Skip to content

JSitter/darknet_cnn_gpu

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Darknet Notebook and Python Script

Installation

Tensorflow GPU Requires python 3.6 and associated version specific dependencies which are included in this module.

To create the required anaconda environment run

conda env create -f conda_environment.yml

This installs the required environment in your default environment path

If you wish to install the conda environment in another location you can use the -p flag and run:

conda env create -f conda_environment.yml -p /home/user/anaconda3/envs/env_name

where env_name is the desired environment name. This is rumored to work, and if it does please let me know.

Trust but Verify

To see if your installation is correct run:

conda env list

This will show all installed environments with an asterisk showing your current one.

Usage

To use the newly installed environment run:

source activate tensorflow_gpu

Debugging

Before training there is a delay of a vew minutes as experienced by this user: tensorflow/tensorflow#18652

But it should still work.