Skip to main content

traice

Tiny yet useful tool for consistent model training logs generation.

Installation

To install through pip use the following command:

pip install traice

The tool requires only pandas package to be installed. However, there is environment.yml file which can be used for the same environment which is used for developing the tool:

conda env create -f environment.yml

Usage

The tool may be used as follows (see examples/dummy.py):

from random import seed, uniform
from time import time, sleep

from traice import Traicer

traicer = Traicer()


def train_step():
    sleep(uniform(0, 1))


seed(17)

init_timestamp = time()

for i in range(1, 5):
    start_timestamp = time()
    train_step()
    traicer.push(i, uniform(0, 1 / i), (time_ := time()) - start_timestamp, time_ - init_timestamp)

print(traicer.df)

Essentially, it accumulates all push arguments in a list which is then converted to a dataframe. The example produces the following log (the last two columns may differ a bit):

   epoch      loss      time  cumulative_time
0      1  0.806691  0.522609         0.522609
1      2  0.144813  0.961565         1.484184
2      3  0.234740  0.767061         2.251254
3      4  0.027541  0.661659         2.912921

Testing

To run test execute the following statement in your terminal:

python -m unittest discover test

Metadata

Release files for traice 0.2.0

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for traice 0.2.0
File Size Uploaded Provenance Actions
traice-0.2.0.tar.gz 6.4 kB Without provenance Details for traice-0.2.0.tar.gz

Release files / traice-0.2.0.tar.gz

Download URL traice-0.2.0.tar.gz
Size 6.4 kB
Tags Source
SHA-256 checksum
How to use checksums
239e0773dc5c88e36a178aac9a712949427a44780a6269530157395b5b6ed7eb
BLAKE2b-256 checksum
How to use checksums
1dcbc4a2a28634360e34a6bd1098841d5fcac2149d4c0874e3a8d9f14051c6c4
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/3.7.1 importlib_metadata/4.11.3 pkginfo/1.8.3 requests/2.28.1 requests-toolbelt/0.9.1 tqdm/4.64.1 CPython/3.10.8

Security for traice 0.2.0

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

No provenance found

None of the files in this release were uploaded with provenance metadata.

A note on package security

PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.

To learn more about protecting yourself from malicious actors, visit our package security documentation.

Release history Release notifications | RSS feed

This release

0.2.0 This release

1 release file No provenance

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page