mirror of
https://github.com/gryf/coach.git
synced 2025-12-17 19:20:19 +01:00
429 lines
42 KiB
HTML
429 lines
42 KiB
HTML
|
|
|
|
<!DOCTYPE html>
|
|
<!--[if IE 8]><html class="no-js lt-ie9" lang="en" > <![endif]-->
|
|
<!--[if gt IE 8]><!--> <html class="no-js" lang="en" > <!--<![endif]-->
|
|
<head>
|
|
<meta charset="utf-8">
|
|
|
|
<meta name="viewport" content="width=device-width, initial-scale=1.0">
|
|
|
|
<title>rl_coach.agents.value_optimization_agent — Reinforcement Learning Coach 0.12.0 documentation</title>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
<script type="text/javascript" src="../../../_static/js/modernizr.min.js"></script>
|
|
|
|
|
|
<script type="text/javascript" id="documentation_options" data-url_root="../../../" src="../../../_static/documentation_options.js"></script>
|
|
<script type="text/javascript" src="../../../_static/jquery.js"></script>
|
|
<script type="text/javascript" src="../../../_static/underscore.js"></script>
|
|
<script type="text/javascript" src="../../../_static/doctools.js"></script>
|
|
<script type="text/javascript" src="../../../_static/language_data.js"></script>
|
|
<script async="async" type="text/javascript" src="https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.5/latest.js?config=TeX-AMS-MML_HTMLorMML"></script>
|
|
|
|
<script type="text/javascript" src="../../../_static/js/theme.js"></script>
|
|
|
|
|
|
|
|
|
|
<link rel="stylesheet" href="../../../_static/css/theme.css" type="text/css" />
|
|
<link rel="stylesheet" href="../../../_static/pygments.css" type="text/css" />
|
|
<link rel="stylesheet" href="../../../_static/css/custom.css" type="text/css" />
|
|
<link rel="index" title="Index" href="../../../genindex.html" />
|
|
<link rel="search" title="Search" href="../../../search.html" />
|
|
<link href="../../../_static/css/custom.css" rel="stylesheet" type="text/css">
|
|
|
|
</head>
|
|
|
|
<body class="wy-body-for-nav">
|
|
|
|
|
|
<div class="wy-grid-for-nav">
|
|
|
|
<nav data-toggle="wy-nav-shift" class="wy-nav-side">
|
|
<div class="wy-side-scroll">
|
|
<div class="wy-side-nav-search" >
|
|
|
|
|
|
|
|
<a href="../../../index.html" class="icon icon-home"> Reinforcement Learning Coach
|
|
|
|
|
|
|
|
|
|
<img src="../../../_static/dark_logo.png" class="logo" alt="Logo"/>
|
|
|
|
</a>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
<div role="search">
|
|
<form id="rtd-search-form" class="wy-form" action="../../../search.html" method="get">
|
|
<input type="text" name="q" placeholder="Search docs" />
|
|
<input type="hidden" name="check_keywords" value="yes" />
|
|
<input type="hidden" name="area" value="default" />
|
|
</form>
|
|
</div>
|
|
|
|
|
|
</div>
|
|
|
|
<div class="wy-menu wy-menu-vertical" data-spy="affix" role="navigation" aria-label="main navigation">
|
|
|
|
|
|
|
|
|
|
|
|
|
|
<p class="caption"><span class="caption-text">Intro</span></p>
|
|
<ul>
|
|
<li class="toctree-l1"><a class="reference internal" href="../../../usage.html">Usage</a></li>
|
|
<li class="toctree-l1"><a class="reference internal" href="../../../dist_usage.html">Usage - Distributed Coach</a></li>
|
|
<li class="toctree-l1"><a class="reference internal" href="../../../features/index.html">Features</a></li>
|
|
<li class="toctree-l1"><a class="reference internal" href="../../../selecting_an_algorithm.html">Selecting an Algorithm</a></li>
|
|
<li class="toctree-l1"><a class="reference internal" href="../../../dashboard.html">Coach Dashboard</a></li>
|
|
</ul>
|
|
<p class="caption"><span class="caption-text">Design</span></p>
|
|
<ul>
|
|
<li class="toctree-l1"><a class="reference internal" href="../../../design/control_flow.html">Control Flow</a></li>
|
|
<li class="toctree-l1"><a class="reference internal" href="../../../design/network.html">Network Design</a></li>
|
|
<li class="toctree-l1"><a class="reference internal" href="../../../design/horizontal_scaling.html">Distributed Coach - Horizontal Scale-Out</a></li>
|
|
</ul>
|
|
<p class="caption"><span class="caption-text">Contributing</span></p>
|
|
<ul>
|
|
<li class="toctree-l1"><a class="reference internal" href="../../../contributing/add_agent.html">Adding a New Agent</a></li>
|
|
<li class="toctree-l1"><a class="reference internal" href="../../../contributing/add_env.html">Adding a New Environment</a></li>
|
|
</ul>
|
|
<p class="caption"><span class="caption-text">Components</span></p>
|
|
<ul>
|
|
<li class="toctree-l1"><a class="reference internal" href="../../../components/agents/index.html">Agents</a></li>
|
|
<li class="toctree-l1"><a class="reference internal" href="../../../components/architectures/index.html">Architectures</a></li>
|
|
<li class="toctree-l1"><a class="reference internal" href="../../../components/data_stores/index.html">Data Stores</a></li>
|
|
<li class="toctree-l1"><a class="reference internal" href="../../../components/environments/index.html">Environments</a></li>
|
|
