + " 3.86106731e-05, -1.97132340e-05, -3.74409021e-04, 0.00000000e+00,\n", + " 0.00000000e+00, -1.13278977e-04, 0.00000000e+00, 0.00000000e+00,\n", + " 1.05968058e-04, 0.00000000e+00, 1.55918504e-04, -8.26295363e-05,\n", + " -2.53444654e-04, -4.88312908e-05, -3.71871050e-04, 0.00000000e+00,\n", + " -2.51845704e-05, 0.00000000e+00, 0.00000000e+00, -1.95649074e-04,\n", + " -5.55169863e-05, 3.69632644e-08, 0.00000000e+00, -7.79392940e-05,\n", + " 0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,\n", + " 0.00000000e+00, 0.00000000e+00, 0.00000000e+00, -4.48481238e-04,\n", + " 0.00000000e+00, 6.39237842e-05, -1.48704552e-04, 0.00000000e+00,\n", + " 1.45096536e-04, 1.64957572e-04, 0.00000000e+00, -2.98823102e-06], dtype=float32), 'kernel': Array([[-0.02561163, 0.03399104, -0.06704379, ..., -0.12532048,\n", + " -0.02462031, 0.08065256],\n", + " [-0.12416931, -0.139428 , -0.13583477, ..., -0.1480573 ,\n", + " 0.00178415, 0.02530298],\n", + " [-0.09367059, -0.12599796, 0.0561337 , ..., 0.01869138,\n", + " 0.0979134 , -0.08274377],\n", + " ...,\n", + " [ 0.00176744, -0.04548926, 0.00608217, ..., -0.04934197,\n", + " 0.06005816, -0.12460606],\n", + " [ 0.08490104, 0.01121706, -0.09934874, ..., 0.04335562,\n", + " 0.1742473 , -0.0600209 ],\n", + " [-0.00646635, 0.03393729, -0.01265469, ..., -0.04704732,\n", + " 0.08156815, 0.05526903]], dtype=float32)}, 'Dense_4': {'bias': Array([-0.00419973, 0.02144877, 0.00450037, -0.01710127, -0.00333009,\n", + " 0.00356246, -0.02329435, -0.00337948, 0.01743394, -0.01690184,\n", + " -0.00378336, -0.00096494, -0.01055289, -0.01592071, -0.00461461,\n", + " 0.00325724, 0.0006841 , -0.0026403 , 0.01686813, 0.00419823,\n", + " -0.00299166, -0.01346661, -0.00564638, 0.00023601, 0.0108557 ,\n", + " 0.00666074, 0.01288243, -0.0186079 , 0.00200332, -0.01379452,\n", + " -0.01773283, 0.01165453, 0.01014983, 0.00133958, -0.00165299,\n", + " 0.01635987, 0.01019435, 0.0065489 , -0.00037118, -0.00397999,\n", + " 0.00284004, 0.00067637, 0.01627364, -0.0078299 , 0.01401774,\n", + " -0.01573489, 0.00123845, 0.01879427, -0.00331776, -0.01645748,\n", + " 0.01143804, 0.00324304, -0.01758391, 0.01165725, -0.00013819,\n", + " -0.00966081, -0.00401079, -0.01114468, 0.01449754, 0.01219868,\n", + " -0.00610742, -0.01529384, -0.00638515, 0.0115498 , 0.00018502,\n", + " -0.00630513, -0.01774968, 0.01080462, 0.00985269, 0.00329219,\n", + " -0.00888223, 0.00862635, -0.00417064, -0.01710633, -0.00388957,\n", + " -0.00096734, -0.00848989, -0.00846955, 0.00176544, -0.00913477,\n", + " -0.01676521, 0.01897634, -0.01933195, -0.02170036, 0.00382782,\n", + " 0.01387727, 0.00211742, 0.00045132, -0.00158703, -0.01627019,\n", + " -0.01727021, -0.00617373, 0.01208867, 0.01809805, -0.0025409 ,\n", + " 0.00676317, -0.01816405, -0.01505812, 0.00675633, 0.00561204,\n", + " 0.00952907, 0.00523477, -0.00641702, -0.00888471, 0.01043094,\n", + " -0.02301796, -0.00906393, -0.02145688, -0.00183525, 0.0053666 ,\n", + " -0.00537894, 0.01132864, 0.00205557, 0.00015121, -0.00068008,\n", + " 0.00564605, 0.00402641, 0.01573794, -0.00125398, -0.01196605,\n", + " 0.02083901, 0.001355 , -0.00012601, -0.00906368, -0.00667637,\n", + " 0.00882754, 0.005579 , 0.0020642 , -0.02276072, -0.00197584,\n", + " -0.02187258, 0.00256457, 0.01032818, 0.00408343, -0.0038473 ,\n", + " 0.00340744, -0.00544693, -0.00072757, -0.00419619, -0.01125627,\n", + " -0.01129378, 0.01304334, -0.00127827, -0.01159526, -0.00451869,\n", + " 0.01206874, 0.0121213 , 0.01013233, -0.00771143, 0.00406143,\n", + " 0.00722241, 0.00493179, 0.00324648, 0.0012184 , 0.00451836,\n", + " 0.00074543, 0.002475 , -0.0102218 , -0.00616384, -0.00154909,\n", + " 0.00523963, 0.0078449 , 0.00397785, -0.00148335, 0.01821327,\n", + " -0.00392404, 0.01884935, -0.00098688, 0.02122537, 0.00128569,\n", + " -0.00076967, -0.01500759, -0.00990747, -0.00027101, 0.00445222,\n", + " -0.00705703, 0.004314 , -0.00305626, -0.00323131, 0.02635842,\n", + " 0.00116516, -0.0139756 , -0.01458777, -0.00274291, 0.00092935,\n", + " 0.0023365 , 0.01803236, 0.00729533, -0.01116792, -0.00257845,\n", + " 0.00634304, 0.00542783, -0.00388852, -0.01765396, -0.00594904,\n", + " -0.00408894, -0.00422952, -0.01045187, 0.00503171, 0.00448202,\n", + " -0.00288425, 0.01044372, -0.00535529, -0.00441791, 0.00134603,\n", + " 0.00075145, -0.00795183, 0.00750755, 0.0156558 , 0.00568503,\n", + " -0.0044635 , 0.00360375, -0.00843096, 0.01188838, -0.02269647,\n", + " -0.0022597 , -0.02019465, 0.00060354, 0.01265363, 0.00845452,\n", + " 0.00879921, 0.00807706, 0.01438782, 0.00859368, 0.00261331,\n", + " -0.02067412, -0.01014025, 0.00590534, 0.01850619, -0.01428847,\n", + " -0.0206866 , -0.00263355, 0.0032291 , -0.01321626, 0.01776856,\n", + " -0.01769381, -0.00120978, 0.0125807 , -0.00196308, -0.00033226,\n", + " 0.00408021, 0.00070809, -0.01705818, -0.00365081, 0.01320496,\n", + " 0.00047279, 0.02105755, -0.01047429, -0.00052492, -0.01004261,\n", + " 0.00748653, 0.00048579, -0.01150925, 0.00665161, -0.01538913,\n", + " 0.01217959], dtype=float32), 'kernel': Array([[ 0.05015327, -0.03041644, -0.03295113, ..., 0.04861935,\n", + " 0.04714565, -0.00957593],\n", + " [-0.07399511, 0.13526867, -0.09466655, ..., -0.06253374,\n", + " -0.08226889, -0.1233295 ],\n", + " [-0.26905268, 0.11402693, 0.00223675, ..., -0.14415021,\n", + " 0.09161562, 0.03695163],\n", + " ...,\n", + " [ 0.04454683, 0.0817698 , -0.10512208, ..., -0.15175393,\n", + " 0.06272368, -0.07536803],\n", + " [-0.14763582, -0.20567636, -0.06718701, ..., 0.06949499,\n", + " -0.09244625, -0.01560377],\n", + " [ 0.17783955, -0.02452615, -0.08177213, ..., 0.11738463,\n", + " 0.05239161, 0.10596669]], dtype=float32)}, 'Dense_5': {'bias': Array([0.0035811], dtype=float32), 'kernel': Array([[ 2.70820758e-03],\n", + " [ 1.22034714e-01],\n", + " [ 2.82488875e-02],\n", + " [ 1.10054217e-01],\n", + " [ 2.88612861e-03],\n", + " [ 1.00878350e-01],\n", + " [-9.54338387e-02],\n", + " [-5.38817490e-04],\n", + " [-7.46907145e-02],\n", + " [ 1.01497479e-01],\n", + " [-2.39448133e-03],\n", + " [-5.07097691e-02],\n", + " [ 2.44271546e-03],\n", + " [-5.56524359e-02],\n", + " [-2.00833529e-04],\n", + " [-2.44829943e-03],\n", + " [ 5.20256609e-02],\n", + " [ 4.98748720e-02],\n", + " [-4.17549051e-02],\n", + " [ 7.25976154e-02],\n", + " [ 1.93015598e-02],\n", + " [ 3.05791646e-02],\n", + " [-3.93271074e-02],\n", + " [-2.87619396e-03],\n", + " [-7.26844668e-02],\n", + " [ 4.09820164e-03],\n", + " [-2.47842744e-02],\n", + " [ 2.37656245e-03],\n", + " [-2.64878366e-02],\n", + " [ 1.08805992e-01],\n", + " [ 7.18410909e-02],\n", + " [ 8.89061242e-02],\n", + " [ 3.34629901e-02],\n", + " [-5.00766002e-02],\n", + " [-7.41747692e-02],\n", + " [ 5.26316576e-02],\n", + " [ 2.47310214e-02],\n", + " [ 2.40735384e-03],\n", + " [ 2.28020432e-03],\n", + " [ 1.67884994e-02],\n", + " [ 4.98004928e-02],\n", + " [-2.18698755e-02],\n", + " [ 4.03015651e-02],\n", + " [-3.65091823e-02],\n", + " [-3.11408862e-02],\n", + " [ 2.81865820e-02],\n", + " [-4.60361242e-02],\n", + " [-1.43801376e-01],\n", + " [-2.95780157e-03],\n", + " [-3.06024291e-02],\n", + " [ 5.84711097e-02],\n", + " [-6.26730844e-02],\n", + " [-5.78055270e-02],\n", + " [-3.30861844e-02],\n", + " [ 6.33575916e-02],\n", + " [ 2.36342289e-03],\n", + " [-1.09188415e-01],\n", + " [ 2.74285823e-02],\n", + " [ 9.99198481e-02],\n", + " [-2.37722322e-02],\n", + " [ 1.54693760e-02],\n", + " [ 3.51395545e-04],\n", + " [-2.49036914e-03],\n", + " [-2.43238173e-03],\n", + " [-2.71108770e-03],\n", + " [-2.81512993e-03],\n", + " [ 7.52533004e-02],\n", + " [-2.30673188e-03],\n", + " [-2.68176068e-02],\n", + " [ 2.34641554e-03],\n", + " [ 6.34463802e-02],\n", + " [-2.49406579e-03],\n", + " [-1.26433372e-01],\n", + " [ 4.81917374e-02],\n", + " [ 2.71963212e-03],\n", + " [ 6.83157369e-02],\n", + " [ 2.66032293e-03],\n", + " [-2.59765540e-03],\n", + " [-1.02545246e-01],\n", + " [-2.67983172e-02],\n", + " [-1.20402068e-01],\n", + " [ 8.17263685e-03],\n", + " [ 3.80963385e-02],\n", + " [-1.16418108e-01],\n", + " [ 6.50902614e-02],\n", + " [ 3.16597298e-02],\n", + " [-1.00745645e-03],\n", + " [ 2.42591905e-03],\n", + " [ 8.84267762e-02],\n", + " [-1.23421356e-01],\n", + " [-8.15720484e-02],\n", + " [-2.29836116e-03],\n", + " [-4.96123694e-02],\n", + " [-9.97869819e-02],\n", + " [ 1.05882205e-01],\n", + " [ 2.38090963e-03],\n", + " [-1.48605332e-01],\n", + " [-6.92568123e-02],\n", + " [-2.40378664e-03],\n", + " [ 2.89715417e-02],\n", + " [ 4.53399196e-02],\n", + " [-2.35003955e-03],\n", + " [ 2.71559134e-03],\n", + " [ 3.02665378e-03],\n", + " [-2.77516176e-03],\n", + " [ 9.67895836e-02],\n", + " [-2.70048380e-02],\n", + " [ 9.07058567e-02],\n", + " [ 6.72250465e-02],\n", + " [ 1.13155596e-01],\n", + " [-6.28051311e-02],\n", + " [ 3.88142392e-02],\n", + " [-2.95326184e-03],\n", + " [ 3.01691471e-03],\n", + " [-4.91913222e-02],\n", + " [-4.36057560e-02],\n", + " [-2.31865980e-03],\n", + " [-1.59075726e-02],\n", + " [-2.36599031e-03],\n", + " [-5.73587306e-02],\n", + " [-6.69424832e-02],\n", + " [-2.31792103e-03],\n", + " [-1.25058293e-02],\n", + " [ 2.36691907e-03],\n", + " [-3.62988077e-02],\n", + " [-7.64152908e-04],\n", + " [-8.65145922e-02],\n", + " [ 6.99884892e-02],\n", + " [-8.82585496e-02],\n", + " [ 6.66770502e-04],\n", + " [-3.02969292e-02],\n", + " [ 5.61803654e-02],\n", + " [-2.67857849e-03],\n", + " [-6.41904399e-02],\n", + " [-2.35566939e-03],\n", + " [-8.04408118e-02],\n", + " [-2.41194596e-03],\n", + " [-2.95872707e-02],\n", + " [ 2.12724190e-04],\n", + " [ 2.06517652e-02],\n", + " [-2.09865719e-02],\n", + " [ 4.59476970e-02],\n", + " [-9.11728367e-02],\n", + " [-2.42488831e-03],\n", + " [ 2.31604790e-03],\n", + " [ 3.38213146e-02],\n", + " [ 2.74549965e-02],\n", + " [-2.82595307e-03],\n", + " [-1.53586334e-02],\n", + " [-7.60907307e-02],\n", + " [ 5.70272142e-03],\n", + " [-2.08155867e-02],\n", + " [-1.59730781e-02],\n", + " [ 6.83187842e-02],\n", + " [ 2.33819662e-03],\n", + " [-2.44431687e-03],\n", + " [ 2.39364873e-03],\n", + " [ 1.71745918e-03],\n", + " [-1.17861658e-01],\n", + " [ 1.97865870e-02],\n", + " [-3.71903144e-02],\n", + " [-6.03415482e-02],\n", + " [-2.15047058e-02],\n", + " [ 1.00045927e-01],\n", + " [ 2.44046352e-03],\n", + " [ 3.01187718e-03],\n", + " [-3.96093018e-02],\n", + " [ 2.33602617e-03],\n", + " [-6.39655069e-02],\n", + " [-2.43620761e-02],\n", + " [-2.40100385e-03],\n", + " [ 4.40300182e-02],\n", + " [-6.22994006e-02],\n", + " [ 7.52893537e-02],\n", + " [-2.30918289e-03],\n", + " [ 6.02973588e-02],\n", + " [-1.56369805e-02],\n", + " [ 9.72255543e-02],\n", + " [-8.82886425e-02],\n", + " [ 3.21217142e-02],\n", + " [-2.34361761e-03],\n", + " [-9.81356762e-03],\n", + " [-8.47740620e-02],\n", + " [ 1.00507773e-01],\n", + " [ 2.28485861e-03],\n", + " [ 2.35938188e-03],\n", + " [ 5.41816652e-02],\n", + " [-5.46566024e-02],\n", + " [ 2.98515125e-03],\n", + " [ 2.35056761e-03],\n", + " [-1.48700150e-02],\n", + " [ 6.53233454e-02],\n", + " [-2.33025430e-03],\n", + " [ 5.09883724e-02],\n", + " [-3.09496559e-02],\n", + " [ 2.34453776e-03],\n", + " [ 5.78722656e-02],\n", + " [-1.19768091e-01],\n", + " [ 1.38532463e-02],\n", + " [-3.57557237e-02],\n", + " [ 2.64675054e-03],\n", + " [-1.10347271e-01],\n", + " [ 6.65369555e-02],\n", + " [-4.75614406e-02],\n", + " [-4.35392261e-02],\n", + " [-1.53100304e-02],\n", + " [-5.26298769e-02],\n", + " [-2.31761602e-03],\n", + " [-4.11217138e-02],\n", + " [-4.69574817e-02],\n", + " [-2.68430705e-03],\n", + " [-9.60814878e-02],\n", + " [ 3.27776968e-02],\n", + " [-9.18827392e-03],\n", + " [ 6.88406900e-02],\n", + " [ 1.17949797e-02],\n", + " [ 1.17942311e-01],\n", + " [-2.40394427e-03],\n", + " [ 7.12209269e-02],\n", + " [ 1.25106133e-03],\n", + " [-2.73891375e-03],\n", + " [ 4.17114347e-02],\n", + " [-3.16448919e-02],\n", + " [ 6.02545952e-05],\n", + " [-2.49948353e-03],\n", + " [ 9.78429690e-02],\n", + " [ 4.67571802e-02],\n", + " [-2.43049371e-03],\n", + " [-4.36916687e-02],\n", + " [ 2.45240722e-02],\n", + " [-6.00878224e-02],\n", + " [ 2.69257021e-03],\n", + " [-7.64835509e-04],\n", + " [ 6.07808419e-02],\n", + " [-7.81335831e-02],\n", + " [-8.25816095e-02],\n", + " [-2.52528535e-03],\n", + " [ 2.51530146e-04],\n", + " [-2.75908224e-03],\n", + " [-4.23399769e-02],\n", + " [-6.91752136e-02],\n", + " [-6.54896572e-02],\n", + " [-9.98401418e-02],\n", + " [ 4.79688719e-02],\n", + " [ 9.98601913e-02],\n", + " [ 5.44612110e-02],\n", + " [ 1.55163601e-01],\n", + " [ 2.63293390e-03],\n", + " [ 9.95034003e-04],\n", + " [-4.56284322e-02],\n", + " [ 2.65756017e-03],\n", + " [-2.34728237e-03],\n", + " [ 1.75290972e-01],\n", + " [ 1.01475142e-01],\n", + " [-1.21499598e-01],\n", + " [-2.89544137e-03]], dtype=float32)}, 'log_std': Array([-4.318472], dtype=float32)}}, tx=GradientTransformationExtraArgs(init=.init_fn at 0x796c5c2a2840>, update=.update_fn at 0x796c5c2a2a20>), opt_state=(EmptyState(), (ScaleByAdamState(count=Array(1874688, dtype=int32), mu={'params': {'Dense_0': {'bias': Array([ 6.e-45, 0.e+00, 0.e+00, 0.e+00, -6.e-45, 0.e+00, -6.e-45,\n", + " 6.e-45, -6.e-45, 0.e+00, -6.e-45, 0.e+00, 0.e+00, -6.e-45,\n", + " 0.e+00, 0.e+00, -6.e-45, 6.e-45, 0.e+00, 6.e-45, 0.e+00,\n", + " 6.e-45, 0.e+00, -6.e-45, 0.e+00, 6.e-45, -6.e-45, 6.e-45,\n", + " 0.e+00, 0.e+00, 0.e+00, 0.e+00, 0.e+00, 0.e+00, 0.e+00,\n", + " 6.e-45, 6.e-45, 0.e+00, 6.e-45, 0.e+00, 0.e+00, 0.e+00,\n", + " 0.e+00, 0.e+00, 0.e+00, 6.e-45, 0.e+00, 0.e+00, 0.e+00,\n", + " -6.e-45, 0.e+00, 0.e+00, 0.e+00, 0.e+00, 6.e-45, -6.e-45,\n", + " 6.e-45, 0.e+00, 0.e+00, 6.e-45, -6.e-45, 6.e-45, -6.e-45,\n", + " 6.e-45, 0.e+00, 0.e+00, 0.e+00, -6.e-45, 6.e-45, 0.e+00,\n", + " 0.e+00, 0.e+00, -6.e-45, 0.e+00, -6.e-45, 0.e+00, 0.e+00,\n", + " 6.e-45, -6.e-45, -6.e-45, 6.e-45, -6.e-45, 0.e+00, -6.e-45,\n", + " 6.e-45, 0.e+00, 0.e+00, 0.e+00, 6.e-45, 0.e+00, 6.e-45,\n", + " 0.e+00, 0.e+00, 0.e+00, 0.e+00, -6.e-45, 0.e+00, 0.e+00,\n", + " -6.e-45, 6.e-45, -6.e-45, 6.e-45, 0.e+00, 0.e+00, -6.e-45,\n", + " 0.e+00, 0.e+00, 0.e+00, 6.e-45, 0.e+00, 0.e+00, 0.e+00,\n", + " 0.e+00, 0.e+00, 0.e+00, 6.e-45, -6.e-45, -6.e-45, 6.e-45,\n", + " -6.e-45, 0.e+00, 6.e-45, -6.e-45, 0.e+00, -6.e-45, 0.e+00,\n", + " 0.e+00, 6.e-45, 0.e+00, 0.e+00, 6.e-45, 6.e-45, -6.e-45,\n", + " -6.e-45, -6.e-45, -6.e-45, 0.e+00, 6.e-45, 6.e-45, 6.e-45,\n", + " 0.e+00, -6.e-45, -6.e-45, -6.e-45, -6.e-45, 0.e+00, 0.e+00,\n", + " 0.e+00, 6.e-45, 0.e+00, 0.e+00, 6.e-45, 6.e-45, -6.e-45,\n", + " -6.e-45, 0.e+00, 6.e-45, 0.e+00, 0.e+00, 0.e+00, 0.e+00,\n", + " 0.e+00, 0.e+00, 0.e+00, 0.e+00, -6.e-45, -6.e-45, 0.e+00,\n", + " 0.e+00, 0.e+00, 6.e-45, 0.e+00, 0.e+00, -6.e-45, 0.e+00,\n", + " 0.e+00, 0.e+00, 0.e+00, 0.e+00, -6.e-45, 6.e-45, 0.e+00,\n", + " -6.e-45, 0.e+00, -6.e-45, 0.e+00, 0.e+00, 0.e+00, -6.e-45,\n", + " 6.e-45, 0.e+00, 0.e+00, 0.e+00, 6.e-45, -6.e-45, 6.e-45,\n", + " -6.e-45, 6.e-45, -6.e-45, -6.e-45, 6.e-45, 0.e+00, -6.e-45,\n", + " 0.e+00, -6.e-45, 0.e+00, -6.e-45, 6.e-45, 0.e+00, 0.e+00,\n", + " 6.e-45, -6.e-45, -6.e-45, 0.e+00, 6.e-45, 6.e-45, 0.e+00,\n", + " 6.e-45, 6.e-45, 0.e+00, 6.e-45, -6.e-45, -6.e-45, 6.e-45,\n", + " 0.e+00, 0.e+00, 6.e-45, 0.e+00, -6.e-45, -6.e-45, 6.e-45,\n", + " -6.e-45, 0.e+00, 0.e+00, 0.e+00, 0.e+00, 0.e+00, 0.e+00,\n", + " 0.e+00, 0.e+00, 0.e+00, 0.e+00, 0.e+00, 0.e+00, 0.e+00,\n", + " 0.e+00, 0.e+00, 0.e+00, 0.e+00, 0.e+00, -6.e-45, 0.e+00,\n", + " 0.e+00, 6.e-45, 0.e+00, 0.e+00], dtype=float32), 'kernel': Array([[-6.e-45, 0.e+00, 0.e+00, ..., 6.e-45, 0.e+00, 0.e+00],\n", + " [ 6.e-45, 0.e+00, 0.e+00, ..., 6.e-45, 0.e+00, 0.e+00],\n", + " [-6.e-45, 0.e+00, 0.e+00, ..., -6.e-45, 0.e+00, 0.e+00],\n", + " ...,\n", + " [ 6.e-45, 0.e+00, 0.e+00, ..., 6.e-45, 0.e+00, 0.e+00],\n", + " [ 6.e-45, 0.e+00, 0.e+00, ..., 6.e-45, 0.e+00, 0.e+00],\n", + " [ 6.e-45, 0.e+00, 0.e+00, ..., 6.e-45, 0.e+00, 0.e+00]], dtype=float32)}, 'Dense_1': {'bias': Array([ 1.03541664e-08, -2.25503545e-08, 1.09754401e-08, 2.26793979e-08,\n", + " -1.08045732e-08, -2.55994550e-08, -2.31924577e-08, 1.04921165e-08,\n", + " 2.39457876e-08, -2.38241267e-08, -9.84679271e-09, 2.40475533e-08,\n", + " 2.26209060e-08, -9.62008251e-09, -9.19785581e-09, 2.29656649e-08,\n", + " -2.29345041e-08, -2.31075994e-08, -2.23311361e-08, -2.37337971e-08,\n", + " -2.60625335e-08, -1.06845848e-08, 2.36961917e-08, 1.07474909e-08,\n", + " 2.26803287e-08, -2.44749021e-08, -1.99485872e-09, -2.37690188e-08,\n", + " -1.04099778e-08, 2.38905891e-08, -1.05689848e-08, -9.92427385e-09,\n", + " -2.47251606e-08, 2.39578704e-08, 2.53755612e-08, 9.96767202e-09,\n", + " -2.39543709e-08, -2.22608314e-08, 1.07252545e-08, 1.07611244e-08,\n", + " -2.35777389e-08, 2.34232349e-08, -2.33633735e-08, 1.01725632e-08,\n", + " -2.31241533e-08, 2.36809523e-08, 2.26009220e-08, -9.67539204e-09,\n", + " 2.40251659e-08, -2.41103955e-08, 9.48442302e-09, -2.45370337e-08,\n", + " -2.28745094e-08, -2.44033593e-08, -1.00123128e-08, 1.03061701e-08,\n", + " 2.28946764e-08, 2.24425740e-08, 2.29827357e-08, 2.25327419e-08,\n", + " -9.98236427e-09, -2.32662103e-08, -1.04185727e-08, -9.28724919e-09,\n", + " 2.14844054e-08, -2.32100703e-08, 2.35621869e-08, -2.37330582e-08,\n", + " 1.09662253e-08, -2.35303794e-08, 2.27551844e-08, 2.43550033e-08,\n", + " -9.78177273e-09, -2.48300562e-08, -2.40180125e-08, -2.29773569e-08,\n", + " 2.36496440e-08, 9.61151869e-09, -2.38107098e-08, -1.03533173e-08,\n", + " 9.93252947e-09, 2.21335767e-08, -1.06222959e-08, -2.35497382e-08,\n", + " 2.31541364e-08, 2.34322943e-08, 2.25955290e-08, -9.75304726e-09,\n", + " -1.13002869e-08, 2.26995169e-08, -2.36510296e-08, -2.32563746e-08,\n", + " -9.97940042e-09, -9.66210578e-09, -2.50858676e-08, 1.04648832e-08,\n", + " 1.07407310e-08, -2.29595720e-08, -9.87424720e-09, 2.33411086e-08,\n", + " -2.42061162e-08, -1.07653841e-08, 2.37902107e-08, -2.36694540e-08,\n", + " 2.39752147e-08, -1.02575592e-08, 2.30720971e-08, -9.81033033e-09,\n", + " -1.00816440e-08, 1.03168682e-08, -2.31944668e-08, 2.22103331e-08,\n", + " 1.12213536e-08, 2.35503119e-08, 2.34428619e-08, -2.37289548e-08,\n", + " -2.31170283e-08, -2.46132217e-08, -1.05486855e-08, 2.45457343e-08,\n", + " 8.75366180e-09, -1.06093188e-08, -2.31070718e-08, 2.34258337e-08,\n", + " -2.47319036e-08, 2.50391921e-08, -2.26980212e-08, 1.01784190e-08,\n", + " -2.39714133e-08, 2.40813680e-08, -9.70173275e-09, 2.38065390e-08,\n", + " -2.38319373e-08, 1.02583231e-08, 2.36317241e-08, 2.38402187e-08,\n", + " 9.70589653e-09, 2.37370621e-08, 2.50829686e-08, 2.45204337e-08,\n", + " -2.37158382e-08, -2.22815331e-08, -2.34159110e-08, 9.60534230e-09,\n", + " -2.38358666e-08, -1.00425810e-08, 2.32160442e-08, -1.05393099e-08,\n", + " -1.00508286e-08, -2.50335965e-08, -9.60210222e-09, -2.31566748e-08,\n", + " 2.35914435e-08, 2.21662475e-08, 2.30486279e-08, 1.07958336e-08,\n", + " 2.39797320e-08, 2.51510350e-08, -2.41677185e-08, 2.30657733e-08,\n", + " 2.26225900e-08, 1.09776863e-08, 2.32889921e-08, -2.39750744e-08,\n", + " -2.41031071e-08, 2.31204851e-08, 2.35466366e-08, 2.35021957e-08,\n", + " -2.36535733e-08, -1.00681845e-08, -9.25054877e-09, -2.26733974e-08,\n", + " -2.43106193e-08, -2.44533993e-08, 2.33814283e-08, -2.29009807e-08,\n", + " 9.55213242e-09, 2.32522748e-08, -1.05972617e-08, -2.22023715e-08,\n", + " 9.42938883e-09, -9.12538400e-09, -2.41387639e-08, -2.46479992e-08,\n", + " -2.40736604e-08, -2.44753497e-08, -2.34043007e-08, -1.05592610e-08,\n", + " 9.71530945e-09, 2.45881857e-08, 2.36805704e-08, -2.41897880e-08,\n", + " 1.04161861e-08, -1.10903278e-08, -1.01811679e-08, -2.27189627e-08,\n", + " -2.30083597e-08, -1.08696794e-08, -1.01657953e-08, 2.19482942e-07,\n", + " -2.39753302e-08, 2.40163764e-08, 1.03893054e-08, -9.03441766e-09,\n", + " -3.46478082e-08, -2.37281910e-08, -2.35396413e-08, -2.40696831e-08,\n", + " 2.33254003e-08, 2.38854891e-08, -2.38879565e-08, -9.85133752e-09,\n", + " 2.34631212e-08, 8.61502603e-09, -2.31104984e-08, -2.31271180e-08,\n", + " 2.36442457e-08, 1.15904211e-08, -1.06073630e-08, 2.39395028e-08,\n", + " 2.36320830e-08, 2.35764475e-08, -9.75345227e-09, -1.04897957e-08,\n", + " 2.49138310e-08, 2.32671695e-08, 2.36456934e-08, -1.04983400e-08,\n", + " -2.43422811e-08, -1.06175531e-08, -9.90782656e-09, 2.46703067e-08,\n", + " -2.40167761e-08, 2.36683544e-08, -1.02754383e-08, 2.26544152e-08,\n", + " 9.47397183e-09, 9.47846956e-09, 2.32374866e-08, 2.45816203e-08,\n", + " -1.06429781e-08, -2.35429987e-08, 2.28810286e-08, 2.35029525e-08,\n", + " -2.46803040e-08, -2.30708466e-08, -1.04155564e-08, 1.02375104e-08,\n", + " -2.34650717e-08, -2.39009310e-08, -3.48849483e-08, 2.34201085e-08,\n", + " 2.29198491e-08, 2.39195010e-08, -2.30075194e-08, 1.00511723e-08], dtype=float32), 'kernel': Array([[ 1.7525373e-08, -2.9312023e-08, 1.7678957e-08, ...,\n", + " 2.7208849e-08, -2.7127996e-08, 1.7897921e-08],\n", + " [ 1.0349746e-08, -2.2549184e-08, 1.0967796e-08, ...,\n", + " 2.3919497e-08, -2.2998300e-08, 1.0045669e-08],\n", + " [ 1.0349746e-08, -2.2549184e-08, 1.0967796e-08, ...,\n", + " 2.3919497e-08, -2.2998300e-08, 1.0045669e-08],\n", + " ...,\n", + " [-1.7525373e-08, 2.9312023e-08, -1.7678957e-08, ...,\n", + " -2.7208849e-08, 2.7127996e-08, -1.7897921e-08],\n", + " [-1.0349746e-08, 2.2549184e-08, -1.0967796e-08, ...,\n", + " -2.3919497e-08, 2.2998300e-08, -1.0045669e-08],\n", + " [-1.0349746e-08, 2.2549184e-08, -1.0967796e-08, ...,\n", + " -2.3919497e-08, 2.2998300e-08, -1.0045669e-08]], dtype=float32)}, 'Dense_2': {'bias': Array([-0.00109201], dtype=float32), 'kernel': Array([[-3.1509115e-03],\n", + " [-1.0919889e-03],\n", + " [-3.1509111e-03],\n", + " [ 1.0919888e-03],\n", + " [ 3.1509115e-03],\n", + " [-1.0919864e-03],\n", + " [-1.0919889e-03],\n", + " [-3.1509108e-03],\n", + " [ 1.0919885e-03],\n", + " [-1.0919881e-03],\n", + " [ 3.1509113e-03],\n", + " [ 1.0919885e-03],\n", + " [ 1.0919904e-03],\n", + " [ 3.1509104e-03],\n", + " [ 3.1509101e-03],\n", + " [ 1.0919895e-03],\n", + " [-1.0919895e-03],\n", + " [-1.0919897e-03],\n", + " [-1.0919897e-03],\n", + " [-1.0919878e-03],\n", + " [-1.0919870e-03],\n", + " [ 3.1509111e-03],\n", + " [ 1.0919881e-03],\n", + " [-3.1509092e-03],\n", + " [ 1.0919888e-03],\n", + " [-1.0919870e-03],\n", + " [-2.2210799e-07],\n", + " [-1.0919896e-03],\n", + " [ 3.1509118e-03],\n", + " [ 1.0919883e-03],\n", + " [ 3.1509104e-03],\n", + " [ 3.1509090e-03],\n", + " [-1.0919872e-03],\n", + " [ 1.0919881e-03],\n", + " [ 1.0919870e-03],\n", + " [-3.1509104e-03],\n", + " [-1.0919889e-03],\n", + " [-1.0919900e-03],\n", + " [-3.1509108e-03],\n", + " [-3.1509111e-03],\n", + " [-1.0919884e-03],\n", + " [ 1.0919883e-03],\n", + " [-1.0919888e-03],\n", + " [-3.1509113e-03],\n", + " [-1.0919895e-03],\n", + " [ 1.0919883e-03],\n", + " [ 1.0919897e-03],\n", + " [ 3.1509115e-03],\n", + " [ 1.0919882e-03],\n", + " [-1.0919881e-03],\n", + " [-3.1509097e-03],\n", + " [-1.0919867e-03],\n", + " [-1.0919886e-03],\n", + " [-1.0919882e-03],\n", + " [ 3.1509094e-03],\n", + " [-3.1509090e-03],\n", + " [ 1.0919892e-03],\n", + " [ 1.0919897e-03],\n", + " [ 1.0919898e-03],\n", + " [ 1.0919893e-03],\n", + " [ 3.1509106e-03],\n", + " [-1.0919895e-03],\n", + " [ 3.1509104e-03],\n", + " [ 3.1509111e-03],\n", + " [ 1.0919906e-03],\n", + " [-1.0919897e-03],\n", + " [ 1.0919895e-03],\n", + " [-1.0919889e-03],\n", + " [-3.1509111e-03],\n", + " [-1.0919890e-03],\n", + " [ 1.0919888e-03],\n", + " [ 1.0919878e-03],\n", + " [ 3.1509111e-03],\n", + " [-1.0919875e-03],\n", + " [-1.0919884e-03],\n", + " [-1.0919888e-03],\n", + " [ 1.0919885e-03],\n", + " [-3.1509113e-03],\n", + " [-1.0919882e-03],\n", + " [ 3.1509087e-03],\n", + " [-3.1509111e-03],\n", + " [ 1.0919902e-03],\n", + " [ 3.1509101e-03],\n", + " [-1.0919888e-03],\n", + " [ 1.0919885e-03],\n", + " [ 1.0919886e-03],\n", + " [ 1.0919897e-03],\n", + " [ 3.1509097e-03],\n", + " [ 2.6822573e-07],\n", + " [ 1.0919892e-03],\n", + " [-1.0919882e-03],\n", + " [-1.0919884e-03],\n", + " [ 3.1509104e-03],\n", + " [ 3.1509111e-03],\n", + " [-1.0919883e-03],\n", + " [-3.1509115e-03],\n", + " [-3.1509106e-03],\n", + " [-1.0919903e-03],\n", + " [ 3.1509115e-03],\n", + " [ 1.0919892e-03],\n", + " [-1.0919874e-03],\n", + " [ 3.1509111e-03],\n", + " [ 1.0919890e-03],\n", + " [-1.0919884e-03],\n", + " [ 1.0919883e-03],\n", + " [ 3.1509104e-03],\n", + " [ 1.0919896e-03],\n", + " [ 3.1509099e-03],\n", + " [ 3.1509115e-03],\n", + " [-3.1509111e-03],\n", + " [-1.0919893e-03],\n", + " [ 1.0919896e-03],\n", + " [-3.1509113e-03],\n", + " [ 1.0919886e-03],\n", + " [ 1.0919884e-03],\n", + " [-1.0919890e-03],\n", + " [-1.0919896e-03],\n", + " [-1.0919884e-03],\n", + " [ 3.1509111e-03],\n", + " [ 1.0919870e-03],\n", + " [-3.1509108e-03],\n", + " [ 3.1509106e-03],\n", + " [-1.0919893e-03],\n", + " [ 1.0919882e-03],\n", + " [-1.0919883e-03],\n", + " [ 1.0919871e-03],\n", + " [-1.0919899e-03],\n", + " [-3.1509111e-03],\n", + " [-1.0919888e-03],\n", + " [ 1.0919874e-03],\n", + " [ 1.0918307e-03],\n", + " [ 1.0919888e-03],\n", + " [-1.0919881e-03],\n", + " [-3.1509115e-03],\n", + " [ 1.0919896e-03],\n", + " [ 1.0919888e-03],\n", + " [-3.1509108e-03],\n", + " [ 1.0919877e-03],\n", + " [ 1.0919876e-03],\n", + " [ 1.0919874e-03],\n", + " [-1.0919891e-03],\n", + " [-1.0919896e-03],\n", + " [-1.0919885e-03],\n", + " [-3.1509115e-03],\n", + " [-1.0919883e-03],\n", + " [ 3.1509106e-03],\n", + " [ 1.0919888e-03],\n", + " [ 3.1509122e-03],\n", + " [ 3.1509111e-03],\n", + " [-1.0919870e-03],\n", + " [ 3.1509111e-03],\n", + " [-1.0919883e-03],\n", + " [ 1.0919889e-03],\n", + " [ 1.0919902e-03],\n", + " [ 1.0919889e-03],\n", + " [-3.1509106e-03],\n", + " [ 1.0919881e-03],\n", + " [ 1.0919869e-03],\n", + " [-1.0919878e-03],\n", + " [ 1.0919889e-03],\n", + " [ 1.0919891e-03],\n", + " [-3.1509099e-03],\n", + " [ 1.0919886e-03],\n", + " [-1.0919890e-03],\n", + " [-1.0919885e-03],\n", + " [ 1.0919881e-03],\n", + " [ 1.0919898e-03],\n", + " [ 1.0919888e-03],\n", + " [-1.0919883e-03],\n", + " [ 3.1509106e-03],\n", + " [ 3.1509090e-03],\n", + " [-1.0919892e-03],\n", + " [-1.0919878e-03],\n", + " [-1.0919874e-03],\n", + " [ 1.0919888e-03],\n", + " [-1.0919897e-03],\n", + " [-3.1509113e-03],\n", + " [ 1.0919884e-03],\n", + " [ 3.1509115e-03],\n", + " [-1.0919906e-03],\n", + " [-3.1509111e-03],\n", + " [ 3.1509113e-03],\n", + " [-1.0919881e-03],\n", + " [-1.0919871e-03],\n", + " [-1.0919885e-03],\n", + " [-1.0919875e-03],\n", + " [-1.0919893e-03],\n", + " [ 3.1509101e-03],\n", + " [-3.1509118e-03],\n", + " [ 1.0919875e-03],\n", + " [ 1.0919888e-03],\n", + " [-1.0919884e-03],\n", + " [-3.1509111e-03],\n", + " [ 3.1509078e-03],\n", + " [ 3.1509118e-03],\n", + " [-1.0919890e-03],\n", + " [-1.0919898e-03],\n", + " [ 3.1509111e-03],\n", + " [ 3.1509111e-03],\n", + " [-1.0640444e-03],\n", + " [-1.0919881e-03],\n", + " [ 1.0919881e-03],\n", + " [-3.1509101e-03],\n", + " [ 3.1509104e-03],\n", + " [-1.0918580e-03],\n", + " [-1.0919890e-03],\n", + " [-1.0919884e-03],\n", + " [-1.0919890e-03],\n", + " [ 1.0919895e-03],\n", + " [ 1.0919879e-03],\n", + " [-1.0919888e-03],\n", + " [ 3.1509113e-03],\n", + " [ 1.0919889e-03],\n", + " [-3.1509094e-03],\n", + " [-1.0919895e-03],\n", + " [-1.0919893e-03],\n", + " [ 1.0919888e-03],\n", + " [-3.1509087e-03],\n", + " [ 3.1509111e-03],\n", + " [ 1.0919888e-03],\n", + " [ 1.0919885e-03],\n", + " [ 1.0919881e-03],\n", + " [ 3.1509111e-03],\n", + " [ 1.0918273e-03],\n", + " [ 1.0919869e-03],\n", + " [ 1.0919883e-03],\n", + " [ 1.0919891e-03],\n", + " [ 3.1509111e-03],\n", + " [-1.0919883e-03],\n", + " [ 3.1509108e-03],\n", + " [ 3.1509111e-03],\n", + " [ 1.0919874e-03],\n", + " [-1.0919878e-03],\n", + " [ 1.0919883e-03],\n", + " [ 3.1509115e-03],\n", + " [ 1.0919895e-03],\n", + " [-3.1509094e-03],\n", + " [-3.1509111e-03],\n", + " [ 1.0919889e-03],\n", + " [ 1.0919872e-03],\n", + " [ 3.1509115e-03],\n", + " [-1.0919891e-03],\n", + " [ 1.0919905e-03],\n", + " [ 1.0919883e-03],\n", + " [-1.0919881e-03],\n", + " [-1.0919878e-03],\n", + " [ 3.1509104e-03],\n", + " [-3.1509122e-03],\n", + " [-1.0919892e-03],\n", + " [-1.0919896e-03],\n", + " [-1.0918557e-03],\n", + " [ 1.0919878e-03],\n", + " [ 1.0919895e-03],\n", + " [ 1.0919883e-03],\n", + " [-1.0919893e-03],\n", + " [-3.1509108e-03]], dtype=float32)}, 'Dense_3': {'bias': Array([ 0.000000e+00, 0.000000e+00, -5.605194e-45, 0.000000e+00,\n", + " -5.605194e-45, 0.000000e+00, 5.605194e-45, 0.000000e+00,\n", + " -5.605194e-45, 0.000000e+00, 0.000000e+00, 0.000000e+00,\n", + " 5.605194e-45, 0.000000e+00, -5.605194e-45, 5.605194e-45,\n", + " 5.605194e-45, -5.605194e-45, 5.605194e-45, 5.605194e-45,\n", + " 0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,\n", + " 5.605194e-45, 5.605194e-45, 5.605194e-45, 0.000000e+00,\n", + " 0.000000e+00, 0.000000e+00, -5.605194e-45, -5.605194e-45,\n", + " 0.000000e+00, 5.605194e-45, 0.000000e+00, 0.000000e+00,\n", + " 5.605194e-45, 0.000000e+00, 0.000000e+00, -5.605194e-45,\n", + " 5.605194e-45, 5.605194e-45, 0.000000e+00, 0.000000e+00,\n", + " 0.000000e+00, 0.000000e+00, 0.000000e+00, -5.605194e-45,\n", + " 5.605194e-45, 0.000000e+00, -5.605194e-45, -5.605194e-45,\n", + " 0.000000e+00, -5.605194e-45, -5.605194e-45, 0.000000e+00,\n", + " 0.000000e+00, 5.605194e-45, 5.605194e-45, 0.000000e+00,\n", + " 0.000000e+00, 5.605194e-45, 0.000000e+00, 0.000000e+00,\n", + " 0.000000e+00, 0.000000e+00, 5.605194e-45, 0.000000e+00,\n", + " 0.000000e+00, 0.000000e+00, 0.000000e+00, 5.605194e-45,\n", + " 5.605194e-45, -5.605194e-45, 0.000000e+00, -5.605194e-45,\n", + " 5.605194e-45, -5.605194e-45, 