 Alexander Korolev
|
ad7069edcd
docker: --entrypoint takes tail[0], rest is argv (no repeat); port convention 30000 in recipes + docs
|
vor 4 Wochen |
|
Alexander Korolev
|
3bcc7760a0
docker: --entrypoint override for entrypoint-driven images; trim cluster to launch set before hooks/sync; recipe.cluster on auto
|
vor 4 Wochen |
|
Alexander Korolev
|
07919e5fc1
exl3 2x: recipe (rc2 image, /model mount, 9g/6cpu cguard, display unlock) + launcher memory_limit/cpus flags
|
vor 4 Wochen |
|
Alexander Korolev
|
2dcf9a3e9e
exl3: temporary 2 GB display-reserve unlock (unlock/restore builtins) + per-recipe device passthrough
|
vor 4 Wochen |
|
Alexander Korolev
|
d5663ec4d9
feat: MiMo-V2.6-Pro-RL 8x Spark recipe + build tree (SGLang a14, DFlash, RoCEnante); dry-run path-ref placeholder
|
vor 1 Monat |
|
Alexander Korolev
|
f4abb53cbf
recipe [env] refs: values can resolve from recipe fields ({ from = "port" })
|
vor 1 Monat |
|
Alexander Korolev
|
bd82f6f9ae
revert generic PORT injection; PORT is an explicit recipe [env] again
|
vor 1 Monat |
|
Alexander Korolev
|
1c53f3c737
sparkring 4x results + recipe cleanup per operator
|
vor 1 Monat |
|
Alexander Korolev
|
9cfed14f41
vision-373 lane benched: 161/176 new 2x record; launcher entrypoint-image support (model='', NODE_RANK/MASTER_ADDR env)
|
vor 1 Monat |
|
Alexander Korolev
|
48bbec1f66
serve: --clear flag to wipe host-side extra_mount folders (sudo)
|
vor 2 Monaten |
|
Alexander Korolev
|
551ea7d8d7
feat(nccl): auto-detect NCCL_IB_GID_INDEX per node, opt-out via config pin
|
vor 2 Monaten |
|
Alexander Korolev
|
5a6ec550d7
fix(ssh): pass through already single-quoted tokens in shq
|
vor 2 Monaten |
|
Alexander Korolev
|
e300668348
feat(docker): authoritative build/sync source, single-node build clusters, chunked upload
|
vor 2 Monaten |
|
Alexander Korolev
|
df4149920b
fix: remediate F1 (image-sync host-key verification) and F2 (docker run quoting)
|
vor 2 Monaten |
|
Alexander Korolev
|
0181645b9e
fix: make IB config optional for single-node clusters
|
vor 2 Monaten |
|
Alexander Korolev
|
1b4d07309a
fix: make NCCL config optional for single-node, capture stderr, support --on for tmux
|
vor 2 Monaten |
|
Alexander Korolev
|
5f646966b8
feat(tmux): add venv_setup field for declarative venv creation
|
vor 2 Monaten |
|
Alexander Korolev
|
d9b922e01b
feat(recipe): add venv field for declarative tmux venv activation
|
vor 2 Monaten |
|
Alexander Korolev
|
c4a74cc152
feat(recipe): add deploy field for docker/tmux variant selection
|
vor 2 Monaten |
|
Alexander Korolev
|
de18759712
feat(serve): add composed_model for multi-repo model directories
|
vor 3 Monaten |
|
Alexander Korolev
|
febdc3b9d7
feat(build): add Solar Open2 NVFP4 recipe with recipe-pinned versions
|
vor 3 Monaten |
|
Alexander Korolev
|
23ae7a6484
feat(serve): add solo mode and fix single-node launch bug
|
vor 3 Monaten |
|
Alexander Korolev
|
65198e60fc
feat(recipe): add built-in script aliases and after_stop hook
|
vor 3 Monaten |
|
Alexander Korolev
|
a7cd759af1
feat(recipe): model-subfolder path refs via {{name}} substitution
|
vor 3 Monaten |
|
Alexander Korolev
|
6cd245f9a8
feat(recipe): model-subfolder mounts resolved per-node from HF cache
|
vor 3 Monaten |
|
Alexander Korolev
|
2b0deeaffa
feat(recipe): extra_mounts with config alias resolution + tilde + env injection
|
vor 3 Monaten |
