Skip to content

Releases: zerfoo/zerfoo

v1.59.0

Choose a tag to compare

@github-actions github-actions released this 10 Aug 12:09

1.59.0 (2026-08-10)

Features

  • inference: demote gemma4/gemma4e/gemma4moe to experimental (T134.2) (60a4ee1)

Bug Fixes

  • attention: disable crash-inducing fused RepeatInterleave GPU path; fix Conv1D parity test (e348a8c)

v1.58.0

Choose a tag to compare

@github-actions github-actions released this 08 Aug 21:09

1.58.0 (2026-08-08)

Features

  • cli: wire --rate-limit/--rate-limit-burst/--keystore serve flags (T145.1) (7a60a62)

Bug Fixes

  • ci: scope benchmark PR-comment write token to the commenting job (CICD-3, T144.3) (a1bf81e)
  • ci: scope govulncheck to allowlist only the no-fix bbolt advisory (CICD-6, T144.6) (70f8ac5)
  • cli: default worker to loopback, add --tls-* flags (T140.2) (9b886c2)
  • cuda: prefer absolute vendor-library paths across native loaders (CUDA-2, T141.2) (252a92d)
  • deploy: digest-pin Dockerfile base images (CICD-4) (1f75836)
  • deps: bump golang.org/x/net to v0.53.0 (GO-2026-4918 HTTP/2 DoS) (941bab8)
  • deps: bump google.golang.org/grpc to v1.82.1 (GO-2026-6061) (33ed17d)
  • distributed: authenticate coordinator worker registration (35db753)
  • gguf: cap metadata array nesting depth to prevent stack overflow (F4) (3c9dae8)
  • gguf: cap tensor dimension count at 8 (F3, T139.3) (7ad5cc0)
  • registry: reject non-https registry URLs unless explicitly opted in (OCI-2, T139.5) (3177eac)
  • registry: support out-of-band expected hash for HF pulls (HF-1, T143.10) (7557dfb)
  • registry: verify blob digest on OCI pull (OCI-1, T139.4) (026ca1f)
  • serve: block unspecified + CGNAT addresses in SSRF guard (5794d09)
  • serve: bound GuardianMiddleware body read, skip SSE buffering (SERVE-4) (ed482ba)
  • serve: bound RateLimiter memory via scheduled cleanup + size cap (4d93744)
  • serve: cap /v1/embeddings input array size (SERVE-3, T143.3) (6ef8841)
  • serve: cap per-request image-fetch fan-out (SERVE-3b, T143.4) (8e8622b)
  • serve: eliminate KeyStore Revoked/ExpiresAt data race (CONC-M2, T142.4) (09e0e8a)
  • serve: log escaped request path to prevent log injection (SERVE-6, T143.8) (eae4e83)
  • serve: reap executeBatch ctx-wait goroutines promptly (CONC-L1) (0e8e7b7)
  • serve: remove dead SSRF validation code (SERVE-7, T143.9) (37d1c8a)
  • serve: validate max_tokens/temperature bounds (SERVE-5, T143.6) (214f06a)

v1.57.0

Choose a tag to compare

@github-actions github-actions released this 06 Jul 18:11

1.57.0 (2026-07-04)

Features

  • cmd: bench_train -seed flag + exact loss bit patterns (T135.5) (d125201)
  • docs: add EXTRA_ENV_PLACEHOLDER marker to validate-arm64 manifest (2fff208)
  • scripts: plumb extra env vars through dgx-validate for env-gated fixtures (40b6653)
  • timeseries: seedable weight init + GPU determinism double-run proof (T135.5) (0f6a316)

Bug Fixes

  • attention: replay-stable FusedSDPA scratch (T133.2) (5093a9b)
  • attention: stream-order flash-decode on the engine stream (T133.1) (db8c7b9)
  • ci: add least-privilege permissions blocks to PR-triggered workflows (5ebc449)
  • ci: pin mutable govulncheck/pip installs to exact versions (CICD-2) (3154617)
  • cuda: bump ztensor to v1.19.1 and guard darwin dlopen probe (T137.1) (215fcef)
  • cuda: guard FP16 graceful test from poisoning the CUDA context (T135.1) (047f2cc)
  • cuda: guard sgemv_m1 float4 vectorized load on row pointer alignment (94834c4)
  • cuda: include cooperative_groups/reduce.h in gemv_q4k_sm121 (d1dd5aa)
  • cuda: port ztensor-required kernels missing from the fork (.so drift) (46c412c)
  • cuda: port ztensor's flash_decode.cu into zerfoo's kernel fork (3c02533)
  • cuda: remove CWD-relative dlopen candidates for the kernel library (CUDA-1, T141.1) (3f0d2fd)
  • distributed: wire mTLS into worker gRPC server, refuse non-loopback without TLS (d6b2562)
  • gguf: bounds-check tensor offset before mmap slice (F2, T139.2) (ab01d5d)
  • gguf: check-before-multiply on tensor element-count overflow (F1, T139.1) (285fafe)
  • gguf: restore missing math import in loader_mmap.go and split_file.go (ca1eb41)
  • inference: serialize SpeculativeGenerate through the graph mutex (5e409a9)
  • layers: FFN GELU backward always differentiated the SwiGLU branch (dd46acd)
  • scripts: fail loudly on zero matched tests (T135.1) (9c8e4e4)
  • serve: bound error-metric label cardinality via route normalization (9922a0b)
  • serve: close model-delete TOCTOU and WaitGroup-misuse race (CONC-H2) (dcf1c90)
  • serve: validate LoRA adapter name against traversal (SERVE-2) (3ad21c3)
  • training: allocation-stable cached loss seed (non-arena device storage) (42b87a9)
  • training: remove #878 containment gate (T133.4) (9968843)