0.000000e+00, 0.000000e+00,\n", + " -5.605194e-45, -5.605194e-45, 5.605194e-45, 0.000000e+00,\n", + " 0.000000e+00, 0.000000e+00, -5.605194e-45, 0.000000e+00,\n", + " 0.000000e+00, 0.000000e+00, 0.000000e+00, -5.605194e-45,\n", + " 0.000000e+00, 0.000000e+00, -5.605194e-45, 0.000000e+00,\n", + " 0.000000e+00, 5.605194e-45, -5.605194e-45, 0.000000e+00,\n", + " -5.605194e-45, -5.605194e-45, 0.000000e+00, -5.605194e-45,\n", + " 0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,\n", + " 0.000000e+00, 5.605194e-45, -5.605194e-45, 5.605194e-45,\n", + " 5.605194e-45, 0.000000e+00, 0.000000e+00, 5.605194e-45,\n", + " 0.000000e+00, 5.605194e-45, 0.000000e+00, 0.000000e+00,\n", + " 0.000000e+00, -5.605194e-45, -5.605194e-45, 0.000000e+00,\n", + " 0.000000e+00, 0.000000e+00, 0.000000e+00, 5.605194e-45,\n", + " 0.000000e+00, 5.605194e-45, 0.000000e+00, 0.000000e+00,\n", + " 0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,\n", + " 0.000000e+00, 5.605194e-45, 5.605194e-45, 5.605194e-45,\n", + " 0.000000e+00, 5.605194e-45, 0.000000e+00, -5.605194e-45,\n", + " -5.605194e-45, 0.000000e+00, 0.000000e+00, 0.000000e+00,\n", + " 5.157024e-20, 0.000000e+00, 0.000000e+00, 5.605194e-45,\n", + " 0.000000e+00, 0.000000e+00, 0.000000e+00, 5.605194e-45,\n", + " 0.000000e+00, 0.000000e+00, 0.000000e+00, -5.605194e-45,\n", + " 0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,\n", + " -5.605194e-45, 0.000000e+00, 5.605194e-45, 5.605194e-45,\n", + " 0.000000e+00, -5.605194e-45, -5.605194e-45, -5.605194e-45,\n", + " 0.000000e+00, -5.605194e-45, -5.605194e-45, 0.000000e+00,\n", + " 0.000000e+00, 0.000000e+00, 0.000000e+00, -5.605194e-45,\n", + " 5.605194e-45, 0.000000e+00, 0.000000e+00, 0.000000e+00,\n", + " 5.605194e-45, 0.000000e+00, 5.605194e-45, -5.605194e-45,\n", + " 0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,\n", + " 5.605194e-45, 0.000000e+00, 5.605194e-45, 0.000000e+00,\n", + " 0.000000e+00, 5.605194e-45, 0.000000e+00, 5.605194e-45,\n", + " 5.605194e-45, 0.000000e+00, 0.000000e+00, 5.605194e-45,\n", + " 0.000000e+00, 0.000000e+00, 0.000000e+00, 5.605194e-45,\n", + " 0.000000e+00, -5.605194e-45, 0.000000e+00, 5.605194e-45,\n", + " -5.605194e-45, 0.000000e+00, 0.000000e+00, 5.605194e-45,\n", + " 5.605194e-45, 5.605194e-45, -5.605194e-45, 0.000000e+00,\n", + " 0.000000e+00, -5.605194e-45, 0.000000e+00, 0.000000e+00,\n", + " -5.605194e-45, 0.000000e+00, 5.605194e-45, 5.605194e-45,\n", + " 5.605194e-45, -5.605194e-45, -5.605194e-45, 0.000000e+00,\n", + " -5.605194e-45, 0.000000e+00, 0.000000e+00, -5.605194e-45,\n", + " -5.605194e-45, -5.605194e-45, 0.000000e+00, 5.605194e-45,\n", + " 0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,\n", + " 0.000000e+00, 0.000000e+00, 0.000000e+00, 5.605194e-45,\n", + " 0.000000e+00, -5.605194e-45, -5.605194e-45, 0.000000e+00,\n", + " -5.605194e-45, -5.605194e-45, 0.000000e+00, 5.605194e-45], dtype=float32), 'kernel': Array([[ 0.e+00, 0.e+00, -6.e-45, ..., -6.e-45, 0.e+00, 0.e+00],\n", + " [ 0.e+00, 0.e+00, -6.e-45, ..., -6.e-45, 0.e+00, 0.e+00],\n", + " [ 0.e+00, 0.e+00, 6.e-45, ..., 6.e-45, 0.e+00, 0.e+00],\n", + " ...,\n", + " [ 0.e+00, 0.e+00, 6.e-45, ..., -6.e-45, 0.e+00, 6.e-45],\n", + " [ 0.e+00, 0.e+00, 6.e-45, ..., -6.e-45, 0.e+00, 6.e-45],\n", + " [ 0.e+00, 0.e+00, 6.e-45, ..., -6.e-45, 0.e+00, 6.e-45]], dtype=float32)}, 'Dense_4': {'bias': Array([-1.68129989e-11, -1.64340308e-12, -2.34251160e-12, -1.37558454e-12,\n", + " -1.79203891e-11, -1.41104142e-12, 1.81627803e-12, 3.31597380e-12,\n", + " 1.50421516e-12, -1.34772467e-12, 1.48133103e-11, 2.25605385e-12,\n", + " -1.51168019e-11, 2.41827153e-12, 1.38962897e-13, 1.51490053e-11,\n", + " -2.24216240e-12, -2.28625213e-12, 2.08522132e-12, -1.51741174e-12,\n", + " -2.21023330e-12, -2.64996055e-12, 3.19520842e-12, 1.78426666e-11,\n", + " 1.44679277e-12, -1.13836562e-11, 3.19035120e-12, -1.47042517e-11,\n", + " 2.45127920e-12, -1.34369816e-12, -1.52956304e-12, -1.85866578e-12,\n", + " -3.77237044e-12, 2.25462010e-12, 1.95828900e-12, -2.53433646e-12,\n", + " -5.05996200e-12, -1.48946688e-11, -1.41007874e-11, -4.52248005e-12,\n", + " -2.27083391e-12, 5.71240971e-12, -3.14632218e-12, 3.46597013e-12,\n", + " 2.59889853e-12, -2.84088400e-12, 2.35422771e-12, 1.23597129e-12,\n", + " 1.83519710e-11, 4.27911699e-12, -2.17041576e-12, 2.06006284e-12,\n", + " 2.36856802e-12, 2.49788818e-12, -2.06548710e-12, -1.46176300e-11,\n", + " 1.39185766e-12, -2.90129466e-12, -1.74847972e-12, 3.31357359e-12,\n", + " -4.87309065e-12, -1.75373301e-12, 1.54117448e-11, 1.50523118e-11,\n", + " 1.68066915e-11, 1.74826213e-11, -1.49217498e-12, 1.42708501e-11,\n", + " 2.40834284e-12, -1.45175174e-11, -1.56042952e-12, 1.54368792e-11,\n", + " 1.71726088e-12, -1.88417164e-12, -1.68614255e-11, -2.01780953e-12,\n", + " -1.64797481e-11, 1.60863510e-11, 1.40228150e-12, 4.78764597e-12,\n", + " 1.65638846e-12, -1.27146930e-11, -2.21389378e-12, 1.67825953e-12,\n", + " -2.20410972e-12, -3.98355524e-12, 6.19858765e-12, -1.50218865e-11,\n", + " -1.44667232e-12, 1.62698934e-12, 1.94670994e-12, 1.42166105e-11,\n", + " 1.92292298e-12, 1.36910600e-12, -1.39693411e-12, -1.47305536e-11,\n", + " 1.56167332e-12, 2.09429696e-12, 1.48727905e-11, -2.31150342e-12,\n", + " -2.81368939e-12, 1.45516429e-11, -1.68236206e-11, -1.88008775e-11,\n", + " 1.72134633e-11, -1.38423095e-12, 4.64812477e-12, -1.40174417e-12,\n", + " -2.02521160e-12, -1.38138654e-12, 2.07544984e-12, -3.25466975e-12,\n", + " 1.83271089e-11, -1.87430939e-11, 2.28413989e-12, 2.43149598e-12,\n", + " 1.43394038e-11, 4.75405045e-12, 1.46364275e-11, 2.35601903e-12,\n", + " 1.57300290e-12, 1.43365944e-11, 3.95476577e-12, -1.46400114e-11,\n", + " 2.00708374e-12, 4.72600865e-12, 1.45893584e-12, -1.99434263e-12,\n", + " 1.86900365e-12, -4.11698974e-12, 4.30520116e-12, -2.16828096e-12,\n", + " 1.65977510e-11, 1.58451019e-12, 1.45723572e-11, 1.91609771e-12,\n", + " 1.49197252e-11, 3.06157443e-12, -9.51521739e-13, -3.74928170e-12,\n", + " 5.89098518e-12, -2.77128147e-12, 1.44175232e-12, 1.50046590e-11,\n", + " -1.43273570e-11, -3.72674157e-12, -4.56247887e-12, 1.75198363e-11,\n", + " 2.90088613e-12, 1.92680854e-12, -3.59396280e-12, 2.90502583e-12,\n", + " 4.78136064e-12, -2.01775554e-12, -1.44605612e-11, 1.51254807e-11,\n", + " -1.48092580e-11, -1.06114145e-11, 1.73609868e-12, -4.02823087e-12,\n", + " 2.59967915e-12, 2.09750620e-12, 5.69491676e-12, -1.40408518e-12,\n", + " -1.51003862e-11, -1.87129860e-11, 2.14852528e-12, -1.44508615e-11,\n", + " 1.61768147e-12, 3.47535325e-12, 1.48539236e-11, -1.99395101e-12,\n", + " 2.22019104e-12, -1.50999309e-12, 1.42853993e-11, -1.62137686e-12,\n", + " 4.72423229e-12, -1.41986249e-12, 1.87225604e-12, -4.11534782e-12,\n", + " 1.45009438e-11, 1.09282982e-11, 1.87759834e-12, -1.40849919e-12,\n", + " -1.41309065e-11, -1.45973476e-11, -2.48948648e-12, 1.62470233e-12,\n", + " -1.85399040e-11, -1.45447855e-11, 3.63228111e-12, -2.03865397e-12,\n", + " 1.44151687e-11, -1.82090736e-12, 4.10953693e-12, -1.45062747e-11,\n", + " -1.63726151e-12, 1.14638854e-12, -8.45686455e-12, 2.63523339e-12,\n", + " -1.63959801e-11, 1.14046766e-13, -1.55685306e-12, 1.73313664e-12,\n", + " 2.41606431e-12, 4.78593077e-12, 2.54235955e-12, 1.43347773e-11,\n", + " 2.10130849e-12, 1.77861686e-12, 1.66376773e-11, 1.82921950e-12,\n", + " -2.86404776e-12, 7.14697355e-12, -1.55668913e-12, -4.02296859e-12,\n", + " -1.30150990e-12, 1.48738851e-11, -2.05342536e-12, -7.71668493e-12,\n", + " 1.69779156e-11, -1.82855857e-12, 2.57383134e-12, -3.70609142e-13,\n", + " 1.54659081e-11, -1.37158340e-12, -1.93308260e-12, 1.50398218e-11,\n", + " 2.00455581e-12, -3.36654337e-12, 2.31098430e-12, -1.66817029e-11,\n", + " 4.70822253e-12, -1.64905600e-12, 1.48161756e-12, 1.91981956e-12,\n", + " 1.56288871e-11, -1.54816199e-12, 1.71096418e-11, 1.85986469e-12,\n", + " 1.52840414e-12, 1.56381071e-12, 1.76934422e-12, -2.30280552e-12,\n", + " -1.74372973e-12, -2.19445816e-12, -1.54470534e-12, -1.63086766e-11,\n", + " -6.12994204e-12, 2.78726695e-12, -1.64549520e-11, 1.45208307e-11,\n", + " -1.17382547e-12, -1.61504729e-12, 1.64195263e-12, 1.79648674e-11], dtype=float32), 'kernel': Array([[-1.6807569e-11, -1.6429605e-12, -2.3418793e-12, ...,\n", + " -1.6142259e-12, 1.6409287e-12, 1.7958557e-11],\n", + " [-1.6807569e-11, -1.6429605e-12, -2.3418793e-12, ...,\n", + " -1.6142259e-12, 1.6409287e-12, 1.7958557e-11],\n", + " [ 1.8002272e-11, 1.7678676e-12, 2.6154382e-12, ...,\n", + " 1.7471987e-12, -1.7676864e-12, -1.9253273e-11],\n", + " ...,\n", + " [ 1.8002272e-11, 1.7678676e-12, 2.6154382e-12, ...,\n", + " 1.7471987e-12, -1.7676864e-12, -1.9253273e-11],\n", + " [ 1.6807569e-11, 1.6429605e-12, 2.3418793e-12, ...,\n", + " 1.6142259e-12, -1.6409287e-12, -1.7958557e-11],\n", + " [-1.8002272e-11, -1.7678676e-12, -2.6154382e-12, ...,\n", + " -1.7471987e-12, 1.7676864e-12, 1.9253273e-11]], dtype=float32)}, 'Dense_5': {'bias': Array([-6.1843397e-09], dtype=float32), 'kernel': Array([[-3.8655559e-10],\n", + " [-6.1776033e-09],\n", + " [ 6.6520154e-09],\n", + " [ 6.1780892e-09],\n", + " [-3.7819989e-10],\n", + " [ 6.6908012e-09],\n", + " [ 6.1748189e-09],\n", + " [ 4.7096027e-10],\n", + " [-6.1742647e-09],\n", + " [ 6.1776988e-09],\n", + " [ 3.4977454e-10],\n", + " [ 6.6747767e-09],\n", + " [-3.4947031e-10],\n", + " [ 6.1625771e-09],\n", + " [ 5.9706617e-09],\n", + " [ 3.7265022e-10],\n", + " [-6.6755206e-09],\n", + " [-6.6740569e-09],\n", + " [-6.1593202e-09],\n", + " [ 6.6869683e-09],\n", + " [ 6.1268306e-09],\n", + " [ 6.1408589e-09],\n", + " [ 6.1435825e-09],\n", + " [ 3.7568490e-10],\n", + " [-6.1743801e-09],\n", + " [-4.5904001e-09],\n", + " [-6.1196417e-09],\n", + " [-3.0053873e-10],\n", + " [-6.6465597e-09],\n", + " [ 6.1781638e-09],\n", + " [ 6.1736865e-09],\n", + " [-6.1738796e-09],\n", + " [-6.1277143e-09],\n", + " [ 6.6744881e-09],\n", + " [ 6.6844414e-09],\n", + " [-6.1602172e-09],\n", + " [-6.0811818e-09],\n", + " [-3.6156822e-10],\n", + " [-3.7010467e-10],\n", + " [ 6.0481544e-09],\n", + " [-6.6741777e-09],\n", + " [ 6.0523346e-09],\n", + " [-6.1451821e-09],\n", + " [ 6.1366912e-09],\n", + " [-6.1424714e-09],\n", + " [ 6.1337397e-09],\n", + " [ 6.6712147e-09],\n", + " [-6.1800414e-09],\n", + " [ 3.8954642e-10],\n", + " [ 6.1140275e-09],\n", + " [-6.1657550e-09],\n", + " [ 6.6809878e-09],\n", + " [ 6.1638197e-09],\n", + " [-6.1464798e-09],\n", + " [-6.6811321e-09],\n", + " [-3.5263170e-10],\n", + " [-6.6914878e-09],\n", + " [ 6.1312262e-09],\n", + " [-6.1755849e-09],\n", + " [-6.1142509e-09],\n", + " [ 6.0247789e-09],\n", + " [ 2.7609612e-09],\n", + " [ 3.9059445e-10],\n", + " [ 3.3516714e-10],\n", + " [ 3.9409370e-10],\n", + " [ 3.7155776e-10],\n", + " [ 6.1744192e-09],\n", + " [ 3.1093988e-10],\n", + " [-6.6517405e-09],\n", + " [-3.0116415e-10],\n", + " [ 6.6849166e-09],\n", + " [ 3.6436976e-10],\n", + " [ 6.6912711e-09],\n", + " [ 6.1647611e-09],\n", + " [-3.8312323e-10],\n", + " [-6.6827650e-09],\n", + " [-3.7707032e-10],\n", + " [ 3.8468503e-10],\n", + " [-6.6909758e-09],\n", + " [ 6.0943597e-09],\n", + " [ 6.1774577e-09],\n", + " [-5.3503202e-09],\n", + " [ 6.1552168e-09],\n", + " [ 6.1771299e-09],\n", + " [-6.1673875e-09],\n", + " [-6.1211054e-09],\n", + " [ 4.5856574e-10],\n", + " [-2.7017391e-10],\n", + " [ 6.6894876e-09],\n", + " [ 6.1777463e-09],\n", + " [ 6.1723977e-09],\n", + " [ 3.3411923e-10],\n", + " [-6.1649326e-09],\n", + " [-6.1774781e-09],\n", + " [ 6.6912418e-09],\n", + " [-3.4201597e-10],\n", + " [ 6.1790848e-09],\n", + " [ 6.1692020e-09],\n", + " [ 3.3980752e-10],\n", + " [ 6.6538179e-09],\n", + " [-6.1532361e-09],\n", + " [ 2.9105487e-10],\n", + " [-3.9288670e-10],\n", + " [-3.9532333e-10],\n", + " [ 3.7320988e-10],\n", + " [ 6.1771868e-09],\n", + " [ 6.0976726e-09],\n", + " [ 6.1766103e-09],\n", + " [-6.6824506e-09],\n", + " [ 6.6917853e-09],\n", + " [ 6.6808989e-09],\n", + " [-6.1422707e-09],\n", + " [ 3.8819764e-10],\n", + " [-3.9496267e-10],\n", + " [ 6.6737331e-09],\n", + " [ 6.6687353e-09],\n", + " [ 3.5650671e-10],\n", + " [-6.0330674e-09],\n", + " [ 3.5956768e-10],\n", + " [ 6.1637682e-09],\n", + " [-6.1725807e-09],\n", + " [ 3.4623948e-10],\n", + " [-6.5151036e-09],\n", + " [-3.6827852e-10],\n", + " [-6.6676780e-09],\n", + " [-5.9607187e-11],\n", + " [-6.6892105e-09],\n", + " [-6.6833223e-09],\n", + " [ 6.1737433e-09],\n", + " [ 2.5892957e-10],\n", + " [ 6.1128778e-09],\n", + " [-6.6779275e-09],\n", + " [ 3.8849327e-10],\n", + " [-6.6848647e-09],\n", + " [ 3.4088171e-10],\n", + " [ 6.6858181e-09],\n", + " [ 3.6210934e-10],\n", + " [ 6.6430235e-09],\n", + " [ 3.8169863e-09],\n", + " [ 6.0928924e-09],\n", + " [ 6.0423604e-09],\n", + " [-6.1541101e-09],\n", + " [-6.6897923e-09],\n", + " [ 3.7626927e-10],\n", + " [-3.0858194e-10],\n", + " [-6.1289978e-09],\n", + " [-6.1006853e-09],\n", + " [ 3.8402945e-10],\n", + " [-6.5876558e-09],\n", + " [ 6.6850130e-09],\n", + " [ 5.8633010e-09],\n", + " [-6.6196324e-09],\n", + " [-6.0328169e-09],\n", + " [-6.6827703e-09],\n", + " [-3.7603515e-10],\n", + " [ 3.7307224e-10],\n", + " [-3.3732298e-10],\n", + " [-4.4634402e-10],\n", + " [ 6.6906591e-09],\n", + " [-6.5883414e-09],\n", + " [ 6.6611330e-09],\n", + " [ 6.6799757e-09],\n", + " [ 6.0504819e-09],\n", + " [ 6.6907759e-09],\n", + " [-3.6563613e-10],\n", + " [-3.8250805e-10],\n", + " [-6.1571601e-09],\n", + " [-3.2388381e-10],\n", + " [-6.1716832e-09],\n", + " [ 6.6217201e-09],\n", + " [ 3.3915809e-10],\n", + " [ 6.1616561e-09],\n", + " [ 6.1664975e-09],\n", + " [ 6.6874382e-09],\n", + " [ 3.2680589e-10],\n", + " [ 6.6836372e-09],\n", + " [ 6.5336008e-09],\n", + " [ 6.6904602e-09],\n", + " [ 6.6872143e-09],\n", + " [-6.1199472e-09],\n", + " [ 3.0335587e-10],\n", + " [ 5.5999756e-09],\n", + " [ 6.1732575e-09],\n", + " [ 6.6907897e-09],\n", + " [-3.4406655e-10],\n", + " [-3.1398401e-10],\n", + " [-6.1613243e-09],\n", + " [-6.6820878e-09],\n", + " [-3.7754083e-10],\n", + " [-2.8464814e-10],\n", + " [-6.5584449e-09],\n", + " [-6.6818768e-09],\n", + " [ 3.2889158e-10],\n", + " [ 6.1664567e-09],\n", + " [ 6.1175909e-09],\n", + " [-3.0087613e-10],\n", + " [ 6.6828512e-09],\n", + " [ 6.1795529e-09],\n", + " [-5.8711915e-09],\n", + " [ 6.6590631e-09],\n", + " [-3.8226583e-10],\n", + " [-6.1838241e-09],\n", + " [ 6.6855770e-09],\n", + " [-6.6782802e-09],\n", + " [ 6.6688735e-09],\n", + " [ 6.5277912e-09],\n", + " [ 6.1601400e-09],\n", + " [ 3.3986441e-10],\n", + " [-6.1587384e-09],\n", + " [-6.6774910e-09],\n", + " [ 3.8561157e-10],\n", + " [ 6.6883641e-09],\n", + " [-6.6517272e-09],\n", + " [-5.7823892e-09],\n", + " [ 6.1730234e-09],\n", + " [ 6.4995875e-09],\n", + " [ 6.1788215e-09],\n", + " [ 3.1742275e-10],\n", + " [-6.1699090e-09],\n", + " [-4.5385315e-10],\n", + " [ 3.8388245e-10],\n", + " [ 6.6740875e-09],\n", + " [-6.1435399e-09],\n", + " [-4.4310564e-10],\n", + " [ 3.8665030e-10],\n", + " [ 6.1773275e-09],\n", + " [ 6.1636358e-09],\n", + " [ 3.5913342e-10],\n", + " [-6.1613599e-09],\n", + " [ 6.1153198e-09],\n", + " [ 6.1650809e-09],\n", + " [-3.7873873e-10],\n", + " [ 4.5789517e-10],\n", + " [ 6.1707599e-09],\n", + " [-6.1748526e-09],\n", + " [ 6.1727072e-09],\n", + " [ 3.8277456e-10],\n", + " [-4.4562393e-10],\n", + " [ 3.9232129e-10],\n", + " [-6.6740924e-09],\n", + " [-6.6863053e-09],\n", + " [-6.6853150e-09],\n", + " [ 6.1754735e-09],\n", + " [-6.6728880e-09],\n", + " [-6.1756054e-09],\n", + " [-6.6770256e-09],\n", + " [-6.1793619e-09],\n", + " [-3.8152137e-10],\n", + " [ 3.8806935e-10],\n", + " [ 6.1537238e-09],\n", + " [-3.7733053e-10],\n", + " [ 3.1123715e-10],\n", + " [ 6.1809926e-09],\n", + " [-6.1763781e-09],\n", + " [ 6.1775802e-09],\n", + " [ 3.7808229e-10]], dtype=float32)}, 'log_std': Array([-0.00014559], dtype=float32)}}, nu={'params': {'Dense_0': {'bias': Array([7.e-43, 0.e+00, 0.e+00, 0.e+00, 7.e-43, 0.e+00, 7.e-43, 7.e-43,\n", + " 7.e-43, 0.e+00, 7.e-43, 0.e+00, 0.e+00, 7.e-43, 0.e+00, 0.e+00,\n", + " 7.e-43, 7.e-43, 0.e+00, 7.e-43, 0.e+00, 7.e-43, 0.e+00, 7.e-43,\n", + " 0.e+00, 7.e-43, 7.e-43, 7.e-43, 0.e+00, 0.e+00, 0.e+00, 0.e+00,\n", + " 0.e+00, 0.e+00, 0.e+00, 7.e-43, 7.e-43, 0.e+00, 7.e-43, 0.e+00,\n", + " 0.e+00, 0.e+00, 0.e+00, 0.e+00, 0.e+00, 7.e-43, 0.e+00, 0.e+00,\n", + " 0.e+00, 7.e-43, 0.e+00, 0.e+00, 0.e+00, 0.e+00, 7.e-43, 7.e-43,\n", + " 7.e-43, 0.e+00, 0.e+00, 7.e-43, 7.e-43, 7.e-43, 7.e-43, 7.e-43,\n", + " 0.e+00, 0.e+00, 0.e+00, 7.e-43, 7.e-43, 0.e+00, 0.e+00, 0.e+00,\n", + " 7.e-43, 0.e+00, 7.e-43, 0.e+00, 0.e+00, 7.e-43, 7.e-43, 7.e-43,\n", + " 7.e-43, 7.e-43, 0.e+00, 7.e-43, 7.e-43, 0.e+00, 0.e+00, 0.e+00,\n", + " 7.e-43, 0.e+00, 7.e-43, 0.e+00, 0.e+00, 0.e+00, 0.e+00, 7.e-43,\n", + " 0.e+00, 0.e+00, 7.e-43, 7.e-43, 7.e-43, 7.e-43, 0.e+00, 0.e+00,\n", + " 7.e-43, 0.e+00, 0.e+00, 0.e+00, 7.e-43, 0.e+00, 0.e+00, 0.e+00,\n", + " 0.e+00, 0.e+00, 0.e+00, 7.e-43, 7.e-43, 7.e-43, 7.e-43, 7.e-43,\n", + " 0.e+00, 7.e-43, 7.e-43, 0.e+00, 7.e-43, 0.e+00, 0.e+00, 7.e-43,\n", + " 0.e+00, 0.e+00, 7.e-43, 7.e-43, 7.e-43, 7.e-43, 7.e-43, 7.e-43,\n", + " 0.e+00, 7.e-43, 7.e-43, 7.e-43, 0.e+00, 7.e-43, 7.e-43, 7.e-43,\n", + " 7.e-43, 0.e+00, 0.e+00, 0.e+00, 7.e-43, 0.e+00, 0.e+00, 7.e-43,\n", + " 7.e-43, 7.e-43, 7.e-43, 0.e+00, 7.e-43, 0.e+00, 0.e+00, 0.e+00,\n", + " 0.e+00, 0.e+00, 0.e+00, 0.e+00, 0.e+00, 7.e-43, 7.e-43, 0.e+00,\n", + " 0.e+00, 0.e+00, 7.e-43, 0.e+00, 0.e+00, 7.e-43, 0.e+00, 0.e+00,\n", + " 0.e+00, 0.e+00, 0.e+00, 7.e-43, 7.e-43, 0.e+00, 7.e-43, 0.e+00,\n", + " 7.e-43, 0.e+00, 0.e+00, 0.e+00, 7.e-43, 7.e-43, 0.e+00, 0.e+00,\n", + " 0.e+00, 7.e-43, 7.e-43, 7.e-43, 7.e-43, 7.e-43, 7.e-43, 7.e-43,\n", + " 7.e-43, 0.e+00, 7.e-43, 0.e+00, 7.e-43, 0.e+00, 7.e-43, 7.e-43,\n", + " 0.e+00, 0.e+00, 7.e-43, 7.e-43, 7.e-43, 0.e+00, 7.e-43, 7.e-43,\n", + " 0.e+00, 7.e-43, 7.e-43, 0.e+00, 7.e-43, 7.e-43, 7.e-43, 7.e-43,\n", + " 0.e+00, 0.e+00, 7.e-43, 0.e+00, 7.e-43, 7.e-43, 7.e-43, 7.e-43,\n", + " 0.e+00, 0.e+00, 0.e+00, 0.e+00, 0.e+00, 0.e+00, 0.e+00, 0.e+00,\n", + " 0.e+00, 0.e+00, 0.e+00, 0.e+00, 0.e+00, 0.e+00, 0.e+00, 0.e+00,\n", + " 0.e+00, 0.e+00, 7.e-43, 0.e+00, 0.e+00, 7.e-43, 0.e+00, 0.e+00], dtype=float32), 'kernel': Array([[7.e-43, 0.e+00, 0.e+00, ..., 7.e-43, 0.e+00, 0.e+00],\n", + " [7.e-43, 0.e+00, 0.e+00, ..., 7.e-43, 0.e+00, 0.e+00],\n", + " [7.e-43, 0.e+00, 0.e+00, ..., 7.e-43, 0.e+00, 0.e+00],\n", + " ...,\n", + " [7.e-43, 0.e+00, 0.e+00, ..., 7.e-43, 0.e+00, 0.e+00],\n", + " [7.e-43, 0.e+00, 0.e+00, ..., 7.e-43, 0.e+00, 0.e+00],\n", + " [7.e-43, 0.e+00, 0.e+00, ..., 7.e-43, 0.e+00, 0.e+00]], dtype=float32)}, 'Dense_1': {'bias': Array([3.1469093e-14, 8.7121838e-14, 3.1524428e-14, 8.1083340e-14,\n", + " 3.1400565e-14, 9.5083995e-14, 8.7520248e-14, 3.0413850e-14,\n", + " 8.3409631e-14, 8.8949932e-14, 3.0898776e-14, 8.4273456e-14,\n", + " 7.8286722e-14, 3.0711541e-14, 3.1281848e-14, 7.9285591e-14,\n", + " 8.2523384e-14, 8.2519515e-14, 8.0005237e-14, 8.4032323e-14,\n", + " 9.4522812e-14, 3.0676721e-14, 8.2428815e-14, 3.4723397e-14,\n", + " 8.1286303e-14, 8.4509575e-14, 2.5317645e-12, 8.6583538e-14,\n", + " 3.0395716e-14, 8.1653136e-14, 3.0561694e-14, 3.1361873e-14,\n", + " 8.6422595e-14, 7.9948851e-14, 8.9926113e-14, 2.9955656e-14,\n", + " 8.6943799e-14, 8.2753350e-14, 3.1579906e-14, 3.1489849e-14,\n", + " 8.5219029e-14, 8.6388646e-14, 7.9661219e-14, 3.1641248e-14,\n", + " 8.4022694e-14, 8.6496687e-14, 7.9054208e-14, 2.9760198e-14,\n", + " 8.6306416e-14, 8.7948779e-14, 3.1711297e-14, 8.8595560e-14,\n", + " 8.0190473e-14, 8.6881708e-14, 3.0019569e-14, 3.3298593e-14,\n", + " 8.1419097e-14, 8.4182220e-14, 8.6198328e-14, 7.7233386e-14,\n", + " 3.1253286e-14, 8.7068068e-14, 3.0019939e-14, 3.0553610e-14,\n", + " 7.9709094e-14, 7.9538874e-14, 8.2772026e-14, 8.4562748e-14,\n", + " 3.0161125e-14, 8.4582169e-14, 7.9997153e-14, 8.3321547e-14,\n", + " 3.0441229e-14, 8.8881762e-14, 8.5223515e-14, 7.6368423e-14,\n", + " 8.6652426e-14, 3.0150670e-14, 8.3026583e-14, 3.3523819e-14,\n", + " 3.0758538e-14, 7.7542119e-14, 3.0456977e-14, 8.1874808e-14,\n", + " 7.9973646e-14, 8.3165225e-14, 8.1575629e-14, 3.1392589e-14,\n", + " 2.8575196e-11, 8.2384240e-14, 8.5764051e-14, 8.3646150e-14,\n", + " 3.0588433e-14, 3.0183958e-14, 9.0230300e-14, 3.0812738e-14,\n", + " 3.0985617e-14, 8.1762762e-14, 2.9775397e-14, 8.4227845e-14,\n", + " 8.7015288e-14, 3.1643850e-14, 8.7046452e-14, 8.4730705e-14,\n", + " 8.2809248e-14, 3.0093461e-14, 8.4697393e-14, 3.1481399e-14,\n", + " 3.1382642e-14, 2.9992227e-14, 8.3107566e-14, 7.7644332e-14,\n", + " 3.1182668e-14, 8.2608345e-14, 8.4545462e-14, 8.6818980e-14,\n", + " 7.9620379e-14, 8.9375264e-14, 3.0522907e-14, 9.0820458e-14,\n", + " 3.0889872e-14, 3.2212978e-14, 8.1036591e-14, 8.3619397e-14,\n", + " 9.1934212e-14, 9.3259798e-14, 8.0280495e-14, 3.0615657e-14,\n", + " 9.3122138e-14, 8.4658314e-14, 2.7671173e-12, 8.3194878e-14,\n", + " 8.7144159e-14, 2.9810051e-14, 8.6270292e-14, 8.5228367e-14,\n", + " 2.9780211e-14, 8.4936425e-14, 9.0915129e-14, 8.3327144e-14,\n", + " 8.6342663e-14, 7.9813841e-14, 8.2344213e-14, 3.0691053e-14,\n", + " 8.6679287e-14, 3.1675715e-14, 8.6835283e-14, 3.0387148e-14,\n", + " 2.9768261e-14, 9.0933622e-14, 3.1604426e-14, 8.3521819e-14,\n", + " 8.2639001e-14, 7.8839550e-14, 7.9093822e-14, 3.1172652e-14,\n", + " 8.7725380e-14, 8.7919106e-14, 8.6635648e-14, 8.4624351e-14,\n", + " 8.1674725e-14, 3.1539235e-14, 8.8392943e-14, 8.3475964e-14,\n", + " 9.0390003e-14, 8.1214867e-14, 8.1143947e-14, 8.2175681e-14,\n", + " 8.1758920e-14, 3.0798060e-14, 3.3724180e-14, 8.1960521e-14,\n", + " 8.9672444e-14, 8.6844919e-14, 8.2994348e-14, 8.1104597e-14,\n", + " 3.0418393e-14, 8.1258168e-14, 3.1523016e-14, 7.6816144e-14,\n", + " 3.0213540e-14, 2.9579363e-14, 8.7230069e-14, 8.3386104e-14,\n", + " 8.4678277e-14, 8.7680717e-14, 8.4607716e-14, 3.1366403e-14,\n", + " 2.9197995e-14, 8.9683781e-14, 8.6310523e-14, 8.4040969e-14,\n", + " 3.1400860e-14, 3.3998845e-14, 2.9987999e-14, 8.5936297e-14,\n", + " 8.7394481e-14, 3.1430116e-14, 3.0393697e-14, 4.8121739e-11,\n", + " 8.4418116e-14, 8.2186773e-14, 3.2397814e-14, 3.0579519e-14,\n", + " 2.2910775e-13, 8.6450439e-14, 8.4291068e-14, 8.6205172e-14,\n", + " 8.8165721e-14, 8.3840053e-14, 9.0104139e-14, 2.9553112e-14,\n", + " 8.3574417e-14, 3.0368787e-14, 8.7539771e-14, 8.3330918e-14,\n", + " 8.3134834e-14, 3.1741059e-14, 3.1806772e-14, 7.9462295e-14,\n", + " 8.4391742e-14, 8.3963577e-14, 3.0999268e-14, 2.2302034e-12,\n", + " 9.0161778e-14, 7.9783307e-14, 8.6268978e-14, 3.0946240e-14,\n", + " 8.9797690e-14, 3.1493207e-14, 3.2540451e-14, 8.8031301e-14,\n", + " 8.9137492e-14, 8.4630741e-14, 3.0880799e-14, 7.9791778e-14,\n", + " 3.1882703e-14, 3.0386226e-14, 8.2147241e-14, 9.0612047e-14,\n", + " 3.0701238e-14, 8.4227703e-14, 8.4848205e-14, 8.3596948e-14,\n", + " 9.1174687e-14, 8.1967893e-14, 3.0689488e-14, 3.0493383e-14,\n", + " 8.6743506e-14, 8.6627767e-14, 2.3575052e-13, 8.4647756e-14,\n", + " 8.2248532e-14, 8.1330918e-14, 8.2987240e-14, 3.1932380e-14], dtype=float32), 'kernel': Array([[3.1910889e-14, 8.7500536e-14, 3.1972512e-14, ..., 8.1760316e-14,\n", + " 8.3333744e-14, 3.2359003e-14],\n", + " [3.1450689e-14, 8.7058344e-14, 3.1506193e-14, ..., 8.1330918e-14,\n", + " 8.2926965e-14, 3.1914067e-14],\n", + " [3.1450689e-14, 8.7058344e-14, 3.1506193e-14, ..., 8.1330918e-14,\n", + " 8.2926965e-14, 3.1914067e-14],\n", + " ...,\n", + " [3.1910889e-14, 8.7500536e-14, 3.1972512e-14, ..., 8.1760316e-14,\n", + " 8.3333744e-14, 3.2359003e-14],\n", + " [3.1450689e-14, 8.7058344e-14, 3.1506193e-14, ..., 8.1330918e-14,\n", + " 8.2926965e-14, 3.1914067e-14],\n", + " [3.1450689e-14, 8.7058344e-14, 3.1506193e-14, ..., 8.1330918e-14,\n", + " 8.2926965e-14, 3.1914067e-14]], dtype=float32)}, 'Dense_2': {'bias': Array([0.00097166], dtype=float32), 'kernel': Array([[9.8969578e-04],\n", + " [9.7163819e-04],\n", + " [9.8969566e-04],\n", + " [9.7163877e-04],\n", + " [9.8969566e-04],\n", + " [9.7163825e-04],\n", + " [9.7163860e-04],\n", + " [9.8969578e-04],\n", + " [9.7163866e-04],\n", + " [9.7163772e-04],\n", + " [9.8969578e-04],\n", + " [9.7163866e-04],\n", + " [9.7163935e-04],\n", + " [9.8969566e-04],\n", + " [9.8969578e-04],\n", + " [9.7163941e-04],\n", + " [9.7163854e-04],\n", + " [9.7163871e-04],\n", + " [9.7163901e-04],\n", + " [9.7163866e-04],\n", + " [9.7163813e-04],\n", + " [9.8969566e-04],\n", + " [9.7163877e-04],\n", + " [9.8969450e-04],\n", + " [9.7163877e-04],\n", + " [9.7163866e-04],\n", + " [1.3174417e-06],\n", + " [9.7163860e-04],\n", + " [9.8969589e-04],\n", + " [9.7163906e-04],\n", + " [9.8969589e-04],\n", + " [9.8969566e-04],\n", + " [9.7163848e-04],\n", + " [9.7163918e-04],\n", + " [9.7163831e-04],\n", + " [9.8969589e-04],\n", + " [9.7163854e-04],\n", + " [9.7163854e-04],\n", + " [9.8969555e-04],\n", + " [9.8969566e-04],\n", + " [9.7163860e-04],\n", + " [9.7163860e-04],\n", + " [9.7163906e-04],\n", + " [9.8969555e-04],\n", + " [9.7163918e-04],\n", + " [9.7163842e-04],\n", + " [9.7163906e-04],\n", + " [9.8969578e-04],\n", + " [9.7163854e-04],\n", + " [9.7163854e-04],\n", + " [9.8969566e-04],\n", + " [9.7163813e-04],\n", + " [9.7163895e-04],\n", + " [9.7163866e-04],\n", + " [9.8969589e-04],\n", + " [9.8969461e-04],\n", + " [9.7163877e-04],\n", + " [9.7163854e-04],\n", + " [9.7163854e-04],\n", + " [9.7163924e-04],\n", + " [9.8969555e-04],\n", + " [9.7163842e-04],\n", + " [9.8969589e-04],\n", + " [9.8969578e-04],\n", + " [9.7163912e-04],\n", + " [9.7163889e-04],\n", + " [9.7163877e-04],\n", + " [9.7163860e-04],\n", + " [9.8969589e-04],\n", + " [9.7163866e-04],\n", + " [9.7163912e-04],\n", + " [9.7163866e-04],\n", + " [9.8969566e-04],\n", + " [9.7163842e-04],\n", + " [9.7163842e-04],\n", + " [9.7163941e-04],\n", + " [9.7163854e-04],\n", + " [9.8969578e-04],\n", + " [9.7163866e-04],\n", + " [9.8969461e-04],\n", + " [9.8969578e-04],\n", + " [9.7163918e-04],\n", + " [9.8969589e-04],\n", + " [9.7163877e-04],\n", + " [9.7163901e-04],\n", + " [9.7163866e-04],\n", + " [9.7163854e-04],\n", + " [9.8969578e-04],\n", + " [3.3374417e-07],\n", + " [9.7163883e-04],\n", + " [9.7163854e-04],\n", + " [9.7163866e-04],\n", + " [9.8969589e-04],\n", + " [9.8969578e-04],\n", + " [9.7163854e-04],\n", + " [9.8969566e-04],\n", + " [9.8969578e-04],\n", + " [9.7163924e-04],\n", + " [9.8969589e-04],\n", + " [9.7163866e-04],\n", + " [9.7163854e-04],\n", + " [9.8969578e-04],\n", + " [9.7163854e-04],\n", + " [9.7163871e-04],\n", + " [9.7163866e-04],\n", + " [9.8969578e-04],\n", + " [9.7163866e-04],\n", + " [9.8969566e-04],\n", + " [9.8969555e-04],\n", + " [9.8969578e-04],\n", + " [9.7163860e-04],\n", + " [9.7163935e-04],\n", + " [9.8969566e-04],\n", + " [9.7163883e-04],\n", + " [9.7163854e-04],\n", + " [9.7163854e-04],\n", + " [9.7163918e-04],\n", + " [9.7163848e-04],\n", + " [9.8969589e-04],\n", + " [9.7163778e-04],\n", + " [9.8969566e-04],\n", + " [9.8969555e-04],\n", + " [9.7163883e-04],\n", + " [9.7163854e-04],\n", + " [9.7163825e-04],\n", + " [9.7163813e-04],\n", + " [9.7163901e-04],\n", + " [9.8969566e-04],\n", + " [9.7163796e-04],\n", + " [9.7163854e-04],\n", + " [9.6905010e-04],\n", + " [9.7163906e-04],\n", + " [9.7163842e-04],\n", + " [9.8969589e-04],\n", + " [9.7163854e-04],\n", + " [9.7163866e-04],\n", + " [9.8969589e-04],\n", + " [9.7163860e-04],\n", + " [9.7163836e-04],\n", + " [9.7163918e-04],\n", + " [9.7163860e-04],\n", + " [9.7163860e-04],\n", + " [9.7163883e-04],\n", + " [9.8969589e-04],\n", + " [9.7163842e-04],\n", + " [9.8969566e-04],\n", + " [9.7163854e-04],\n", + " [9.8969578e-04],\n", + " [9.8969589e-04],\n", + " [9.7163842e-04],\n", + " [9.8969566e-04],\n", + " [9.7163842e-04],\n", + " [9.7163866e-04],\n", + " [9.7163930e-04],\n", + " [9.7163935e-04],\n", + " [9.8969566e-04],\n", + " [9.7163860e-04],\n", + " [9.7163854e-04],\n", + " [9.7163854e-04],\n", + " [9.7163854e-04],\n", + " [9.7163854e-04],\n", + " [9.8969589e-04],\n", + " [9.7163854e-04],\n", + " [9.7163889e-04],\n", + " [9.7163848e-04],\n", + " [9.7163871e-04],\n", + " [9.7163895e-04],\n", + " [9.7163866e-04],\n", + " [9.7163889e-04],\n", + " [9.8969578e-04],\n", + " [9.8969461e-04],\n", + " [9.7163883e-04],\n", + " [9.7163842e-04],\n", + " [9.7163825e-04],\n", + " [9.7163860e-04],\n", + " [9.7163889e-04],\n", + " [9.8969578e-04],\n", + " [9.7163860e-04],\n", + " [9.8969566e-04],\n", + " [9.7163947e-04],\n", + " [9.8969589e-04],\n", + " [9.8969589e-04],\n", + " [9.7163854e-04],\n", + " [9.7163866e-04],\n", + " [9.7163871e-04],\n", + " [9.7163860e-04],\n", + " [9.7163854e-04],\n", + " [9.8969566e-04],\n", + " [9.8969589e-04],\n", + " [9.7163854e-04],\n", + " [9.7163854e-04],\n", + " [9.7163854e-04],\n", + " [9.8969566e-04],\n", + " [9.8969450e-04],\n", + " [9.8969589e-04],\n", + " [9.7163854e-04],\n", + " [9.7163854e-04],\n", + " [9.8969578e-04],\n", + " [9.8969589e-04],\n", + " [9.0059126e-04],\n", + " [9.7163854e-04],\n", + " [9.7163877e-04],\n", + " [9.8969555e-04],\n", + " [9.8969578e-04],\n", + " [9.7154133e-04],\n", + " [9.7163854e-04],\n", + " [9.7163860e-04],\n", + " [9.7163854e-04],\n", + " [9.7163860e-04],\n", + " [9.7163854e-04],\n", + " [9.7163831e-04],\n", + " [9.8969589e-04],\n", + " [9.7163889e-04],\n", + " [9.8969589e-04],\n", + " [9.7163854e-04],\n", + " [9.7163854e-04],\n", + " [9.7163906e-04],\n", + " [9.8969555e-04],\n", + " [9.8969578e-04],\n", + " [9.7163935e-04],\n", + " [9.7163854e-04],\n", + " [9.7163854e-04],\n", + " [9.8969578e-04],\n", + " [9.6969621e-04],\n", + " [9.7163831e-04],\n", + " [9.7163866e-04],\n", + " [9.7163854e-04],\n", + " [9.8969578e-04],\n", + " [9.7163842e-04],\n", + " [9.8969566e-04],\n", + " [9.8969543e-04],\n", + " [9.7163854e-04],\n", + " [9.7163842e-04],\n", + " [9.7163854e-04],\n", + " [9.8969566e-04],\n", + " [9.7163895e-04],\n", + " [9.8969555e-04],\n", + " [9.8969566e-04],\n", + " [9.7163883e-04],\n", + " [9.7163836e-04],\n", + " [9.8969589e-04],\n", + " [9.7163854e-04],\n", + " [9.7163866e-04],\n", + " [9.7163854e-04],\n", + " [9.7163854e-04],\n", + " [9.7163860e-04],\n", + " [9.8969578e-04],\n", + " [9.8969578e-04],\n", + " [9.7163866e-04],\n", + " [9.7163866e-04],\n", + " [9.7153842e-04],\n", + " [9.7163860e-04],\n", + " [9.7163866e-04],\n", + " [9.7163866e-04],\n", + " [9.7163906e-04],\n", + " [9.8969566e-04]], dtype=float32)}, 