This is a companion piece to my instructions on building TensorFlow from source. In particular, the aim is to install the following pieces of software
on an Ubuntu Linux system, in particular Ubuntu 20.04.
| # install | |
| !npx degit facebookresearch/demucs -f | |
| !pip install lameenc==1.2.2 musdb==0.3.1 museval==0.3.0 treetable==0.2.3 | |
| # separate | |
| !python3 -m demucs.separate --dl -n demucs_extra song.m4a | |
| # get separated/demucs_extra/song/vocals.wav | |
| 'an easy-to-use function' | |
| def get_voc(fname): | |
| from pathlib import Path |
| {"constructors":[{"id":"-1132882121","predicate":"boolFalse","params":[],"type":"Bool"},{"id":"-1720552011","predicate":"boolTrue","params":[],"type":"Bool"},{"id":"1072550713","predicate":"true","params":[],"type":"True"},{"id":"481674261","predicate":"vector","params":[],"type":"Vector t"},{"id":"-994444869","predicate":"error","params":[{"name":"code","type":"int"},{"name":"text","type":"string"}],"type":"Error"},{"id":"1450380236","predicate":"null","params":[],"type":"Null"},{"id":"2134579434","predicate":"inputPeerEmpty","params":[],"type":"InputPeer"},{"id":"2107670217","predicate":"inputPeerSelf","params":[],"type":"InputPeer"},{"id":"396093539","predicate":"inputPeerChat","params":[{"name":"chat_id","type":"int"}],"type":"InputPeer"},{"id":"-1182234929","predicate":"inputUserEmpty","params":[],"type":"InputUser"},{"id":"-138301121","predicate":"inputUserSelf","params":[],"type":"InputUser"},{"id":"-208488460","predicate":"inputPhoneContact","params":[{"name":"client_id","type":"long"},{"name":"phone" |
| #!/usr/bin/env bash | |
| set -e | |
| # See also: https://github.com/frost-nzcr4/find_forks (same thing but in python) | |
| origin_url="$(git remote show origin | grep 'Fetch URL:' | sed 's+.*: ++')" | |
| full_repo_name="$(echo "$origin_url" | sed 's+.*github.com/++ ; s+\.git$++')" | |
| forks_url="https://api.github.com/repos/${full_repo_name}/forks" |
| """Tensorflow implementation of VGG perceptual distance network.""" | |
| import numpy as np | |
| import tensorflow as tf | |
| tft=tf.transpose | |
| def _i(x): return tft(x,[0,2,3,1]) | |
| def _o(x): return tft(x,[0,3,1,2]) | |
| # Initialize dict which maps TF variable names to pre-trained weight dict keys. | |
| _vgg16_params_dict = dict() |
A good way to get a taste of Swift for Tensorflow language and tools is to set it up with Jupyter with the fastai Swift notebooks. I wanted a quick setup, which the Mac install experience currently not, so instead I installed the release binaries in a Ubuntu container via Docker. The setup process for this scenario is not well documented, so here it is for you / future me.
What we're about to do is install the S4TF 0.4 release and the fastai v3 Swift notebooks on Ubuntu 18.04. Generally we follow the swift-jupyter docker file, but install cpu-only release versions of the packages.
Below are some of the references I looked at:
https://github.com/tensorflow/swift/blob/master/docs/WhySwiftForTensorFlow.md https://github.com/tensorflow/swift/blob/master/docs/DifferentiableFunctions.md
| # (forked and adapted from Gitlab's https://gitlab.com/gitlab-data/analytics/blob/master/admin/onboarding_script.sh) | |
| ## install homebrew | |
| echo "Installing Homebrew.." | |
| /usr/bin/ruby -e "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/master/install)" | |
| echo "Homebrew successfully installed" | |
| ## install brew cask | |
| echo "Installing brew cask.." | |
| brew tap homebrew/cask |
| @objc func keyboardWillShow(_ notification: NSNotification) { | |
| let keyboardAnimationDetail = notification.userInfo | |
| let animationCurve: Int = { | |
| if let keyboardAnimationCurve = keyboardAnimationDetail?[UIResponder.keyboardAnimationCurveUserInfoKey] as? Int { | |
| let curve: Int? = UIView.AnimationCurve(rawValue: keyboardAnimationCurve)?.rawValue | |
| return curve ?? 0 | |
| } else { | |
| return 0 | |
| } |
