Issues

Reject log encoding on a categorically-sampled env_param: the combination is contradictory and degenerates in practice
NVIDIA/cloudai#1037 · opened on Sep 17, 2026
Support relative-path test references and lazy-load only referenced test/hook tomls
NVIDIA/cloudai#985 · opened on Jul 29, 2026
7
Sleep workload never checks whether the command actually succeeded; always reports PASSED
NVIDIA/cloudai#980 · opened on Jul 27, 2026
Infer DSE param encoding (categorical/ordinal/log) at parse time in EnvParamSpec
NVIDIA/cloudai#956 · opened on Jul 1, 2026
env_params: support full dot-path keys from TestDefinition (cmd_args.foo.bar, bench_cmd_args.*)
NVIDIA/cloudai#950 · opened on Jun 29, 2026
DSE: single-element list dims ([1]) aren't scalarized before command-gen (no central invariant)
NVIDIA/cloudai#946 · opened on Jun 26, 2026
1
Rust-style source-located diagnostics for TOML config validation errors
NVIDIA/cloudai#945 · opened on Jun 26, 2026
No single source of truth for "agent-driven" run; --single-sbatch conflates scheduling with search strategy
NVIDIA/cloudai#944 · opened on Jun 26, 2026
single_sbatch routing: rework agent-driven (DSE/live-RL) handling and mixed-job guard ordering
NVIDIA/cloudai#937 · opened on Jun 20, 2026
Pin GitHub Actions to full commit SHAs for supply-chain security
NVIDIA/cloudai#926 · opened on Jun 15, 2026
Make trajectory cache opt-in per workload (cache_safe contract)
NVIDIA/cloudai#886 · opened on May 8, 2026
Propagate exclude_nodes to all get_nodes_by_spec callers (nemo-run, nemo-launcher, single-sbatch-runner, and others)
NVIDIA/cloudai#832 · opened on Mar 11, 2026
Refactor hardcoded container mount paths to use constants
NVIDIA/cloudai#819 · opened on Feb 27, 2026
Implement consistent model merging across the codebase
NVIDIA/cloudai#806 · opened on Feb 12, 2026
Refactor private attributes in Pydantic models to use PrivateAttr
NVIDIA/cloudai#800 · opened on Feb 10, 2026
Pass --kubeconfig to all kubectl subprocess invocations
NVIDIA/cloudai#794 · opened on Jan 30, 2026