'Dense_3': {'bias': Array([0.0000000e+00, 0.0000000e+00, 7.0064923e-43, 0.0000000e+00,\n", + " 7.0064923e-43, 0.0000000e+00, 7.0064923e-43, 0.0000000e+00,\n", + " 7.0064923e-43, 0.0000000e+00, 0.0000000e+00, 0.0000000e+00,\n", + " 7.0064923e-43, 0.0000000e+00, 7.0064923e-43, 7.0064923e-43,\n", + " 7.0064923e-43, 7.0064923e-43, 7.0064923e-43, 7.0064923e-43,\n", + " 0.0000000e+00, 0.0000000e+00, 0.0000000e+00, 0.0000000e+00,\n", + " 7.0064923e-43, 7.0064923e-43, 7.0064923e-43, 0.0000000e+00,\n", + " 0.0000000e+00, 0.0000000e+00, 7.0064923e-43, 7.0064923e-43,\n", + " 0.0000000e+00, 7.0064923e-43, 0.0000000e+00, 0.0000000e+00,\n", + " 7.0064923e-43, 0.0000000e+00, 0.0000000e+00, 7.0064923e-43,\n", + " 7.0064923e-43, 7.0064923e-43, 0.0000000e+00, 0.0000000e+00,\n", + " 0.0000000e+00, 0.0000000e+00, 0.0000000e+00, 7.0064923e-43,\n", + " 7.0064923e-43, 0.0000000e+00, 7.0064923e-43, 7.0064923e-43,\n", + " 0.0000000e+00, 7.0064923e-43, 7.0064923e-43, 0.0000000e+00,\n", + " 0.0000000e+00, 7.0064923e-43, 7.0064923e-43, 0.0000000e+00,\n", + " 0.0000000e+00, 7.0064923e-43, 0.0000000e+00, 0.0000000e+00,\n", + " 0.0000000e+00, 0.0000000e+00, 7.0064923e-43, 0.0000000e+00,\n", + " 0.0000000e+00, 0.0000000e+00, 0.0000000e+00, 7.0064923e-43,\n", + " 7.0064923e-43, 7.0064923e-43, 0.0000000e+00, 7.0064923e-43,\n", + " 7.0064923e-43, 7.0064923e-43, 0.0000000e+00, 0.0000000e+00,\n", + " 7.0064923e-43, 7.0064923e-43, 7.0064923e-43, 0.0000000e+00,\n", + " 0.0000000e+00, 0.0000000e+00, 7.0064923e-43, 0.0000000e+00,\n", + " 0.0000000e+00, 0.0000000e+00, 0.0000000e+00, 7.0064923e-43,\n", + " 0.0000000e+00, 0.0000000e+00, 7.0064923e-43, 0.0000000e+00,\n", + " 0.0000000e+00, 7.0064923e-43, 7.0064923e-43, 0.0000000e+00,\n", + " 7.0064923e-43, 7.0064923e-43, 0.0000000e+00, 7.0064923e-43,\n", + " 0.0000000e+00, 0.0000000e+00, 0.0000000e+00, 0.0000000e+00,\n", + " 0.0000000e+00, 7.0064923e-43, 7.0064923e-43, 7.0064923e-43,\n", + " 7.0064923e-43, 0.0000000e+00, 0.0000000e+00, 7.0064923e-43,\n", + " 0.0000000e+00, 7.0064923e-43, 0.0000000e+00, 0.0000000e+00,\n", + " 0.0000000e+00, 7.0064923e-43, 7.0064923e-43, 0.0000000e+00,\n", + " 0.0000000e+00, 0.0000000e+00, 0.0000000e+00, 7.0064923e-43,\n", + " 0.0000000e+00, 7.0064923e-43, 0.0000000e+00, 0.0000000e+00,\n", + " 0.0000000e+00, 0.0000000e+00, 0.0000000e+00, 0.0000000e+00,\n", + " 0.0000000e+00, 7.0064923e-43, 7.0064923e-43, 7.0064923e-43,\n", + " 0.0000000e+00, 7.0064923e-43, 0.0000000e+00, 7.0064923e-43,\n", + " 7.0064923e-43, 0.0000000e+00, 0.0000000e+00, 0.0000000e+00,\n", + " 1.3990084e-31, 0.0000000e+00, 0.0000000e+00, 7.0064923e-43,\n", + " 0.0000000e+00, 0.0000000e+00, 0.0000000e+00, 7.0064923e-43,\n", + " 0.0000000e+00, 0.0000000e+00, 0.0000000e+00, 7.0064923e-43,\n", + " 0.0000000e+00, 0.0000000e+00, 0.0000000e+00, 0.0000000e+00,\n", + " 7.0064923e-43, 0.0000000e+00, 7.0064923e-43, 7.0064923e-43,\n", + " 0.0000000e+00, 7.0064923e-43, 7.0064923e-43, 7.0064923e-43,\n", + " 0.0000000e+00, 7.0064923e-43, 7.0064923e-43, 0.0000000e+00,\n", + " 0.0000000e+00, 0.0000000e+00, 0.0000000e+00, 7.0064923e-43,\n", + " 7.0064923e-43, 0.0000000e+00, 0.0000000e+00, 0.0000000e+00,\n", + " 7.0064923e-43, 0.0000000e+00, 7.0064923e-43, 7.0064923e-43,\n", + " 0.0000000e+00, 0.0000000e+00, 0.0000000e+00, 0.0000000e+00,\n", + " 7.0064923e-43, 0.0000000e+00, 7.0064923e-43, 0.0000000e+00,\n", + " 0.0000000e+00, 7.0064923e-43, 0.0000000e+00, 7.0064923e-43,\n", + " 7.0064923e-43, 0.0000000e+00, 0.0000000e+00, 7.0064923e-43,\n", + " 0.0000000e+00, 0.0000000e+00, 0.0000000e+00, 7.0064923e-43,\n", + " 0.0000000e+00, 7.0064923e-43, 0.0000000e+00, 7.0064923e-43,\n", + " 7.0064923e-43, 0.0000000e+00, 0.0000000e+00, 7.0064923e-43,\n", + " 7.0064923e-43, 7.0064923e-43, 7.0064923e-43, 0.0000000e+00,\n", + " 0.0000000e+00, 7.0064923e-43, 0.0000000e+00, 0.0000000e+00,\n", + " 7.0064923e-43, 0.0000000e+00, 7.0064923e-43, 7.0064923e-43,\n", + " 7.0064923e-43, 7.0064923e-43, 7.0064923e-43, 0.0000000e+00,\n", + " 7.0064923e-43, 0.0000000e+00, 0.0000000e+00, 7.0064923e-43,\n", + " 7.0064923e-43, 7.0064923e-43, 0.0000000e+00, 7.0064923e-43,\n", + " 0.0000000e+00, 0.0000000e+00, 0.0000000e+00, 0.0000000e+00,\n", + " 0.0000000e+00, 0.0000000e+00, 0.0000000e+00, 7.0064923e-43,\n", + " 0.0000000e+00, 7.0064923e-43, 7.0064923e-43, 0.0000000e+00,\n", + " 7.0064923e-43, 7.0064923e-43, 0.0000000e+00, 7.0064923e-43], dtype=float32), 'kernel': Array([[0.e+00, 0.e+00, 7.e-43, ..., 7.e-43, 0.e+00, 0.e+00],\n", + " [0.e+00, 0.e+00, 7.e-43, ..., 7.e-43, 0.e+00, 0.e+00],\n", + " [0.e+00, 0.e+00, 7.e-43, ..., 7.e-43, 0.e+00, 0.e+00],\n", + " ...,\n", + " [0.e+00, 0.e+00, 7.e-43, ..., 7.e-43, 0.e+00, 7.e-43],\n", + " [0.e+00, 0.e+00, 7.e-43, ..., 7.e-43, 0.e+00, 7.e-43],\n", + " [0.e+00, 0.e+00, 7.e-43, ..., 7.e-43, 0.e+00, 7.e-43]], dtype=float32)}, 'Dense_4': {'bias': Array([1.11147006e-13, 1.14333926e-15, 5.01215187e-15, 4.51753693e-16,\n", + " 1.24167129e-13, 1.32680183e-15, 1.38136907e-15, 3.14773503e-14,\n", + " 1.02126826e-15, 8.05438283e-16, 1.22144563e-13, 1.63302298e-15,\n", + " 1.23096493e-13, 2.43947733e-15, 8.06308365e-15, 1.21173565e-13,\n", + " 1.60967113e-15, 1.67378602e-15, 2.08144815e-15, 1.61125614e-15,\n", + " 8.70114002e-15, 3.62691461e-15, 4.18669624e-15, 1.34334465e-13,\n", + " 1.04127878e-15, 9.61887159e-14, 5.55652724e-15, 1.26615366e-13,\n", + " 5.58106663e-15, 6.67521467e-16, 1.05202604e-15, 1.39298221e-15,\n", + " 5.55633496e-15, 1.63374825e-15, 1.20252833e-15, 2.66218622e-15,\n", + " 9.94882887e-15, 1.19610349e-13, 1.10831070e-13, 1.21089187e-14,\n", + " 1.66470636e-15, 1.24741102e-14, 4.05156569e-15, 4.69808137e-15,\n", + " 3.47577281e-15, 4.25318494e-15, 1.79768576e-15, 6.69750169e-16,\n", + " 1.37947013e-13, 6.92806798e-15, 2.27587906e-15, 1.34198723e-15,\n", + " 2.33558324e-15, 3.26473794e-15, 1.34586611e-15, 1.20851950e-13,\n", + " 1.27619479e-15, 4.52999742e-15, 1.31318906e-15, 6.05969785e-15,\n", + " 1.43606423e-14, 5.08057900e-15, 1.19047337e-13, 1.25855909e-13,\n", + " 1.20304781e-13, 1.16084537e-13, 1.00067107e-15, 1.20600375e-13,\n", + " 8.32181813e-16, 1.24271511e-13, 1.73354282e-15, 1.23314756e-13,\n", + " 9.07148464e-16, 1.67129066e-15, 1.23213641e-13, 1.28431233e-15,\n", + " 1.29811933e-13, 1.23925826e-13, 1.30627769e-15, 8.65812854e-15,\n", + " 1.16579421e-15, 6.09477001e-14, 2.39697894e-15, 1.18334145e-15,\n", + " 1.54219977e-15, 6.24974663e-15, 3.89225429e-14, 1.26377831e-13,\n", + " 1.42212978e-15, 1.14297821e-15, 1.57845606e-15, 1.17426007e-13,\n", + " 8.39788645e-16, 8.30107006e-16, 1.28629682e-15, 1.20845770e-13,\n", + " 1.03154817e-15, 1.84780195e-15, 1.24067179e-13, 4.80924640e-15,\n", + " 3.29096991e-15, 1.16399891e-13, 1.29364862e-13, 1.32772008e-13,\n", + " 1.24708485e-13, 8.45288218e-16, 8.39788274e-15, 8.71083971e-16,\n", + " 1.29553065e-15, 1.25092759e-15, 1.35118272e-15, 4.31502088e-15,\n", + " 1.35790562e-13, 1.30850869e-13, 1.68146279e-15, 1.93179882e-15,\n", + " 1.16471503e-13, 1.36290718e-14, 1.17019499e-13, 2.33501297e-15,\n", + " 1.12036730e-15, 1.17656766e-13, 2.22182112e-14, 1.17177603e-13,\n", + " 3.37921740e-15, 2.50024593e-14, 1.44682121e-15, 1.25104268e-15,\n", + " 1.46812060e-15, 2.98356819e-14, 6.98916785e-15, 1.49618841e-15,\n", + " 1.23805968e-13, 1.79598640e-15, 1.21200481e-13, 1.14091103e-15,\n", + " 1.21506537e-13, 3.28142787e-15, 2.55043844e-16, 8.04658684e-15,\n", + " 1.35907139e-14, 3.23509898e-15, 1.39872573e-15, 1.17672175e-13,\n", + " 1.23305392e-13, 5.54425585e-15, 8.14029833e-15, 1.35223294e-13,\n", + " 1.48698718e-14, 1.16512632e-15, 5.81296757e-14, 8.63803522e-15,\n", + " 1.27647145e-14, 1.27805572e-15, 1.14998207e-13, 1.19000743e-13,\n", + " 1.23659397e-13, 7.30826600e-14, 9.29546662e-16, 6.51455623e-15,\n", + " 2.26462962e-15, 1.39792571e-15, 1.28625866e-14, 1.30987176e-15,\n", + " 1.20761962e-13, 1.29759715e-13, 2.23604226e-15, 1.20007669e-13,\n", + " 1.18605227e-15, 4.50020643e-15, 1.24981256e-13, 1.89436843e-15,\n", + " 2.11049990e-15, 1.58556743e-15, 1.17945557e-13, 1.90461817e-15,\n", + " 1.00407379e-14, 1.35031197e-15, 1.08059868e-15, 6.36802588e-15,\n", + " 1.23453412e-13, 4.66601279e-14, 1.51196154e-15, 1.31245796e-15,\n", + " 1.15348364e-13, 1.23728149e-13, 2.56502286e-15, 1.86687756e-15,\n", + " 1.30860763e-13, 1.27511703e-13, 1.63286827e-14, 1.30722627e-15,\n", + " 1.20218519e-13, 1.54296824e-15, 6.39236749e-15, 1.24716914e-13,\n", + " 1.96577585e-15, 1.10161198e-15, 2.77457517e-14, 2.37774388e-15,\n", + " 1.25074864e-13, 6.19380772e-16, 1.72980846e-15, 2.33123986e-15,\n", + " 1.91562684e-15, 1.03251326e-14, 2.40200311e-15, 1.18654246e-13,\n", + " 2.09975444e-15, 2.43577749e-15, 1.21498433e-13, 1.04311620e-15,\n", + " 2.80282165e-15, 3.44417420e-14, 1.09318443e-15, 2.45724461e-14,\n", + " 7.45257756e-16, 1.26091385e-13, 1.80462318e-15, 5.51285464e-14,\n", + " 1.25968084e-13, 2.76142079e-15, 3.39748346e-15, 2.21840351e-14,\n", + " 1.22972514e-13, 8.29267702e-16, 1.63695427e-15, 1.20667988e-13,\n", + " 1.90551369e-15, 6.36976061e-15, 2.22418189e-15, 1.29284468e-13,\n", + " 3.29301880e-14, 1.23827217e-15, 9.81244047e-16, 1.56630792e-15,\n", + " 1.25498394e-13, 2.61702959e-14, 1.22535066e-13, 2.74570177e-15,\n", + " 1.64503984e-15, 1.73674788e-15, 1.32765861e-15, 1.71886935e-15,\n", + " 1.31337689e-15, 1.52677096e-15, 1.00561043e-15, 1.25612411e-13,\n", + " 4.44012875e-14, 3.28541994e-15, 1.34587410e-13, 1.23902109e-13,\n", + " 6.01474867e-16, 1.26386199e-15, 1.16123696e-15, 1.33933554e-13], dtype=float32), 'kernel': Array([[1.11066314e-13, 1.14253903e-15, 5.00865108e-15, ...,\n", + " 1.26299886e-15, 1.16050682e-15, 1.33837548e-13],\n", + " [1.11066314e-13, 1.14253903e-15, 5.00865108e-15, ...,\n", + " 1.26299886e-15, 1.16050682e-15, 1.33837548e-13],\n", + " [5.83394213e-14, 6.39162486e-16, 3.37381926e-15, ...,\n", + " 7.34622834e-16, 6.52683884e-16, 7.18212314e-14],\n", + " ...,\n", + " [5.83394213e-14, 6.39162486e-16, 3.37381926e-15, ...,\n", + " 7.34622834e-16, 6.52683884e-16, 7.18212314e-14],\n", + " [1.11066314e-13, 1.14253903e-15, 5.00865108e-15, ...,\n", + " 1.26299886e-15, 1.16050682e-15, 1.33837548e-13],\n", + " [5.83394213e-14, 6.39162486e-16, 3.37381926e-15, ...,\n", + " 7.34622834e-16, 6.52683884e-16, 7.18212314e-14]], dtype=float32)}, 'Dense_5': {'bias': Array([1.8588816e-08], dtype=float32), 'kernel': Array([[1.77343906e-09],\n", + " [1.85511713e-08],\n", + " [1.06154072e-08],\n", + " [1.85626678e-08],\n", + " [1.77615245e-09],\n", + " [1.07988063e-08],\n", + " [1.85358182e-08],\n", + " [2.77409123e-10],\n", + " [1.85305069e-08],\n", + " [1.85507467e-08],\n", + " [8.39492587e-10],\n", + " [1.07695124e-08],\n", + " [9.13987219e-10],\n", + " [1.84680111e-08],\n", + " [9.68565339e-10],\n", + " [9.59194169e-10],\n", + " [1.07713829e-08],\n", + " [1.07677387e-08],\n", + " [1.84396853e-08],\n", + " [1.07827489e-08],\n", + " [1.79307857e-08],\n", + " [1.83198221e-08],\n", + " [1.83649735e-08],\n", + " [1.54407265e-09],\n", + " [1.85282509e-08],\n", + " [9.41930800e-09],\n", + " [1.81778432e-08],\n", + " [7.20907889e-10],\n", + " [1.05883631e-08],\n", + " [1.85564524e-08],\n", + " [1.85272757e-08],\n", + " [1.85316598e-08],\n", + " [1.82859026e-08],\n", + " [1.07687086e-08],\n", + " [1.07940501e-08],\n", + " [1.84554345e-08],\n", + " [1.80410673e-08],\n", + " [9.08479403e-10],\n", + " [8.23337787e-10],\n", + " [1.76927948e-08],\n", + " [1.07677307e-08],\n", + " [1.78959834e-08],\n", + " [1.83739051e-08],\n", + " [1.83334699e-08],\n", + " [1.83302227e-08],\n", + " [1.82727664e-08],\n", + " [1.07601341e-08],\n", + " [1.85643412e-08],\n", + " [1.62061142e-09],\n", + " [1.82190973e-08],\n", + " [1.84777100e-08],\n", + " [1.07852971e-08],\n", + " [1.84750633e-08],\n", + " [1.83529991e-08],\n", + " [1.07857625e-08],\n", + " [7.96473276e-10],\n", + " [1.08016254e-08],\n", + " [1.82535853e-08],\n", + " [1.85395184e-08],\n", + " [1.81413284e-08],\n", + " [1.75299384e-08],\n", + " [8.07640121e-10],\n", + " [1.06576625e-09],\n", + " [8.56901383e-10],\n", + " [1.47650048e-09],\n", + " [1.81374438e-09],\n", + " [1.85315159e-08],\n", + " [7.28875404e-10],\n", + " [1.07648788e-08],\n", + " [7.21797122e-10],\n", + " [1.07740705e-08],\n", + " [9.99326288e-10],\n", + " [1.08110418e-08],\n", + " [1.84730649e-08],\n", + " [1.45370471e-09],\n", + " [1.07897717e-08],\n", + " [1.22695976e-09],\n", + " [1.18892507e-09],\n", + " [1.07995053e-08],\n", + " [1.81170847e-08],\n", + " [1.85501943e-08],\n", + " [1.46006212e-08],\n", + " [1.84133775e-08],\n", + " [1.85485796e-08],\n", + " [1.85067908e-08],\n", + " [1.82493309e-08],\n", + " [6.89084345e-11],\n", + " [8.28668578e-10],\n", + " [1.07933635e-08],\n", + " [1.85515319e-08],\n", + " [1.85225328e-08],\n", + " [7.60147223e-10],\n", + " [1.85095050e-08],\n", + " [1.85494926e-08],\n", + " [1.08006990e-08],\n", + " [8.32950930e-10],\n", + " [1.85594402e-08],\n", + " [1.85043572e-08],\n", + " [8.32338920e-10],\n", + " [1.06252411e-08],\n", + " [1.84166460e-08],\n", + " [8.87902418e-10],\n", + " [1.30626399e-09],\n", + " [1.90158911e-09],\n", + " [1.52603730e-09],\n", + " [1.85479436e-08],\n", + " [1.81277979e-08],\n", + " [1.85445028e-08],\n", + " [1.07889999e-08],\n", + " [1.08028475e-08],\n", + " [1.07853069e-08],\n", + " [1.83585698e-08],\n", + " [1.66110670e-09],\n", + " [1.93041028e-09],\n", + " [1.07667084e-08],\n", + " [1.07535723e-08],\n", + " [7.87128696e-10],\n", + " [1.75856734e-08],\n", + " [8.67216854e-10],\n", + " [1.84741378e-08],\n", + " [1.85206339e-08],\n", + " [7.79589948e-10],\n", + " [9.78929027e-09],\n", + " [8.62705463e-10],\n", + " [1.06933911e-08],\n", + " [1.13852150e-10],\n", + " [1.07922782e-08],\n", + " [1.07912497e-08],\n", + " [1.85296862e-08],\n", + " [4.91609288e-11],\n", + " [1.82137416e-08],\n", + " [1.07775824e-08],\n", + " [1.34694544e-09],\n", + " [1.07737055e-08],\n", + " [7.94075472e-10],\n", + " [1.07976224e-08],\n", + " [8.57651339e-10],\n", + " [1.06828555e-08],\n", + " [1.21420962e-08],\n", + " [1.79951183e-08],\n", + " [1.78354647e-08],\n", + " [1.84203852e-08],\n", + " [1.07947162e-08],\n", + " [9.85018733e-10],\n", + " [7.12441828e-10],\n", + " [1.82894304e-08],\n", + " [1.81422433e-08],\n", + " [1.42385170e-09],\n", + " [1.01230144e-08],\n", + " [1.07954108e-08],\n", + " [1.29947022e-08],\n", + " [1.04455360e-08],\n", + " [1.76217583e-08],\n", + " [1.07899165e-08],\n", + " [8.64446792e-10],\n", + " [9.82921522e-10],\n", + " [8.10801482e-10],\n", + " [9.68508940e-10],\n", + " [1.08095408e-08],\n", + " [1.05149960e-08],\n", + " [1.07329514e-08],\n", + " [1.07827107e-08],\n", + " [1.78733828e-08],\n", + " [1.07987734e-08],\n", + " [9.37642186e-10],\n", + " [1.91362859e-09],\n", + " [1.84258564e-08],\n", + " [7.46206763e-10],\n", + " [1.85154487e-08],\n", + " [1.06191873e-08],\n", + " [8.05266964e-10],\n", + " [1.84539317e-08],\n", + " [1.84884428e-08],\n", + " [1.07847216e-08],\n", + " [7.65444708e-10],\n", + " [1.07683444e-08],\n", + " [1.03283115e-08],\n", + " [1.07973843e-08],\n", + " [1.08011964e-08],\n", + " [1.82511020e-08],\n", + " [7.32881700e-10],\n", + " [1.56540079e-08],\n", + " [1.85264000e-08],\n", + " [1.07988729e-08],\n", + " [7.56698038e-10],\n", + " [7.45338846e-10],\n", + " [1.84616251e-08],\n", + " [1.07629408e-08],\n", + " [1.81938387e-09],\n", + " [7.08268000e-10],\n", + " [1.00834638e-08],\n", + " [1.07876668e-08],\n", + " [7.60466301e-10],\n", + " [1.84836892e-08],\n", + " [1.82375786e-08],\n", + " [7.23514970e-10],\n", + " [1.07647251e-08],\n", + " [1.85511713e-08],\n", + " [1.69684284e-08],\n", + " [1.07262093e-08],\n", + " [1.25731936e-09],\n", + " [1.85585662e-08],\n", + " [1.07766001e-08],\n", + " [1.07434923e-08],\n", + " [1.07536975e-08],\n", + " [1.03105053e-08],\n", + " [1.84621616e-08],\n", + " [7.61398389e-10],\n", + " [1.84367686e-08],\n", + " [1.07405791e-08],\n", + " [1.40466716e-09],\n", + " [1.08037677e-08],\n", + " [1.07074678e-08],\n", + " [1.58409605e-08],\n", + " [1.85233624e-08],\n", + " [9.66643476e-09],\n", + " [1.85572375e-08],\n", + " [7.64926233e-10],\n", + " [1.85075919e-08],\n", + " [6.48838872e-10],\n", + " [1.41721423e-09],\n", + " [1.07223652e-08],\n", + " [1.83372801e-08],\n", + " [1.72032610e-10],\n", + " [1.01132713e-09],\n", + " [1.85487288e-08],\n", + " [1.84707591e-08],\n", + " [9.31928090e-10],\n", + " [1.84524911e-08],\n", + " [1.81442097e-08],\n", + " [1.84819910e-08],\n", + " [1.26733934e-09],\n", + " [2.92671998e-10],\n", + " [1.85098088e-08],\n", + " [1.85341698e-08],\n", + " [1.85235578e-08],\n", + " [1.03954267e-09],\n", + " [2.07360185e-10],\n", + " [1.53739332e-09],\n", + " [1.07246176e-08],\n", + " [1.07799094e-08],\n", + " [1.07757181e-08],\n", + " [1.85391471e-08],\n", + " [1.07643547e-08],\n", + " [1.85394491e-08],\n", + " [1.07754712e-08],\n", + " [1.85609785e-08],\n", + " [1.23019894e-09],\n", + " [1.40486844e-10],\n", + " [1.84178468e-08],\n", + " [1.11376808e-09],\n", + " [7.20859372e-10],\n", + " [1.85698052e-08],\n", + " [1.85411118e-08],\n", + " [1.85506632e-08],\n", + " [1.58691948e-09]], dtype=float32)}, 'log_std': Array([3.184018e-06], dtype=float32)}}), EmptyState())))" + ] + }, + "execution_count": 8, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "out['runner_state'][0]" + ] + }, + { + "cell_type": "code", + "execution_count": 9, + "metadata": {}, + "outputs": [ + { + "ename": "ValueError", + "evalue": "Destination /tmp/flax_ckpt/orbax/single_save already exists.", + "output_type": "error", + "traceback": [ + "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m", + "\u001b[0;31mValueError\u001b[0m Traceback (most recent call last)", + "Cell \u001b[0;32mIn[9], line 6\u001b[0m\n\u001b[1;32m 4\u001b[0m ckpt \u001b[38;5;241m=\u001b[39m {\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mmodel\u001b[39m\u001b[38;5;124m'\u001b[39m: out[\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mrunner_state\u001b[39m\u001b[38;5;124m'\u001b[39m][\u001b[38;5;241m0\u001b[39m]}\n\u001b[1;32m 5\u001b[0m save_args \u001b[38;5;241m=\u001b[39m orbax_utils\u001b[38;5;241m.\u001b[39msave_args_from_target(ckpt)\n\u001b[0;32m----> 6\u001b[0m \u001b[43morbax_checkpointer\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43msave\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;124;43m'\u001b[39;49m\u001b[38;5;124;43m/tmp/flax_ckpt/orbax/single_save\u001b[39;49m\u001b[38;5;124;43m'\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mckpt\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43msave_args\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43msave_args\u001b[49m\u001b[43m)\u001b[49m\n", + "File \u001b[0;32m~/Documents/Code/solarcarsim/.venv/lib/python3.12/site-packages/orbax/checkpoint/_src/checkpointers/checkpointer.py:201\u001b[0m, in \u001b[0;36mCheckpointer.save\u001b[0;34m(self, directory, force, *args, **kwargs)\u001b[0m\n\u001b[1;32m 199\u001b[0m directory\u001b[38;5;241m.\u001b[39mrmtree() \u001b[38;5;66;03m# Post-sync handled by create_tmp_directory.\u001b[39;00m\n\u001b[1;32m 200\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[0;32m--> 201\u001b[0m \u001b[38;5;28;01mraise\u001b[39;00m \u001b[38;5;167;01mValueError\u001b[39;00m(\u001b[38;5;124mf\u001b[39m\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mDestination \u001b[39m\u001b[38;5;132;01m{\u001b[39;00mdirectory\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m already exists.\u001b[39m\u001b[38;5;124m'\u001b[39m)\n\u001b[1;32m 202\u001b[0m ckpt_args \u001b[38;5;241m=\u001b[39m construct_checkpoint_args(\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_handler, \u001b[38;5;28;01mTrue\u001b[39;00m, \u001b[38;5;241m*\u001b[39margs, \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mkwargs)\n\u001b[1;32m 203\u001b[0m tmpdir \u001b[38;5;241m=\u001b[39m asyncio_utils\u001b[38;5;241m.\u001b[39mrun_sync(\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mcreate_temporary_path(directory))\n", + "\u001b[0;31mValueError\u001b[0m: Destination /tmp/flax_ckpt/orbax/single_save already exists." + ] + } + ], + "source": [ + "import orbax.checkpoint\n", + "from flax.training import orbax_utils\n", + "orbax_checkpointer = orbax.checkpoint.PyTreeCheckpointer()\n", + "ckpt = {'model': out['runner_state'][0]}\n", + "save_args = orbax_utils.save_args_from_target(ckpt)\n", + "orbax_checkpointer.save('/tmp/flax_ckpt/orbax/single_save', ckpt, save_args=save_args)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [] } ], "metadata": { diff --git a/notebooks/v1gym.ipynb b/notebooks/v1gym.ipynb index fb0c02d..06d4d5d 100644 --- a/notebooks/v1gym.ipynb +++ b/notebooks/v1gym.ipynb @@ -10,8 +10,11 @@ "from gymnasium.wrappers.jax_to_numpy import JaxToNumpy\n", "from gymnasium.wrappers.vector import JaxToNumpy as VJaxToNumpy\n", "from solarcarsim.simv1 import SolarRaceV1\n", - "from stable_baselines3.common.env_checker import check_env\n", "from gymnasium.utils.env_checker import check_env as gym_check_env\n", + "from stable_baselines3 import TD3, PPO\n", + "from stable_baselines3.common.env_checker import check_env\n", + "import matplotlib.pyplot as plt\n", + "import jax.numpy as jnp\n", "env = SolarRaceV1()\n", "wrapped_env = JaxToNumpy(env)" ] @@ -25,8 +28,6 @@ "name": "stderr", "output_type": "stream", "text": [ - "/home/saji/Documents/Code/solarcarsim/.venv/lib/python3.12/site-packages/stable_baselines3/common/env_checker.py:271: UserWarning: Your observation wind has an unconventional shape (neither an image, nor a 1D vector). We recommend you to flatten the observation to have only a 1D vector or use a custom policy to properly process the data.\n", - " warnings.warn(\n", "/home/saji/Documents/Code/solarcarsim/.venv/lib/python3.12/site-packages/gymnasium/utils/env_checker.py:384: UserWarning: \u001b[33mWARN: The environment (>) is different from the unwrapped version (). This could effect the environment checker as the environment most likely has a wrapper applied to it. We recommend using the raw environment for `check_env` using `env.unwrapped`.\u001b[0m\n", " logger.warn(\n", "/home/saji/Documents/Code/solarcarsim/.venv/lib/python3.12/site-packages/gymnasium/utils/env_checker.py:434: UserWarning: \u001b[33mWARN: Not able to test alternative render modes due to the environment not having a spec. Try instantiating the environment through `gymnasium.make`\u001b[0m\n", @@ -35,31 +36,2517 @@ } ], "source": [ - "env.reset()\n", - "check_env(wrapped_env)\n", "gym_check_env(wrapped_env)" ] }, { "cell_type": "code", - "execution_count": 3, + "execution_count": null, "metadata": {}, "outputs": [ - { - "name": "stderr", - "output_type": "stream", - "text": [ - "/home/saji/Documents/Code/solarcarsim/.venv/lib/python3.12/site-packages/stable_baselines3/common/buffers.py:605: UserWarning: This system does not have apparently enough memory to store the complete replay buffer 80.85GB > 53.66GB\n", - " warnings.warn(\n" - ] - }, { "name": "stdout", "output_type": "stream", "text": [ "Using cuda device\n", "Wrapping the env with a `Monitor` wrapper\n", - "Wrapping the env in a DummyVecEnv.\n" + "Wrapping the env in a DummyVecEnv.\n", + "---------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -293 |\n", + "| time/ | |\n", + "| fps | 181 |\n", + "| iterations | 1 |\n", + "| time_elapsed | 11 |\n", + "| total_timesteps | 2048 |\n", + "---------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -273 |\n", + "| time/ | |\n", + "| fps | 174 |\n", + "| iterations | 2 |\n", + "| time_elapsed | 23 |\n", + "| total_timesteps | 4096 |\n", + "| train/ | |\n", + "| approx_kl | 0.0054363105 |\n", + "| clip_fraction | 0.036 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.42 |\n", + "| explained_variance | -0.000109 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 3.72e+03 |\n", + "| n_updates | 10 |\n", + "| policy_gradient_loss | 0.00132 |\n", + "| std | 1 |\n", + "| value_loss | 3.03e+03 |\n", + "------------------------------------------\n", + "-----------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -269 |\n", + "| time/ | |\n", + "| fps | 174 |\n", + "| iterations | 3 |\n", + "| time_elapsed | 35 |\n", + "| total_timesteps | 6144 |\n", + "| train/ | |\n", + "| approx_kl | 0.011383371 |\n", + "| clip_fraction | 0.119 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.42 |\n", + "| explained_variance | 8.55e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 2.63e+03 |\n", + "| n_updates | 20 |\n", + "| policy_gradient_loss | -0.00406 |\n", + "| std | 0.998 |\n", + "| value_loss | 3.05e+03 |\n", + "-----------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -267 |\n", + "| time/ | |\n", + "| fps | 173 |\n", + "| iterations | 4 |\n", + "| time_elapsed | 47 |\n", + "| total_timesteps | 8192 |\n", + "| train/ | |\n", + "| approx_kl | 0.0013208076 |\n", + "| clip_fraction | 0.002 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.41 |\n", + "| explained_variance | 0.000122 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 931 |\n", + "| n_updates | 30 |\n", + "| policy_gradient_loss | 8.3e-05 |\n", + "| std | 0.99 |\n", + "| value_loss | 4.03e+03 |\n", + "------------------------------------------\n", + "-----------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -272 |\n", + "| time/ | |\n", + "| fps | 172 |\n", + "| iterations | 5 |\n", + "| time_elapsed | 59 |\n", + "| total_timesteps | 10240 |\n", + "| train/ | |\n", + "| approx_kl | 0.012045372 |\n", + "| clip_fraction | 0.0221 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.4 |\n", + "| explained_variance | 5.51e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 1.17e+03 |\n", + "| n_updates | 40 |\n", + "| policy_gradient_loss | 0.00043 |\n", + "| std | 0.973 |\n", + "| value_loss | 3.06e+03 |\n", + "-----------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -272 |\n", + "| time/ | |\n", + "| fps | 172 |\n", + "| iterations | 6 |\n", + "| time_elapsed | 71 |\n", + "| total_timesteps | 12288 |\n", + "| train/ | |\n", + "| approx_kl | 0.0049332893 |\n", + "| clip_fraction | 0.0111 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.39 |\n", + "| explained_variance | 0.000114 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 7.39e+03 |\n", + "| n_updates | 50 |\n", + "| policy_gradient_loss | -0.00083 |\n", + "| std | 0.973 |\n", + "| value_loss | 4.05e+03 |\n", + "------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -275 |\n", + "| time/ | |\n", + "| fps | 171 |\n", + "| iterations | 7 |\n", + "| time_elapsed | 83 |\n", + "| total_timesteps | 14336 |\n", + "| train/ | |\n", + "| approx_kl | 0.0038162381 |\n", + "| clip_fraction | 0.0192 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.39 |\n", + "| explained_variance | 7.37e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 1.5e+03 |\n", + "| n_updates | 60 |\n", + "| policy_gradient_loss | -0.000316 |\n", + "| std | 0.971 |\n", + "| value_loss | 3.06e+03 |\n", + "------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -274 |\n", + "| time/ | |\n", + "| fps | 170 |\n", + "| iterations | 8 |\n", + "| time_elapsed | 96 |\n", + "| total_timesteps | 16384 |\n", + "| train/ | |\n", + "| approx_kl | 0.0039417995 |\n", + "| clip_fraction | 0.0062 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.39 |\n", + "| explained_variance | 7.75e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 2.76e+03 |\n", + "| n_updates | 70 |\n", + "| policy_gradient_loss | -0.000468 |\n", + "| std | 0.973 |\n", + "| value_loss | 3.08e+03 |\n", + "------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -275 |\n", + "| time/ | |\n", + "| fps | 170 |\n", + "| iterations | 9 |\n", + "| time_elapsed | 108 |\n", + "| total_timesteps | 18432 |\n", + "| train/ | |\n", + "| approx_kl | 0.0017004285 |\n", + "| clip_fraction | 0.0129 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.4 |\n", + "| explained_variance | 0.000155 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 756 |\n", + "| n_updates | 80 |\n", + "| policy_gradient_loss | -0.000469 |\n", + "| std | 0.98 |\n", + "| value_loss | 4.04e+03 |\n", + "------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -277 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 10 |\n", + "| time_elapsed | 121 |\n", + "| total_timesteps | 20480 |\n", + "| train/ | |\n", + "| approx_kl | 0.0034604114 |\n", + "| clip_fraction | 0.0167 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.41 |\n", + "| explained_variance | 0.000104 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 2.09e+03 |\n", + "| n_updates | 90 |\n", + "| policy_gradient_loss | -0.00122 |\n", + "| std | 0.995 |\n", + "| value_loss | 3.07e+03 |\n", + "------------------------------------------\n", + "-----------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -276 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 