Skip to content

v1.0.0

Latest

Choose a tag to compare

@FareedKhan-dev FareedKhan-dev released this 07 Aug 16:39
ff11dce

Verified end to end on the full released 1.56 TB checkpoint, and made substantially faster, with byte-identical output preserved at every step. The first-run experience, broken on a clean clone, now works.

Faster (output unchanged)

  • Fused matmul kernels (fp32, bf16, MXFP4): about 8x less per-token compute, scalar and AVX2 paths bitwise identical.
  • --preset auto: sizes the trunk and expert-cache budgets from the machine's own free RAM, trunk-first.
  • Chunk-union prefill: fetches each unique routed expert once per chunk, about half the expert bytes on a prompt, generated token bit-identical.
  • Conversation resume (--save-state / --load-state): a second turn resumes instead of re-reading the prompt, 3.9x faster.
  • KDA recurrence parallelised over heads; --spec N speculative decode, exact by construction.

Correctness and first run

  • All shell scripts are committed executable; the first documented command works on a clean clone.
  • download-model.sh pins an immutable revision with checksum verification and refuses without free space.
  • k3-doctor.sh no longer fails a machine that can build and test the engine.
  • The config-refusal fixtures are gated in make test, ctest, and CI.
  • Two silent-corruption paths (the MLA KV overflow and the single-slot trunk reader) now abort or are prevented.

Verification

Token ids byte-identical to the reference from 8 GB to 128 GB+, oracle gates untouched, make test and ctest green.

Full log in CHANGELOG.md; measured speed in docs/data/speed-2026-08.md.