Skip to main content
Join the official 2020 Python Developers SurveyStart the survey!

Machine learning file and parameter version control SDK for

Project description

tracker_ml SDK and CLI


pip install tracker-ml


Install build tools

python -m pip install --user --upgrade setuptools wheel

Create distribution archives

python sdist bdist_wheel

Install from archives

python -m pip install dist/tracker_ml-X.X.X.tar.gz

Python SDK


First, initialize tracker using the CLI and update the files you want to track. Then import tracker_ml.tml anywhere and everywhere. Use is easy:

import tracker_ml.tml as tml

tml.login("username", "password")
tml.model("Logistic Regression")

# <machine learning code>

# record int, float, or str
tml.record("accuracy", 0.42)
tml.record("model", "Logistic Regression")

# record multiple values under the same key
tml.mrecord("epoch", 1)
tml.mrecord("epoch", 2)
tml.mrecord("epoch", 3)

# data will be saved locally and to the API on exit

All changes since the previous run and all recorded values will be automatically saved. The CLI can be used to view/compare trials and undo changes.

Command Line Interface command line interface for locally tracking/reverting file changes and tracking results for each change. Similar to git, but works with the SDK to track every time a new model is trained/tested.


Use the help command. (Not all commands displayed work yet)

$ tracker --help
$ tracker status --help

Initialize in the project root.

$ tracker init -u <username> -p <password> -n <project name>

Add file(s)/directory(s) that will be saved every run.

$ tracker add .

Stop recording file(s)/directory(s) that would be saved every run.

$ tracker remove .

View past trials and sort them

$ tracker status
 Total trials: 4
 Sorted by: id

  Id  |  Accuracy  |         Model
  4   |     63     |  Logistic Regression
  3   |     74     |  Logistic Regression
  2   |     50     |  Logistic Regression
  1   |     92     |  Logistic Regression
$ python status -k accuracy -l 2 -r
 Total trials: 4
 Reverse sorted by: accuracy
 Only displaying 2 results

  Id  |  Accuracy  |         Model
  2   |     50     |  Logistic Regression
  4   |     63     |  Logistic Regression

Project details

Download files

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

Files for tracker-ml, version 0.0.6
Filename, size File type Python version Upload date Hashes
Filename, size tracker_ml-0.0.6-py3-none-any.whl (10.2 kB) File type Wheel Python version py3 Upload date Hashes View
Filename, size tracker_ml-0.0.6.tar.gz (7.8 kB) File type Source Python version None Upload date Hashes View

Supported by

Pingdom Pingdom Monitoring Google Google Object Storage and Download Analytics Sentry Sentry Error logging AWS AWS Cloud computing DataDog DataDog Monitoring Fastly Fastly CDN DigiCert DigiCert EV certificate StatusPage StatusPage Status page