Python

cpprb v10.1.0 can save & load transitions

From cpprb version 10.1.0, ReplayBuffer and its sub-classes can save and load transitions. from cpprb import ReplayBuffer rb1 = ReplayBuffer(256, {"obs": {"shape": 3}, "act": {}, "rew": {}, "done": {}}, next_of="obs")

ValueError: numpy.ndarray size changed, may indicate binary incompatibility. Expected 88 from C header, got 80 from PyObject

When import cpprb, I got ValueError: numpy.ndarray size changed, may indicate binary incompatibility. Expected 88 from C header, got 80 from PyObject. Recently, NumPy has changed its ABI at version

Speed up Ape-X implementation on single machine

Alghough neural network is optimized for GPU, environments for reinforcement learning (e.g. simulator) are not always GPU friendly. One of the method to speed up reinforcement learning is to run

Snippet: Binary Classification with LightGBM

Following snippet executes binary classification with LightGBM. The binary_gbm_cv function runs cross validation on training data and returns prediction function composing boosters used at the cross validation. The function also

b4tf

Bayes Neural Network for TensorFlow

b4tf: Introducing new Bayes Neural Net library

I started a new project b4tf “Bayes Neural Net for TensorFlow”. I plan to implement many Bayes neural net algorithms based on TensorFlow 2 and TensorFlow Probability. The first and

DeepMind's reinforcement learning framework: Acme

As I mentioned at the previous post, DeepMind published a reinforcement learning framework Acme. 1. Introduction Acme provides higher level API, a simple training code can be something like following;

[gnwrapper] Support KeyboardInterrupt on Google Colab

Gym-Notebook-Wrapper (aka. gnwrapper) is a Python package to render OpenAI Gym on Google Colaboratory (or Jupyter Notebook at Linux). A class gnwrapper.Monitor wrapps gym.wrappers.Monitor and adds usuful features (e.g. starting

Specifying package version with pip

When installing Python packages with pip, we can specify the package version explicitly. >, <, <=, >=, ==, and != operators are well known and understandable easily. Additionally, pip support

TF2.3 starts warning of tf.function's problematic usage

This post describes what I noticed when I made PR for TF2RL. From version 2.3, TensorFlow changes warning for tf.function’s problematic usage. Input Python object are also considered. You may