Issues

Detector silently keeps checkpoint hyper-parameters when config values update at resume.
AI-ModCon/BaseSIM_APEIRON#136 · opened on Oct 2, 2026
Prepare for formal public release
AI-ModCon/BaseSIM_APEIRON#133 · opened on Sep 22, 2026
import-time RNG consumption in MNIST_CNN's default Cnn() argument invalidated the numerical reference
AI-ModCon/BaseSIM_APEIRON#129 · opened on Sep 10, 2026
Add SKILLs for recommending a CL method
AI-ModCon/BaseSIM_APEIRON#117 · opened on Aug 8, 2026
Add tests for the drift/CL codes in src
AI-ModCon/BaseSIM_APEIRON#111 · opened on Jul 29, 2026
Filter checkpoints based on how better they are compared to previous ones
AI-ModCon/BaseSIM_APEIRON#108 · opened on Jul 22, 2026
Extend metric_index drift-trigger logic beyond single metric selection
AI-ModCon/BaseSIM_APEIRON#86 · opened on Mar 4, 2026
Test BaseSim with prometheus
AI-ModCon/BaseSIM_APEIRON#63 · opened on Feb 4, 2026
Design a better way to deal with metrics for model evaluation (for the drift detector)
AI-ModCon/BaseSIM_APEIRON#56 · opened on Dec 30, 2025
First implementation of the metadata management system
AI-ModCon/BaseSIM_APEIRON#54 · opened on Dec 18, 2025
CI on personal workstations
AI-ModCon/BaseSIM_APEIRON#52 · opened on Dec 18, 2025
Unit tests using a pre-trained mnist model
AI-ModCon/BaseSIM_APEIRON#51 · opened on Dec 18, 2025
1
First performance numbers on Frontier/Perlmutter/Aurora
AI-ModCon/BaseSIM_APEIRON#50 · opened on Dec 18, 2025
Create scripts to run things on NERSC/Frontier/Aurora
AI-ModCon/BaseSIM_APEIRON#49 · opened on Dec 18, 2025
2