<li class="toctree-l1"><a class="reference internal" href="../../../components/exploration_policies/index.html">Exploration Policies</a></li>
|
|
<li class="toctree-l1"><a class="reference internal" href="../../../components/filters/index.html">Filters</a></li>
|
|
<li class="toctree-l1"><a class="reference internal" href="../../../components/memories/index.html">Memories</a></li>
|
|
<li class="toctree-l1"><a class="reference internal" href="../../../components/memory_backends/index.html">Memory Backends</a></li>
|
|
<li class="toctree-l1"><a class="reference internal" href="../../../components/orchestrators/index.html">Orchestrators</a></li>
|
|
<li class="toctree-l1"><a class="reference internal" href="../../../components/core_types.html">Core Types</a></li>
|
|
<li class="toctree-l1"><a class="reference internal" href="../../../components/spaces.html">Spaces</a></li>
|
|
<li class="toctree-l1"><a class="reference internal" href="../../../components/additional_parameters.html">Additional Parameters</a></li>
|
|
</ul>
|
|
|
|
|
|
|
|
</div>
|
|
</div>
|
|
</nav>
|
|
|
|
<section data-toggle="wy-nav-shift" class="wy-nav-content-wrap">
|
|
|
|
|
|
<nav class="wy-nav-top" aria-label="top navigation">
|
|
|
|
<i data-toggle="wy-nav-top" class="fa fa-bars"></i>
|
|
<a href="../../../index.html">Reinforcement Learning Coach</a>
|
|
|
|
</nav>
|
|
|
|
|
|
<div class="wy-nav-content">
|
|
|
|
<div class="rst-content">
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
<div role="navigation" aria-label="breadcrumbs navigation">
|
|
|
|
<ul class="wy-breadcrumbs">
|
|
|
|
<li><a href="../../../index.html">Docs</a> »</li>
|
|
|
|
<li><a href="../../index.html">Module code</a> »</li>
|
|
|
|
<li>rl_coach.agents.value_optimization_agent</li>
|
|
|
|
|
|
<li class="wy-breadcrumbs-aside">
|
|
|
|
</li>
|
|
|
|
</ul>
|
|
|
|
|
|
<hr/>
|
|
</div>
|
|
<div role="main" class="document" itemscope="itemscope" itemtype="http://schema.org/Article">
|
|
<div itemprop="articleBody">
|
|
|
|
<h1>Source code for rl_coach.agents.value_optimization_agent</h1><div class="highlight"><pre>
|
|
<span></span><span class="c1">#</span>
|
|
<span class="c1"># Copyright (c) 2017 Intel Corporation</span>
|
|
<span class="c1">#</span>
|
|
<span class="c1"># Licensed under the Apache License, Version 2.0 (the "License");</span>
|
|
<span class="c1"># you may not use this file except in compliance with the License.</span>
|
|
<span class="c1"># You may obtain a copy of the License at</span>
|
|
<span class="c1">#</span>
|
|
<span class="c1"># http://www.apache.org/licenses/LICENSE-2.0</span>
|
|
<span class="c1">#</span>
|
|
<span class="c1"># Unless required by applicable law or agreed to in writing, software</span>
|
|
<span class="c1"># distributed under the License is distributed on an "AS IS" BASIS,</span>
|
|
<span class="c1"># WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.</span>
|
|
<span class="c1"># See the License for the specific language governing permissions and</span>
|
|
<span class="c1"># limitations under the License.</span>
|
|
<span class="c1">#</span>
|
|
<span class="kn">from</span> <span class="nn">collections</span> <span class="k">import</span> <span class="n">OrderedDict</span>
|
|
<span class="kn">from</span> <span class="nn">typing</span> <span class="k">import</span> <span class="n">Union</span><span class="p">,</span> <span class="n">List</span>
|
|
|
|
<span class="kn">import</span> <span class="nn">numpy</span> <span class="k">as</span> <span class="nn">np</span>
|
|
|
|
<span class="kn">from</span> <span class="nn">rl_coach.agents.agent</span> <span class="k">import</span> <span class="n">Agent</span>
|
|
<span class="kn">from</span> <span class="nn">rl_coach.core_types</span> <span class="k">import</span> <span class="n">ActionInfo</span><span class="p">,</span> <span class="n">StateType</span><span class="p">,</span> <span class="n">Batch</span>
|
|
<span class="kn">from</span> <span class="nn">rl_coach.filters.filter</span> <span class="k">import</span> <span class="n">NoInputFilter</span>
|
|
<span class="kn">from</span> <span class="nn">rl_coach.logger</span> <span class="k">import</span> <span class="n">screen</span>
|
|
<span class="kn">from</span> <span class="nn">rl_coach.memories.non_episodic.prioritized_experience_replay</span> <span class="k">import</span> <span class="n">PrioritizedExperienceReplay</span>
|
|
<span class="kn">from</span> <span class="nn">rl_coach.spaces</span> <span class="k">import</span> <span class="n">DiscreteActionSpace</span>
|
|
<span class="kn">from</span> <span class="nn">copy</span> <span class="k">import</span> <span class="n">deepcopy</span><span class="p">,</span> <span class="n">copy</span>
|
|
|
|
|
|
<span class="c1">## This is an abstract agent - there is no learn_from_batch method ##</span>
|
|
|
|
|
|
<span class="k">class</span> <span class="nc">ValueOptimizationAgent</span><span class="p">(</span><span class="n">Agent</span><span class="p">):</span>
|
|
