From 1ae73db0846242bc693a7bf30c065667bca82937 Mon Sep 17 00:00:00 2001 From: Frank Date: Sun, 20 Sep 2026 22:06:26 -0700 Subject: [PATCH] =?UTF-8?q?test(frank-medium):=20A/B=20=E2=80=94=20flip=20?= =?UTF-8?q?to=20DFlash2=20drafter=20(SPEC=5FMETHOD=3Ddflash,=20SPEC=3D7)?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- .../frank-medium/quadlet/frank-medium.container | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) diff --git a/dankserver/frank-medium/quadlet/frank-medium.container b/dankserver/frank-medium/quadlet/frank-medium.container index e8704ca..a74af63 100644 --- a/dankserver/frank-medium/quadlet/frank-medium.container +++ b/dankserver/frank-medium/quadlet/frank-medium.container @@ -110,11 +110,12 @@ Environment=MAXLEN=262144 # 400 here. Vision traffic routes to frank-small on dankclaw (harness # modelRoles vision -> frank-small). Environment=GPU_UTIL=0.98 -# Speculative decoding: MTP self-speculation at depth 8 (the model's own -# single-layer MTP head, applied autoregressively). The documented plain-FP8 -# path (upstream docker-compose) uses num_speculative_tokens=8. -Environment=SPEC_METHOD=mtp -Environment=SPEC=8 +# Speculative decoding: A/B test — DFlash2 external drafter at depth 7 (the +# documented DFlash2 depth), to compare against the MTP self-speculation +# baseline (depth 8, mean accept 5.66/8). The DFlash2 drafter is a separate +# quantization-agnostic checkpoint at /models/Qwen3.8-27B-DFlash2-FP8. +Environment=SPEC_METHOD=dflash +Environment=SPEC=7 # KV cache: 0 = profile the shape (the documented plain-FP8 path's default; # the upstream docker-compose leaves the pin commented out). The measured pin # for this shape is 15989735424 (14.89 GiB, upstream, 2x R9700 32GB); after