Add pre-flight memory validation for Metal agent#204
Merged
Conversation
Estimate model memory requirements (weights + KV cache + overhead) before spawning llama-server processes on Apple Silicon. Refuses to start models that exceed the system's memory budget, preventing OOM crashes and macOS force-kills. - MemoryProvider interface with DarwinMemoryProvider (sysctl/vm_stat) and non-darwin stub for CI - KV cache estimation from GGUF metadata with file-size heuristic fallback - Auto-detected memory fraction (67% for ≤36GB, 75% for larger systems) - Sets InferenceService SchedulingStatus to InsufficientMemory on rejection - --memory-fraction flag for manual override - 15 unit tests covering estimation, budget checks, formatting, parsing Closes #185 Signed-off-by: Christopher Maher <chris@mahercode.io>
Document --memory-fraction flag, InsufficientMemory troubleshooting, and memory budget behavior in the Metal Agent guide and quickstart. Closes #185 Signed-off-by: Christopher Maher <chris@mahercode.io>
This was referenced Mar 4, 2026
Merged
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
llama-serveron Apple Silicon, preventing OOM crashes and macOS force-kills--memory-fractionflag for manual overrideSchedulingStatus: InsufficientMemoryon the InferenceServiceTest plan
make test— all existing + 15 new unit tests passmake vet && make fmt— cleanmake lint— 0 issuesmake build— builds on darwin (real provider) and linux (stub)--memory-fraction 0.5on a small system, verify rejection with actionable error messageCloses #185