11 |\n", + "| time_elapsed | 133 |\n", + "| total_timesteps | 22528 |\n", + "| train/ | |\n", + "| approx_kl | 0.005835003 |\n", + "| clip_fraction | 0.0289 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.41 |\n", + "| explained_variance | 0.000224 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 1.44e+03 |\n", + "| n_updates | 100 |\n", + "| policy_gradient_loss | -0.00135 |\n", + "| std | 0.985 |\n", + "| value_loss | 4.06e+03 |\n", + "-----------------------------------------\n", + "-------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -276 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 12 |\n", + "| time_elapsed | 145 |\n", + "| total_timesteps | 24576 |\n", + "| train/ | |\n", + "| approx_kl | 0.00068298285 |\n", + "| clip_fraction | 0 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.4 |\n", + "| explained_variance | 0.000121 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 3.33e+03 |\n", + "| n_updates | 110 |\n", + "| policy_gradient_loss | 7.82e-05 |\n", + "| std | 0.982 |\n", + "| value_loss | 3.08e+03 |\n", + "-------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -274 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 13 |\n", + "| time_elapsed | 158 |\n", + "| total_timesteps | 26624 |\n", + "| train/ | |\n", + "| approx_kl | 0.0048444057 |\n", + "| clip_fraction | 0.00918 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.4 |\n", + "| explained_variance | 0.000109 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 1.07e+03 |\n", + "| n_updates | 120 |\n", + "| policy_gradient_loss | 0.000241 |\n", + "| std | 0.973 |\n", + "| value_loss | 3.09e+03 |\n", + "------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -272 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 14 |\n", + "| time_elapsed | 170 |\n", + "| total_timesteps | 28672 |\n", + "| train/ | |\n", + "| approx_kl | 0.0024140258 |\n", + "| clip_fraction | 0.0194 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.39 |\n", + "| explained_variance | 0.000147 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 4.1e+03 |\n", + "| n_updates | 130 |\n", + "| policy_gradient_loss | -0.000116 |\n", + "| std | 0.974 |\n", + "| value_loss | 4.07e+03 |\n", + "------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -271 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 15 |\n", + "| time_elapsed | 182 |\n", + "| total_timesteps | 30720 |\n", + "| train/ | |\n", + "| approx_kl | 0.0012023712 |\n", + "| clip_fraction | 0.0306 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.39 |\n", + "| explained_variance | 1.91e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 2.61e+03 |\n", + "| n_updates | 140 |\n", + "| policy_gradient_loss | 0.000104 |\n", + "| std | 0.964 |\n", + "| value_loss | 3.09e+03 |\n", + "------------------------------------------\n", + "-----------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -270 |\n", + "| time/ | |\n", + "| fps | 167 |\n", + "| iterations | 16 |\n", + "| time_elapsed | 195 |\n", + "| total_timesteps | 32768 |\n", + "| train/ | |\n", + "| approx_kl | 0.005513249 |\n", + "| clip_fraction | 0.0216 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.37 |\n", + "| explained_variance | 6.94e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 647 |\n", + "| n_updates | 150 |\n", + "| policy_gradient_loss | -0.00111 |\n", + "| std | 0.948 |\n", + "| value_loss | 4.06e+03 |\n", + "-----------------------------------------\n", + "-----------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -272 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 17 |\n", + "| time_elapsed | 207 |\n", + "| total_timesteps | 34816 |\n", + "| train/ | |\n", + "| approx_kl | 0.008722976 |\n", + "| clip_fraction | 0.0229 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.37 |\n", + "| explained_variance | 4.42e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 2.42e+03 |\n", + "| n_updates | 160 |\n", + "| policy_gradient_loss | -0.000433 |\n", + "| std | 0.945 |\n", + "| value_loss | 3.08e+03 |\n", + "-----------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -276 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 18 |\n", + "| time_elapsed | 219 |\n", + "| total_timesteps | 36864 |\n", + "| train/ | |\n", + "| approx_kl | 0.0060544205 |\n", + "| clip_fraction | 0.0893 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.37 |\n", + "| explained_variance | 2.23e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 593 |\n", + "| n_updates | 170 |\n", + "| policy_gradient_loss | -0.00357 |\n", + "| std | 0.952 |\n", + "| value_loss | 3.1e+03 |\n", + "------------------------------------------\n", + "-----------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -277 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 19 |\n", + "| time_elapsed | 231 |\n", + "| total_timesteps | 38912 |\n", + "| train/ | |\n", + "| approx_kl | 0.006287749 |\n", + "| clip_fraction | 0.00957 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.37 |\n", + "| explained_variance | 9.32e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 3.89e+03 |\n", + "| n_updates | 180 |\n", + "| policy_gradient_loss | -0.000475 |\n", + "| std | 0.952 |\n", + "| value_loss | 4.06e+03 |\n", + "-----------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -278 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 20 |\n", + "| time_elapsed | 243 |\n", + "| total_timesteps | 40960 |\n", + "| train/ | |\n", + "| approx_kl | 0.0006410396 |\n", + "| clip_fraction | 0.00317 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.38 |\n", + "| explained_variance | 1.19e-06 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 609 |\n", + "| n_updates | 190 |\n", + "| policy_gradient_loss | 0.000116 |\n", + "| std | 0.963 |\n", + "| value_loss | 3.09e+03 |\n", + "------------------------------------------\n", + "-------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -279 |\n", + "| time/ | |\n", + "| fps | 167 |\n", + "| iterations | 21 |\n", + "| time_elapsed | 256 |\n", + "| total_timesteps | 43008 |\n", + "| train/ | |\n", + "| approx_kl | 0.00017864068 |\n", + "| clip_fraction | 0.0233 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.38 |\n", + "| explained_variance | 1.61e-06 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 1.26e+03 |\n", + "| n_updates | 200 |\n", + "| policy_gradient_loss | 1.53e-05 |\n", + "| std | 0.961 |\n", + "| value_loss | 4.07e+03 |\n", + "-------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -280 |\n", + "| time/ | |\n", + "| fps | 167 |\n", + "| iterations | 22 |\n", + "| time_elapsed | 268 |\n", + "| total_timesteps | 45056 |\n", + "| train/ | |\n", + "| approx_kl | 0.0052862475 |\n", + "| clip_fraction | 0.0678 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.38 |\n", + "| explained_variance | 4.77e-07 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 154 |\n", + "| n_updates | 210 |\n", + "| policy_gradient_loss | 0.000592 |\n", + "| std | 0.962 |\n", + "| value_loss | 3.1e+03 |\n", + "------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -280 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 23 |\n", + "| time_elapsed | 280 |\n", + "| total_timesteps | 47104 |\n", + "| train/ | |\n", + "| approx_kl | 0.0017830351 |\n", + "| clip_fraction | 0.0224 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.38 |\n", + "| explained_variance | 5.36e-07 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 719 |\n", + "| n_updates | 220 |\n", + "| policy_gradient_loss | -0.000802 |\n", + "| std | 0.961 |\n", + "| value_loss | 3.11e+03 |\n", + "------------------------------------------\n", + "-------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -281 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 24 |\n", + "| time_elapsed | 292 |\n", + "| total_timesteps | 49152 |\n", + "| train/ | |\n", + "| approx_kl | 0.00015185933 |\n", + "| clip_fraction | 0 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.37 |\n", + "| explained_variance | 1.01e-06 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 2.93e+03 |\n", + "| n_updates | 230 |\n", + "| policy_gradient_loss | -3.24e-05 |\n", + "| std | 0.951 |\n", + "| value_loss | 4.01e+03 |\n", + "-------------------------------------------\n", + "----------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -282 |\n", + "| time/ | |\n", + "| fps | 167 |\n", + "| iterations | 25 |\n", + "| time_elapsed | 305 |\n", + "| total_timesteps | 51200 |\n", + "| train/ | |\n", + "| approx_kl | 0.00283485 |\n", + "| clip_fraction | 0.0281 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.37 |\n", + "| explained_variance | 4.17e-07 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 1.26e+03 |\n", + "| n_updates | 240 |\n", + "| policy_gradient_loss | -0.00137 |\n", + "| std | 0.954 |\n", + "| value_loss | 3.11e+03 |\n", + "----------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -282 |\n", + "| time/ | |\n", + "| fps | 167 |\n", + "| iterations | 26 |\n", + "| time_elapsed | 317 |\n", + "| total_timesteps | 53248 |\n", + "| train/ | |\n", + "| approx_kl | 0.0021635226 |\n", + "| clip_fraction | 0.0132 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.37 |\n", + "| explained_variance | 8.34e-07 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 321 |\n", + "| n_updates | 250 |\n", + "| policy_gradient_loss | -0.000334 |\n", + "| std | 0.955 |\n", + "| value_loss | 4.07e+03 |\n", + "------------------------------------------\n", + "-----------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -283 |\n", + "| time/ | |\n", + "| fps | 167 |\n", + "| iterations | 27 |\n", + "| time_elapsed | 330 |\n", + "| total_timesteps | 55296 |\n", + "| train/ | |\n", + "| approx_kl | 0.012732552 |\n", + "| clip_fraction | 0.0278 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.38 |\n", + "| explained_variance | 2.98e-07 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 250 |\n", + "| n_updates | 260 |\n", + "| policy_gradient_loss | -0.00149 |\n", + "| std | 0.963 |\n", + "| value_loss | 3.1e+03 |\n", + "-----------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -284 |\n", + "| time/ | |\n", + "| fps | 167 |\n", + "| iterations | 28 |\n", + "| time_elapsed | 342 |\n", + "| total_timesteps | 57344 |\n", + "| train/ | |\n", + "| approx_kl | 0.0035805362 |\n", + "| clip_fraction | 0.0155 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.38 |\n", + "| explained_variance | 8.34e-07 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 4.12e+03 |\n", + "| n_updates | 270 |\n", + "| policy_gradient_loss | -0.000792 |\n", + "| std | 0.966 |\n", + "| value_loss | 4.09e+03 |\n", + "------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -285 |\n", + "| time/ | |\n", + "| fps | 167 |\n", + "| iterations | 29 |\n", + "| time_elapsed | 354 |\n", + "| total_timesteps | 59392 |\n", + "| train/ | |\n", + "| approx_kl | 0.0018168361 |\n", + "| clip_fraction | 0.000488 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.39 |\n", + "| explained_variance | 4.77e-07 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 2.16e+03 |\n", + "| n_updates | 280 |\n", + "| policy_gradient_loss | 4.89e-05 |\n", + "| std | 0.971 |\n", + "| value_loss | 3.1e+03 |\n", + "------------------------------------------\n", + "-------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -285 |\n", + "| time/ | |\n", + "| fps | 167 |\n", + "| iterations | 30 |\n", + "| time_elapsed | 366 |\n", + "| total_timesteps | 61440 |\n", + "| train/ | |\n", + "| approx_kl | 0.00029722328 |\n", + "| clip_fraction | 0.000635 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.38 |\n", + "| explained_variance | 1.79e-07 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 1.02e+03 |\n", + "| n_updates | 290 |\n", + "| policy_gradient_loss | -0.00093 |\n", + "| std | 0.957 |\n", + "| value_loss | 3.1e+03 |\n", + "-------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -285 |\n", + "| time/ | |\n", + "| fps | 167 |\n", + "| iterations | 31 |\n", + "| time_elapsed | 378 |\n", + "| total_timesteps | 63488 |\n", + "| train/ | |\n", + "| approx_kl | 0.0036160094 |\n", + "| clip_fraction | 0.00591 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.37 |\n", + "| explained_variance | 7.15e-07 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 1.69e+03 |\n", + "| n_updates | 300 |\n", + "| policy_gradient_loss | -0.000226 |\n", + "| std | 0.953 |\n", + "| value_loss | 4.08e+03 |\n", + "------------------------------------------\n", + "-------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -286 |\n", + "| time/ | |\n", + "| fps | 167 |\n", + "| iterations | 32 |\n", + "| time_elapsed | 390 |\n", + "| total_timesteps | 65536 |\n", + "| train/ | |\n", + "| approx_kl | 0.00017739431 |\n", + "| clip_fraction | 0.0329 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.38 |\n", + "| explained_variance | 1.79e-07 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 3.15e+03 |\n", + "| n_updates | 310 |\n", + "| policy_gradient_loss | 0.000172 |\n", + "| std | 0.965 |\n", + "| value_loss | 3.11e+03 |\n", + "-------------------------------------------\n", + "-----------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -287 |\n", + "| time/ | |\n", + "| fps | 167 |\n", + "| iterations | 33 |\n", + "| time_elapsed | 402 |\n", + "| total_timesteps | 67584 |\n", + "| train/ | |\n", + "| approx_kl | 0.004563484 |\n", + "| clip_fraction | 0.0295 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.39 |\n", + "| explained_variance | 9.54e-07 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 1.55e+03 |\n", + "| n_updates | 320 |\n", + "| policy_gradient_loss | -0.00134 |\n", + "| std | 0.972 |\n", + "| value_loss | 4.09e+03 |\n", + "-----------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -287 |\n", + "| time/ | |\n", + "| fps | 167 |\n", + "| iterations | 34 |\n", + "| time_elapsed | 414 |\n", + "| total_timesteps | 69632 |\n", + "| train/ | |\n", + "| approx_kl | 0.0049857013 |\n", + "| clip_fraction | 0.018 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.39 |\n", + "| explained_variance | 7.15e-07 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 2.61e+03 |\n", + "| n_updates | 330 |\n", + "| policy_gradient_loss | -0.0015 |\n", + "| std | 0.966 |\n", + "| value_loss | 3.1e+03 |\n", + "------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -286 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 35 |\n", + "| time_elapsed | 426 |\n", + "| total_timesteps | 71680 |\n", + "| train/ | |\n", + "| approx_kl | 0.0012287534 |\n", + "| clip_fraction | 0.000879 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.38 |\n", + "| explained_variance | 3.46e-06 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 2.31e+03 |\n", + "| n_updates | 340 |\n", + "| policy_gradient_loss | -0.00019 |\n", + "| std | 0.969 |\n", + "| value_loss | 3.11e+03 |\n", + "------------------------------------------\n", + "-------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -287 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 36 |\n", + "| time_elapsed | 438 |\n", + "| total_timesteps | 73728 |\n", + "| train/ | |\n", + "| approx_kl | 3.8835948e-05 |\n", + "| clip_fraction | 0 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.38 |\n", + "| explained_variance | 4.41e-06 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 2.76e+03 |\n", + "| n_updates | 350 |\n", + "| policy_gradient_loss | 0.000106 |\n", + "| std | 0.967 |\n", + "| value_loss | 4.07e+03 |\n", + "-------------------------------------------\n", + "-----------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -285 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 37 |\n", + "| time_elapsed | 450 |\n", + "| total_timesteps | 75776 |\n", + "| train/ | |\n", + "| approx_kl | 0.004455052 |\n", + "| clip_fraction | 0.0113 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.39 |\n", + "| explained_variance | 5.01e-06 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 4.84e+03 |\n", + "| n_updates | 360 |\n", + "| policy_gradient_loss | -0.00126 |\n", + "| std | 0.976 |\n", + "| value_loss | 3.11e+03 |\n", + "-----------------------------------------\n", + "-----------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -285 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 38 |\n", + "| time_elapsed | 462 |\n", + "| total_timesteps | 77824 |\n", + "| train/ | |\n", + "| approx_kl | 0.004241547 |\n", + "| clip_fraction | 0.0108 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.4 |\n", + "| explained_variance | 7.57e-06 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 565 |\n", + "| n_updates | 370 |\n", + "| policy_gradient_loss | -0.000582 |\n", + "| std | 0.98 |\n", + "| value_loss | 4.09e+03 |\n", + "-----------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -285 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 39 |\n", + "| time_elapsed | 474 |\n", + "| total_timesteps | 79872 |\n", + "| train/ | |\n", + "| approx_kl | 0.0017373057 |\n", + "| clip_fraction | 0.00103 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.4 |\n", + "| explained_variance | 5.19e-06 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 1.68e+03 |\n", + "| n_updates | 380 |\n", + "| policy_gradient_loss | 7.21e-05 |\n", + "| std | 0.981 |\n", + "| value_loss | 3.11e+03 |\n", + "------------------------------------------\n", + "-------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -285 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 40 |\n", + "| time_elapsed | 486 |\n", + "| total_timesteps | 81920 |\n", + "| train/ | |\n", + "| approx_kl | 0.00016679132 |\n", + "| clip_fraction | 0.0324 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.4 |\n", + "| explained_variance | 6.26e-06 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 2.55e+03 |\n", + "| n_updates | 390 |\n", + "| policy_gradient_loss | 4.64e-06 |\n", + "| std | 0.991 |\n", + "| value_loss | 3.11e+03 |\n", + "-------------------------------------------\n", + "-------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -286 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 41 |\n", + "| time_elapsed | 497 |\n", + "| total_timesteps | 83968 |\n", + "| train/ | |\n", + "| approx_kl | 4.9029622e-05 |\n", + "| clip_fraction | 0 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.41 |\n", + "| explained_variance | 1.26e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 3.1e+03 |\n", + "| n_updates | 400 |\n", + "| policy_gradient_loss | -0.000107 |\n", + "| std | 0.987 |\n", + "| value_loss | 4.08e+03 |\n", + "-------------------------------------------\n", + "-----------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -288 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 42 |\n", + "| time_elapsed | 509 |\n", + "| total_timesteps | 86016 |\n", + "| train/ | |\n", + "| approx_kl | 0.008285521 |\n", + "| clip_fraction | 0.0146 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.4 |\n", + "| explained_variance | 4.95e-06 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 5.63e+03 |\n", + "| n_updates | 410 |\n", + "| policy_gradient_loss | -0.000514 |\n", + "| std | 0.983 |\n", + "| value_loss | 3.11e+03 |\n", + "-----------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -290 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 43 |\n", + "| time_elapsed | 521 |\n", + "| total_timesteps | 88064 |\n", + "| train/ | |\n", + "| approx_kl | 0.0044103963 |\n", + "| clip_fraction | 0.0221 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.41 |\n", + "| explained_variance | 1.2e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 3.19e+03 |\n", + "| n_updates | 420 |\n", + "| policy_gradient_loss | 0.000946 |\n", + "| std | 0.989 |\n", + "| value_loss | 4.09e+03 |\n", + "------------------------------------------\n", + "-----------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -292 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 44 |\n", + "| time_elapsed | 533 |\n", + "| total_timesteps | 90112 |\n", + "| train/ | |\n", + "| approx_kl | 0.005043611 |\n", + "| clip_fraction | 0.0923 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.41 |\n", + "| explained_variance | 4.23e-06 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 983 |\n", + "| n_updates | 430 |\n", + "| policy_gradient_loss | -0.00239 |\n", + "| std | 0.989 |\n", + "| value_loss | 3.11e+03 |\n", + "-----------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -295 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 45 |\n", + "| time_elapsed | 546 |\n", + "| total_timesteps | 92160 |\n", + "| train/ | |\n", + "| approx_kl | 0.0047482466 |\n", + "| clip_fraction | 0.0449 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.4 |\n", + "| explained_variance | 4.65e-06 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 1.78e+03 |\n", + "| n_updates | 440 |\n", + "| policy_gradient_loss | -0.000126 |\n", + "| std | 0.98 |\n", + "| value_loss | 3.11e+03 |\n", + "------------------------------------------\n", + "----------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -296 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 46 |\n", + "| time_elapsed | 558 |\n", + "| total_timesteps | 94208 |\n", + "| train/ | |\n", + "| approx_kl | 0.00606206 |\n", + "| clip_fraction | 0.0219 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.4 |\n", + "| explained_variance | 9.48e-06 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 3.98e+03 |\n", + "| n_updates | 450 |\n", + "| policy_gradient_loss | -0.000753 |\n", + "| std | 0.985 |\n", + "| value_loss | 4.09e+03 |\n", + "----------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -296 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 47 |\n", + "| time_elapsed | 570 |\n", + "| total_timesteps | 96256 |\n", + "| train/ | |\n", + "| approx_kl | 0.0008331981 |\n", + "| clip_fraction | 0.013 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.41 |\n", + "| explained_variance | 3.7e-06 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 2.69e+03 |\n", + "| n_updates | 460 |\n", + "| policy_gradient_loss | 2.58e-05 |\n", + "| std | 0.997 |\n", + "| value_loss | 3.11e+03 |\n", + "------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -295 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 48 |\n", + "| time_elapsed | 582 |\n", + "| total_timesteps | 98304 |\n", + "| train/ | |\n", + "| approx_kl | 0.0005460837 |\n", + "| clip_fraction | 0.00146 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.41 |\n", + "| explained_variance | 8.94e-06 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 578 |\n", + "| n_updates | 470 |\n", + "| policy_gradient_loss | 0.00032 |\n", + "| std | 0.998 |\n", + "| value_loss | 4.09e+03 |\n", + "------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -295 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 49 |\n", + "| time_elapsed | 595 |\n", + "| total_timesteps | 100352 |\n", + "| train/ | |\n", + "| approx_kl | 0.0009762709 |\n", + "| clip_fraction | 0.0345 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.41 |\n", + "| explained_variance | 2.98e-06 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 2.52e+03 |\n", + "| n_updates | 480 |\n", + "| policy_gradient_loss | -0.00181 |\n", + "| std | 0.997 |\n", + "| value_loss | 3.11e+03 |\n", + "------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -296 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 50 |\n", + "| time_elapsed | 607 |\n", + "| total_timesteps | 102400 |\n", + "| train/ | |\n", + "| approx_kl | 0.0003773085 |\n", + "| clip_fraction | 0.00215 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.42 |\n", + "| explained_variance | 4.05e-06 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 1.89e+03 |\n", + "| n_updates | 490 |\n", + "| policy_gradient_loss | 0.000501 |\n", + "| std | 1.01 |\n", + "| value_loss | 3.12e+03 |\n", + "------------------------------------------\n", + "-----------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -297 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 51 |\n", + "| time_elapsed | 619 |\n", + "| total_timesteps | 104448 |\n", + "| train/ | |\n", + "| approx_kl | 0.007983657 |\n", + "| clip_fraction | 0.0524 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.43 |\n", + "| explained_variance | 1.47e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 4.23e+03 |\n", + "| n_updates | 500 |\n", + "| policy_gradient_loss | -0.000208 |\n", + "| std | 1.01 |\n", + "| value_loss | 3.96e+03 |\n", + "-----------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -298 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 52 |\n", + "| time_elapsed | 631 |\n", + "| total_timesteps | 106496 |\n", + "| train/ | |\n", + "| approx_kl | 0.0004374912 |\n", + "| clip_fraction | 0.0302 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.43 |\n", + "| explained_variance | 9.12e-06 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 1.55e+03 |\n", + "| n_updates | 510 |\n", + "| policy_gradient_loss | -3.42e-05 |\n", + "| std | 1.01 |\n", + "| value_loss | 3.11e+03 |\n", + "------------------------------------------\n", + "-----------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -300 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 53 |\n", + "| time_elapsed | 643 |\n", + "| total_timesteps | 108544 |\n", + "| train/ | |\n", + "| approx_kl | 0.005380518 |\n", + "| clip_fraction | 0.0136 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.43 |\n", + "| explained_variance | 1.67e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 1.03e+03 |\n", + "| n_updates | 520 |\n", + "| policy_gradient_loss | 0.000128 |\n", + "| std | 1.01 |\n", + "| value_loss | 4.09e+03 |\n", + "-----------------------------------------\n", + "----------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -303 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 54 |\n", + "| time_elapsed | 655 |\n", + "| total_timesteps | 110592 |\n", + "| train/ | |\n", + "| approx_kl | 0.00812779 |\n", + "| clip_fraction | 0.0241 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.43 |\n", + "| explained_variance | 4.89e-06 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 563 |\n", + "| n_updates | 530 |\n", + "| policy_gradient_loss | -0.00053 |\n", + "| std | 1.01 |\n", + "| value_loss | 3.11e+03 |\n", + "----------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -305 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 55 |\n", + "| time_elapsed | 668 |\n", + "| total_timesteps | 112640 |\n", + "| train/ | |\n", + "| approx_kl | 0.0014875259 |\n", + "| clip_fraction | 0.00767 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.43 |\n", + "| explained_variance | 8.34e-06 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 2.67e+03 |\n", + "| n_updates | 540 |\n", + "| policy_gradient_loss | -0.000391 |\n", + "| std | 1.01 |\n", + "| value_loss | 4.1e+03 |\n", + "------------------------------------------\n", + "----------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -308 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 56 |\n", + "| time_elapsed | 680 |\n", + "| total_timesteps | 114688 |\n", + "| train/ | |\n", + "| approx_kl | 0.00965928 |\n", + "| clip_fraction | 0.092 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.43 |\n", + "| explained_variance | 3.28e-06 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 2.47e+03 |\n", + "| n_updates | 550 |\n", + "| policy_gradient_loss | -0.00266 |\n", + "| std | 1.02 |\n", + "| value_loss | 3.11e+03 |\n", + "----------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -311 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 57 |\n", + "| time_elapsed | 692 |\n", + "| total_timesteps | 116736 |\n", + "| train/ | |\n", + "| approx_kl | 0.0022691623 |\n", + "| clip_fraction | 0.0141 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.44 |\n", + "| explained_variance | 3.4e-06 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 936 |\n", + "| n_updates | 560 |\n", + "| policy_gradient_loss | -0.000543 |\n", + "| std | 1.01 |\n", + "| value_loss | 3.12e+03 |\n", + "------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -313 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 58 |\n", + "| time_elapsed | 704 |\n", + "| total_timesteps | 118784 |\n", + "| train/ | |\n", + "| approx_kl | 0.0005463155 |\n", + "| clip_fraction | 0.00444 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.43 |\n", + "| explained_variance | 5.13e-06 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 2.02e+03 |\n", + "| n_updates | 570 |\n", + "| policy_gradient_loss | -0.000174 |\n", + "| std | 1.02 |\n", + "| value_loss | 4.1e+03 |\n", + "------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -317 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 59 |\n", + "| time_elapsed | 716 |\n", + "| total_timesteps | 120832 |\n", + "| train/ | |\n", + "| approx_kl | 0.0037943618 |\n", + "| clip_fraction | 0.0239 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.44 |\n", + "| explained_variance | 2.09e-06 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 372 |\n", + "| n_updates | 580 |\n", + "| policy_gradient_loss | -2.05e-05 |\n", + "| std | 1.02 |\n", + "| value_loss | 3.12e+03 |\n", + "------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -321 