3 people 3c84fe5a1f Promote vllm/qwen-35b-a3b-dual-nvfp4-fast: experimental -> caveats (#676) ...

Full gate complete on the reference 2x3090: verify-stress 8/8 (NIAH
clean to 240,635 tok = 91% of 262K, ceiling margin 1,801 MB, zero drift
across the ladder) + soak-continuous PASS (0 errors, 0/25 silent-empty,
0 VRAM growth, p50 215.9, 100% retention), joining the existing bench
(179.5/179.4 decode) and 8-pack (103/150 tie) results.

Caveats (why not full Production): streaming tool-calls + thinking-on
hit finish=length (known family class, verify-full check 6); native-FP4
quality remains unvalidated (all numbers are the Ampere W4A16 bound).


Claude-Session: https://claude.ai/code/session_01EfF565T9eSLaqGzidyJ1Pm

Co-authored-by: noonghunna <[email protected]>
Co-authored-by: Claude Fable 5 <[email protected]>

2026-07-11 17:55:41 +05:00

..

calibration

Wire gemma-4-26b-a4b AWQ on stock vLLM v0.22.0; retire AutoRound (#326)

2026-06-06 00:46:13 +00:00

drafters

Add Tess-4-27B to the catalog (llama.cpp dual, external MTP, 262K) (#661)

2026-07-09 21:45:38 +05:00

engines

Add Tess-4-27B to the catalog (llama.cpp dual, external MTP, 262K) (#661)

2026-07-09 21:45:38 +05:00

hardware

Phase 2: PRO 6000 + DGX Spark envelopes, fix Blackwell detector

2026-07-05 06:16:59 +00:00

models

Add vllm/qwen-35b-a3b-dual-nvfp4-fast: Ampere-validated unsloth NVFP4 MoE slug (#675)

2026-07-11 17:10:34 +05:00

workloads

feat(profiles): ship v0.7.0 data layer

2026-05-14 14:46:42 +00:00

__init__.py

feat(profiles): ship v0.7.0 data layer

2026-05-14 14:46:42 +00:00

arch_patches.yml

patches: de-register vllm-marlin-pad (merged upstream, native in v0.24.0)

2026-07-02 10:03:44 +00:00

baselines.yml

preflight: warn on single-card GPU_MEMORY_UTILIZATION override above default (#641)

2026-07-08 23:18:33 +05:00

booter.py

v0.8.0 [F] F3: G6-A 3-part additive [E] touch (pt1.predicted_b_breakdown, pt3.failure_log_excerpt+actual, container-log capture) + §6.1 Tier-1 (CONTRACT-2)

2026-05-17 04:19:53 +00:00

calibration_seed.yml

Add v0.8 Phase A patch attribution data

2026-05-16 05:13:00 +00:00

canonical_scenarios.py

feat(compat): add profile validator and estate self-test

2026-05-14 15:25:28 +00:00

capture.py

feat(pull): v0.8.2 STEP V1 — capture-on-hard-block pt1-gate emitter + BaseCaptureBundle protocol lift

2026-05-18 16:43:28 +00:00

classifier.py

feat(pull): v0.8.2 STEP V1 — capture-on-hard-block pt1-gate emitter + BaseCaptureBundle protocol lift

2026-05-18 16:43:28 +00:00

compat.py

NVFP4: unhide on fallback-capable hardware (fallback_sm + hw_fallback badge) (#670)

2026-07-11 05:48:00 +05:00

compose_registry.py

Promote vllm/qwen-35b-a3b-dual-nvfp4-fast: experimental -> caveats (#676)

2026-07-11 17:55:41 +05:00

dedup.py

feat(pull): v0.8.2 STEP V1 — capture-on-hard-block pt1-gate emitter + BaseCaptureBundle protocol lift

2026-05-18 16:43:28 +00:00

deriver.py

deriver: resolve HF_HOME from MODEL_DIR so bare pull.sh lands on the model disk (#646)

2026-07-09 04:40:20 +05:00

diagnose_profile_cli.py

dgemma: adopt official vllm/vllm-openai:gemma image (drop 120-file sideload)

2026-06-11 05:31:36 +00:00

downloader.py

c3+pull: detect in-progress downloads + per-repo lock (#617) (#644)

2026-07-09 02:53:50 +05:00

einput.py

v0.8.0 [E] E1: generate_from_profile + derived-vllm template + EInput + CONTRACT-5 gate

2026-05-17 00:12:26 +00:00

envelopes.yml

Phase 2: PRO 6000 + DGX Spark envelopes, fix Blackwell detector

2026-07-05 06:16:59 +00:00

estate_cli.py

refactor(pods): rename cluster → pod (#610) + heterogeneous-rig guidance

2026-07-07 02:10:17 +00:00

failure_fingerprints.yml

feat(pull): v0.8.2 STEP V1 — capture-on-hard-block pt1-gate emitter + BaseCaptureBundle protocol lift

2026-05-18 16:43:28 +00:00

gates.py

NVFP4: unhide on fallback-capable hardware (fallback_sm + hw_fallback badge) (#670)

2026-07-11 05:48:00 +05:00

hwdetect.py

feat(pull): v0.8.2 STEP V4 — optional whichllm hw-detect subprocess (CONTRACT-3, hw-detect-only)

2026-05-18 19:29:43 +00:00

launch_compat.py

fix(nvfp4): disable deepgemm on consumer cards

2026-07-07 22:05:09 +00:00

loop_input.py

feat(pull): v0.8.2 STEP V1 — capture-on-hard-block pt1-gate emitter + BaseCaptureBundle protocol lift

2026-05-18 16:43:28 +00:00

measurement_record.py

catalog-baselines slice 2c: canonical two-depth prefill/TTFT probe + anchor calibration

2026-07-04 16:22:48 +00:00

patch_attribution.py

feat(pull): v0.8.2 STEP V3 — arch-registry expansion + chat-template attribution/drift_guard

2026-05-18 18:29:58 +00:00

patches.yml

patches: de-register vllm-marlin-pad (merged upstream, native in v0.24.0)

2026-07-02 10:03:44 +00:00

profile_runtime.yml

Pull gate: actionable messages for uncurated derives of curated models (#424)

2026-06-18 01:10:01 +05:00

pull.py

feat(byo): apply-swap — download + serve a Route-C fine-tune (#628)

2026-07-08 05:50:16 +05:00

submit_pull.py

fix(pull): v0.8.2 STEP V2 — gh-less issue body must not carry the absolute capture path

2026-05-18 17:49:00 +00:00

swap_apply.py

fix(c3): route-C swap compose — absolute mounts + runtime-dir location (#655)

2026-07-09 15:39:54 +05:00

trust_pipeline.py

v0.8.0 [review] pre-tag fixes: scrub internal-path leaks from shipped source + make .pull-captures-corpus tests CI-safe (skip-when-absent)

2026-05-17 12:01:25 +00:00

weights.py

feat(qwen-moe): NVFP4 35B-A3B community slugs — the unified-memory pairing

2026-07-06 22:10:21 +00:00