Building Tensorflow from source on Ubuntu 16.04LTS for maximum performance:
TensorFlow is now distributed under an Apache v2 open source license on GitHub.
On Ubuntu 16.04LTS+:
Step 1. Install NVIDIA CUDA:
To use TensorFlow with NVIDIA GPUs, the first step is to install the CUDA Toolkit as shown:
| """Downloads ArXiv papers linked to in a Google Doc converted to HTML. | |
| Notes: | |
| ------ | |
| * tested on OS X with Python 3 | |
| * Requires arxiv (`pip install arxiv`) | |
| * Names of PDFs will be the papers' titles on ArXiv (with some | |
| slight formatting changes). | |
| """ |
| """Scrape the definition of a word (or phrase) from dictionary.com | |
| Usage: | |
| ====== | |
| python define.py onomatopoeia | |
| Warning: | |
| ======== | |
| Sometimes (at least with phrases) you'll be unexpectedly redirected | |
| to the definition of another only loosely related word, |
| """The goal of this gist is to show how to compute many points on a path | |
| quickly using NumPy arrays. I.e. there's a much faster way than using, say | |
| [some_path.point(t) for t in many_tvals]. The example below assumes the | |
| `Path` object is composed entirely of `CubicBezier` objects, but this can | |
| easily be generalized to paths containing `Line` and `QuadraticBezier` objects | |
| also. | |
| Note: The relevant matrix transformation for quadratics can be found in the | |
| svgpathtools.bezier module.""" | |
| import numpy as np |
| from svgpathtools import * | |
| # create some example paths | |
| path1 = CubicBezier(1,2+3j,3-5j,4+1j) | |
| path2 = path1.rotated(60).translated(3) | |
| # find minimizer | |
| from scipy.optimize import fminbound | |
| def dist(t): | |
| return path1.radialrange(path2.point(t))[0][0] |
| """ | |
| Description: | |
| Takes in a color image and creates a two CSV files storing downsized | |
| matrix (grayscale) and 3-tensor representations of the image. | |
| Instructions: | |
| Put your image in the same folder as this file (tool4c.py), | |
| then open a terminal in that folder and enter: | |
| python tool4c <your_image> |