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 60 |\n", + "| time_elapsed | 728 |\n", + "| total_timesteps | 122880 |\n", + "| train/ | |\n", + "| approx_kl | 0.0015846763 |\n", + "| clip_fraction | 0.0468 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.45 |\n", + "| explained_variance | 3.22e-06 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 1.53e+03 |\n", + "| n_updates | 590 |\n", + "| policy_gradient_loss | -0.000768 |\n", + "| std | 1.03 |\n", + "| value_loss | 4.11e+03 |\n", + "------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -325 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 61 |\n", + "| time_elapsed | 740 |\n", + "| total_timesteps | 124928 |\n", + "| train/ | |\n", + "| approx_kl | 0.0014858413 |\n", + "| clip_fraction | 0.0124 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.45 |\n", + "| explained_variance | 1.67e-06 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 2.44e+03 |\n", + "| n_updates | 600 |\n", + "| policy_gradient_loss | 0.000545 |\n", + "| std | 1.04 |\n", + "| value_loss | 3.12e+03 |\n", + "------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -331 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 62 |\n", + "| time_elapsed | 752 |\n", + "| total_timesteps | 126976 |\n", + "| train/ | |\n", + "| approx_kl | 0.0038123443 |\n", + "| clip_fraction | 0.0339 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.45 |\n", + "| explained_variance | 1.97e-06 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 2.05e+03 |\n", + "| n_updates | 610 |\n", + "| policy_gradient_loss | -0.00259 |\n", + "| std | 1.03 |\n", + "| value_loss | 3.13e+03 |\n", + "------------------------------------------\n", + "-----------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -335 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 63 |\n", + "| time_elapsed | 764 |\n", + "| total_timesteps | 129024 |\n", + "| train/ | |\n", + "| approx_kl | 0.003941954 |\n", + "| clip_fraction | 0.00273 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.45 |\n", + "| explained_variance | 3.7e-06 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 2.89e+03 |\n", + "| n_updates | 620 |\n", + "| policy_gradient_loss | 0.0002 |\n", + "| std | 1.02 |\n", + "| value_loss | 4.11e+03 |\n", + "-----------------------------------------\n", + "-----------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -341 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 64 |\n", + "| time_elapsed | 776 |\n", + "| total_timesteps | 131072 |\n", + "| train/ | |\n", + "| approx_kl | 0.007216826 |\n", + "| clip_fraction | 0.0402 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.44 |\n", + "| explained_variance | 1.61e-06 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 1.17e+03 |\n", + "| n_updates | 630 |\n", + "| policy_gradient_loss | -0.000444 |\n", + "| std | 1.02 |\n", + "| value_loss | 3.13e+03 |\n", + "-----------------------------------------\n", + "-----------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -345 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 65 |\n", + "| time_elapsed | 788 |\n", + "| total_timesteps | 133120 |\n", + "| train/ | |\n", + "| approx_kl | 0.001702552 |\n", + "| clip_fraction | 0.0259 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.44 |\n", + "| explained_variance | 2.15e-06 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 1.62e+03 |\n", + "| n_updates | 640 |\n", + "| policy_gradient_loss | 3.53e-05 |\n", + "| std | 1.02 |\n", + "| value_loss | 4.12e+03 |\n", + "-----------------------------------------\n", + "----------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -350 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 66 |\n", + "| time_elapsed | 800 |\n", + "| total_timesteps | 135168 |\n", + "| train/ | |\n", + "| approx_kl | 0.01116517 |\n", + "| clip_fraction | 0.128 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.44 |\n", + "| explained_variance | 1.19e-06 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 370 |\n", + "| n_updates | 650 |\n", + "| policy_gradient_loss | -0.000384 |\n", + "| std | 1.02 |\n", + "| value_loss | 3.13e+03 |\n", + "----------------------------------------\n", + "-------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -356 |\n", + "| time/ | |\n", + "| fps | 168 |\n", + "| iterations | 67 |\n", + "| time_elapsed | 812 |\n", + "| total_timesteps | 137216 |\n", + "| train/ | |\n", + "| approx_kl | 0.00059924903 |\n", + "| clip_fraction | 0.0156 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.44 |\n", + "| explained_variance | 2.8e-06 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 2.21e+03 |\n", + "| n_updates | 660 |\n", + "| policy_gradient_loss | 0.000519 |\n", + "| std | 1.02 |\n", + "| value_loss | 3.14e+03 |\n", + "-------------------------------------------\n", + "-----------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -360 |\n", + "| time/ | |\n", + "| fps | 169 |\n", + "| iterations | 68 |\n", + "| time_elapsed | 823 |\n", + "| total_timesteps | 139264 |\n", + "| train/ | |\n", + "| approx_kl | 0.008353274 |\n", + "| clip_fraction | 0.0397 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.43 |\n", + "| explained_variance | 8.05e-06 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 6.21e+03 |\n", + "| n_updates | 670 |\n", + "| policy_gradient_loss | -0.00075 |\n", + "| std | 1.01 |\n", + "| value_loss | 4.12e+03 |\n", + "-----------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -365 |\n", + "| time/ | |\n", + "| fps | 169 |\n", + "| iterations | 69 |\n", + "| time_elapsed | 835 |\n", + "| total_timesteps | 141312 |\n", + "| train/ | |\n", + "| approx_kl | 0.0058903834 |\n", + "| clip_fraction | 0.0345 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.43 |\n", + "| explained_variance | 3.28e-06 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 1.21e+03 |\n", + "| n_updates | 680 |\n", + "| policy_gradient_loss | -0.000968 |\n", + "| std | 1.02 |\n", + "| value_loss | 3.14e+03 |\n", + "------------------------------------------\n", + "-------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -370 |\n", + "| time/ | |\n", + "| fps | 169 |\n", + "| iterations | 70 |\n", + "| time_elapsed | 847 |\n", + "| total_timesteps | 143360 |\n", + "| train/ | |\n", + "| approx_kl | 0.00016515396 |\n", + "| clip_fraction | 0.0125 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.44 |\n", + "| explained_variance | 1.1e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 1.36e+03 |\n", + "| n_updates | 690 |\n", + "| policy_gradient_loss | 7.53e-05 |\n", + "| std | 1.02 |\n", + "| value_loss | 4.12e+03 |\n", + "-------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -373 |\n", + "| time/ | |\n", + "| fps | 169 |\n", + "| iterations | 71 |\n", + "| time_elapsed | 859 |\n", + "| total_timesteps | 145408 |\n", + "| train/ | |\n", + "| approx_kl | 0.0013749554 |\n", + "| clip_fraction | 0.0172 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.44 |\n", + "| explained_variance | 3.99e-06 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 1.63e+03 |\n", + "| n_updates | 700 |\n", + "| policy_gradient_loss | -0.00127 |\n", + "| std | 1.02 |\n", + "| value_loss | 3.14e+03 |\n", + "------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -377 |\n", + "| time/ | |\n", + "| fps | 169 |\n", + "| iterations | 72 |\n", + "| time_elapsed | 870 |\n", + "| total_timesteps | 147456 |\n", + "| train/ | |\n", + "| approx_kl | 0.0012910418 |\n", + "| clip_fraction | 0.0167 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.44 |\n", + "| explained_variance | 1.08e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 882 |\n", + "| n_updates | 710 |\n", + "| policy_gradient_loss | -4.58e-05 |\n", + "| std | 1.03 |\n", + "| value_loss | 3.14e+03 |\n", + "------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -380 |\n", + "| time/ | |\n", + "| fps | 169 |\n", + "| iterations | 73 |\n", + "| time_elapsed | 882 |\n", + "| total_timesteps | 149504 |\n", + "| train/ | |\n", + "| approx_kl | 0.0010234144 |\n", + "| clip_fraction | 0.000293 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.44 |\n", + "| explained_variance | 2.32e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 109 |\n", + "| n_updates | 720 |\n", + "| policy_gradient_loss | 0.000243 |\n", + "| std | 1.02 |\n", + "| value_loss | 4.12e+03 |\n", + "------------------------------------------\n", + "-----------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -384 |\n", + "| time/ | |\n", + "| fps | 169 |\n", + "| iterations | 74 |\n", + "| time_elapsed | 894 |\n", + "| total_timesteps | 151552 |\n", + "| train/ | |\n", + "| approx_kl | 0.000599641 |\n", + "| clip_fraction | 0.0156 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.44 |\n", + "| explained_variance | 8.64e-06 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 1.79e+03 |\n", + "| n_updates | 730 |\n", + "| policy_gradient_loss | 0.000339 |\n", + "| std | 1.02 |\n", + "| value_loss | 3.14e+03 |\n", + "-----------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -387 |\n", + "| time/ | |\n", + "| fps | 169 |\n", + "| iterations | 75 |\n", + "| time_elapsed | 906 |\n", + "| total_timesteps | 153600 |\n", + "| train/ | |\n", + "| approx_kl | 0.0004998102 |\n", + "| clip_fraction | 0.0404 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.44 |\n", + "| explained_variance | 3.27e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 980 |\n", + "| n_updates | 740 |\n", + "| policy_gradient_loss | 0.00127 |\n", + "| std | 1.01 |\n", + "| value_loss | 4.12e+03 |\n", + "------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -390 |\n", + "| time/ | |\n", + "| fps | 169 |\n", + "| iterations | 76 |\n", + "| time_elapsed | 918 |\n", + "| total_timesteps | 155648 |\n", + "| train/ | |\n", + "| approx_kl | 0.0055045467 |\n", + "| clip_fraction | 0.0171 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.43 |\n", + "| explained_variance | 1.28e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 3.81e+03 |\n", + "| n_updates | 750 |\n", + "| policy_gradient_loss | 0.000433 |\n", + "| std | 1.01 |\n", + "| value_loss | 3.14e+03 |\n", + "------------------------------------------\n", + "-------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -395 |\n", + "| time/ | |\n", + "| fps | 169 |\n", + "| iterations | 77 |\n", + "| time_elapsed | 930 |\n", + "| total_timesteps | 157696 |\n", + "| train/ | |\n", + "| approx_kl | 0.00067343883 |\n", + "| clip_fraction | 0.0165 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.43 |\n", + "| explained_variance | 1.79e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 888 |\n", + "| n_updates | 760 |\n", + "| policy_gradient_loss | 0.000467 |\n", + "| std | 1.02 |\n", + "| value_loss | 3.14e+03 |\n", + "-------------------------------------------\n", + "-----------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -400 |\n", + "| time/ | |\n", + "| fps | 169 |\n", + "| iterations | 78 |\n", + "| time_elapsed | 942 |\n", + "| total_timesteps | 159744 |\n", + "| train/ | |\n", + "| approx_kl | 0.005202517 |\n", + "| clip_fraction | 0.104 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.43 |\n", + "| explained_variance | 2.89e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 662 |\n", + "| n_updates | 770 |\n", + "| policy_gradient_loss | -0.00199 |\n", + "| std | 1.02 |\n", + "| value_loss | 3.9e+03 |\n", + "-----------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -406 |\n", + "| time/ | |\n", + "| fps | 169 |\n", + "| iterations | 79 |\n", + "| time_elapsed | 954 |\n", + "| total_timesteps | 161792 |\n", + "| train/ | |\n", + "| approx_kl | 0.0017581655 |\n", + "| clip_fraction | 0.00181 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.44 |\n", + "| explained_variance | 1.59e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 1.07e+03 |\n", + "| n_updates | 780 |\n", + "| policy_gradient_loss | 0.000173 |\n", + "| std | 1.03 |\n", + "| value_loss | 3.15e+03 |\n", + "------------------------------------------\n", + "-------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -410 |\n", + "| time/ | |\n", + "| fps | 169 |\n", + "| iterations | 80 |\n", + "| time_elapsed | 966 |\n", + "| total_timesteps | 163840 |\n", + "| train/ | |\n", + "| approx_kl | 0.00069459836 |\n", + "| clip_fraction | 0.0629 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.45 |\n", + "| explained_variance | 2.23e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 2.34e+03 |\n", + "| n_updates | 790 |\n", + "| policy_gradient_loss | -2.99e-05 |\n", + "| std | 1.03 |\n", + "| value_loss | 4.14e+03 |\n", + "-------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -414 |\n", + "| time/ | |\n", + "| fps | 169 |\n", + "| iterations | 81 |\n", + "| time_elapsed | 978 |\n", + "| total_timesteps | 165888 |\n", + "| train/ | |\n", + "| approx_kl | 0.0035149038 |\n", + "| clip_fraction | 0.036 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.45 |\n", + "| explained_variance | 1.04e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 1.69e+03 |\n", + "| n_updates | 800 |\n", + "| policy_gradient_loss | 0.000948 |\n", + "| std | 1.03 |\n", + "| value_loss | 3.14e+03 |\n", + "------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -418 |\n", + "| time/ | |\n", + "| fps | 169 |\n", + "| iterations | 82 |\n", + "| time_elapsed | 990 |\n", + "| total_timesteps | 167936 |\n", + "| train/ | |\n", + "| approx_kl | 0.0068787774 |\n", + "| clip_fraction | 0.0504 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.44 |\n", + "| explained_variance | 2.04e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 3.49e+03 |\n", + "| n_updates | 810 |\n", + "| policy_gradient_loss | -0.000511 |\n", + "| std | 1.02 |\n", + "| value_loss | 4.14e+03 |\n", + "------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -421 |\n", + "| time/ | |\n", + "| fps | 169 |\n", + "| iterations | 83 |\n", + "| time_elapsed | 1001 |\n", + "| total_timesteps | 169984 |\n", + "| train/ | |\n", + "| approx_kl | 0.0018102819 |\n", + "| clip_fraction | 0.00742 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.44 |\n", + "| explained_variance | 1.06e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 5.99e+03 |\n", + "| n_updates | 820 |\n", + "| policy_gradient_loss | 0.000792 |\n", + "| std | 1.02 |\n", + "| value_loss | 3.15e+03 |\n", + "------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -424 |\n", + "| time/ | |\n", + "| fps | 169 |\n", + "| iterations | 84 |\n", + "| time_elapsed | 1013 |\n", + "| total_timesteps | 172032 |\n", + "| train/ | |\n", + "| approx_kl | 0.0044906293 |\n", + "| clip_fraction | 0.00903 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.45 |\n", + "| explained_variance | 2.19e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 536 |\n", + "| n_updates | 830 |\n", + "| policy_gradient_loss | 8.28e-05 |\n", + "| std | 1.03 |\n", + "| value_loss | 3.15e+03 |\n", + "------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -427 |\n", + "| time/ | |\n", + "| fps | 169 |\n", + "| iterations | 85 |\n", + "| time_elapsed | 1025 |\n", + "| total_timesteps | 174080 |\n", + "| train/ | |\n", + "| approx_kl | 0.0013765441 |\n", + "| clip_fraction | 0.000635 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.45 |\n", + "| explained_variance | 3.98e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 2.56e+03 |\n", + "| n_updates | 840 |\n", + "| policy_gradient_loss | -1.69e-05 |\n", + "| std | 1.03 |\n", + "| value_loss | 4.14e+03 |\n", + "------------------------------------------\n", + "-------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -430 |\n", + "| time/ | |\n", + "| fps | 169 |\n", + "| iterations | 86 |\n", + "| time_elapsed | 1037 |\n", + "| total_timesteps | 176128 |\n", + "| train/ | |\n", + "| approx_kl | 0.00035555626 |\n", + "| clip_fraction | 0.00757 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.44 |\n", + "| explained_variance | 1.58e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 3.53e+03 |\n", + "| n_updates | 850 |\n", + "| policy_gradient_loss | -0.000221 |\n", + "| std | 1.01 |\n", + "| value_loss | 3.15e+03 |\n", + "-------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -432 |\n", + "| time/ | |\n", + "| fps | 169 |\n", + "| iterations | 87 |\n", + "| time_elapsed | 1049 |\n", + "| total_timesteps | 178176 |\n", + "| train/ | |\n", + "| approx_kl | 0.0026123272 |\n", + "| clip_fraction | 0.0108 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.43 |\n", + "| explained_variance | 3.12e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 2.08e+03 |\n", + "| n_updates | 860 |\n", + "| policy_gradient_loss | 0.000388 |\n", + "| std | 1.02 |\n", + "| value_loss | 4.15e+03 |\n", + "------------------------------------------\n", + "-------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -434 |\n", + "| time/ | |\n", + "| fps | 169 |\n", + "| iterations | 88 |\n", + "| time_elapsed | 1061 |\n", + "| total_timesteps | 180224 |\n", + "| train/ | |\n", + "| approx_kl | 0.00018668428 |\n", + "| clip_fraction | 0.0435 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.43 |\n", + "| explained_variance | 1.5e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 1.38e+03 |\n", + "| n_updates | 870 |\n", + "| policy_gradient_loss | 0.000264 |\n", + "| std | 1.01 |\n", + "| value_loss | 3.15e+03 |\n", + "-------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -436 |\n", + "| time/ | |\n", + "| fps | 169 |\n", + "| iterations | 89 |\n", + "| time_elapsed | 1072 |\n", + "| total_timesteps | 182272 |\n", + "| train/ | |\n", + "| approx_kl | 0.0069202585 |\n", + "| clip_fraction | 0.0126 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.43 |\n", + "| explained_variance | 2.89e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 732 |\n", + "| n_updates | 880 |\n", + "| policy_gradient_loss | -0.000634 |\n", + "| std | 1.01 |\n", + "| value_loss | 3.15e+03 |\n", + "------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -437 |\n", + "| time/ | |\n", + "| fps | 169 |\n", + "| iterations | 90 |\n", + "| time_elapsed | 1084 |\n", + "| total_timesteps | 184320 |\n", + "| train/ | |\n", + "| approx_kl | 0.0013296772 |\n", + "| clip_fraction | 0.0431 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.43 |\n", + "| explained_variance | 4.08e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 3.81e+03 |\n", + "| n_updates | 890 |\n", + "| policy_gradient_loss | 0.00076 |\n", + "| std | 1.01 |\n", + "| value_loss | 4.14e+03 |\n", + "------------------------------------------\n", + "-----------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -439 |\n", + "| time/ | |\n", + "| fps | 169 |\n", + "| iterations | 91 |\n", + "| time_elapsed | 1096 |\n", + "| total_timesteps | 186368 |\n", + "| train/ | |\n", + "| approx_kl | 0.005249043 |\n", + "| clip_fraction | 0.0232 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.44 |\n", + "| explained_variance | 1.92e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 3.29e+03 |\n", + "| n_updates | 900 |\n", + "| policy_gradient_loss | -0.000256 |\n", + "| std | 1.02 |\n", + "| value_loss | 3.15e+03 |\n", + "-----------------------------------------\n", + "-------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -439 |\n", + "| time/ | |\n", + "| fps | 170 |\n", + "| iterations | 92 |\n", + "| time_elapsed | 1108 |\n", + "| total_timesteps | 188416 |\n", + "| train/ | |\n", + "| approx_kl | 3.8214377e-05 |\n", + "| clip_fraction | 0.0542 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.44 |\n", + "| explained_variance | 4.1e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 2.05e+03 |\n", + "| n_updates | 910 |\n", + "| policy_gradient_loss | -9.36e-05 |\n", + "| std | 1.02 |\n", + "| value_loss | 4.14e+03 |\n", + "-------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -440 |\n", + "| time/ | |\n", + "| fps | 170 |\n", + "| iterations | 93 |\n", + "| time_elapsed | 1120 |\n", + "| total_timesteps | 190464 |\n", + "| train/ | |\n", + "| approx_kl | 0.0019262378 |\n", + "| clip_fraction | 0.0352 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.44 |\n", + "| explained_variance | 2.12e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 2.11e+03 |\n", + "| n_updates | 920 |\n", + "| policy_gradient_loss | -0.000944 |\n", + "| std | 1.03 |\n", + "| value_loss | 3.15e+03 |\n", + "------------------------------------------\n", + "-------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -441 |\n", + "| time/ | |\n", + "| fps | 170 |\n", + "| iterations | 94 |\n", + "| time_elapsed | 1132 |\n", + "| total_timesteps | 192512 |\n", + "| train/ | |\n", + "| approx_kl | 0.00047276722 |\n", + "| clip_fraction | 0.036 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.44 |\n", + "| explained_variance | 2.87e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 163 |\n", + "| n_updates | 930 |\n", + "| policy_gradient_loss | 0.00054 |\n", + "| std | 1.02 |\n", + "| value_loss | 3.15e+03 |\n", + "-------------------------------------------\n", + "-------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -442 |\n", + "| time/ | |\n", + "| fps | 170 |\n", + "| iterations | 95 |\n", + "| time_elapsed | 1144 |\n", + "| total_timesteps | 194560 |\n", + "| train/ | |\n", + "| approx_kl | 0.00018265905 |\n", + "| clip_fraction | 0.00132 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.44 |\n", + "| explained_variance | 4.35e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 1.17e+03 |\n", + "| n_updates | 940 |\n", + "| policy_gradient_loss | 0.000267 |\n", + "| std | 1.03 |\n", + "| value_loss | 4.15e+03 |\n", + "-------------------------------------------\n", + "-----------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -444 |\n", + "| time/ | |\n", + "| fps | 170 |\n", + "| iterations | 96 |\n", + "| time_elapsed | 1156 |\n", + "| total_timesteps | 196608 |\n", + "| train/ | |\n", + "| approx_kl | 0.014808972 |\n", + "| clip_fraction | 0.0734 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.45 |\n", + "| explained_variance | 2.44e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 1.85e+03 |\n", + "| n_updates | 950 |\n", + "| policy_gradient_loss | -0.000946 |\n", + "| std | 1.04 |\n", + "| value_loss | 3.16e+03 |\n", + "-----------------------------------------\n", + "-----------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -446 |\n", + "| time/ | |\n", + "| fps | 170 |\n", + "| iterations | 97 |\n", + "| time_elapsed | 1167 |\n", + "| total_timesteps | 198656 |\n", + "| train/ | |\n", + "| approx_kl | 0.000299958 |\n", + "| clip_fraction | 0.0442 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.46 |\n", + "| explained_variance | 4.51e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 1.81e+03 |\n", + "| n_updates | 960 |\n", + "| policy_gradient_loss | -0.000387 |\n", + "| std | 1.04 |\n", + "| value_loss | 4.14e+03 |\n", + "-----------------------------------------\n", + "-------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -447 |\n", + "| time/ | |\n", + "| fps | 170 |\n", + "| iterations | 98 |\n", + "| time_elapsed | 1179 |\n", + "| total_timesteps | 200704 |\n", + "| train/ | |\n", + "| approx_kl | 6.6622015e-05 |\n", + "| clip_fraction | 0.0193 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.46 |\n", + "| explained_variance | 2.52e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 2.14e+03 |\n", + "| n_updates | 970 |\n", + "| policy_gradient_loss | 0.000193 |\n", + "| std | 1.04 |\n", + "| value_loss | 3.16e+03 |\n", + "-------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -448 |\n", + "| time/ | |\n", + "| fps | 170 |\n", + "| iterations | 99 |\n", + "| time_elapsed | 1191 |\n", + "| total_timesteps | 202752 |\n", + "| train/ | |\n", + "| approx_kl | 0.0016583821 |\n", + "| clip_fraction | 0.068 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.47 |\n", + "| explained_variance | 3.03e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 620 |\n", + "| n_updates | 980 |\n", + "| policy_gradient_loss | -0.0011 |\n", + "| std | 1.05 |\n", + "| value_loss | 3.16e+03 |\n", + "------------------------------------------\n", + "-----------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -449 |\n", + "| time/ | |\n", + "| fps | 170 |\n", + "| iterations | 100 |\n", + "| time_elapsed | 1203 |\n", + "| total_timesteps | 204800 |\n", + "| train/ | |\n", + "| approx_kl | 0.014906235 |\n", + "| clip_fraction | 0.0531 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.47 |\n", + "| explained_variance | 4.49e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 188 |\n", + "| n_updates | 990 |\n", + "| policy_gradient_loss | -0.00305 |\n", + "| std | 1.06 |\n", + "| value_loss | 4.14e+03 |\n", + "-----------------------------------------\n", + "-------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -450 |\n", + "| time/ | |\n", + "| fps | 170 |\n", + "| iterations | 101 |\n", + "| time_elapsed | 1215 |\n", + "| total_timesteps | 206848 |\n", + "| train/ | |\n", + "| approx_kl | 0.00044346167 |\n", + "| clip_fraction | 0.00498 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.49 |\n", + "| explained_variance | 2.75e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 634 |\n", + "| n_updates | 1000 |\n", + "| policy_gradient_loss | -0.000378 |\n", + "| std | 1.08 |\n", + "| value_loss | 3.16e+03 |\n", + "-------------------------------------------\n", + "----------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -451 |\n", + "| time/ | |\n", + "| fps | 170 |\n", + "| iterations | 102 |\n", + "| time_elapsed | 1227 |\n", + "| total_timesteps | 208896 |\n", + "| train/ | |\n", + "| approx_kl | 0.00400657 |\n", + "| clip_fraction | 0.074 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.49 |\n", + "| explained_variance | 4.66e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 4.04e+03 |\n", + "| n_updates | 1010 |\n", + "| policy_gradient_loss | -0.00208 |\n", + "| std | 1.08 |\n", + "| value_loss | 4.15e+03 |\n", + "----------------------------------------\n", + "--------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -452 |\n", + "| time/ | |\n", + "| fps | 170 |\n", + "| iterations | 103 |\n", + "| time_elapsed | 1238 |\n", + "| total_timesteps | 210944 |\n", + "| train/ | |\n", + "| approx_kl | 0.000119188306 |\n", + "| clip_fraction | 0.0681 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.49 |\n", + "| explained_variance | 2.54e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 130 |\n", + "| n_updates | 1020 |\n", + "| policy_gradient_loss | 0.000329 |\n", + "| std | 1.08 |\n", + "| value_loss | 3.16e+03 |\n", + "--------------------------------------------\n", + "----------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -454 |\n", + "| time/ | |\n", + "| fps | 170 |\n", + "| iterations | 104 |\n", + "| time_elapsed | 1251 |\n", + "| total_timesteps | 212992 |\n", + "| train/ | |\n", + "| approx_kl | 0.02515565 |\n", + "| clip_fraction | 0.0928 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.49 |\n", + "| explained_variance | 3.2e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 507 |\n", + "| n_updates | 1030 |\n", + "| policy_gradient_loss | -0.00217 |\n", + "| std | 1.08 |\n", + "| value_loss | 3.16e+03 |\n", + "----------------------------------------\n", + "-------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -455 |\n", + "| time/ | |\n", + "| fps | 170 |\n", + "| iterations | 105 |\n", + "| time_elapsed | 1263 |\n", + "| total_timesteps | 215040 |\n", + "| train/ | |\n", + "| approx_kl | 0.00012812333 |\n", + "| clip_fraction | 0.00112 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.5 |\n", + "| explained_variance | 5.17e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 1.92e+03 |\n", + "| n_updates | 1040 |\n", + "| policy_gradient_loss | 0.000452 |\n", + "| std | 1.09 |\n", + "| value_loss | 3.75e+03 |\n", + "-------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -457 |\n", + "| time/ | |\n", + "| fps | 170 |\n", + "| iterations | 106 |\n", + "| time_elapsed | 1275 |\n", + "| total_timesteps | 217088 |\n", + "| train/ | |\n", + "| approx_kl | 0.0017133974 |\n", + "| clip_fraction | 0.00947 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.5 |\n", + "| explained_variance | 4.11e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 2.63e+03 |\n", + "| n_updates | 1050 |\n", + "| policy_gradient_loss | 0.000314 |\n", + "| std | 1.09 |\n", + "| value_loss | 3.16e+03 |\n", + "------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -457 |\n", + "| time/ | |\n", + "| fps | 170 |\n", + "| iterations | 107 |\n", + "| time_elapsed | 1288 |\n", + "| total_timesteps | 219136 |\n", + "| train/ | |\n", + "| approx_kl | 0.0031836042 |\n", + "| clip_fraction | 0.00962 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.5 |\n", + "| explained_variance | 6.03e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 3.78e+03 |\n", + "| n_updates | 1060 |\n", + "| policy_gradient_loss | -0.000246 |\n", + "| std | 1.08 |\n", + "| value_loss | 4.14e+03 |\n", + "------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -458 |\n", + "| time/ | |\n", + "| fps | 170 |\n", + "| iterations | 108 |\n", + "| time_elapsed | 1300 |\n", + "| total_timesteps | 221184 |\n", + "| train/ | |\n", + "| approx_kl | 0.0026518258 |\n", + "| clip_fraction | 0.0826 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.5 |\n", + "| explained_variance | 3.83e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 895 |\n", + "| n_updates | 1070 |\n", + "| policy_gradient_loss | -0.000444 |\n", + "| std | 1.08 |\n", + "| value_loss | 3.16e+03 |\n", + "------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -459 |\n", + "| time/ | |\n", + "| fps | 170 |\n", + "| iterations | 109 |\n", + "| time_elapsed | 1312 |\n", + "| total_timesteps | 223232 |\n", + "| train/ | |\n", + "| approx_kl | 0.0048480523 |\n", + "| clip_fraction | 0.0352 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.5 |\n", + "| explained_variance | 5.64e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 1.95e+03 |\n", + "| n_updates | 1080 |\n", + "| policy_gradient_loss | -0.000337 |\n", + "| std | 1.08 |\n", + "| value_loss | 4.15e+03 |\n", + "------------------------------------------\n", + "------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -459 |\n", + "| time/ | |\n", + "| fps | 170 |\n", + "| iterations | 110 |\n", + "| time_elapsed | 1324 |\n", + "| total_timesteps | 225280 |\n", + "| train/ | |\n", + "| approx_kl | 0.0028444673 |\n", + "| clip_fraction | 0.00771 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.5 |\n", + "| explained_variance | 3.26e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 1.62e+03 |\n", + "| n_updates | 1090 |\n", + "| policy_gradient_loss | -1.15e-05 |\n", + "| std | 1.08 |\n", + "| value_loss | 3.16e+03 |\n", + "------------------------------------------\n", + "-------------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -459 |\n", + "| time/ | |\n", + "| fps | 170 |\n", + "| iterations | 111 |\n", + "| time_elapsed | 1336 |\n", + "| total_timesteps | 227328 |\n", + "| train/ | |\n", + "| approx_kl | 0.00074651965 |\n", + "| clip_fraction | 0.000146 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.5 |\n", + "| explained_variance | 4.02e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 1.73e+03 |\n", + "| n_updates | 1100 |\n", + "| policy_gradient_loss | 7.33e-05 |\n", + "| std | 1.09 |\n", + "| value_loss | 3.16e+03 |\n", + "-------------------------------------------\n", + "----------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -459 |\n", + "| time/ | |\n", + "| fps | 170 |\n", + "| iterations | 112 |\n", + "| time_elapsed | 1348 |\n", + "| total_timesteps | 229376 |\n", + "| train/ | |\n", + "| approx_kl | 0.00920542 |\n", + "| clip_fraction | 0.0695 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.5 |\n", + "| explained_variance | 7.62e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 857 |\n", + "| n_updates | 1110 |\n", + "| policy_gradient_loss | -0.00293 |\n", + "| std | 1.09 |\n", + "| value_loss | 4.14e+03 |\n", + "----------------------------------------\n", + "-----------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -459 |\n", + "| time/ | |\n", + "| fps | 170 |\n", + "| iterations | 113 |\n", + "| time_elapsed | 1360 |\n", + "| total_timesteps | 231424 |\n", + "| train/ | |\n", + "| approx_kl | 0.003430673 |\n", + "| clip_fraction | 0.0324 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.51 |\n", + "| explained_variance | 4e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 2.51e+03 |\n", + "| n_updates | 1120 |\n", + "| policy_gradient_loss | 0.000274 |\n", + "| std | 1.1 |\n", + "| value_loss | 3.16e+03 |\n", + "-----------------------------------------\n", + "-----------------------------------------\n", + "| rollout/ | |\n", + "| ep_len_mean | 601 |\n", + "| ep_rew_mean | -458 |\n", + "| time/ | |\n", + "| fps | 170 |\n", + "| iterations | 114 |\n", + "| time_elapsed | 1372 |\n", + "| total_timesteps | 233472 |\n", + "| train/ | |\n", + "| approx_kl | 0.010919753 |\n", + "| clip_fraction | 0.0383 |\n", + "| clip_range | 0.2 |\n", + "| entropy_loss | -1.52 |\n", + "| explained_variance | 7.24e-05 |\n", + "| learning_rate | 0.0003 |\n", + "| loss | 3.47e+03 |\n", + "| n_updates | 1130 |\n", + "| policy_gradient_loss | -0.0024 |\n", + "| std | 1.11 |\n", + "| value_loss | 4.15e+03 |\n", + "-----------------------------------------\n" ] }, { @@ -69,19 +2556,41 @@ "traceback": [ "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m", "\u001b[0;31mKeyboardInterrupt\u001b[0m Traceback (most recent call last)", - "Cell \u001b[0;32mIn[3], line 4\u001b[0m\n\u001b[1;32m 2\u001b[0m \u001b[38;5;28;01mfrom\u001b[39;00m \u001b[38;5;21;01mstable_baselines3\u001b[39;00m \u001b[38;5;28;01mimport\u001b[39;00m TD3\n\u001b[1;32m 3\u001b[0m model \u001b[38;5;241m=\u001b[39m TD3(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mMultiInputPolicy\u001b[39m\u001b[38;5;124m\"\u001b[39m, wrapped_env, verbose\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m1\u001b[39m)\n\u001b[0;32m----> 4\u001b[0m \u001b[43mmodel\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mlearn\u001b[49m\u001b[43m(\u001b[49m\u001b[43mtotal_timesteps\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;241;43m30_000\u001b[39;49m\u001b[43m)\u001b[49m\n", - "File \u001b[0;32m~/Documents/Code/solarcarsim/.venv/lib/python3.12/site-packages/stable_baselines3/td3/td3.py:222\u001b[0m, in \u001b[0;36mTD3.learn\u001b[0;34m(self, total_timesteps, callback, log_interval, tb_log_name, reset_num_timesteps, progress_bar)\u001b[0m\n\u001b[1;32m 213\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mlearn\u001b[39m(\n\u001b[1;32m 214\u001b[0m \u001b[38;5;28mself\u001b[39m: SelfTD3,\n\u001b[1;32m 215\u001b[0m total_timesteps: \u001b[38;5;28mint\u001b[39m,\n\u001b[0;32m (...)\u001b[0m\n\u001b[1;32m 220\u001b[0m progress_bar: \u001b[38;5;28mbool\u001b[39m \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;01mFalse\u001b[39;00m,\n\u001b[1;32m 221\u001b[0m ) \u001b[38;5;241m-\u001b[39m\u001b[38;5;241m>\u001b[39m SelfTD3:\n\u001b[0;32m--> 222\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43msuper\u001b[39;49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mlearn\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m 223\u001b[0m \u001b[43m \u001b[49m\u001b[43mtotal_timesteps\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mtotal_timesteps\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 224\u001b[0m \u001b[43m \u001b[49m\u001b[43mcallback\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mcallback\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 225\u001b[0m \u001b[43m \u001b[49m\u001b[43mlog_interval\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mlog_interval\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 226\u001b[0m \u001b[43m \u001b[49m\u001b[43mtb_log_name\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mtb_log_name\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 227\u001b[0m \u001b[43m \u001b[49m\u001b[43mreset_num_timesteps\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mreset_num_timesteps\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 228\u001b[0m \u001b[43m \u001b[49m\u001b[43mprogress_bar\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mprogress_bar\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 229\u001b[0m \u001b[43m \u001b[49m\u001b[43m)\u001b[49m\n", - "File \u001b[0;32m~/Documents/Code/solarcarsim/.venv/lib/python3.12/site-packages/stable_baselines3/common/off_policy_algorithm.py:347\u001b[0m, in \u001b[0;36mOffPolicyAlgorithm.learn\u001b[0;34m(self, total_timesteps, callback, log_interval, tb_log_name, reset_num_timesteps, progress_bar)\u001b[0m\n\u001b[1;32m 345\u001b[0m \u001b[38;5;66;03m# Special case when the user passes `gradient_steps=0`\u001b[39;00m\n\u001b[1;32m 346\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m gradient_steps \u001b[38;5;241m>\u001b[39m \u001b[38;5;241m0\u001b[39m:\n\u001b[0;32m--> 347\u001b[0m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mtrain\u001b[49m\u001b[43m(\u001b[49m\u001b[43mbatch_size\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mbatch_size\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mgradient_steps\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mgradient_steps\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 349\u001b[0m callback\u001b[38;5;241m.\u001b[39mon_training_end()\n\u001b[1;32m 351\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28mself\u001b[39m\n", - "File \u001b[0;32m~/Documents/Code/solarcarsim/.venv/lib/python3.12/site-packages/stable_baselines3/td3/td3.py:184\u001b[0m, in \u001b[0;36mTD3.train\u001b[0;34m(self, gradient_steps, batch_size)\u001b[0m\n\u001b[1;32m 182\u001b[0m critic_loss \u001b[38;5;241m=\u001b[39m \u001b[38;5;28msum\u001b[39m(F\u001b[38;5;241m.\u001b[39mmse_loss(current_q, target_q_values) \u001b[38;5;28;01mfor\u001b[39;00m current_q \u001b[38;5;129;01min\u001b[39;00m current_q_values)\n\u001b[1;32m 183\u001b[0m \u001b[38;5;28;01massert\u001b[39;00m \u001b[38;5;28misinstance\u001b[39m(critic_loss, th\u001b[38;5;241m.\u001b[39mTensor)\n\u001b[0;32m--> 184\u001b[0m critic_losses\u001b[38;5;241m.\u001b[39mappend(\u001b[43mcritic_loss\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mitem\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m)\n\u001b[1;32m 186\u001b[0m \u001b[38;5;66;03m# Optimize the critics\u001b[39;00m\n\u001b[1;32m 187\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mcritic\u001b[38;5;241m.\u001b[39moptimizer\u001b[38;5;241m.\u001b[39mzero_grad()\n", + "Cell \u001b[0;32mIn[3], line 3\u001b[0m\n\u001b[1;32m 1\u001b[0m \u001b[38;5;66;03m# import a model and try it out!\u001b[39;00m\n\u001b[1;32m 2\u001b[0m model \u001b[38;5;241m=\u001b[39m PPO(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mMultiInputPolicy\u001b[39m\u001b[38;5;124m\"\u001b[39m, wrapped_env, verbose\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m1\u001b[39m)\n\u001b[0;32m----> 3\u001b[0m \u001b[43mmodel\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mlearn\u001b[49m\u001b[43m(\u001b[49m\u001b[43mtotal_timesteps\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;241;43m1_000_000\u001b[39;49m\u001b[43m)\u001b[49m\n", + "File \u001b[0;32m~/Documents/Code/solarcarsim/.venv/lib/python3.12/site-packages/stable_baselines3/ppo/ppo.py:311\u001b[0m, in \u001b[0;36mPPO.learn\u001b[0;34m(self, total_timesteps, callback, log_interval, tb_log_name, reset_num_timesteps, progress_bar)\u001b[0m\n\u001b[1;32m 302\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mlearn\u001b[39m(\n\u001b[1;32m 303\u001b[0m \u001b[38;5;28mself\u001b[39m: SelfPPO,\n\u001b[1;32m 304\u001b[0m total_timesteps: \u001b[38;5;28mint\u001b[39m,\n\u001b[0;32m (...)\u001b[0m\n\u001b[1;32m 309\u001b[0m progress_bar: \u001b[38;5;28mbool\u001b[39m \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;01mFalse\u001b[39;00m,\n\u001b[1;32m 310\u001b[0m ) \u001b[38;5;241m-\u001b[39m\u001b[38;5;241m>\u001b[39m SelfPPO:\n\u001b[0;32m--> 311\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43msuper\u001b[39;49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mlearn\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m 312\u001b[0m \u001b[43m \u001b[49m\u001b[43mtotal_timesteps\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mtotal_timesteps\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 313\u001b[0m \u001b[43m \u001b[49m\u001b[43mcallback\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mcallback\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 314\u001b[0m \u001b[43m \u001b[49m\u001b[43mlog_interval\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mlog_interval\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 315\u001b[0m \u001b[43m \u001b[49m\u001b[43mtb_log_name\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mtb_log_name\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 316\u001b[0m \u001b[43m \u001b[49m\u001b[43mreset_num_timesteps\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mreset_num_timesteps\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 317\u001b[0m \u001b[43m \u001b[49m\u001b[43mprogress_bar\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mprogress_bar\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 318\u001b[0m \u001b[43m \u001b[49m\u001b[43m)\u001b[49m\n", + "File \u001b[0;32m~/Documents/Code/solarcarsim/.venv/lib/python3.12/site-packages/stable_baselines3/common/on_policy_algorithm.py:323\u001b[0m, in \u001b[0;36mOnPolicyAlgorithm.learn\u001b[0;34m(self, total_timesteps, callback, log_interval, tb_log_name, reset_num_timesteps, progress_bar)\u001b[0m\n\u001b[1;32m 320\u001b[0m \u001b[38;5;28;01massert\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39menv \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m\n\u001b[1;32m 322\u001b[0m \u001b[38;5;28;01mwhile\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mnum_timesteps \u001b[38;5;241m<\u001b[39m total_timesteps:\n\u001b[0;32m--> 323\u001b[0m continue_training \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mcollect_rollouts\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43menv\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mcallback\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mrollout_buffer\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mn_rollout_steps\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mn_steps\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 325\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m continue_training:\n\u001b[1;32m 326\u001b[0m \u001b[38;5;28;01mbreak\u001b[39;00m\n", + "File \u001b[0;32m~/Documents/Code/solarcarsim/.venv/lib/python3.12/site-packages/stable_baselines3/common/on_policy_algorithm.py:218\u001b[0m, in \u001b[0;36mOnPolicyAlgorithm.collect_rollouts\u001b[0;34m(self, env, callback, rollout_buffer, n_rollout_steps)\u001b[0m\n\u001b[1;32m 213\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[1;32m 214\u001b[0m \u001b[38;5;66;03m# Otherwise, clip the actions to avoid out of bound error\u001b[39;00m\n\u001b[1;32m 215\u001b[0m \u001b[38;5;66;03m# as we are sampling from an unbounded Gaussian distribution\u001b[39;00m\n\u001b[1;32m 216\u001b[0m clipped_actions \u001b[38;5;241m=\u001b[39m np\u001b[38;5;241m.\u001b[39mclip(actions, \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39maction_space\u001b[38;5;241m.\u001b[39mlow, \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39maction_space\u001b[38;5;241m.\u001b[39mhigh)\n\u001b[0;32m--> 218\u001b[0m new_obs, rewards, dones, infos \u001b[38;5;241m=\u001b[39m \u001b[43menv\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mstep\u001b[49m\u001b[43m(\u001b[49m\u001b[43mclipped_actions\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 220\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mnum_timesteps \u001b[38;5;241m+\u001b[39m\u001b[38;5;241m=\u001b[39m env\u001b[38;5;241m.\u001b[39mnum_envs\n\u001b[1;32m 222\u001b[0m \u001b[38;5;66;03m# Give access to local variables\u001b[39;00m\n", + "File \u001b[0;32m~/Documents/Code/solarcarsim/.venv/lib/python3.12/site-packages/stable_baselines3/common/vec_env/base_vec_env.py:206\u001b[0m, in \u001b[0;36mVecEnv.step\u001b[0;34m(self, actions)\u001b[0m\n\u001b[1;32m 199\u001b[0m \u001b[38;5;250m\u001b[39m\u001b[38;5;124;03m\"\"\"\u001b[39;00m\n\u001b[1;32m 200\u001b[0m \u001b[38;5;124;03mStep the environments with the given action\u001b[39;00m\n\u001b[1;32m 201\u001b[0m \n\u001b[1;32m 202\u001b[0m \u001b[38;5;124;03m:param actions: the action\u001b[39;00m\n\u001b[1;32m 203\u001b[0m \u001b[38;5;124;03m:return: observation, reward, done, information\u001b[39;00m\n\u001b[1;32m 204\u001b[0m \u001b[38;5;124;03m\"\"\"\u001b[39;00m\n\u001b[1;32m 205\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mstep_async(actions)\n\u001b[0;32m--> 206\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mstep_wait\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\n", + "File \u001b[0;32m~/Documents/Code/solarcarsim/.venv/lib/python3.12/site-packages/stable_baselines3/common/vec_env/dummy_vec_env.py:58\u001b[0m, in \u001b[0;36mDummyVecEnv.step_wait\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m 55\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mstep_wait\u001b[39m(\u001b[38;5;28mself\u001b[39m) \u001b[38;5;241m-\u001b[39m\u001b[38;5;241m>\u001b[39m VecEnvStepReturn:\n\u001b[1;32m 56\u001b[0m \u001b[38;5;66;03m# Avoid circular imports\u001b[39;00m\n\u001b[1;32m 57\u001b[0m \u001b[38;5;28;01mfor\u001b[39;00m env_idx \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28mrange\u001b[39m(\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mnum_envs):\n\u001b[0;32m---> 58\u001b[0m obs, \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mbuf_rews[env_idx], terminated, truncated, \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mbuf_infos[env_idx] \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43menvs\u001b[49m\u001b[43m[\u001b[49m\u001b[43menv_idx\u001b[49m\u001b[43m]\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mstep\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m 59\u001b[0m \u001b[43m \u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mactions\u001b[49m\u001b[43m[\u001b[49m\u001b[43menv_idx\u001b[49m\u001b[43m]\u001b[49m\n\u001b[1;32m 60\u001b[0m \u001b[43m \u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 61\u001b[0m \u001b[38;5;66;03m# convert to SB3 VecEnv api\u001b[39;00m\n\u001b[1;32m 62\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mbuf_dones[env_idx] \u001b[38;5;241m=\u001b[39m terminated \u001b[38;5;129;01mor\u001b[39;00m truncated\n", + "File \u001b[0;32m~/Documents/Code/solarcarsim/.venv/lib/python3.12/site-packages/stable_baselines3/common/monitor.py:94\u001b[0m, in \u001b[0;36mMonitor.step\u001b[0;34m(self, action)\u001b[0m\n\u001b[1;32m 92\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mneeds_reset:\n\u001b[1;32m 93\u001b[0m \u001b[38;5;28;01mraise\u001b[39;00m \u001b[38;5;167;01mRuntimeError\u001b[39;00m(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mTried to step environment that needs reset\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n\u001b[0;32m---> 94\u001b[0m observation, reward, terminated, truncated, info \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43menv\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mstep\u001b[49m\u001b[43m(\u001b[49m\u001b[43maction\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 95\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mrewards\u001b[38;5;241m.\u001b[39mappend(\u001b[38;5;28mfloat\u001b[39m(reward))\n\u001b[1;32m 96\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m terminated \u001b[38;5;129;01mor\u001b[39;00m truncated:\n", + "File \u001b[0;32m~/Documents/Code/solarcarsim/.venv/lib/python3.12/site-packages/gymnasium/wrappers/jax_to_numpy.py:166\u001b[0m, in \u001b[0;36mJaxToNumpy.step\u001b[0;34m(self, action)\u001b[0m\n\u001b[1;32m 157\u001b[0m \u001b[38;5;250m\u001b[39m\u001b[38;5;124;03m\"\"\"Transforms the action to a jax array .\u001b[39;00m\n\u001b[1;32m 158\u001b[0m \n\u001b[1;32m 159\u001b[0m \u001b[38;5;124;03mArgs:\u001b[39;00m\n\u001b[0;32m (...)\u001b[0m\n\u001b[1;32m 163\u001b[0m \u001b[38;5;124;03m A tuple containing numpy versions of the next observation, reward, termination, truncation, and extra info.\u001b[39;00m\n\u001b[1;32m 164\u001b[0m \u001b[38;5;124;03m\"\"\"\u001b[39;00m\n\u001b[1;32m 165\u001b[0m jax_action \u001b[38;5;241m=\u001b[39m numpy_to_jax(action)\n\u001b[0;32m--> 166\u001b[0m obs, reward, terminated, truncated, info \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43menv\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mstep\u001b[49m\u001b[43m(\u001b[49m\u001b[43mjax_action\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 168\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m (\n\u001b[1;32m 169\u001b[0m jax_to_numpy(obs),\n\u001b[1;32m 170\u001b[0m \u001b[38;5;28mfloat\u001b[39m(reward),\n\u001b[0;32m (...)\u001b[0m\n\u001b[1;32m 173\u001b[0m jax_to_numpy(info),\n\u001b[1;32m 174\u001b[0m )\n", + "File \u001b[0;32m~/Documents/Code/solarcarsim/src/solarcarsim/simv1.py:123\u001b[0m, in \u001b[0;36mSolarRaceV1.step\u001b[0;34m(self, action)\u001b[0m\n\u001b[1;32m 120\u001b[0m reward \u001b[38;5;241m-\u001b[39m\u001b[38;5;241m=\u001b[39m \u001b[38;5;241m500\u001b[39m\n\u001b[1;32m 121\u001b[0m truncated \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;01mTrue\u001b[39;00m\n\u001b[0;32m--> 123\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_get_obs\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m, reward, terminated, truncated, {}\n", + "File \u001b[0;32m~/Documents/Code/solarcarsim/src/solarcarsim/simv1.py:64\u001b[0m, in \u001b[0;36mSolarRaceV1._get_obs\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m 63\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21m_get_obs\u001b[39m(\u001b[38;5;28mself\u001b[39m):\n\u001b[0;32m---> 64\u001b[0m slope_view, wind_view \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_vision_function\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 65\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m {\n\u001b[1;32m 66\u001b[0m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mposition\u001b[39m\u001b[38;5;124m\"\u001b[39m: \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_state[\u001b[38;5;241m0\u001b[39m],\n\u001b[1;32m 67\u001b[0m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mtime\u001b[39m\u001b[38;5;124m\"\u001b[39m: \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_state[\u001b[38;5;241m1\u001b[39m],\n\u001b[0;32m (...)\u001b[0m\n\u001b[1;32m 72\u001b[0m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mwind\u001b[39m\u001b[38;5;124m\"\u001b[39m: wind_view,\n\u001b[1;32m 73\u001b[0m }\n", + "File \u001b[0;32m~/Documents/Code/solarcarsim/src/solarcarsim/simv1.py:59\u001b[0m, in \u001b[0;36mSolarRaceV1._vision_function\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m 57\u001b[0m pos \u001b[38;5;241m=\u001b[39m jnp\u001b[38;5;241m.\u001b[39mastype(jnp\u001b[38;5;241m.\u001b[39mround(\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_state[\u001b[38;5;241m0\u001b[39m]), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mint32\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n\u001b[1;32m 58\u001b[0m time \u001b[38;5;241m=\u001b[39m jnp\u001b[38;5;241m.\u001b[39mastype(jnp\u001b[38;5;241m.\u001b[39mround(\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_state[\u001b[38;5;241m1\u001b[39m]), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mint32\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n\u001b[0;32m---> 59\u001b[0m wind_view \u001b[38;5;241m=\u001b[39m \u001b[43mslookup\u001b[49m\u001b[43m(\u001b[49m\u001b[43mjnp\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mhstack\u001b[49m\u001b[43m(\u001b[49m\u001b[43m[\u001b[49m\u001b[43mpos\u001b[49m\u001b[43m,\u001b[49m\u001b[43mtime\u001b[49m\u001b[43m]\u001b[49m\u001b[43m)\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 60\u001b[0m slope_view \u001b[38;5;241m=\u001b[39m jax\u001b[38;5;241m.\u001b[39mlax\u001b[38;5;241m.\u001b[39mdynamic_slice(\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_environment[\u001b[38;5;241m2\u001b[39m], pos, (\u001b[38;5;241m100\u001b[39m,))\n\u001b[1;32m 61\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m slope_view, wind_view\n", + "File \u001b[0;32m~/Documents/Code/solarcarsim/src/solarcarsim/simv1.py:56\u001b[0m, in \u001b[0;36mSolarRaceV1._vision_function..slookup\u001b[0;34m(x)\u001b[0m\n\u001b[1;32m 55\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mslookup\u001b[39m(x):\n\u001b[0;32m---> 56\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mjax\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mlax\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mdynamic_slice\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_environment\u001b[49m\u001b[43m[\u001b[49m\u001b[38;5;241;43m0\u001b[39;49m\u001b[43m]\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mx\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m100\u001b[39;49m\u001b[43m,\u001b[49m\u001b[38;5;241;43m100\u001b[39;49m\u001b[43m)\u001b[49m\u001b[43m)\u001b[49m\n", + "File \u001b[0;32m~/Documents/Code/solarcarsim/.venv/lib/python3.12/site-packages/jax/_src/lax/slicing.py:160\u001b[0m, in \u001b[0;36mdynamic_slice\u001b[0;34m(operand, start_indices, slice_sizes)\u001b[0m\n\u001b[1;32m 112\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mdynamic_slice\u001b[39m(\n\u001b[1;32m 113\u001b[0m operand: Array \u001b[38;5;241m|\u001b[39m np\u001b[38;5;241m.\u001b[39mndarray,\n\u001b[1;32m 114\u001b[0m start_indices: Array \u001b[38;5;241m|\u001b[39m np\u001b[38;5;241m.\u001b[39mndarray \u001b[38;5;241m|\u001b[39m Sequence[ArrayLike],\n\u001b[1;32m 115\u001b[0m slice_sizes: Shape,\n\u001b[1;32m 116\u001b[0m ) \u001b[38;5;241m-\u001b[39m\u001b[38;5;241m>\u001b[39m Array:\n\u001b[1;32m 117\u001b[0m \u001b[38;5;250m \u001b[39m\u001b[38;5;124;03m\"\"\"Wraps XLA's `DynamicSlice\u001b[39;00m\n\u001b[1;32m 118\u001b[0m \u001b[38;5;124;03m `_\u001b[39;00m\n\u001b[1;32m 119\u001b[0m \u001b[38;5;124;03m operator.\u001b[39;00m\n\u001b[0;32m (...)\u001b[0m\n\u001b[1;32m 158\u001b[0m \u001b[38;5;124;03m - :func:`jax.lax.dynamic_index_in_dim`\u001b[39;00m\n\u001b[1;32m 159\u001b[0m \u001b[38;5;124;03m \"\"\"\u001b[39;00m\n\u001b[0;32m--> 160\u001b[0m start_indices \u001b[38;5;241m=\u001b[39m \u001b[43m_dynamic_slice_indices\u001b[49m\u001b[43m(\u001b[49m\u001b[43moperand\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mstart_indices\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 161\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m config\u001b[38;5;241m.\u001b[39mdynamic_shapes\u001b[38;5;241m.\u001b[39mvalue:\n\u001b[1;32m 162\u001b[0m dynamic_sizes, static_sizes \u001b[38;5;241m=\u001b[39m lax\u001b[38;5;241m.\u001b[39m_extract_tracers_dyn_shape(slice_sizes)\n", + "File \u001b[0;32m~/Documents/Code/solarcarsim/.venv/lib/python3.12/site-packages/jax/_src/lax/slicing.py:3057\u001b[0m, in \u001b[0;36m_dynamic_slice_indices\u001b[0;34m(operand, start_indices)\u001b[0m\n\u001b[1;32m 3055\u001b[0m \u001b[38;5;28;01mcontinue\u001b[39;00m\n\u001b[1;32m 3056\u001b[0m d_arr \u001b[38;5;241m=\u001b[39m lax\u001b[38;5;241m.\u001b[39mconvert_element_type(d, _dtype(i))\n\u001b[0;32m-> 3057\u001b[0m result\u001b[38;5;241m.\u001b[39mappend(lax\u001b[38;5;241m.\u001b[39mselect(i \u001b[38;5;241m<\u001b[39m \u001b[38;5;241m0\u001b[39m, \u001b[43mi\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m+\u001b[39;49m\u001b[43m \u001b[49m\u001b[43md_arr\u001b[49m, i))\n\u001b[1;32m 3058\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m result\n", + "File \u001b[0;32m~/Documents/Code/solarcarsim/.venv/lib/python3.12/site-packages/jax/_src/numpy/array_methods.py:573\u001b[0m, in \u001b[0;36m_defer_to_unrecognized_arg..deferring_binary_op\u001b[0;34m(self, other)\u001b[0m\n\u001b[1;32m 571\u001b[0m args \u001b[38;5;241m=\u001b[39m (other, \u001b[38;5;28mself\u001b[39m) \u001b[38;5;28;01mif\u001b[39;00m swap \u001b[38;5;28;01melse\u001b[39;00m (\u001b[38;5;28mself\u001b[39m, other)\n\u001b[1;32m 572\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28misinstance\u001b[39m(other, _accepted_binop_types):\n\u001b[0;32m--> 573\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mbinary_op\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 574\u001b[0m \u001b[38;5;66;03m# Note: don't use isinstance here, because we don't want to raise for\u001b[39;00m\n\u001b[1;32m 575\u001b[0m \u001b[38;5;66;03m# subclasses, e.g. NamedTuple objects that may override operators.\u001b[39;00m\n\u001b[1;32m 576\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mtype\u001b[39m(other) \u001b[38;5;129;01min\u001b[39;00m _rejected_binop_types:\n", + "File \u001b[0;32m~/Documents/Code/solarcarsim/.venv/lib/python3.12/site-packages/jax/_src/numpy/ufunc_api.py:179\u001b[0m, in \u001b[0;36mufunc.__call__\u001b[0;34m(self, out, where, *args)\u001b[0m\n\u001b[1;32m 177\u001b[0m \u001b[38;5;28;01mraise\u001b[39;00m \u001b[38;5;167;01mNotImplementedError\u001b[39;00m(\u001b[38;5;124mf\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mwhere argument of \u001b[39m\u001b[38;5;132;01m{\u001b[39;00m\u001b[38;5;28mself\u001b[39m\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m\"\u001b[39m)\n\u001b[1;32m 178\u001b[0m call \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m__static_props[\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mcall\u001b[39m\u001b[38;5;124m'\u001b[39m] \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_call_vectorized\n\u001b[0;32m--> 179\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mcall\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m)\u001b[49m\n", "\u001b[0;31mKeyboardInterrupt\u001b[0m: " ] + }, + { + "ename": "", + "evalue": "", + "output_type": "error", + "traceback": [ + "\u001b[1;31mThe Kernel crashed while executing code in the current cell or a previous cell. \n", + "\u001b[1;31mPlease review the code in the cell(s) to identify a possible cause of the failure. \n", + "\u001b[1;31mClick here for more info. \n", + "\u001b[1;31mView Jupyter log for further details." + ] } ], "source": [ "# import a model and try it out!\n", - "from sbx import TD3\n", - "model = TD3(\"MultiInputPolicy\", env, verbose=1)\n", - "model.learn(total_timesteps=30_000)\n" + "model = PPO(\"MultiInputPolicy\", wrapped_env, verbose=1)\n", + "model.learn(total_timesteps=1_000_000)\n" ] }, { @@ -91,8 +2600,6 @@ "outputs": [], "source": [ "vec_env = model.get_env()\n", - "import matplotlib.pyplot as plt\n", - "import jax.numpy as jnp\n", "obs = vec_env.reset()\n", "actions = []\n", "obs_list = []\n", @@ -123,7 +2630,7 @@ { "data": { "text/plain": [ - "[]" + "[]" ] }, "execution_count": 5, @@ -132,7 +2639,7 @@ }, { "data": { - "image/png": "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", + "image/png": "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", "text/plain": [ "