<span class="k">def</span> <span class="nf">__init__</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">agent_parameters</span><span class="p">,</span> <span class="n">parent</span><span class="p">:</span> <span class="n">Union</span><span class="p">[</span><span class="s1">'LevelManager'</span><span class="p">,</span> <span class="s1">'CompositeAgent'</span><span class="p">]</span><span class="o">=</span><span class="kc">None</span><span class="p">):</span>
|
|
<span class="nb">super</span><span class="p">()</span><span class="o">.</span><span class="fm">__init__</span><span class="p">(</span><span class="n">agent_parameters</span><span class="p">,</span> <span class="n">parent</span><span class="p">)</span>
|
|
<span class="bp">self</span><span class="o">.</span><span class="n">q_values</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">register_signal</span><span class="p">(</span><span class="s2">"Q"</span><span class="p">)</span>
|
|
<span class="bp">self</span><span class="o">.</span><span class="n">q_value_for_action</span> <span class="o">=</span> <span class="p">{}</span>
|
|
|
|
<span class="c1"># currently we use softmax action probabilities only in batch-rl,</span>
|
|
<span class="c1"># but we might want to extend this later at some point.</span>
|
|
<span class="bp">self</span><span class="o">.</span><span class="n">should_get_softmax_probabilities</span> <span class="o">=</span> \
|
|
<span class="nb">hasattr</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">ap</span><span class="o">.</span><span class="n">network_wrappers</span><span class="p">[</span><span class="s1">'main'</span><span class="p">],</span> <span class="s1">'should_get_softmax_probabilities'</span><span class="p">)</span> <span class="ow">and</span> \
|
|
<span class="bp">self</span><span class="o">.</span><span class="n">ap</span><span class="o">.</span><span class="n">network_wrappers</span><span class="p">[</span><span class="s1">'main'</span><span class="p">]</span><span class="o">.</span><span class="n">should_get_softmax_probabilities</span>
|
|
|
|
<span class="k">def</span> <span class="nf">init_environment_dependent_modules</span><span class="p">(</span><span class="bp">self</span><span class="p">):</span>
|
|
<span class="nb">super</span><span class="p">()</span><span class="o">.</span><span class="n">init_environment_dependent_modules</span><span class="p">()</span>
|
|
<span class="k">if</span> <span class="nb">isinstance</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">spaces</span><span class="o">.</span><span class="n">action</span><span class="p">,</span> <span class="n">DiscreteActionSpace</span><span class="p">):</span>
|
|
<span class="k">for</span> <span class="n">i</span> <span class="ow">in</span> <span class="nb">range</span><span class="p">(</span><span class="nb">len</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">spaces</span><span class="o">.</span><span class="n">action</span><span class="o">.</span><span class="n">actions</span><span class="p">)):</span>
|
|
<span class="bp">self</span><span class="o">.</span><span class="n">q_value_for_action</span><span class="p">[</span><span class="n">i</span><span class="p">]</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">register_signal</span><span class="p">(</span><span class="s2">"Q for action </span><span class="si">{}</span><span class="s2">"</span><span class="o">.</span><span class="n">format</span><span class="p">(</span><span class="n">i</span><span class="p">),</span>
|
|
<span class="n">dump_one_value_per_episode</span><span class="o">=</span><span class="kc">False</span><span class="p">,</span>
|
|
<span class="n">dump_one_value_per_step</span><span class="o">=</span><span class="kc">True</span><span class="p">)</span>
|
|
|
|
<span class="c1"># Algorithms for which q_values are calculated from predictions will override this function</span>
|
|
<span class="k">def</span> <span class="nf">get_all_q_values_for_states</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">states</span><span class="p">:</span> <span class="n">StateType</span><span class="p">):</span>
|
|
<span class="n">actions_q_values</span> <span class="o">=</span> <span class="kc">None</span>
|
|
<span class="k">if</span> <span class="bp">self</span><span class="o">.</span><span class="n">exploration_policy</span><span class="o">.</span><span class="n">requires_action_values</span><span class="p">():</span>
|
|
<span class="n">actions_q_values</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">get_prediction</span><span class="p">(</span><span class="n">states</span><span class="p">)</span>
|
|
|
|
<span class="k">return</span> <span class="n">actions_q_values</span>
|
|
|
|
<span class="k">def</span> <span class="nf">get_all_q_values_for_states_and_softmax_probabilities</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">states</span><span class="p">:</span> <span class="n">StateType</span><span class="p">):</span>