v1.56.0

Choose a tag to compare

@github-actions github-actions released this 03 Jul 00:17

1.56.0 (2026-07-03)

Features

  • bench: add validate-arm64 Spark pod manifest (T131.1) (f95633a)
  • scripts: -pkgs override for targeted GPU test scoping (T131.3) (2b96e9e)
  • scripts: add -delete mode to dgx-validate for pod cleanup (T131.3) (3135db9)
  • scripts: add dgx-validate.sh one-command GPU validation submit (T131.1) (9b500f6)
  • scripts: committed in-pod validation stage dgx-validate-inpod.sh (T131.3) (518ef61)
  • training: loud-fail gate on capture-replay training pending zerfoo#878 (T129.2) (72cfb48)

Bug Fixes

  • bench: drop unprovisioned hostPath mounts from validate-arm64 pod (T131.3) (2b79ca4)
  • bench: parameterize test scope via TEST_PKGS substitution (T131.3) (697ea89)
  • bench: purego package set for validate-arm64 env (T131.3) (1e82694)
  • bench: single-line bootstrap args -- Spark YAML drops block scalars (T131.3) (14c3141)
  • cuda: skip dlsym invalid-handle test on linux -- handle 0 is RTLD_DEFAULT on glibc (T131.3) (7e29ef6)
  • scripts: dgx-validate diagnosis hardening -- image pre-pull, events on failure, -keep flag (T131.3) (d6e8b0d)
  • scripts: in-pod CGo header paths + QUALITY.md vet policy for purego warnings (T131.3) (7b6daa3)
  • scripts: scope standing gate to purego GPU path; -tags cuda tracked in #921 (T131.3) (87465c5)

v1.55.1

Choose a tag to compare

@github-actions github-actions released this 30 Jun 22:20

1.55.1 (2026-06-30)

Bug Fixes

  • timeseries: seed iTransformer init for deterministic gradient checks (7115733)
  • training: reserve captured-graph arena footprint under capture-replay (ztensor#167) (6cb6c02)

v1.55.0

Choose a tag to compare

@github-actions github-actions released this 21 Jun 00:18

1.55.0 (2026-06-21)

Features

  • regularization: add optional seeded RNG to Dropout and FeatureDropout (ec7dc54)
  • regularization: capture-safe engine-op dropout path (WithEngineDropout) (8cedbc1)

v1.54.0

Choose a tag to compare

@github-actions github-actions released this 18 Jun 03:32

1.54.0 (2026-06-18)

Features

  • bench: bench_gemm + ltx2-fp8-spike manifest (E127/T127.1.0b) (92909be)
  • layers: Conv3d + ConvTranspose3d forward (E127/T127.4.2-4.3) (374b807)
  • layers: GroupNormalization layer + LTX-2 VAE decoder skeleton (E127) (7a5ffc8)

Bug Fixes

  • optimizer: keep AdamW grad-norm reduction on-device (T11.8) (39750c7)
  • parity: build GPU parity container with CGO, no cuda tag (#902) (58731e9)

v1.53.1

Choose a tag to compare

@github-actions github-actions released this 17 Jun 05:45

1.53.1 (2026-06-17)

Bug Fixes

  • training: bf16 device-resident gradient accumulation uses the engine (f8570db)

v1.53.0

Choose a tag to compare

@github-actions github-actions released this 17 Jun 05:14

1.53.0 (2026-06-17)

Features

  • layers/core: use MatMulTransposeA for Linear dW gradient (bf16 GPU path) (a56365d)

v1.52.0

Choose a tag to compare

@github-actions github-actions released this 16 Jun 23:34

1.52.0 (2026-06-16)

Features

  • layers: make Softmax, Linear, Conv1D, VarSelection, CE-loss bf16-instantiable (138c233)

Bug Fixes

  • optimizer: keep AdamW hyperparameters in float64 for reduced-precision T (9083452)