gitwtfhub

wtf is tensorflow_ddpg?

sykwer/tensorflow_ddpg — explained in plain English

Analysis updated 2026-07-17 · repo last pushed 2018-04-18

★ 1PythonAudience · researcherComplexity · 3/5DormantSetup · moderate

TL;DR

A learning project implementing DDPG, an AI technique that trains simulated robots to make smooth, fine-grained decisions instead of picking between a few fixed choices.

Mindmap

mindmap
  root((DDPG repo))
    What it does
      Continuous control learning
      Smooth action decisions
      Simulated physics training
    Tech stack
      Python
      TensorFlow
      OpenAI Gym
    Use cases
      Reference implementation
      Balance a pendulum
      Learn RL basics
    Audience
      Researchers
      Students
      Hobbyists

Code map

Detail Auto

An interactive map of this repo's files and how they connect — its source is parsed live in your browser. Click Visualize to build it.

filefunction / class

Why would anyone build with this?

REASON 1

Study a reference implementation of DDPG to understand how continuous-control reinforcement learning works.

REASON 2

Train a simulated pendulum to balance itself using the included InvertedPendulum example.

REASON 3

Use as a starting point to experiment with other continuous control problems in OpenAI Gym.

REASON 4

Follow along with the author's Japanese explanatory articles to learn DDPG step-by-step.

What's in the stack?

PythonTensorFlowOpenAI Gym

How it stacks up

sykwer/tensorflow_ddpga-bissell/unleash-liteabhiinnovates/whatsapp-hr-assistant
Stars111
LanguagePythonPythonPython
Last pushed2018-04-18——
MaintenanceDormant——
Setup difficultymoderatehardhard
Complexity3/54/53/5
Audienceresearcherresearcherdeveloper

Figures from each repo's GitHub metadata at analysis time.

How do you spin it up?

Difficulty · moderate Time to first run · 30min

Requires TensorFlow and OpenAI Gym installed to run the training example.

Yoink these prompts

Prompt 1
Explain how the two neural networks in this DDPG implementation work together to choose and evaluate actions.
Prompt 2
Walk me through how to run the InvertedPendulum training example in this repo step by step.
Prompt 3
Help me modify this DDPG code to work with a different OpenAI Gym continuous-control environment.
Prompt 4
What does the reward graph in this repo tell me about how training progress was measured?

Frequently asked questions

wtf is tensorflow_ddpg?

A learning project implementing DDPG, an AI technique that trains simulated robots to make smooth, fine-grained decisions instead of picking between a few fixed choices.

What language is tensorflow_ddpg written in?

Mainly Python. The stack also includes Python, TensorFlow, OpenAI Gym.

Is tensorflow_ddpg actively maintained?

Dormant — no commits in 2+ years (last push 2018-04-18).

How hard is tensorflow_ddpg to set up?

Setup difficulty is rated moderate, with roughly 30min to a first successful run.

Who is tensorflow_ddpg for?

Mainly researcher.

View the repo → Decode another repo

This repo across BitVibe Labs

Don't trust strangers blindly. Verify against the repo.

Other repos in this lane