|
|
<span class="n">actions_q_values</span><span class="p">,</span> <span class="n">softmax_probabilities</span> <span class="o">=</span> <span class="kc">None</span><span class="p">,</span> <span class="kc">None</span>
|
|
<span class="k">if</span> <span class="bp">self</span><span class="o">.</span><span class="n">exploration_policy</span><span class="o">.</span><span class="n">requires_action_values</span><span class="p">():</span>
|
|
<span class="n">outputs</span> <span class="o">=</span> <span class="n">copy</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">networks</span><span class="p">[</span><span class="s1">'main'</span><span class="p">]</span><span class="o">.</span><span class="n">online_network</span><span class="o">.</span><span class="n">outputs</span><span class="p">)</span>
|
|
<span class="n">outputs</span><span class="o">.</span><span class="n">append</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">networks</span><span class="p">[</span><span class="s1">'main'</span><span class="p">]</span><span class="o">.</span><span class="n">online_network</span><span class="o">.</span><span class="n">output_heads</span><span class="p">[</span><span class="mi">0</span><span class="p">]</span><span class="o">.</span><span class="n">softmax</span><span class="p">)</span>
|
|
|
|
<span class="n">actions_q_values</span><span class="p">,</span> <span class="n">softmax_probabilities</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">get_prediction</span><span class="p">(</span><span class="n">states</span><span class="p">,</span> <span class="n">outputs</span><span class="o">=</span><span class="n">outputs</span><span class="p">)</span>
|
|
<span class="k">return</span> <span class="n">actions_q_values</span><span class="p">,</span> <span class="n">softmax_probabilities</span>
|
|
|
|
<span class="k">def</span> <span class="nf">get_prediction</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">states</span><span class="p">,</span> <span class="n">outputs</span><span class="o">=</span><span class="kc">None</span><span class="p">):</span>
|
|
<span class="k">return</span> <span class="bp">self</span><span class="o">.</span><span class="n">networks</span><span class="p">[</span><span class="s1">'main'</span><span class="p">]</span><span class="o">.</span><span class="n">online_network</span><span class="o">.</span><span class="n">predict</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">prepare_batch_for_inference</span><span class="p">(</span><span class="n">states</span><span class="p">,</span> <span class="s1">'main'</span><span class="p">),</span>
|
|
<span class="n">outputs</span><span class="o">=</span><span class="n">outputs</span><span class="p">)</span>
|
|
|
|
<span class="k">def</span> <span class="nf">update_transition_priorities_and_get_weights</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">TD_errors</span><span class="p">,</span> <span class="n">batch</span><span class="p">):</span>
|
|
<span class="c1"># update errors in prioritized replay buffer</span>
|
|
<span class="n">importance_weights</span> <span class="o">=</span> <span class="kc">None</span>
|
|
<span class="k">if</span> <span class="nb">isinstance</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">memory</span><span class="p">,</span> <span class="n">PrioritizedExperienceReplay</span><span class="p">):</span>
|
|
<span class="bp">self</span><span class="o">.</span><span class="n">call_memory</span><span class="p">(</span><span class="s1">'update_priorities'</span><span class="p">,</span> <span class="p">(</span><span class="n">batch</span><span class="o">.</span><span class="n">info</span><span class="p">(</span><span class="s1">'idx'</span><span class="p">),</span> <span class="n">TD_errors</span><span class="p">))</span>
|
|
<span class="n">importance_weights</span> <span class="o">=</span> <span class="n">batch</span><span class="o">.</span><span class="n">info</span><span class="p">(</span><span class="s1">'weight'</span><span class="p">)</span>
|
|
<span class="k">return</span> <span class="n">importance_weights</span>
|
|
|
|
<span class="k">def</span> <span class="nf">_validate_action</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">policy</span><span class="p">,</span> <span class="n">action</span><span class="p">):</span>
|
|
<span class="k">if</span> <span class="n">np</span><span class="o">.</span><span class="n">array</span><span class="p">(</span><span class="n">action</span><span class="p">)</span><span class="o">.</span><span class="n">shape</span> <span class="o">!=</span> <span class="p">():</span>
|
|
<span class="k">raise</span> <span class="ne">ValueError</span><span class="p">((</span>
|
|
<span class="s1">'The exploration_policy </span><span class="si">{}</span><span class="s1"> returned a vector of actions '</span>
|
|
<span class="s1">'instead of a single action. ValueOptimizationAgents '</span>
|
|
