MA

AGI-Arena/MARS

The official implementation of MARS: Unleashing the Power of Variance Reduction for Training Large Models

717 49 -1/wk
GitHub
fine-tuning large-language-models optimization-algorithms optimizer pretraining
Trend 0

Star & Fork Trend (19 data points)

Stars
Forks

Multi-Source Signals

Growth Velocity

AGI-Arena/MARS has -1 stars this period . 7-day velocity: -0.1%.

Deep analysis is being generated for this repository.

Signal-backed technical analysis will be available soon.

Metric MARS SqueezeLLM autocontext InstructIR
Stars 717 717716718
Forks 49 505145
Weekly Growth -1 +0+10+0
Language Python PythonPythonJupyter Notebook
Sources 1 111
License Apache-2.0 MITApache-2.0MIT

Capability Radar vs SqueezeLLM

MARS
SqueezeLLM
Maintenance Activity 96

Last code push 14 days ago.

Community Engagement 34

Fork-to-star ratio: 6.8%. Lower fork ratio may indicate passive usage.

Issue Burden 70

Issue data not yet available.

Growth Momentum 30

No measurable growth in the current period (first-day cold start expected).

License Clarity 95

Licensed under Apache-2.0. Permissive — safe for commercial use.

Risk scores are computed from real-time repository data. Higher scores indicate healthier metrics.