Commit Graph
2 Commits
Author SHA1 Message Date
osobhandClaude Sonnet 5 4aaa36a57a style: cargo fmt --workspace (whitespace/wrapping only, no semantic change)
Whole-workspace rustfmt pass picked up while iterating on Mamba GPU
backward work. Verified formatting-only via diff sampling; no logic
changed.

Co-Authored-By: Claude Sonnet 5 <[email protected]>
2026-08-10 07:09:36 -07:00
Omar SobhandClaude Sonnet 4.6 bff27c302f feat(batch19): Medusa heads, TIES+DARE model merging, Mixture of Depths
CI / Format Check (push) Failing after 13s
CI / Clippy Check (push) Failing after 43s
GPU Tests / CUDA Tests (12.1) (push) Has been skipped
CI / Python Bindings (maturin) (macos-latest) (push) Has been skipped
CI / Python Bindings (maturin) (ubuntu-latest) (push) Has been skipped
CI / WASM Build + Size Check (push) Has been skipped
CI / Distributed Training Tests (push) Has been skipped
CI / Build CPU-Only (Explicit) (push) Failing after 1m40s
CI / CI Success (push) Failing after 0s
Performance Benchmarks / Run Benchmarks (push) Successful in 2m10s
CI / Build (macos-latest) (push) Failing after 29s
Documentation / Build User Guide (push) Successful in 6s
GPU Tests / Check GPU Availability (push) Successful in 0s
GPU Tests / CUDA Tests (11.8) (push) Has been skipped
CI / Build (ubuntu-latest) (push) Failing after 1m12s
CI / Test (macos-latest) (push) Has been skipped
CI / Test (ubuntu-latest) (push) Has been skipped
Documentation / Build API Documentation (push) Failing after 1m26s
GPU Tests / Metal Tests (push) Has been skipped
- MedusaHeads: K FFN draft heads (SiLU 2-layer); tree candidate generation
  via cartesian product of per-head top-k; path verification with oracle;
  CE training loss per head (arXiv:2401.10774); 23 tests
- ModelMerger: TIES (task-vector trim+elect-sign+disjoint-merge,
  arXiv:2306.01708) + DARE sparse rescaling (arXiv:2311.03099); linear
  merge baseline; 29 tests
- MoDLayer/MoDStack: per-token capacity routing (top-k by router score);
  residual bypass for skipped tokens; load-balancing aux loss; flops_reduction
  = product of capacity_fractions (arXiv:2404.02258); 22 tests

Co-Authored-By: Claude Sonnet 4.6 <[email protected]>
2026-06-27 13:07:39 +00:00