<span class="s1">'require exploration policies which return a single action.'</span>
|
|
<span class="p">)</span><span class="o">.</span><span class="n">format</span><span class="p">(</span><span class="n">policy</span><span class="o">.</span><span class="vm">__class__</span><span class="o">.</span><span class="vm">__name__</span><span class="p">))</span>
|
|
|
|
<span class="k">def</span> <span class="nf">choose_action</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">curr_state</span><span class="p">):</span>
|
|
<span class="k">if</span> <span class="bp">self</span><span class="o">.</span><span class="n">should_get_softmax_probabilities</span><span class="p">:</span>
|
|
<span class="n">actions_q_values</span><span class="p">,</span> <span class="n">softmax_probabilities</span> <span class="o">=</span> \
|
|
<span class="bp">self</span><span class="o">.</span><span class="n">get_all_q_values_for_states_and_softmax_probabilities</span><span class="p">(</span><span class="n">curr_state</span><span class="p">)</span>
|
|
<span class="k">else</span><span class="p">:</span>
|
|
<span class="n">actions_q_values</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">get_all_q_values_for_states</span><span class="p">(</span><span class="n">curr_state</span><span class="p">)</span>
|
|
|
|
<span class="c1"># choose action according to the exploration policy and the current phase (evaluating or training the agent)</span>
|
|
<span class="n">action</span><span class="p">,</span> <span class="n">action_probabilities</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">exploration_policy</span><span class="o">.</span><span class="n">get_action</span><span class="p">(</span><span class="n">actions_q_values</span><span class="p">)</span>
|
|
<span class="k">if</span> <span class="bp">self</span><span class="o">.</span><span class="n">should_get_softmax_probabilities</span> <span class="ow">and</span> <span class="n">softmax_probabilities</span> <span class="ow">is</span> <span class="ow">not</span> <span class="kc">None</span><span class="p">:</span>
|
|
<span class="c1"># override the exploration policy's generated probabilities when an action was taken</span>
|
|
<span class="c1"># with the agent's actual policy</span>
|
|
<span class="n">action_probabilities</span> <span class="o">=</span> <span class="n">softmax_probabilities</span>
|
|
|
|
<span class="bp">self</span><span class="o">.</span><span class="n">_validate_action</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">exploration_policy</span><span class="p">,</span> <span class="n">action</span><span class="p">)</span>
|
|
|
|
<span class="k">if</span> <span class="n">actions_q_values</span> <span class="ow">is</span> <span class="ow">not</span> <span class="kc">None</span><span class="p">:</span>
|
|
<span class="c1"># this is for bootstrapped dqn</span>
|
|
<span class="k">if</span> <span class="nb">type</span><span class="p">(</span><span class="n">actions_q_values</span><span class="p">)</span> <span class="o">==</span> <span class="nb">list</span> <span class="ow">and</span> <span class="nb">len</span><span class="p">(</span><span class="n">actions_q_values</span><span class="p">)</span> <span class="o">></span> <span class="mi">0</span><span class="p">:</span>
|
|
<span class="n">actions_q_values</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">exploration_policy</span><span class="o">.</span><span class="n">last_action_values</span>
|
|
|
|
<span class="c1"># store the q values statistics for logging</span>
|
|
<span class="bp">self</span><span class="o">.</span><span class="n">q_values</span><span class="o">.</span><span class="n">add_sample</span><span class="p">(</span><span class="n">actions_q_values</span><span class="p">)</span>
|
|
|
|
<span class="n">actions_q_values</span> <span class="o">=</span> <span class="n">actions_q_values</span><span class="o">.</span><span class="n">squeeze</span><span class="p">()</span>
|
|
<span class="n">action_probabilities</span> <span class="o">=</span> <span class="n">action_probabilities</span><span class="o">.</span><span class="n">squeeze</span><span class="p">()</span>
|
|
|
|
<span class="k">for</span> <span class="n">i</span><span class="p">,</span> <span class="n">q_value</span> <span class="ow">in</span> <span class="nb">enumerate</span><span class="p">(</span><span class="n">actions_q_values</span><span class="p">):</span>
|
|
<span class="bp">self</span><span class="o">.</span><span class="n">q_value_for_action</span><span class="p">[</span><span class="n">i</span><span class="p">]</span><span class="o">.</span><span class="n">add_sample</span><span class="p">(</span><span class="n">q_value</span><span class="p">)</span>
|
|
|
|
<span class="n">action_info</span> <span class="o">=</span> <span class="n">ActionInfo</span><span class="p">(</span><span class="n">action</span><span class="o">=</span><span class="n">action</span><span class="p">,</span>