" ] @@ -145,7 +2652,7 @@ "fig, (ax1, ax2, ax3) = plt.subplots(3,1, figsize=(12,6))\n", "ax1.plot(position, label=\"position\")\n", "ax2.plot(actions, label=\"energy\")\n", - "ax3.plot(rewards[0:250])\n", + "ax3.plot(rewards)\n", "# plt.legend()" ] }, @@ -157,53 +2664,22 @@ { "data": { "text/plain": [ - "Array([-1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.,\n", - " -1., -1., -1.], dtype=float32)" + "Array([1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,\n", + " 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,\n", + " 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,\n", + " 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,\n", + " 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,\n", + " 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,\n", + " 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,\n", + " 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,\n", + " 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,\n", + " 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,\n", + " 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,\n", + " 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,\n", + " 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,\n", + " 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,\n", + " 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,\n", + " 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.], dtype=float32)" ] }, "execution_count": 6, @@ -214,13 +2690,6 @@ "source": [ "actions" ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": {}, - "outputs": [], - "source": [] } ], "metadata": { diff --git a/notebooks/v1sim.ipynb b/notebooks/v1sim.ipynb index d076f90..d852456 100644 --- a/notebooks/v1sim.ipynb +++ b/notebooks/v1sim.ipynb @@ -477,24 +477,14 @@ }, { "cell_type": "code", - "execution_count": 18, + "execution_count": 13, "metadata": {}, "outputs": [ { "data": { + "image/png": "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", "text/plain": [ - "Text(0, 0.5, 'Slope (rad)')" - ] - }, - "execution_count": 18, - "metadata": {}, - "output_type": "execute_result" - }, - { - "data": { - "image/png": "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", - "text/plain": [ - "
" + "
" ] }, "metadata": {}, @@ -504,20 +494,32 @@ "source": [ "%run ../src/solarcarsim/physsim.py\n", "from jax import random\n", + "import matplotlib.pyplot as plt\n", + "plt.rcParams.update({\n", + " \"text.usetex\": True,\n", + "})\n", "wind, elevation, slope = make_environment(random.key(123))\n", - "fig, (ax1, ax2) = plt.subplots(1,2, figsize=(12,6))\n", + "fig, (ax1, ax2) = plt.subplots(1,2, figsize=(10,4))\n", "fig.set_tight_layout('auto')\n", "fig.suptitle(\"Generated Environment\")\n", "\n", "ax1.imshow(wind, aspect='auto')\n", + "ax1.set_title(\"Wind Map\")\n", "ax1.set_ylabel(\"Time (sec)\")\n", "ax1.set_xlabel(\"Distance (m)\")\n", + "\n", + "\n", + "ax2.set_title(\"Terrain\")\n", "ax_slope = ax2.twinx()\n", - "ax2.plot(elevation)\n", + "\n", + "ax2.plot(elevation, label=\"Elevation\")\n", "ax2.set_ylabel(\"Elevation (m)\")\n", "ax2.set_xlabel(\"Distance (m)\")\n", - "ax_slope.plot(slope, color='r')\n", - "ax_slope.set_ylabel(\"Slope (rad)\")" + "ax_slope.plot(slope, color='r', label='Slope')\n", + "ax_slope.set_ylabel(\"Slope (rad)\")\n", + "ax2.legend(loc=2)\n", + "ax_slope.legend(loc=1)\n", + "fig.savefig(\"environment.pdf\")" ] }, { @@ -548,6 +550,7 @@ ], "source": [ "# testing the indexing into the wind array.\n", + "ax2.legend()\n", "# given an array of shape (10,2)\n", "# return an array of (10,100,100)\n", "key = random.key(0)\n", diff --git a/pdm.lock b/pdm.lock index 011a0cf..c19b2b4 100644 --- a/pdm.lock +++ b/pdm.lock @@ -5,7 +5,7 @@ groups = ["default", "dev"] strategy = ["inherit_metadata"] lock_version = "4.5.0" -content_hash = "sha256:a3b65f863c554725c33d452fd759776141740661fa3555d306ed08563a7e16e2" +content_hash = "sha256:2f7c4bee801973a3b7856ba0707891eb01fd05659948707f44be4aa302e5dabd" [[metadata.targets]] requires_python = ">=3.12,<3.13" @@ -237,6 +237,27 @@ files = [ {file = "decorator-5.1.1.tar.gz", hash = "sha256:637996211036b6385ef91435e4fae22989472f9d571faba8927ba8253acbc330"}, ] +[[package]] +name = "distrax" +version = "0.1.5" +requires_python = ">=3.9" +summary = "Distrax: Probability distributions in JAX." +groups = ["default"] +marker = "python_version >= \"3.12\" and python_version < \"3.13\"" +dependencies = [ + "absl-py>=0.9.0", + "chex>=0.1.8", + "jax>=0.1.55", + "jaxlib>=0.1.67", + "numpy>=1.23.0", + "setuptools; python_version >= \"3.12\"", + "tensorflow-probability>=0.15.0", +] +files = [ + {file = "distrax-0.1.5-py3-none-any.whl", hash = "sha256:5020f4b53a9a480d019c12e44292fbacb7de857cce478bc594dacf29519c61b7"}, + {file = "distrax-0.1.5.tar.gz", hash = "sha256:ec41522d389af69efedc8d475a7e6d8f229429c00f2140dcd641feacf7e21948"}, +] + [[package]] name = "dm-tree" version = "0.1.8" @@ -254,6 +275,18 @@ files = [ {file = "dm_tree-0.1.8-cp312-cp312-win_amd64.whl", hash = "sha256:96a548a406a6fb15fe58f6a30a57ff2f2aafbf25f05afab00c8f5e5977b6c715"}, ] +[[package]] +name = "docstring-parser" +version = "0.16" +requires_python = ">=3.6,<4.0" +summary = "Parse Python docstrings in reST, Google and Numpydoc format" +groups = ["default"] +marker = "python_version >= \"3.12\" and python_version < \"3.13\"" +files = [ + {file = "docstring_parser-0.16-py3-none-any.whl", hash = "sha256:bf0a1387354d3691d102edef7ec124f219ef639982d096e26e3b60aeffa90637"}, + {file = "docstring_parser-0.16.tar.gz", hash = "sha256:538beabd0af1e2db0146b6bd3caa526c35a34d61af9fd2887f3a8a27a739aa6e"}, +] + [[package]] name = "etils" version = "1.11.0" @@ -408,6 +441,26 @@ files = [ {file = "gast-0.6.0.tar.gz", hash = "sha256:88fc5300d32c7ac6ca7b515310862f71e6fdf2c029bbec7c66c0f5dd47b6b1fb"}, ] +[[package]] +name = "grpcio" +version = "1.68.1" +requires_python = ">=3.8" +summary = "HTTP/2-based RPC framework" +groups = ["default"] +marker = "python_version >= \"3.12\" and python_version < \"3.13\"" +files = [ + {file = "grpcio-1.68.1-cp312-cp312-linux_armv7l.whl", hash = "sha256:8829924fffb25386995a31998ccbbeaa7367223e647e0122043dfc485a87c666"}, + {file = "grpcio-1.68.1-cp312-cp312-macosx_10_9_universal2.whl", hash = "sha256:3aed6544e4d523cd6b3119b0916cef3d15ef2da51e088211e4d1eb91a6c7f4f1"}, + {file = "grpcio-1.68.1-cp312-cp312-manylinux_2_17_aarch64.whl", hash = "sha256:4efac5481c696d5cb124ff1c119a78bddbfdd13fc499e3bc0ca81e95fc573684"}, + {file = "grpcio-1.68.1-cp312-cp312-manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:6ab2d912ca39c51f46baf2a0d92aa265aa96b2443266fc50d234fa88bf877d8e"}, + {file = "grpcio-1.68.1-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:95c87ce2a97434dffe7327a4071839ab8e8bffd0054cc74cbe971fba98aedd60"}, + {file = "grpcio-1.68.1-cp312-cp312-musllinux_1_1_i686.whl", hash = "sha256:e4842e4872ae4ae0f5497bf60a0498fa778c192cc7a9e87877abd2814aca9475"}, + {file = "grpcio-1.68.1-cp312-cp312-musllinux_1_1_x86_64.whl", hash = "sha256:255b1635b0ed81e9f91da4fcc8d43b7ea5520090b9a9ad9340d147066d1d3613"}, + {file = "grpcio-1.68.1-cp312-cp312-win32.whl", hash = "sha256:7dfc914cc31c906297b30463dde0b9be48e36939575eaf2a0a22a8096e69afe5"}, + {file = "grpcio-1.68.1-cp312-cp312-win_amd64.whl", hash = "sha256:a0c8ddabef9c8f41617f213e527254c41e8b96ea9d387c632af878d05db9229c"}, + {file = "grpcio-1.68.1.tar.gz", hash = "sha256:44a8502dd5de653ae6a73e2de50a401d84184f0331d0ac3daeb044e66d5c5054"}, +] + [[package]] name = "gym" version = "0.26.2" @@ -622,6 +675,76 @@ files = [ {file = "jax-0.4.37.tar.gz", hash = "sha256:7774f3d9e23fe199c65589c680c5a5be87a183b89598421a632d8245222b637b"}, ] +[[package]] +name = "jax-cuda12-pjrt" +version = "0.4.36" +summary = "JAX XLA PJRT Plugin for NVIDIA GPUs" +groups = ["default"] +marker = "python_version >= \"3.12\" and python_version < \"3.13\"" +files = [ + {file = "jax_cuda12_pjrt-0.4.36-py3-none-manylinux2014_aarch64.whl", hash = "sha256:1dfc0bec0820ba801b61e9421064b6e58238c430b4ad8f54043323d93c0217c6"}, + {file = "jax_cuda12_pjrt-0.4.36-py3-none-manylinux2014_x86_64.whl", hash = "sha256:e3c3705d8db7d63da9abfaebf06f5cd0667f5acb0748a5c5eb00d80041e922ed"}, +] + +[[package]] +name = "jax-cuda12-plugin" +version = "0.4.36" +requires_python = ">=3.10" +summary = "JAX Plugin for NVIDIA GPUs" +groups = ["default"] +marker = "python_version >= \"3.12\" and python_version < \"3.13\"" +dependencies = [ + "jax-cuda12-pjrt==0.4.36", +] +files = [ + {file = "jax_cuda12_plugin-0.4.36-cp312-cp312-manylinux2014_aarch64.whl", hash = "sha256:6a0b0c2bdc1da2eea2c20723a1e8f39b3cda67d24c665de936647e8091f5790d"}, + {file = "jax_cuda12_plugin-0.4.36-cp312-cp312-manylinux2014_x86_64.whl", hash = "sha256:5d4727fb519fedc06a9a984d5a0714804d81ef126a2cb60cefd5cbc4a3ea2627"}, +] + +[[package]] +name = "jax-cuda12-plugin" +version = "0.4.36" +extras = ["with_cuda"] +requires_python = ">=3.10" +summary = "JAX Plugin for NVIDIA GPUs" +groups = ["default"] +marker = "python_version >= \"3.12\" and python_version < \"3.13\"" +dependencies = [ + "jax-cuda12-plugin==0.4.36", + "nvidia-cublas-cu12>=12.1.3.1", + "nvidia-cuda-cupti-cu12>=12.1.105", + "nvidia-cuda-nvcc-cu12>=12.6.85", + "nvidia-cuda-runtime-cu12>=12.1.105", + "nvidia-cudnn-cu12<10.0,>=9.1", + "nvidia-cufft-cu12>=11.0.2.54", + "nvidia-cusolver-cu12>=11.4.5.107", + "nvidia-cusparse-cu12>=12.1.0.106", + "nvidia-nccl-cu12>=2.18.1", + "nvidia-nvjitlink-cu12>=12.1.105", +] +files = [ + {file = "jax_cuda12_plugin-0.4.36-cp312-cp312-manylinux2014_aarch64.whl", hash = "sha256:6a0b0c2bdc1da2eea2c20723a1e8f39b3cda67d24c665de936647e8091f5790d"}, + {file = "jax_cuda12_plugin-0.4.36-cp312-cp312-manylinux2014_x86_64.whl", hash = "sha256:5d4727fb519fedc06a9a984d5a0714804d81ef126a2cb60cefd5cbc4a3ea2627"}, +] + +[[package]] +name = "jax" +version = "0.4.37" +extras = ["cuda12"] +requires_python = ">=3.10" +summary = "Differentiate, compile, and transform Numpy code." +groups = ["default"] +marker = "python_version >= \"3.12\" and python_version < \"3.13\"" +dependencies = [ + "jax-cuda12-plugin[with_cuda]<=0.4.37,>=0.4.36", + "jax==0.4.37", + "jaxlib==0.4.36", +] +files = [ + {file = "jax-0.4.37-py3-none-any.whl", hash = "sha256:bdc0686d7e5a944e2d38026eae632214d98dd2d91869cbcedbf1c11298ae3e3e"}, + {file = "jax-0.4.37.tar.gz", hash = "sha256:7774f3d9e23fe199c65589c680c5a5be87a183b89598421a632d8245222b637b"}, +] + [[package]] name = "jaxlib" version = "0.4.36" @@ -737,6 +860,21 @@ files = [ {file = "kiwisolver-1.4.7.tar.gz", hash = "sha256:9893ff81bd7107f7b685d3017cc6583daadb4fc26e4a888350df530e41980a60"}, ] +[[package]] +name = "markdown" +version = "3.7" +requires_python = ">=3.8" +summary = "Python implementation of John Gruber's Markdown." +groups = ["default"] +marker = "python_version >= \"3.12\" and python_version < \"3.13\"" +dependencies = [ + "importlib-metadata>=4.4; python_version < \"3.10\"", +] +files = [ + {file = "Markdown-3.7-py3-none-any.whl", hash = "sha256:7eb6df5690b81a1d7942992c97fad2938e956e79df20cbc6186e9c3a77b1c803"}, + {file = "markdown-3.7.tar.gz", hash = "sha256:2ae2471477cfd02dbbf038d5d9bc226d40def84b4fe2986e49b59b6b472bbed2"}, +] + [[package]] name = "markdown-it-py" version = "3.0.0" @@ -937,7 +1075,7 @@ version = "12.4.5.8" requires_python = ">=3" summary = "CUBLAS native runtime libraries" groups = ["default"] -marker = "platform_system == \"Linux\" and platform_machine == \"x86_64\" and python_version >= \"3.12\" and python_version < \"3.13\"" +marker = "python_version >= \"3.12\" and python_version < \"3.13\"" files = [ {file = "nvidia_cublas_cu12-12.4.5.8-py3-none-manylinux2014_aarch64.whl", hash = "sha256:0f8aa1706812e00b9f19dfe0cdb3999b092ccb8ca168c0db5b8ea712456fd9b3"}, {file = "nvidia_cublas_cu12-12.4.5.8-py3-none-manylinux2014_x86_64.whl", hash = "sha256:2fc8da60df463fdefa81e323eef2e36489e1c94335b5358bcb38360adf75ac9b"}, @@ -950,13 +1088,26 @@ version = "12.4.127" requires_python = ">=3" summary = "CUDA profiling tools runtime libs." groups = ["default"] -marker = "platform_system == \"Linux\" and platform_machine == \"x86_64\" and python_version >= \"3.12\" and python_version < \"3.13\"" +marker = "python_version >= \"3.12\" and python_version < \"3.13\"" files = [ {file = "nvidia_cuda_cupti_cu12-12.4.127-py3-none-manylinux2014_aarch64.whl", hash = "sha256:79279b35cf6f91da114182a5ce1864997fd52294a87a16179ce275773799458a"}, {file = "nvidia_cuda_cupti_cu12-12.4.127-py3-none-manylinux2014_x86_64.whl", hash = "sha256:9dec60f5ac126f7bb551c055072b69d85392b13311fcc1bcda2202d172df30fb"}, {file = "nvidia_cuda_cupti_cu12-12.4.127-py3-none-win_amd64.whl", hash = "sha256:5688d203301ab051449a2b1cb6690fbe90d2b372f411521c86018b950f3d7922"}, ] +[[package]] +name = "nvidia-cuda-nvcc-cu12" +version = "12.6.85" +requires_python = ">=3" +summary = "CUDA nvcc" +groups = ["default"] +marker = "python_version >= \"3.12\" and python_version < \"3.13\"" +files = [ + {file = "nvidia_cuda_nvcc_cu12-12.6.85-py3-none-manylinux1_x86_64.manylinux_2_5_x86_64.whl", hash = "sha256:d75d9d74599f4d7c0865df19ed21b739e6cb77a6497a3f73d6f61e8038a765e4"}, + {file = "nvidia_cuda_nvcc_cu12-12.6.85-py3-none-manylinux2014_aarch64.manylinux_2_17_aarch64.whl", hash = "sha256:5d2edd5531b13e3daac8ffee9fc2b70a147e6088b2af2565924773d63d36d294"}, + {file = "nvidia_cuda_nvcc_cu12-12.6.85-py3-none-win_amd64.whl", hash = "sha256:aa04742337973dcb5bcccabb590edc8834c60ebfaf971847888d24ffef6c46b5"}, +] + [[package]] name = "nvidia-cuda-nvrtc-cu12" version = "12.4.127" @@ -976,7 +1127,7 @@ version = "12.4.127" requires_python = ">=3" summary = "CUDA Runtime native Libraries" groups = ["default"] -marker = "platform_system == \"Linux\" and platform_machine == \"x86_64\" and python_version >= \"3.12\" and python_version < \"3.13\"" +marker = "python_version >= \"3.12\" and python_version < \"3.13\"" files = [ {file = "nvidia_cuda_runtime_cu12-12.4.127-py3-none-manylinux2014_aarch64.whl", hash = "sha256:961fe0e2e716a2a1d967aab7caee97512f71767f852f67432d572e36cb3a11f3"}, {file = "nvidia_cuda_runtime_cu12-12.4.127-py3-none-manylinux2014_x86_64.whl", hash = "sha256:64403288fa2136ee8e467cdc9c9427e0434110899d07c779f25b5c068934faa5"}, @@ -989,7 +1140,7 @@ version = "9.1.0.70" requires_python = ">=3" summary = "cuDNN runtime libraries" groups = ["default"] -marker = "platform_system == \"Linux\" and platform_machine == \"x86_64\" and python_version >= \"3.12\" and python_version < \"3.13\"" +marker = "python_version >= \"3.12\" and python_version < \"3.13\"" dependencies = [ "nvidia-cublas-cu12", ] @@ -1004,7 +1155,7 @@ version = "11.2.1.3" requires_python = ">=3" summary = "CUFFT native runtime libraries" groups = ["default"] -marker = "platform_system == \"Linux\" and platform_machine == \"x86_64\" and python_version >= \"3.12\" and python_version < \"3.13\"" +marker = "python_version >= \"3.12\" and python_version < \"3.13\"" dependencies = [ "nvidia-nvjitlink-cu12", ] @@ -1033,7 +1184,7 @@ version = "11.6.1.9" requires_python = ">=3" summary = "CUDA solver native runtime libraries" groups = ["default"] -marker = "platform_system == \"Linux\" and platform_machine == \"x86_64\" and python_version >= \"3.12\" and python_version < \"3.13\"" +marker = "python_version >= \"3.12\" and python_version < \"3.13\"" dependencies = [ "nvidia-cublas-cu12", "nvidia-cusparse-cu12", @@ -1051,7 +1202,7 @@ version = "12.3.1.170" requires_python = ">=3" summary = "CUSPARSE native runtime libraries" groups = ["default"] -marker = "platform_system == \"Linux\" and platform_machine == \"x86_64\" and python_version >= \"3.12\" and python_version < \"3.13\"" +marker = "python_version >= \"3.12\" and python_version < \"3.13\"" dependencies = [ "nvidia-nvjitlink-cu12", ] @@ -1067,7 +1218,7 @@ version = "2.21.5" requires_python = ">=3" summary = "NVIDIA Collective Communication Library (NCCL) Runtime" groups = ["default"] -marker = "platform_system == \"Linux\" and platform_machine == \"x86_64\" and python_version >= \"3.12\" and python_version < \"3.13\"" +marker = "python_version >= \"3.12\" and python_version < \"3.13\"" files = [ {file = "nvidia_nccl_cu12-2.21.5-py3-none-manylinux2014_x86_64.whl", hash = "sha256:8579076d30a8c24988834445f8d633c697d42397e92ffc3f63fa26766d25e0a0"}, ] @@ -1078,7 +1229,7 @@ version = "12.4.127" requires_python = ">=3" summary = "Nvidia JIT LTO Library" groups = ["default"] -marker = "platform_system == \"Linux\" and platform_machine == \"x86_64\" and python_version >= \"3.12\" and python_version < \"3.13\"" +marker = "python_version >= \"3.12\" and python_version < \"3.13\"" files = [ {file = "nvidia_nvjitlink_cu12-12.4.127-py3-none-manylinux2014_aarch64.whl", hash = "sha256:4abe7fef64914ccfa909bc2ba39739670ecc9e820c83ccc7a6ed414122599b83"}, {file = "nvidia_nvjitlink_cu12-12.4.127-py3-none-manylinux2014_x86_64.whl", hash = "sha256:06b3b9b25bf3f8af351d664978ca26a16d2c5127dbd53c0497e28d1fb9611d57"}, @@ -1759,6 +1910,18 @@ files = [ {file = "shiboken6-6.8.0.2-cp39-abi3-win_amd64.whl", hash = "sha256:b11e750e696bb565d897e0f5836710edfb86bd355f87b09988bd31b2aad404d3"}, ] +[[package]] +name = "shtab" +version = "1.7.1" +requires_python = ">=3.7" +summary = "Automagic shell tab completion for Python CLI applications" +groups = ["default"] +marker = "python_version >= \"3.12\" and python_version < \"3.13\"" +files = [ + {file = "shtab-1.7.1-py3-none-any.whl", hash = "sha256:32d3d2ff9022d4c77a62492b6ec875527883891e33c6b479ba4d41a51e259983"}, + {file = "shtab-1.7.1.tar.gz", hash = "sha256:4e4bcb02eeb82ec45920a5d0add92eac9c9b63b2804c9196c1f1fdc2d039243c"}, +] + [[package]] name = "simplejson" version = "3.19.3" @@ -1847,6 +2010,42 @@ files = [ {file = "sympy-1.13.1.tar.gz", hash = "sha256:9cebf7e04ff162015ce31c9c6c9144daa34a93bd082f54fd8f12deca4f47515f"}, ] +[[package]] +name = "tensorboard" +version = "2.18.0" +requires_python = ">=3.9" +summary = "TensorBoard lets you watch Tensors Flow" +groups = ["default"] +marker = "python_version >= \"3.12\" and python_version < \"3.13\"" +dependencies = [ + "absl-py>=0.4", + "grpcio>=1.48.2", + "markdown>=2.6.8", + "numpy>=1.12.0", + "packaging", + "protobuf!=4.24.0,>=3.19.6", + "setuptools>=41.0.0", + "six>1.9", + "tensorboard-data-server<0.8.0,>=0.7.0", + "werkzeug>=1.0.1", +] +files = [ + {file = "tensorboard-2.18.0-py3-none-any.whl", hash = "sha256:107ca4821745f73e2aefa02c50ff70a9b694f39f790b11e6f682f7d326745eab"}, +] + +[[package]] +name = "tensorboard-data-server" +version = "0.7.2" +requires_python = ">=3.7" +summary = "Fast data loading for TensorBoard" +groups = ["default"] +marker = "python_version >= \"3.12\" and python_version < \"3.13\"" +files = [ + {file = "tensorboard_data_server-0.7.2-py3-none-any.whl", hash = "sha256:7e0610d205889588983836ec05dc098e80f97b7e7bbff7e994ebb78f578d0ddb"}, + {file = "tensorboard_data_server-0.7.2-py3-none-macosx_10_9_x86_64.whl", hash = "sha256:9fe5d24221b29625dbc7328b0436ca7fc1c23de4acf4d272f1180856e32f9f60"}, + {file = "tensorboard_data_server-0.7.2-py3-none-manylinux_2_31_x86_64.whl", hash = "sha256:ef687163c24185ae9754ed5650eb5bc4d84ff257aabdc33f0cc6f74d8ba54530"}, +] + [[package]] name = "tensorflow-probability" version = "0.25.0" @@ -1997,6 +2196,22 @@ files = [ {file = "triton-3.1.0-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:c8182f42fd8080a7d39d666814fa36c5e30cc00ea7eeeb1a2983dbb4c99a0fdc"}, ] +[[package]] +name = "typeguard" +version = "4.4.1" +requires_python = ">=3.9" +summary = "Run-time type checker for Python" +groups = ["default"] +marker = "python_version >= \"3.12\" and python_version < \"3.13\"" +dependencies = [ + "importlib-metadata>=3.6; python_version < \"3.10\"", + "typing-extensions>=4.10.0", +] +files = [ + {file = "typeguard-4.4.1-py3-none-any.whl", hash = "sha256:9324ec07a27ec67fc54a9c063020ca4c0ae6abad5e9f0f9804ca59aee68c6e21"}, + {file = "typeguard-4.4.1.tar.gz", hash = "sha256:0d22a89d00b453b47c49875f42b6601b961757541a2e1e0ef517b6e24213c21b"}, +] + [[package]] name = "typing-extensions" version = "4.12.2" @@ -2009,6 +2224,29 @@ files = [ {file = "typing_extensions-4.12.2.tar.gz", hash = "sha256:1a7ead55c7e559dd4dee8856e3a88b41225abfe1ce8df57b7c13915fe121ffb8"}, ] +[[package]] +name = "tyro" +version = "0.9.2" +requires_python = ">=3.7" +summary = "CLI interfaces & config objects, from types" +groups = ["default"] +marker = "python_version >= \"3.12\" and python_version < \"3.13\"" +dependencies = [ + "backports-cached-property>=1.0.2; python_version < \"3.8\"", + "colorama>=0.4.0; platform_system == \"Windows\"", + "docstring-parser>=0.16", + "eval-type-backport>=0.1.3; python_version < \"3.10\"", + "rich>=11.1.0", + "shtab>=1.5.6", + "typeguard>=4.0.0", + "typing-extensions>=4.7.0; python_version < \"3.8\"", + "typing-extensions>=4.9.0; python_version >= \"3.8\"", +] +files = [ + {file = "tyro-0.9.2-py3-none-any.whl", hash = "sha256:f7c301b30b1ac7b18672f234e45013786c494d64c0e3621b25b8414637af8f90"}, + {file = "tyro-0.9.2.tar.gz", hash = "sha256:692687e07c1ed35cc3a841e8c4a188424023f16bdef37f2d9c23cbeb8a3b51aa"}, +] + [[package]] name = "tzdata" version = "2024.2" @@ -2063,6 +2301,21 @@ files = [ {file = "wcwidth-0.2.13.tar.gz", hash = "sha256:72ea0c06399eb286d978fdedb6923a9eb47e1c486ce63e9b4e64fc18303972b5"}, ] +[[package]] +name = "werkzeug" +version = "3.1.3" +requires_python = ">=3.9" +summary = "The comprehensive WSGI web application library." +groups = ["default"] +marker = "python_version >= \"3.12\" and python_version < \"3.13\"" +dependencies = [ + "MarkupSafe>=2.1.1", +] +files = [ + {file = "werkzeug-3.1.3-py3-none-any.whl", hash = "sha256:54b78bf3716d19a65be4fceccc0d1d7b89e608834989dfae50ea87564639213e"}, + {file = "werkzeug-3.1.3.tar.gz", hash = "sha256:60723ce945c19328679790e3282cc758aa4a6040e4bb330f53d30fa546d44746"}, +] + [[package]] name = "zipp" version = "3.21.0" diff --git a/pyproject.toml b/pyproject.toml index 1f5c82b..4c0e2a8 100644 --- a/pyproject.toml +++ b/pyproject.toml @@ -5,7 +5,7 @@ description = "A solar car racing simulation library and GUI tool" authors = [ {name = "saji", email = "saji@saji.dev"}, ] -dependencies = ["pyqtgraph>=0.13.7", "jax>=0.4.37", "pytest>=8.3.3", "pyside6>=6.8.0.2", "matplotlib>=3.9.2", "gymnasium[jax]>=1.0.0", "pyvista>=0.44.2", "pyvistaqt>=0.11.1", "stable-baselines3>=2.4.0", "gymnax>=0.0.8", "sbx-rl>=0.18.0"] +dependencies = ["pyqtgraph>=0.13.7", "jax[cuda12]>=0.4.37", "pytest>=8.3.3", "pyside6>=6.8.0.2", "matplotlib>=3.9.2", "gymnasium[jax]>=1.0.0", "pyvista>=0.44.2", "pyvistaqt>=0.11.1", "stable-baselines3>=2.4.0", "gymnax>=0.0.8", "sbx-rl>=0.18.0", "tyro>=0.9.2", "tensorboard>=2.18.0", "distrax>=0.1.5"] requires-python = ">=3.10,<3.13" readme = "README.md" license = {text = "MIT"} diff --git a/report/report.tex b/report/report.tex index 4cddc6c..2839178 100644 --- a/report/report.tex +++ b/report/report.tex @@ -7,21 +7,23 @@ \usepackage{graphicx} \usepackage[margin=1in]{geometry} \usepackage{hyperref} +\usepackage{pdfpages} \usepackage{algorithm} \usepackage{algorithmic} \usepackage{float} \usepackage{booktabs} \usepackage{caption} \usepackage{subcaption} +\usepackage{tikz} % Custom commands \newcommand{\sectionheading}[1]{\noindent\textbf{#1}} % Title and author information -\title{\Large{Your Project Title: A Study in Optimal Control and Reinforcement Learning}} -\author{Your Name\\ - Course Name\\ - Institution} +\title{\Large{Solarcarsim: A Solar Racing Environment for RL Agents}} +\author{Saji Champlin\\ + EE5241 + } \date{\today} \begin{document} @@ -33,102 +35,273 @@ Solar Racing is a competition with the goal of creating highly efficient solar-a requires awareness and complex decision making to determine optimal speeds to exploit the environmental conditions, such as winds, cloud cover, and changes in elevation. We present an environment modelled on the dynamics involved for a race, including generated elevation and wind profiles. The model uses the \texttt{gymnasium} interface to allow it to be used by a variety of algorithms. -We demonstrate a method of designing reward functions for multi-objective problems. The environment shows to be solvable by modern -reinforcement learning algorithms. +We demonstrate a method of designing reward functions for multi-objective problems. We show learning using an Jax-based PPO model. \end{abstract} \section{Introduction} -Start with a broad context of your problem area in optimal control/reinforcement learning. Then narrow down to your specific focus. Include: -\begin{itemize} - \item Problem motivation - \item Brief overview of existing approaches - \item Your specific contributions - \item Paper organization -\end{itemize} Solar racing was invented in the early 90s as a technology incubator for high-efficiency motor vehicles. The first solar races were speed focused, however a style of race that focused on minimal energy use within a given route was developed to push focus towards vehicle efficiency. The goal of these races is to arrive at a destination within a given time frame, while using as little grid (non-solar) energy as possible. -Aerodynamic drag is one of the most significant sources of energy consumption, along with elevation changes. The simplest policy to meet -the constraints of on-time arrival is: -$$ -V_{\text{avg}} = \frac{D}{T} -$$ -Where $D$ is the distance needed to travel, and $T$ is the maximum allowed time. +Optimal driving is a complex policy based on terrain slope, wind forecasts, and solar forecasts. + +Direct solutions to find the global minima of the energy usage on a route +segment are difficult to compute. Instead, we present a reinforcement learning +environment that can be used to train RL agents to race efficiently given +limited foresight. The environment simulates key components of the race, such +as terrain and wind, as well as car dynamics. The simulator is written using +the Jax\cite{jax2018github} library which enables computations to be offloaded to the GPU. We +provide wrappers for the \texttt{gymnasium} API as well as a +\texttt{purejaxrl}\cite{lu2022discovered} implementation which can train a PPO +agent with millions of timesteps in several minutes. We present an exploration of reward function design with regards to sparsity, +magnitude, and learning efficiency. \section{Background} -Provide necessary background on: -\begin{itemize} - \item Your specific application domain - \item Relevant algorithms and methods - \item Previous work in this area -\end{itemize} +Performance evaluation for solar races typically take the form of +$$ + S = D/E \times T +$$ +Where $S$ is the score, $D$ is the distance travelled, $E$ is the energy consumed, and $T$ is the speed derating. +The speed derate is calculated based on a desired average speed throughout the race. If average velocity is at or above the target, +$T=1$, however $T$ approaches $0$ exponentially as the average velocity goes below the target. +Based on this metric we conclude that the optimal score: +\begin{enumerate} + \item Maintains an average speed $V_{\text{avg}}$ as required by the derate. + \item Minimizes energy usage otherwise. +\end{enumerate} +The simplest control to meet the constraints of on-time arrival is: +$$ +V_{\text{avg}} = \frac{D_{goal}}{T_{goal}} +$$ +Where $D_{goal}$ is the distance needed to travel, and $T_{goal}$ is the maximum allowed time. The average speed is nearly-optimal in most cases, but +is not a globally optimal solution. Consider a small - there is much more energy being used from the battery when going uphill, +but the same energy is returned to the car going downhill. Losses in the vehicle dictate that it is more effective to drive slowly +up the hill, and speed up on the descent. The decision is further complicated by wind and cloud cover, which can aid or neuter the +performance of the car. It is therefore of great interest for solar racing teams to have advanced strategies that can effectively +traverse the terrain while minimizing environmental resistances. + +Existing research on this subject is limited, as advanced solar car strategy is a competitive differentiator and is usually kept secret. +However, the author knows that most of the work on this subject involves use of Modelica or similar acausal system simulators, and +non-linear solvers that use multi-starts to attempt to find the global optimum. Other methods include exhaustive search, genetic algorithms, +and Big Bang-Big Crunch optimization\cite{heuristicsolar}. + +We start by analyzing a simple force-based model of the car, and then connect this to an energy system using motor equations. We generate +a simulated environment including terrain and wind. Then, we develop a reward system that encapsulates the goals of the environment. +Finally, we train off-the-shelf RL models from Stable Baselines3 and purejaxrl to show learning on the environment. \section{Methodology} -Describe