Alexander Korolev
|
23ae7a6484
feat(serve): add solo mode and fix single-node launch bug
|
hai 3 meses |
Alexander Korolev
|
65198e60fc
feat(recipe): add built-in script aliases and after_stop hook
|
hai 3 meses |
Alexander Korolev
|
1ba1d99e05
fix: resolve review findings across modules
|
hai 3 meses |
Alexander Korolev
|
2b0deeaffa
feat(recipe): extra_mounts with config alias resolution + tilde + env injection
|
hai 3 meses |
Alexander Korolev
|
9e28b49887
feat(recipes): add volumes support and GLM-5.2 Hybrid FP8-MXFP4 recipe
|
hai 3 meses |
Alexander Korolev
|
731937a625
feat: --build-node, docker_image_dir recipe field
|
hai 3 meses |
Alexander Korolev
|
acadda1259
feat: auto-sync Docker image (ANY source node) + --sync-image / --sync-image-source flags
|
hai 3 meses |
Alexander Korolev
|
ff23a80a81
feat: extend NcclConfig with per-cluster NCCL overrides
|
hai 3 meses |
Alexander Korolev
|
0ae454c259
feat: add --node parameter to logs command
|
hai 4 meses |
Alexander Korolev
|
4936530710
feat: RecipeArgs enum — inline string or list for recipe args
|
hai 4 meses |
Alexander Korolev
|
5931869d69
refactor: DistributedBackend as Option — remove None_ variant
|
hai 4 meses |
Alexander Korolev
|
dbdbc7ed62
refactor: replace String distributed_backend with DistributedBackend enum
|
hai 4 meses |
Alexander Korolev
|
6081ccf443
feat: mods system — entrypoint wrapping + global/recipe mods with volume mount
|
hai 4 meses |
Alexander Korolev
|
f55cd70082
refactor: replace #[allow] with structs for start_cluster and build_docker_command
|
hai 4 meses |
Alexander Korolev
|
04413a8991
feat: auto-scale TP when recipe runs on larger cluster
|
hai 4 meses |
Alexander Korolev
|
9d54d3f94c
gemma4-12b: working recipe with distributed_backend=none
|
hai 4 meses |
Alexander Korolev
|
7b01e19ec2
fix: HF_HOME -> /root/.cache/huggingface (fixes stale cache)
|
hai 4 meses |
Alexander Korolev
|
9241bdc682
fix: don't pass --distributed-executor-backend mp (auto-detected)
|
hai 4 meses |
Alexander Korolev
|
6f2c33a82f
fix: only add --shm-size when recipe explicitly sets it
|
hai 4 meses |
Alexander Korolev
|
d9c3f845c2
fix(deepseek): align parameters with working system
|
hai 4 meses |
Alexander Korolev
|
7ba894be67
feat: per-recipe shm_size + DeepSeek V4 Flash recipe
|
hai 4 meses |
Alexander Korolev
|
7b9f963036
feat: optional cluster-level extra_entities for power commands
|
hai 4 meses |
Alexander Korolev
|
7b8651cead
feat: Home Assistant integration for power on/off
|
hai 4 meses |
Alexander Korolev
|
9c0e7680a4
feat: set safe NCCL defaults (IGNORE_CPU_AFFINITY, NET, CROSS_NIC, DEBUG)
|
hai 4 meses |
Alexander Korolev
|
1b02f53b07
feat: set GLOO_SOCKET_IFNAME and TP_SOCKET_IFNAME for vLLM
|
hai 4 meses |
Alexander Korolev
|
bfe94fa39e
feat: pre-serve hooks + exec subcommand for cluster mass-execution
|
hai 4 meses |
Alexander Korolev
|
5097d55c77
fix: redirect stderr in `cargo run logs --follow` docker pipeline
|
hai 4 meses |
Alexander Korolev
|
d92da29eb5
fix: shell-quote entrypoint script so 'sh -c SCRIPT' survives bash word-split
|
hai 4 meses |
Alexander Korolev
|
75860da26d
feat: test-sleep recipe streams logs; add --timestamps to docker logs
|
hai 4 meses |
Alexander Korolev
|
2b0be29648
feat: cluster defaults, configurable init_port, --headless for workers
|
hai 4 meses |