|
|
<span class="n">action_value</span><span class="o">=</span><span class="n">actions_q_values</span><span class="p">[</span><span class="n">action</span><span class="p">],</span>
|
|
<span class="n">max_action_value</span><span class="o">=</span><span class="n">np</span><span class="o">.</span><span class="n">max</span><span class="p">(</span><span class="n">actions_q_values</span><span class="p">),</span>
|
|
<span class="n">all_action_probabilities</span><span class="o">=</span><span class="n">action_probabilities</span><span class="p">)</span>
|
|
|
|
<span class="k">else</span><span class="p">:</span>
|
|
<span class="n">action_info</span> <span class="o">=</span> <span class="n">ActionInfo</span><span class="p">(</span><span class="n">action</span><span class="o">=</span><span class="n">action</span><span class="p">,</span> <span class="n">all_action_probabilities</span><span class="o">=</span><span class="n">action_probabilities</span><span class="p">)</span>
|
|
|
|
<span class="k">return</span> <span class="n">action_info</span>
|
|
|
|
<span class="k">def</span> <span class="nf">learn_from_batch</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">batch</span><span class="p">):</span>
|
|
<span class="k">raise</span> <span class="ne">NotImplementedError</span><span class="p">(</span><span class="s2">"ValueOptimizationAgent is an abstract agent. Not to be used directly."</span><span class="p">)</span>
|
|
|
|
<span class="k">def</span> <span class="nf">run_off_policy_evaluation</span><span class="p">(</span><span class="bp">self</span><span class="p">):</span>
|
|
<span class="sd">"""</span>
|
|
<span class="sd"> Run the off-policy evaluation estimators to get a prediction for the performance of the current policy based on</span>
|
|
<span class="sd"> an evaluation dataset, which was collected by another policy(ies).</span>
|
|
<span class="sd"> :return: None</span>
|
|
<span class="sd"> """</span>
|
|
<span class="k">assert</span> <span class="bp">self</span><span class="o">.</span><span class="n">ope_manager</span>
|
|
|
|
<span class="k">if</span> <span class="ow">not</span> <span class="nb">isinstance</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">pre_network_filter</span><span class="p">,</span> <span class="n">NoInputFilter</span><span class="p">)</span> <span class="ow">and</span> <span class="nb">len</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">pre_network_filter</span><span class="o">.</span><span class="n">reward_filters</span><span class="p">)</span> <span class="o">!=</span> <span class="mi">0</span><span class="p">:</span>
|
|
<span class="k">raise</span> <span class="ne">ValueError</span><span class="p">(</span><span class="s2">"Defining a pre-network reward filter when OPEs are calculated will result in a mismatch "</span>
|
|
<span class="s2">"between q values (which are scaled), and actual rewards, which are not. It is advisable "</span>
|
|
<span class="s2">"to use an input_filter, if possible, instead, which will filter the transitions directly "</span>
|
|
<span class="s2">"in the replay buffer, affecting both the q_values and the rewards themselves. "</span><span class="p">)</span>
|
|
|
|
<span class="n">ips</span><span class="p">,</span> <span class="n">dm</span><span class="p">,</span> <span class="n">dr</span><span class="p">,</span> <span class="n">seq_dr</span><span class="p">,</span> <span class="n">wis</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">ope_manager</span><span class="o">.</span><span class="n">evaluate</span><span class="p">(</span>
|
|
<span class="n">evaluation_dataset_as_episodes</span><span class="o">=</span><span class="bp">self</span><span class="o">.</span><span class="n">memory</span><span class="o">.</span><span class="n">evaluation_dataset_as_episodes</span><span class="p">,</span>
|
|
<span class="n">evaluation_dataset_as_transitions</span><span class="o">=</span><span class="bp">self</span><span class="o">.</span><span class="n">memory</span><span class="o">.</span><span class="n">evaluation_dataset_as_transitions</span><span class="p">,</span>
|
|
<span class="n">batch_size</span><span class="o">=</span><span class="bp">self</span><span class="o">.</span><span class="n">ap</span><span class="o">.</span><span class="n">network_wrappers</span><span class="p">[</span><span class="s1">'main'</span><span class="p">]</span><span class="o">.</span><span class="n">batch_size</span><span class="p">,</span>
|
|
<span class="n">discount_factor</span><span class="o">=</span><span class="bp">self</span><span class="o">.</span><span class="n">ap</span><span class="o">.</span><span class="n">algorithm</span><span class="o">.</span><span class="n">discount</span><span class="p">,</span>
|
|
<span class="n">q_network</span><span class="o">=</span><span class="bp">self</span><span class="o">.</span><span class="n">networks</span><span class="p">[</span><span class="s1">'main'</span><span class="p">]</span><span class="o">.</span><span class="n">online_network</span><span class="p">,</span>
