diff options
| author | Paul Buetow <paul@buetow.org> | 2026-03-21 14:56:38 +0200 |
|---|---|---|
| committer | Paul Buetow <paul@buetow.org> | 2026-03-21 14:56:38 +0200 |
| commit | 1acb65324e8d7d520483535843a753757b0dd4a0 (patch) | |
| tree | fccaf3d2e10144489e4a480c3e70523133238fb6 /hyperstack.fish | |
| parent | 535f0adad3b0237b7d465bbd8231d5559c7be7d9 (diff) | |
Fix Nemotron OOM; add VM lifecycle fish abbrs; document automated setup
- hyperstack-vm1/vm2.toml: reduce nemotron-super max_model_len 262144→131072
and add --enforce-eager to disable CUDA graph capture (~3-4 GB overhead).
Nemotron 120B weights (~60 GB) leave too little VRAM headroom for KV cache
allocation and CUDA graph buffers at 262K context on a single A100 80GB.
131K context with eager mode is stable. README VRAM table updated to match.
- hyperstack.fish: add hyperstack-create/delete/test and hyperstack-create/delete-both
abbreviations for VM lifecycle management alongside the existing pi-* aliases.
- README.md: add "Automated setup reference" section with single-VM and two-VM
command flows before the manual vLLM Docker setup section.
End-to-end tested: single VM (GPT-OSS 120B), dual VM (Nemotron + Qwen3-Coder),
pi queries on all three models — all passed.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Diffstat (limited to 'hyperstack.fish')
| -rw-r--r-- | hyperstack.fish | 12 |
1 files changed, 10 insertions, 2 deletions
diff --git a/hyperstack.fish b/hyperstack.fish index 1b32a37..09706b5 100644 --- a/hyperstack.fish +++ b/hyperstack.fish @@ -1,3 +1,11 @@ +# Single-VM setup (hyperstack-vm.toml → hyperstack.wg1) abbr pi-hyperstack pi --model hyperstack/openai/gpt-oss-120b -abbr pi-hyperstack-nemotron pi --model hyperstack1/cyankiwi/NVIDIA-Nemotron-3-Super-120B-A12B-AWQ-4bit -abbr pi-hyperstack-coder pi --model hyperstack2/bullpoint/Qwen3-Coder-Next-AWQ-4bit +abbr hyperstack-create ruby ~/git/hyperstack/hyperstack.rb create +abbr hyperstack-delete ruby ~/git/hyperstack/hyperstack.rb delete +abbr hyperstack-test ruby ~/git/hyperstack/hyperstack.rb test + +# Dual-VM setup (hyperstack-vm1/vm2.toml → hyperstack1/2.wg1) +abbr pi-hyperstack-nemotron pi --model hyperstack1/cyankiwi/NVIDIA-Nemotron-3-Super-120B-A12B-AWQ-4bit +abbr pi-hyperstack-coder pi --model hyperstack2/bullpoint/Qwen3-Coder-Next-AWQ-4bit +abbr hyperstack-create-both ruby ~/git/hyperstack/hyperstack.rb create-both +abbr hyperstack-delete-both ruby ~/git/hyperstack/hyperstack.rb delete-both |