Quants are hard-fought and we keep re-paying for the same lessons. A survey
found quant knowledge scattered across 18 files in four trees, with three
documents having independently discovered and recorded overlapping
"landmines" sections — and one of them now actively misleading.
Adds docs/pfi/model-quantization-playbook.md as the single home for the
TRANSFERABLE lessons, with per-model artifacts demoted to worked examples
that link up to it. Contents:
- scheme decision table, incl. that a literal "W4A8" NVFP4 checkpoint is
unservable on vLLM (two legal activation settings, FP8 is not one)
- the reference mixed-precision recipe and the three parts of it that are
load-bearing and easy to drop
- the recurring landmines, ordered by cost: the loader-class trap
(rediscovered THREE times), the three separate ways to lose the MTP head,
toolchain deadlocks, vision configs, memory/device placement
- pipeline shape: prove targets before spending GPU time; mandatory
post-steps that verify rather than assume
- the acceptance gate, and the three ways measurement has lied to us —
prefix caching faking both speed metrics, prompt_logprobs going uniform
under speculative decoding, and a 0600 .env making compose silently no-op
- hardware/co-residency, including that a SMALLER model can starve its
neighbour because gpu-memory-utilization is a fraction of the whole card
- a superseded-claims table, and measured negatives not to re-chase
The superseded table earns its place immediately: the heretic2 runbook tells
readers to use modelopt because "compressed-tensors can't load the BF16 MTP
head, 0% acceptance". That symptom was real but the cause was not the format
-- it was the missing re:^mtp.* ignore entry. compressed-tensors gives
47.7-83.2% acceptance in production. A fresh session following that doc would
be sent down the modelopt path that current memory calls dependency hell, so
the runbook now carries a stale-warning header pointing here.
Wires discovery: an orientation.md "Where to look for what" row, pointers
from the gen-seat / heretic2 / mistral artifacts, and a CLAUDE.md maintenance
rule so the playbook gets fed instead of going stale -- model-agnostic
lessons land in the playbook, model-specific ones stay put, and a wrong
claim earns a dated superseded row rather than a silent edit.
Motivated by Qwen3.8 having just released: the next model swap will need a
requant, and this is what that session should read first.
Closes the one axis of the original premise left unverified. Measured
cold (cache-busted) on both builds under matching serve configs:
~6.7k-token prompt 3,206 -> 6,334 tok/s prefill (+98%)
~27k-token prompt 2,862 -> 5,085 tok/s prefill (+78%)
TTFT on a ~27k doc 9.43 -> 5.31 s (-44%)
Prefill gains far exceed the +18% decode gain, and that ordering is the
expected one: decode at bs=1 is memory-bandwidth-bound and the weights
are 4-bit under either scheme, so little changes; prefill is
compute-bound, which is where native Blackwell FP4 tensor cores replace
the Marlin dequant-to-BF16 path. The summarizer aliases are the
consumers that feel this.
Adds bench/prefill_bench.py plus the raw JSON. The harness deliberately
uses SystemRandom: a seeded nonce regenerates the previous run's prompts
verbatim, prefix caching then serves them, and the first attempt read
~41k tok/s of cache-hit rather than ~5k of actual prefill.
Re-quantizes the fleet `gen` seat from weight-only NVFP4A16 to a
mixed-precision build: NVFP4 W4A4 for layers 0-55 MLPs, FP8 W8A8 for the
attention projections / linear_attn / lm_head / layers 56-63 MLPs, FP8 KV
cache. Replicates the scheme of unsloth/Qwen3.8-27B-NVFP4 on the
abliterated weights.
The queued task named this "W4A8" (NVFP4 weights + FP8 activations). That
checkpoint cannot be served: vLLM 0.24's compressed-tensors dispatcher
(compressed_tensors.py:704-713) accepts NVFP4 weights with either no input
quantization (W4A16, which forces the Marlin kernel) or NVFP4 input
quantization (W4A4) -- anything else, FP8 included, raises ValueError at
load. CompressedTensorsW4A8Fp8 is INT4 weights gated on an exact-sm90
check, so it is closed on Blackwell twice over. The ~20% intuition was
correct; the scheme name was not. Getting FP8 into the mix has to be done
per-layer-group.
Established the gain before spending GPU time: unsloth's build was already
on-box, so serving it as a probe measured +19.1% over our seat at identical
MTP acceptance -- a kernel-level result, no requant needed to learn it.
Measured, cache-busted, bs=1:
decode 80.12 -> 94.53 tok/s (+18.0%)
MTP acceptance 47.8% -> 47.7% (unchanged)
perplexity (n=6) 6.941 -> 7.059 (+1.7%)
abliteration 4/4 -> 4/4 (preserved)
weights on disk 27.7 -> 22.5 GB (-19%)
Surface test green on the live seat: plain chat, vision, tool calling,
thinking split, 36K-token needle retrieval, streaming. All 7 LiteLLM
aliases verified routing.
GEN_GPU_MEM_UTIL 0.45 -> 0.43: the new weights are 5.2 GB smaller, and at
0.45 the seat absorbed that slack as KV, leaving meromero-charrp 0.18 GiB
short of its budget on the shared GPU0 -- it crash-looped. Handing the
space back leaves gen 422K tokens of KV (1.6x its 262K context) and both
seats co-resident at 89.8/97.9 GB.
Also records two measured negatives so they are not re-chased:
GEN_SPEC_TOKENS is already optimal at 3 (swept 2/3/4/5 -> 77.1/80.1/78.7/
75.9 tok/s), and vLLM's prompt_logprobs are ~uniform while speculative
decoding is on, so perplexity must be measured with spec off.
Pipeline, acceptance harness and raw measurements land in
services/gen-seat-mixed-quant/. Rollback is one .env line; the previous
build is untouched at /tank/aimodels/qwen38-27b-uncensored-nvfp4.
vh