Trackle
Trackle is a lightweight, local-first, git-aware ML experiment tracker.
Goals
- Local filesystem storage only
- No servers or networking
- Simple Python API and CLI
- Git-based reproducibility
Installation
pip install trackle-ml # from PyPI
# or for development
pip install -e .
# optional extras
# pip install trackle-ml[viz] # matplotlib plots
# pip install trackle-ml[rich] # richer CLI output (if extended)
Quick start
import trackle_ml as trackle
trackle.init(run_name="demo")
trackle.log_params({"lr": 0.01})
trackle.log_metric("loss", 0.5, step=1)
trackle.finish()
Storage layout
Runs are stored locally (default ./trackle_experiments):
trackle_experiments/
runs/
<run_id>/
run.json # metadata: name, created_at, finished_at, duration_s, tags
params.json # merged params
metrics.csv # step,key,value,timestamp
context.json # env info (python/os/hw/torch)
git.json # git commit/branch/dirty/remote
artifacts/ # copied artifacts
notes.md # optional note
CLI
trackle-ml list-runs [--base DIR] [--limit N]
trackle-ml diff-runs <run_a> <run_b> [--base DIR]
trackle-ml plot-metrics <run_id> [--key metric] [--base DIR]
1 maintainer
Metadata
1 maintainer
Release files for trackle-ml 0.1.2
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | |
|---|---|---|---|
| trackle_ml-0.1.2.tar.gz | 15.2 kB | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| trackle_ml-0.1.2-py3-none-any.whl | Python 3 | none | any | Details |
Total release size: 34.6 kB
Release files / trackle_ml-0.1.2.tar.gz
| Download URL | trackle_ml-0.1.2.tar.gz |
|---|---|
| Size | 15.2 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
58061e012807534c3067bc344d34e20ba7725f8040ca2aab845c1293203418f6
|
|
BLAKE2b-256 checksum How to use checksums |
b0b77b62ee81f9607bdf4e4bbafc8f2285046ebab0c26f757f049d5ee6a47244
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.11.4
|
Release files / trackle_ml-0.1.2-py3-none-any.whl
| Download URL | trackle_ml-0.1.2-py3-none-any.whl |
|---|---|
| Size | 19.3 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
f29d6b971e4822d41e2d671f65f0a01e4a7b309dc2e7e71e20221010f86db6ca
|
|
BLAKE2b-256 checksum How to use checksums |
8c665ddc273e2855a5d73d66b6c232652a18a4309fa02cd627a34b335e1ecfdc
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.11.4
|