|
|
<span class="n">network_keys</span><span class="o">=</span><span class="nb">list</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">ap</span><span class="o">.</span><span class="n">network_wrappers</span><span class="p">[</span><span class="s1">'main'</span><span class="p">]</span><span class="o">.</span><span class="n">input_embedders_parameters</span><span class="o">.</span><span class="n">keys</span><span class="p">()))</span>
|
|
|
|
<span class="c1"># get the estimators out to the screen</span>
|
|
<span class="n">log</span> <span class="o">=</span> <span class="n">OrderedDict</span><span class="p">()</span>
|
|
<span class="n">log</span><span class="p">[</span><span class="s1">'Epoch'</span><span class="p">]</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">training_epoch</span>
|
|
<span class="n">log</span><span class="p">[</span><span class="s1">'IPS'</span><span class="p">]</span> <span class="o">=</span> <span class="n">ips</span>
|
|
<span class="n">log</span><span class="p">[</span><span class="s1">'DM'</span><span class="p">]</span> <span class="o">=</span> <span class="n">dm</span>
|
|
<span class="n">log</span><span class="p">[</span><span class="s1">'DR'</span><span class="p">]</span> <span class="o">=</span> <span class="n">dr</span>
|
|
<span class="n">log</span><span class="p">[</span><span class="s1">'WIS'</span><span class="p">]</span> <span class="o">=</span> <span class="n">wis</span>
|
|
<span class="n">log</span><span class="p">[</span><span class="s1">'Sequential-DR'</span><span class="p">]</span> <span class="o">=</span> <span class="n">seq_dr</span>
|
|
<span class="n">screen</span><span class="o">.</span><span class="n">log_dict</span><span class="p">(</span><span class="n">log</span><span class="p">,</span> <span class="n">prefix</span><span class="o">=</span><span class="s1">'Off-Policy Evaluation'</span><span class="p">)</span>
|
|
|
|
<span class="c1"># get the estimators out to dashboard</span>
|
|
<span class="bp">self</span><span class="o">.</span><span class="n">agent_logger</span><span class="o">.</span><span class="n">set_current_time</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">get_current_time</span><span class="p">()</span> <span class="o">+</span> <span class="mi">1</span><span class="p">)</span>
|
|
<span class="bp">self</span><span class="o">.</span><span class="n">agent_logger</span><span class="o">.</span><span class="n">create_signal_value</span><span class="p">(</span><span class="s1">'Inverse Propensity Score'</span><span class="p">,</span> <span class="n">ips</span><span class="p">)</span>
|
|
<span class="bp">self</span><span class="o">.</span><span class="n">agent_logger</span><span class="o">.</span><span class="n">create_signal_value</span><span class="p">(</span><span class="s1">'Direct Method Reward'</span><span class="p">,</span> <span class="n">dm</span><span class="p">)</span>
|
|
<span class="bp">self</span><span class="o">.</span><span class="n">agent_logger</span><span class="o">.</span><span class="n">create_signal_value</span><span class="p">(</span><span class="s1">'Doubly Robust'</span><span class="p">,</span> <span class="n">dr</span><span class="p">)</span>
|
|
<span class="bp">self</span><span class="o">.</span><span class="n">agent_logger</span><span class="o">.</span><span class="n">create_signal_value</span><span class="p">(</span><span class="s1">'Sequential Doubly Robust'</span><span class="p">,</span> <span class="n">seq_dr</span><span class="p">)</span>
|
|
<span class="bp">self</span><span class="o">.</span><span class="n">agent_logger</span><span class="o">.</span><span class="n">create_signal_value</span><span class="p">(</span><span class="s1">'Weighted Importance Sampling'</span><span class="p">,</span> <span class="n">wis</span><span class="p">)</span>
|
|
|
|
<span class="k">def</span> <span class="nf">get_reward_model_loss</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">batch</span><span class="p">:</span> <span class="n">Batch</span><span class="p">):</span>
|
|
<span class="n">network_keys</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">ap</span><span class="o">.</span><span class="n">network_wrappers</span><span class="p">[</span><span class="s1">'reward_model'</span><span class="p">]</span><span class="o">.</span><span class="n">input_embedders_parameters</span><span class="o">.</span><span class="n">keys</span><span class="p">()</span>
|
|
<span class="n">current_rewards_prediction_for_all_actions</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">networks</span><span class="p">[</span><span class="s1">'reward_model'</span><span class="p">]</span><span class="o">.</span><span class="n">online_network</span><span class="o">.</span><span class="n">predict</span><span class="p">(</span>
|
|
