Commit History

Auteur SHA1 Bericht Datum
  Alexander Korolev 31aaf141e5 refactor: split ends (HA/SSH) from control logic with traits and orchestrator 3 maanden geleden
  Alexander Korolev 56bcdf4bf5 feat(recipe): update deepseek-v4-flash-dspark from NVIDIA reference 3 maanden geleden
  Alexander Korolev 66b9879240 fix: auto-power-on now powers on per-node entities, not just cluster extras 3 maanden geleden
  Alexander Korolev 7158e63649 refactor: separate cluster state from SSH probing for testability 3 maanden geleden
  Alexander Korolev f54e192ef7 fix: mark unreachable nodes as unavailable in cluster selection 3 maanden geleden
  Alexander Korolev ff23a80a81 feat: extend NcclConfig with per-cluster NCCL overrides 3 maanden geleden
  Alexander Korolev 9123df86cd fix: use original cluster name for solo instances instead of 'solo/<node>' 4 maanden geleden
  Alexander Korolev 27c02331d8 refactor: rename --power-on-if-off to --auto-power-on 4 maanden geleden
  Alexander Korolev 811b8b0848 feat: add --power-on-if-off and --power-off flags 4 maanden geleden
  Alexander Korolev 98abe64703 fix: address soundness issues from audit 4 maanden geleden
  Alexander Korolev 0ae454c259 feat: add --node parameter to logs command 4 maanden geleden
  Alexander Korolev cefdfa2368 feat: deployment discovery + stop ambiguity resolution 4 maanden geleden
  Alexander Korolev 3774e765e5 feat: cluster selection with occupancy awareness 4 maanden geleden
  Alexander Korolev 3d986923c7 fix: drop incompatible -q flag from container label query (docker -q + --format conflict) 4 maanden geleden
  Alexander Korolev 68087045df fix: resolve_cluster_by_container_label now actually reads the sm121.cluster label from container metadata instead of inferring cluster from node ownership 4 maanden geleden
  Alexander Korolev 2946e07f79 fix: stop_cluster filters by cluster label to avoid cross-cluster interference 4 maanden geleden
  Alexander Korolev 5931869d69 refactor: DistributedBackend as Option — remove None_ variant 4 maanden geleden
  Alexander Korolev dbdbc7ed62 refactor: replace String distributed_backend with DistributedBackend enum 4 maanden geleden
  Alexander Korolev 6081ccf443 feat: mods system — entrypoint wrapping + global/recipe mods with volume mount 4 maanden geleden
  Alexander Korolev f55cd70082 refactor: replace #[allow] with structs for start_cluster and build_docker_command 4 maanden geleden
  Alexander Korolev ce60c5c59f feat: serve follows logs by default, returns Cluster for follow-up 4 maanden geleden
  Alexander Korolev e84b676397 feat: serve follows logs by default, --no-follow to skip 4 maanden geleden
  Alexander Korolev d6ccc9af9a fix: use docker stop only, no rm step 4 maanden geleden
  Alexander Korolev be65b44ddb fix: stop containers in single rm -f step 4 maanden geleden
  Alexander Korolev 6ff1e710aa feat: record cluster label on containers for auto-detection 4 maanden geleden
  Alexander Korolev ef03b8c784 feat: auto-scale PP when pipeline_parallel_size == required_nodes 4 maanden geleden
  Alexander Korolev 04413a8991 feat: auto-scale TP when recipe runs on larger cluster 4 maanden geleden
  Alexander Korolev 25540b4f48 fix: allow recipe with fewer required_nodes to run on larger cluster 4 maanden geleden
  Alexander Korolev 9d54d3f94c gemma4-12b: working recipe with distributed_backend=none 4 maanden geleden
  Alexander Korolev 03c293c4ca refactor(docker): move health_check to submodule 4 maanden geleden