your approach in detail: -\begin{itemize} - \item Problem formulation - \item Algorithm description - \item Implementation details -\end{itemize} +% \begin{tikzpicture}[scale=1.5] +% % Define slope angle +% \def\angle{30} +% +% % Draw ground/slope +% \draw[thick] (-3,0) -- (3,0); +% \draw[thick] (-2,0) -- (2.5,2); +% +% % Draw angle arc and label +% \draw (-,0) arc (0:\angle:0.5); +% \node at (0.4,0.3) {$\theta$}; +% +% % Draw simplified car (rectangle) +% \begin{scope}[rotate=\angle,shift={(0,1)}] +% \draw[thick] (-0.8,-0.4) rectangle (0.8,0.4); +% % Add wheels +% \fill[black] (-0.6,-0.4) circle (0.1); +% \fill[black] (0.6,-0.4) circle (0.1); +% \end{scope} +% +% % Draw forces +% % Weight force +% \draw[->,thick,red] (0,1) -- (0,0) node[midway,right] {$W$}; +% +% % Normal force +% \draw[->,thick,blue] (0,1) -- ({-sin(\angle)},{cos(\angle)}) node[midway,above left] {$N$}; +% +% % Downslope component +% \draw[->,thick,green] (0,1) -- ({cos(\angle)},{sin(\angle)}) node[midway,below right] {$W\sin\theta$}; +% \end{tikzpicture} -% Example of how to include an algorithm -\begin{algorithm}[H] -\caption{Your Algorithm Name} -\begin{algorithmic}[1] -\STATE Initialize parameters -\WHILE{not converged} - \STATE Update step -\ENDWHILE -\RETURN Result -\end{algorithmic} -\end{algorithm} - -\section{Experiments and Results} -Present your findings: -\begin{itemize} - \item Experimental setup - \item Results and analysis - \item Comparison with baselines (if applicable) -\end{itemize} - -% Example of how to include figures \begin{figure}[H] +\begin{tikzpicture}[scale=1.5] + % Define slope angle + \def\angle{30} + + % Calculate some points for consistent geometry + \def\slopeStart{-2} + \def\slopeEnd{2} + \def\slopeHeight{2.309} % tan(30°) * 2 + + % Draw ground (horizontal line) + \draw[thick] (-3,0) -- (3,0); + + % Draw slope + \draw[thick] (\slopeStart,0) -- (\slopeEnd,\slopeHeight); + + + % Calculate car center position on slope + \def\carX{0} % Center position along x-axis + \def\carY{1.6} % tan(30°) * carX + appropriate offset + + % Draw car (rectangle) exactly on slope + \begin{scope}[shift={(\carX,\carY)}, rotate=\angle] + \draw[thick] (-0.6,-0.3) rectangle (0.6,0.3); + % Add wheels aligned with slope + \fill[black] (-0.45,-0.3) circle (0.08); + \fill[black] (0.45,-0.3) circle (0.08); + \draw[->,thick] (0,0) -- ++(-0.8, 0) node[left] {$F_{slope} + F_{drag} + F_{rolling}$}; + \draw[->,thick] (0,0) -- ++(0.8, 0) node[right] {$F_{motor}$}; + \node at (0,0) [circle,fill,inner sep=1.5pt]{}; + \end{scope} + + % Draw forces from center of car + % Center point of car for forces + \coordinate (carCenter) at (\carX,\carY); + + +\end{tikzpicture} \centering -\caption{Description of your figure} -\label{fig:example} +\caption{Free body diagram showing relevant forces on a 2-dimensional car} +\label{fig:freebody} \end{figure} -% Example of how to include tables -\begin{table}[H] +To model the vehicle dynamics, we simplify the system to a 2d plane. As seen in Figure~\ref{fig:freebody}, the forces on the car +are due to intrinsic vehicle properties, current velocity, and environment conditions like slope and wind. If the velocity is held +constant, we can assert that the sum of the forces on the car is zero: +\begin{align} + F_{drag} + F_{slope} + F_{rolling} + F_{motor} &= 0 \\ + F_{drag} &= \frac{1}{2} \rho v^2 C_dA \\ + F_{slope} &= mg\sin {\theta} \\ + F_{rolling} &= mg\cos {\theta} C_{rr} +\end{align} +The $F_{motor}$ term is modulated by the driver. In our case, we give the agent a simpler control mechanism with a normalized +velocity control instead of a force-based control. This is written as $v = \alpha v_{max}$ where $\alpha$ is the action taken +by the agent in $\left[-1,1\right]$. +From the velocity, and the forces acting on the car, we can determine +the power of the car using a simple $K_t$ model: +\begin{align} + \tau &= \left(F_{drag} + F_{slope} + F_{rolling}\right) r \\ + P_{motor} &= \tau v + R_{motor} \left(\frac{\tau}{K_t}\right)^2 +\end{align} +The torque of motor is the sum of the outstanding forces times the wheel radius. $K_t$ is a motor parameter, as is $R_{motor}$. +Both can be extracted from physical motors to simulate them, but simple "rule-of-thumb" numbers were used during development. +The power of the motor is given in watts. Based on the time-step of the simulation, we can determine the energy consumed in joules +with $W \times s = J$. A time-step of 1 second was chosen to accelerate simulation. Lower values result in reduced integration +errors over time at the cost of longer episodes. + + +\subsection{Environment Generation} +It is important that our agent learns not just the optimal policy for a fixed course, but an approximate optimal policy +for any course. To this end we must be able to generate a wide variety of terrain and wind scenarios. Perlin noise +is typically used in this context. We use a 1D Perlin noise to generate the slope of the terrain, and then integrate the slope to create +the elevation profile. Currently the elevation profile is unused, but it can be important for drag force due to changes in air pressure. +This was done because differentiated Perlin noise is not smooth, and is not an accurate representation of slope. The wind was +generated with a 2D Perlin noise, where one axis was time, and the other was position. The noise was blurred in the time-axis +to ease the changes in wind at any given point. +An example of the environment can be seen in Figure~\ref{fig:env_vis}. + + +\begin{figure}[H] \centering -\caption{Your Table Caption} -\begin{tabular}{lcc} -\toprule -Method & Metric 1 & Metric 2 \\ -\midrule -Approach 1 & Value & Value \\ -Approach 2 & Value & Value \\ -\bottomrule -\end{tabular} -\label{tab:results} -\end{table} +\includegraphics[width=\textwidth]{environment.pdf} +\caption{Visualization of the generated environment} +\label{fig:env_vis} +\end{figure} + +\subsection{Performance Evaluation} + +To quantify agent performance, we must produce a single value reward. While multi-objective learning is an interesting +subject\footnote{I especially wanted to do meta-rl using a neural net to compute a single reward from inputs} it is out of the scope +of this project. Additionally, sparse rewards can significantly slow down learning. A poor reward function can prevent +agents from approaching optimal policy. With these factors in mind, we use the following: +\[ + R = x/D_{goal} + (x > D_{goal}) * \left(100 - E - 10(t - T_{goal})\right) + (t > T_{goal}) * -500 +\] +To understand this, there are three major components: the continuous reward, which is rewarded at every step, and is the position of the car +relative to the goal distance. The victory reward is a constant, minus the energy used and the early arrival penalty. +This was added to help guide the agent towards arriving with as little time left as possible. Finally, there's a penalty for the time +going above the goal time, as after that point the car is disqualified from the race. + +It took a few iterations to find a reward metric that promoted fast learning. Some of these issues were exacerbated by the initially low +performance when using stable baselines. A crucial part of the improvement was the energy loss only being applied during wins. +This allowed the model to quickly learn to go forward to finish, after which refinement of speed could take +place\footnote{I looked into Q-initialization but couldn't figure out a way to implement it easily.}. + + +\subsection{State and Observation Spaces} + +The complete state of the simulator is the position, velocity, and energy of the car, as well as the entire environment. +These parameters are sufficient for a deterministic snapshot of the simulator. However, one goal of the project +was to enable partial-observation of the system. To this end, we separate the observation space into a small snippet +of the upcoming wind and slope. This also simplifies the agent calculations since the view of the environment is +relative to its current position. The size of the view can be controlled as a parameter. + + +\section{Experiments and Results} + +An implementation of the aforementioned simulator was developed with Jax. Jax was chosen as it enables +vectorization and optimization to improve performance. Additionally, Jax allows for gradients of any function +to be computed, which is useful for certain classes of reinforcement learning. In our case, we didn't +use this as there seemed to be very little available off the shelf. + +Initially Stable Baselines was used since it is one of the most popular implemntations of common RL algorithms. +Stable Baselines3\cite{stable-baselines3} is written in PyTorch\cite{Ansel_PyTorch_2_Faster_2024}, and uses the Gym\cite{gymnasium} format for environments. A basic Gym wrapper +was created to connect SB3 to our environment. +PPO was chosen as the RL algorithm as it is very simple, while still being effective \cite{proximalpolicyoptimization} +The performance and convergence was very bad. This made +it difficult to diagnose as the model would need potentially millions of steps before it would learn anything interesting. +The primary performance loss was in the Jax/Numpy/PyTorch conversion, as this requires a CPU roundtrip. +To combat this I found a Jax-based implementation of PPO called \texttt{purejaxrl}. This library is +written in the style of CleanRL but instead uses pure Jax and an environment library called \texttt{gymnax}\cite{gymnax2022github}. +The primary advantage of writing everything in Jax is that both the RL agent and the environment can be offloaded to the GPU. +Additionally, the infrastructure provided by \texttt{gymnax} allows for environments to be vectorized. The speedup from +using this library cannot be understated. The SB3 PPO implementation ran at around 150 actions per second. After rewriting +some of the code to make it work with \texttt{purejaxrl}, the effective action rate\footnote{I ran 2048 environments in parallel} +was nearly$238000$ actions per second\footnote{It's likely that performance with SB3 could have been improved, but I was struggling to figure out exactly how.}. + + +The episode returns after 50 million timesteps with a PPO model can be seen in Figure~\ref{fig:returns}. Each update step +is performed after collecting minibatches of rollouts based on the current policy. We can see a clean ascent at the start of training, +this is the agent learning to drive forward. After a certain point, the returns become noisy. This is likely due to energy scoring +being random based on the terrain. A solution to this, which wasn't pursued due to lack of time, would be to compute the +"nominal energy" use based on travelling at $v_{avg}$. Energy consumption that was above the nominal use would be penalized, and +below would be heavily rewarded. Despite this, performance continued to improve, which is a good sign for the agent being +able to learn the underlying dynamics. + +\begin{figure}[H] +\centering +\includegraphics[width=0.8\textwidth]{PPO_results.pdf} +\caption{Episodic Returns during PPO training} +\label{fig:returns} +\end{figure} + + +Initially I thought that this was actually pretty impressive, but I looked at an individual level +and it seemed to just drive forward too fast. Reworking the reward function might cause this to converge better. +I wish I had a graph, but I keep running out of memory when I try to capture a rollout. + \section{Discussion} -Analyze your results: -\begin{itemize} - \item Interpretation of findings - \item Limitations and challenges - \item Potential improvements -\end{itemize} + +While the PPO performance was decent, it still had a significant amount of improvement on the table. Tuning the reward function +would probably help it find a solution better. One strategy that would help significantly is to pre-tune the model to output the +average speed by default, so the model doesn't have to learn that at the beginning. This is called Q-initialization and is a common +trick for problem spaces where an initial estimate exists and is easy to define. Perhaps the most important takeaway from this +work is the power of end-to-end Jax RL. \texttt{purejaxrl} is CleanRL levels of code clarity, with everything for an agent +being contained in one file, but surpassing Stable Baselines3 significantly in terms of performance. One drawback is that +the ecosystem is very new, so there was very little to reference when I was developing my simulator. Often there would be +an opaque error message that would yield no results on search engines, and would require digging into the Jax source code to diagnose. +Typically this was some misunderstanding about the inner works of Jax. Future work on this project would involve trying out +other agents, and comparing different reward functions. Adjusting the actor-critic network would also be an interesting avenue, +especially since a CNN will likely work well with wind and cloud information, which have both a spatial and temporal +axis\footnote{You can probably tell that the quality dropped off near the end - bunch of life things got in the way, so this didn't go as well as I'd hoped. Learned a lot though.}. + + \section{Conclusion} -Summarize your work: -\begin{itemize} - \item Key contributions - \item Practical implications - \item Future work directions -\end{itemize} + +We outline the design of a physics based model of solar car races. We implement this model and create a simulation environment +for use with popular RL algorithm packages. We demonstrate the performance and learning ability of these algorithms on our model. +Further work includes more accurate modelling, improved reward functions, and hyperparameter tuning. \bibliography{references} \bibliographystyle{plain} diff --git a/src/solarcarsim/cleanrl_td3_jax.py b/src/solarcarsim/cleanrl_td3_jax.py new file mode 100644 index 0000000..82ed58d --- /dev/null +++ b/src/solarcarsim/cleanrl_td3_jax.py @@ -0,0 +1,368 @@ +# docs and experiment results can be found at https://docs.cleanrl.dev/rl-algorithms/td3/#td3_continuous_action_jaxpy +import os +import random +import time +from dataclasses import dataclass + +import flax +import flax.linen as nn +import gymnasium as gym +import jax +import jax.numpy as jnp +import numpy as np +import optax +import tyro +from flax.training.train_state import TrainState +from stable_baselines3.common.buffers import ReplayBuffer +from torch.utils.tensorboard.writer import SummaryWriter + + +@dataclass +class Args: + exp_name: str = os.path.basename(__file__)[: -len(".py")] + """the name of this experiment""" + seed: int = 1 + """seed of the experiment""" + track: bool = False + """if toggled, this experiment will be tracked with Weights and Biases""" + wandb_project_name: str = "cleanRL" + """the wandb's project name""" + wandb_entity: str = None + """the entity (team) of wandb's project""" + capture_video: bool = False + """whether to capture videos of the agent performances (check out `videos` folder)""" + save_model: bool = False + """whether to save model into the `runs/{run_name}` folder""" + upload_model: bool = False + """whether to upload the saved model to huggingface""" + hf_entity: str = "" + """the user or org name of the model repository from the Hugging Face Hub""" + + # Algorithm specific arguments + env_id: str = "MountainCarContinuous-v0" + """the id of the environment""" + total_timesteps: int = 1000000 + """total timesteps of the experiments""" + learning_rate: float = 3e-4 + """the learning rate of the optimizer""" + buffer_size: int = int(1e6) + """the replay memory buffer size""" + gamma: float = 0.99 + """the discount factor gamma""" + tau: float = 0.005 + """target smoothing coefficient (default: 0.005)""" + batch_size: int = 256 + """the batch size of sample from the reply memory""" + policy_noise: float = 0.2 + """the scale of policy noise""" + exploration_noise: float = 0.1 + """the scale of exploration noise""" + learning_starts: int = 25e3 + """timestep to start learning""" + policy_frequency: int = 2 + """the frequency of training policy (delayed)""" + noise_clip: float = 0.5 + """noise clip parameter of the Target Policy Smoothing Regularization""" + + +def make_env(env_id, seed, idx, capture_video, run_name): + def thunk(): + if capture_video and idx == 0: + env = gym.make(env_id, render_mode="rgb_array") + env = gym.wrappers.RecordVideo(env, f"videos/{run_name}") + else: + env = gym.make(env_id) + env = gym.wrappers.RecordEpisodeStatistics(env) + env.action_space.seed(seed) + return env + + return thunk + + +# ALGO LOGIC: initialize agent here: +class QNetwork(nn.Module): + @nn.compact + def __call__(self, x: jnp.ndarray, a: jnp.ndarray): + x = jnp.concatenate([x, a], -1) + x = nn.Dense(256)(x) + x = nn.relu(x) + x = nn.Dense(256)(x) + x = nn.relu(x) + x = nn.Dense(1)(x) + return x + + +class Actor(nn.Module): + action_dim: int + action_scale: jnp.ndarray + action_bias: jnp.ndarray + + @nn.compact + def __call__(self, x): + x = nn.Dense(256)(x) + x = nn.relu(x) + x = nn.Dense(256)(x) + x = nn.relu(x) + x = nn.Dense(self.action_dim)(x) + x = nn.tanh(x) + x = x * self.action_scale + self.action_bias + return x + + +class TrainState(TrainState): + target_params: flax.core.FrozenDict + + +if __name__ == "__main__": + import stable_baselines3 as sb3 + + if sb3.__version__ < "2.0": + raise ValueError( + """Ongoing migration: run the following command to install the new dependencies: +poetry run pip install "stable_baselines3==2.0.0a1" +""" + ) + args = tyro.cli(Args) + run_name = f"{args.env_id}__{args.exp_name}__{args.seed}__{int(time.time())}" + if args.track: + import wandb + + wandb.init( + project=args.wandb_project_name, + entity=args.wandb_entity, + sync_tensorboard=True, + config=vars(args), + name=run_name, + monitor_gym=True, + save_code=True, + ) + writer = SummaryWriter(f"runs/{run_name}") + writer.add_text( + "hyperparameters", + "|param|value|\n|-|-|\n%s" % ("\n".join([f"|{key}|{value}|" for key, value in vars(args).items()])), + ) + + # TRY NOT TO MODIFY: seeding + random.seed(args.seed) + np.random.seed(args.seed) + key = jax.random.PRNGKey(args.seed) + key, actor_key, qf1_key, qf2_key = jax.random.split(key, 4) + + # env setup + envs = gym.vector.SyncVectorEnv([make_env(args.env_id, args.seed, 0, args.capture_video, run_name)]) + assert isinstance(envs.single_action_space, gym.spaces.Box), "only continuous action space is supported" + + max_action = float(envs.single_action_space.high[0]) + envs.single_observation_space.dtype = np.float32 + rb = ReplayBuffer( + args.buffer_size, + envs.single_observation_space, + envs.single_action_space, + device="cpu", + handle_timeout_termination=False, + ) + + # TRY NOT TO MODIFY: start the game + obs, _ = envs.reset(seed=args.seed) + + actor = Actor( + action_dim=np.prod(envs.single_action_space.shape), + action_scale=jnp.array((envs.action_space.high - envs.action_space.low) / 2.0), + action_bias=jnp.array((envs.action_space.high + envs.action_space.low) / 2.0), + ) + actor_state = TrainState.create( + apply_fn=actor.apply, + params=actor.init(actor_key, obs), + target_params=actor.init(actor_key, obs), + tx=optax.adam(learning_rate=args.learning_rate), + ) + qf = QNetwork() + qf1_state = TrainState.create( + apply_fn=qf.apply, + params=qf.init(qf1_key, obs, envs.action_space.sample()), + target_params=qf.init(qf1_key, obs, envs.action_space.sample()), + tx=optax.adam(learning_rate=args.learning_rate), + ) + qf2_state = TrainState.create( + apply_fn=qf.apply, + params=qf.init(qf2_key, obs, envs.action_space.sample()), + target_params=qf.init(qf2_key, obs, envs.action_space.sample()), + tx=optax.adam(learning_rate=args.learning_rate), + ) + actor.apply = jax.jit(actor.apply) + qf.apply = jax.jit(qf.apply) + + @jax.jit + def update_critic( + actor_state: TrainState, + qf1_state: TrainState, + qf2_state: TrainState, + observations: np.ndarray, + actions: np.ndarray, + next_observations: np.ndarray, + rewards: np.ndarray, + terminations: np.ndarray, + key: jnp.ndarray, + ): + # TODO Maybe pre-generate a lot of random keys + # also check https://jax.readthedocs.io/en/latest/jax.random.html + key, noise_key = jax.random.split(key, 2) + clipped_noise = ( + jnp.clip( + (jax.random.normal(noise_key, actions.shape) * args.policy_noise), + -args.noise_clip, + args.noise_clip, + ) + * actor.action_scale + ) + next_state_actions = jnp.clip( + actor.apply(actor_state.target_params, next_observations) + clipped_noise, + envs.single_action_space.low, + envs.single_action_space.high, + ) + qf1_next_target = qf.apply(qf1_state.target_params, next_observations, next_state_actions).reshape(-1) + qf2_next_target = qf.apply(qf2_state.target_params, next_observations, next_state_actions).reshape(-1) + min_qf_next_target = jnp.minimum(qf1_next_target, qf2_next_target) + next_q_value = (rewards + (1 - terminations) * args.gamma * (min_qf_next_target)).reshape(-1) + + def mse_loss(params): + qf_a_values = qf.apply(params, observations, actions).squeeze() + return ((qf_a_values - next_q_value) ** 2).mean(), qf_a_values.mean() + + (qf1_loss_value, qf1_a_values), grads1 = jax.value_and_grad(mse_loss, has_aux=True)(qf1_state.params) + (qf2_loss_value, qf2_a_values), grads2 = jax.value_and_grad(mse_loss, has_aux=True)(qf2_state.params) + qf1_state = qf1_state.apply_gradients(grads=grads1) + qf2_state = qf2_state.apply_gradients(grads=grads2) + + return (qf1_state, qf2_state), (qf1_loss_value, qf2_loss_value), (qf1_a_values, qf2_a_values), key + + @jax.jit + def update_actor( + actor_state: TrainState, + qf1_state: TrainState, + qf2_state: TrainState, + observations: np.ndarray, + ): + def actor_loss(params): + return -qf.apply(qf1_state.params, observations, actor.apply(params, observations)).mean() + + actor_loss_value, grads = jax.value_and_grad(actor_loss)(actor_state.params) + actor_state = actor_state.apply_gradients(grads=grads) + actor_state = actor_state.replace( + target_params=optax.incremental_update(actor_state.params, actor_state.target_params, args.tau) + ) + + qf1_state = qf1_state.replace( + target_params=optax.incremental_update(qf1_state.params, qf1_state.target_params, args.tau) + ) + qf2_state = qf2_state.replace( + target_params=optax.incremental_update(qf2_state.params, qf2_state.target_params, args.tau) + ) + return actor_state, (qf1_state, qf2_state), actor_loss_value + + start_time = time.time() + for global_step in range(args.total_timesteps): + # ALGO LOGIC: put action logic here + if global_step < args.learning_starts: + actions = np.array([envs.single_action_space.sample() for _ in range(envs.num_envs)]) + else: + actions = actor.apply(actor_state.params, obs) + actions = np.array( + [ + ( + jax.device_get(actions)[0] + + np.random.normal(0, max_action * args.exploration_noise, size=envs.single_action_space.shape) + ).clip(envs.single_action_space.low, envs.single_action_space.high) + ] + ) + + # TRY NOT TO MODIFY: execute the game and log data. + next_obs, rewards, terminations, truncations, infos = envs.step(actions) + + # TRY NOT TO MODIFY: record rewards for plotting purposes + if "final_info" in infos: + for info in infos["final_info"]: + print(f"global_step={global_step}, episodic_return={info['episode']['r']}") + writer.add_scalar("charts/episodic_return", info["episode"]["r"], global_step) + writer.add_scalar("charts/episodic_length", info["episode"]["l"], global_step) + break + + # TRY NOT TO MODIFY: save data to replay buffer; handle `final_observation` + real_next_obs = next_obs.copy() + for idx, trunc in enumerate(truncations): + if trunc: + real_next_obs[idx] = infos["final_observation"][idx] + rb.add(obs, real_next_obs, actions, rewards, terminations, infos) + + # TRY NOT TO MODIFY: CRUCIAL step easy to overlook + obs = next_obs + + # ALGO LOGIC: training. + if global_step > args.learning_starts: + data = rb.sample(args.batch_size) + + (qf1_state, qf2_state), (qf1_loss_value, qf2_loss_value), (qf1_a_values, qf2_a_values), key = update_critic( + actor_state, + qf1_state, + qf2_state, + data.observations.numpy(), + data.actions.numpy(), + data.next_observations.numpy(), + data.rewards.flatten().numpy(), + data.dones.flatten().numpy(), + key, + ) + + if global_step % args.policy_frequency == 0: + actor_state, (qf1_state, qf2_state), actor_loss_value = update_actor( + actor_state, + qf1_state, + qf2_state, + data.observations.numpy(), + ) + + if global_step % 100 == 0: + writer.add_scalar("losses/qf1_loss", qf1_loss_value.item(), global_step) + writer.add_scalar("losses/qf2_loss", qf2_loss_value.item(), global_step) + writer.add_scalar("losses/qf1_values", qf1_a_values.item(), global_step) + writer.add_scalar("losses/qf2_values", qf2_a_values.item(), global_step) + writer.add_scalar("losses/actor_loss", actor_loss_value.item(), global_step) + print("SPS:", int(global_step / (time.time() - start_time))) + writer.add_scalar("charts/SPS", int(global_step / (time.time() - start_time)), global_step) + + if args.save_model: + model_path = f"runs/{run_name}/{args.exp_name}.cleanrl_model" + with open(model_path, "wb") as f: + f.write( + flax.serialization.to_bytes( + [ + actor_state.params, + qf1_state.params, + qf2_state.params, + ] + ) + ) + print(f"model saved to {model_path}") + from cleanrl_utils.evals.td3_jax_eval import evaluate + + episodic_returns = evaluate( + model_path, + make_env, + args.env_id, + eval_episodes=10, + run_name=f"{run_name}-eval", + Model=(Actor, QNetwork), + exploration_noise=args.exploration_noise, + ) + for idx, episodic_return in enumerate(episodic_returns): + writer.add_scalar("eval/episodic_return", episodic_return, idx) + + if args.upload_model: + from cleanrl_utils.huggingface import push_to_hub + + repo_name = f"{args.env_id}-{args.exp_name}-seed{args.seed}" + repo_id = f"{args.hf_entity}/{repo_name}" if args.hf_entity else repo_name + push_to_hub(args, episodic_returns, repo_id, "TD3", f"runs/{run_name}", f"videos/{run_name}-eval") + + envs.close() + writer.close() \ No newline at end of file diff --git a/src/solarcarsim/physsim.py b/src/solarcarsim/physsim.py index da14ba1..c9a2da8 100644 --- a/src/solarcarsim/physsim.py +++ b/src/solarcarsim/physsim.py @@ -66,12 +66,12 @@ def downslope_force(mass, theta): return mass * 9.8 * jnp.sin(theta) -@partial(jit, static_argnames=["crr"]) +@jit def rolling_force(mass, theta, crr): return normal_force(mass, theta) * crr -@partial(jit, static_argnames=["area", "cd", "rho"]) +@jit def drag_force(u, area, cd, rho): return 0.5 * rho * jnp.pow(u, 2) * cd * area diff --git a/src/solarcarsim/simv1.py b/src/solarcarsim/simv1.py index 47fe60c..5444b94 100644 --- a/src/solarcarsim/simv1.py +++ b/src/solarcarsim/simv1.py @@ -93,7 +93,7 @@ class SolarRaceV1(gym.Env): } ) - self.action_space = gym.spaces.Box(-1.0, 1.0, shape=(1,)) # velocity, m/s + self.action_space = gym.spaces.Box(0.0, 1.0, shape=(1,)) # velocity, m/s def reset(self, *, seed = None, options = None): @@ -117,7 +117,7 @@ class SolarRaceV1(gym.Env): reward -= 600 - self._state[1][0] reward += 1e-6 * (self._state[2][0]) # net energy is negative. if jnp.all(self._state[1] > 600): - reward -= 50000 + reward -= 500 truncated = True return self._get_obs(), reward, terminated, truncated, {} diff --git a/src/solarcarsim/simv2.py b/src/solarcarsim/simv2.py index fffe8c0..a79320b 100644 --- a/src/solarcarsim/simv2.py +++ b/src/solarcarsim/simv2.py @@ -5,7 +5,7 @@ import jax import jax.numpy as jnp import chex from flax import struct -from jax import lax +from jax import lax, vmap from gymnax.environments import environment from gymnax.environments import spaces @@ -59,19 +59,6 @@ class Snax(environment.Environment[SimState, SimParams]): dtype=jnp.float32, ) return spaces.Box(low, high, shape=(shape,)) - # return spaces.Dict( - # { - # "position": spaces.Box(0.0, params.map_size, (), jnp.float32), - # "realtime": spaces.Box(0.0, params.goal_time + 100, (), jnp.float32), - # "energy": spaces.Box(-1e11, 0.0, (), jnp.float32), - # "dist_to_goal": spaces.Box(0.0, params.goal_dist, (), jnp.float32), - # "time_remaining": spaces.Box(0.0, params.goal_time, (), jnp.float32), - # "upcoming_terrain": spaces.Box( - # -1.0, 1.0, shape=(100,), dtype=jnp.float32 - # ), - # # skip wind for now - # } - # ) def state_space(self, params: Optional[SimParams] = None) -> spaces.Dict: if params is None: @@ -170,9 +157,15 @@ class Snax(environment.Environment[SimState, SimParams]): # ): # reward -= 500 - # we have to vectorize that. + # # we have to vectorize that. + # reward = new_state.position / params.goal_dist # constant reward for moving forward + # # reward for finishing + # reward += (new_state.position >= params.goal_dist) * (100 + params.goal_time - new_state.realtime + 1e-7*new_state.energy) + # # reward for failure + # reward += (new_state.realtime >= params.goal_time) * -500 + reward = new_state.position / params.goal_dist + \ - (new_state.position >= params.goal_dist) * (100 + params.goal_time - new_state.realtime + 1e-7*new_state.energy) + \ + (new_state.position >= params.goal_dist) * (100 + params.goal_time - new_state.realtime + 1e-6*new_state.energy) + \ (new_state.realtime >= params.goal_time) * -500 reward = reward.squeeze() terminal = self.is_terminal(state, params)