<span class="n">batch</span><span class="o">.</span><span class="n">states</span><span class="p">(</span><span class="n">network_keys</span><span class="p">))</span>
|
|
<span class="n">current_rewards_prediction_for_all_actions</span><span class="p">[</span><span class="nb">range</span><span class="p">(</span><span class="n">batch</span><span class="o">.</span><span class="n">size</span><span class="p">),</span> <span class="n">batch</span><span class="o">.</span><span class="n">actions</span><span class="p">()]</span> <span class="o">=</span> <span class="n">batch</span><span class="o">.</span><span class="n">rewards</span><span class="p">()</span>
|
|
|
|
<span class="k">return</span> <span class="bp">self</span><span class="o">.</span><span class="n">networks</span><span class="p">[</span><span class="s1">'reward_model'</span><span class="p">]</span><span class="o">.</span><span class="n">train_and_sync_networks</span><span class="p">(</span>
|
|
<span class="n">batch</span><span class="o">.</span><span class="n">states</span><span class="p">(</span><span class="n">network_keys</span><span class="p">),</span> <span class="n">current_rewards_prediction_for_all_actions</span><span class="p">)[</span><span class="mi">0</span><span class="p">]</span>
|
|
|
|
<span class="k">def</span> <span class="nf">improve_reward_model</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">epochs</span><span class="p">:</span> <span class="nb">int</span><span class="p">):</span>
|
|
<span class="sd">"""</span>
|
|
<span class="sd"> Train a reward model to be used by the doubly-robust estimator</span>
|
|
|
|
<span class="sd"> :param epochs: The total number of epochs to use for training a reward model</span>
|
|
<span class="sd"> :return: None</span>
|
|
<span class="sd"> """</span>
|
|
<span class="n">batch_size</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">ap</span><span class="o">.</span><span class="n">network_wrappers</span><span class="p">[</span><span class="s1">'reward_model'</span><span class="p">]</span><span class="o">.</span><span class="n">batch_size</span>
|
|
|
|
<span class="c1"># this is fitted from the training dataset</span>
|
|
<span class="k">for</span> <span class="n">epoch</span> <span class="ow">in</span> <span class="nb">range</span><span class="p">(</span><span class="n">epochs</span><span class="p">):</span>
|
|
<span class="n">loss</span> <span class="o">=</span> <span class="mi">0</span>
|
|
<span class="n">total_transitions_processed</span> <span class="o">=</span> <span class="mi">0</span>
|
|
<span class="k">for</span> <span class="n">i</span><span class="p">,</span> <span class="n">batch</span> <span class="ow">in</span> <span class="nb">enumerate</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">call_memory</span><span class="p">(</span><span class="s1">'get_shuffled_training_data_generator'</span><span class="p">,</span> <span class="n">batch_size</span><span class="p">)):</span>
|
|
<span class="n">batch</span> <span class="o">=</span> <span class="n">Batch</span><span class="p">(</span><span class="n">batch</span><span class="p">)</span>
|
|
<span class="n">loss</span> <span class="o">+=</span> <span class="bp">self</span><span class="o">.</span><span class="n">get_reward_model_loss</span><span class="p">(</span><span class="n">batch</span><span class="p">)</span>
|
|
<span class="n">total_transitions_processed</span> <span class="o">+=</span> <span class="n">batch</span><span class="o">.</span><span class="n">size</span>
|
|
|
|
<span class="n">log</span> <span class="o">=</span> <span class="n">OrderedDict</span><span class="p">()</span>
|
|
<span class="n">log</span><span class="p">[</span><span class="s1">'Epoch'</span><span class="p">]</span> <span class="o">=</span> <span class="n">epoch</span>
|
|
<span class="n">log</span><span class="p">[</span><span class="s1">'loss'</span><span class="p">]</span> <span class="o">=</span> <span class="n">loss</span> <span class="o">/</span> <span class="n">total_transitions_processed</span>
|
|
<span class="n">screen</span><span class="o">.</span><span class="n">log_dict</span><span class="p">(</span><span class="n">log</span><span class="p">,</span> <span class="n">prefix</span><span class="o">=</span><span class="s1">'Training Reward Model'</span><span class="p">)</span>
|
|
|
|
</pre></div>
|
|
|
|
</div>
|
|
|
|
</div>
|
|
<footer>
|
|
|
|
|
|
<hr/>
|
|
|
|
<div role="contentinfo">
|
|
<p>
|
|
© Copyright 2018-2019, Intel AI Lab
|
|
|
|
</p>
|
|
</div>
|
|
Built with <a href="http://sphinx-doc.org/">Sphinx</a> using a <a href="https://github.com/rtfd/sphinx_rtd_theme">theme</a> provided by <a href="https://readthedocs.org">Read the Docs</a>.
|
|
|
|
</footer>
|
|
|
|
</div>
|
|
</div>
|
|
|
|
</section>
|
|
|
|
</div>
|
|
|
|
|
|
|
|
<script type="text/javascript">
|
|
jQuery(function () {
|
|
SphinxRtdTheme.Navigation.enable(true);
|
|
});
|
|
</script>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
</body>
|
|
</html> |