Skip to main content

Active Learning Environments

Project description

Dinos

Dinos is a simulation environment for active learning algorithms.

Getting started

First of all, install the package either using pip:

pip install dinos

Or from the git repository:

pip install -r ./requirements.txt
pip install -e .

Examples are provided in the examples folder from the git repository.

How does it works

To run a Dinos experiment you need an Environment and an Agent.
For instance, an environment may be initialized as follow:

from dinos.environments.playground import PlaygroundEnvironment
env = PlaygroundEnvironment()

From there you can either use your own code and use low level API to interact with the environment: env.step(self, action, actionParameters=[], config=None) as detailed later on. The second option is to use the Dinos Agent system to manage your algorithm.

For instance to create an agent that will perform a random action at each step:

from dinos.agents.random import RandomAgent
agent = RandomAgent(env.world.findHost())

env.world.findHost() let you find an entity in the environment that can be controlled by your learner (we call such entity an host)

Each Agent has a reach(self, configOrGoal) method that can be used to tell the agent to reach a specific goal.

Additionally a specific type of agent exists: Learner. This class is designed to be used with a dataset or a memory to learn from its interactions with the environment.

Each Learner has a train(self, iterations=None, untilIteration=None, episodes=None, untilEpisode=None) method used to train your learner for a given number of iterations or episodes.

More details are present in the examples folder from the git repository.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

dinos-0.1.1.tar.gz (235.1 kB view details)

Uploaded Source

Built Distribution

dinos-0.1.1-cp38-cp38-win_amd64.whl (431.0 kB view details)

Uploaded CPython 3.8 Windows x86-64

File details

Details for the file dinos-0.1.1.tar.gz.

File metadata

  • Download URL: dinos-0.1.1.tar.gz
  • Upload date:
  • Size: 235.1 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/3.4.2 importlib_metadata/4.8.1 pkginfo/1.7.1 requests/2.26.0 requests-toolbelt/0.9.1 tqdm/4.62.3 CPython/3.8.3

File hashes

Hashes for dinos-0.1.1.tar.gz
Algorithm Hash digest
SHA256 a44c7f765aa20e3d1307817407f3328b0aff743fe667333d6b06e15ca81ce051
MD5 ee15d703a0197e2f3cc044c93b7aeda7
BLAKE2b-256 04b4aeead96f8c5f6d16c338741c729e3665076dd193a4751fc9cfff9a71b3ae

See more details on using hashes here.

File details

Details for the file dinos-0.1.1-cp38-cp38-win_amd64.whl.

File metadata

  • Download URL: dinos-0.1.1-cp38-cp38-win_amd64.whl
  • Upload date:
  • Size: 431.0 kB
  • Tags: CPython 3.8, Windows x86-64
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/3.4.2 importlib_metadata/4.8.1 pkginfo/1.7.1 requests/2.26.0 requests-toolbelt/0.9.1 tqdm/4.62.3 CPython/3.8.3

File hashes

Hashes for dinos-0.1.1-cp38-cp38-win_amd64.whl
Algorithm Hash digest
SHA256 07b17e6ecfc40021ae2f681fed6f3f98a9c272826a78d8bf728b9650830a5d20
MD5 dd760168dd1708220d4da2fb085d3311
BLAKE2b-256 d5fe7b52bbe5598bc58fa9dee2a5bab219e9e74d3a01b606ad2e06433ccae67a

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page