From fbe8c6f1e9cb409a6ca3eaef82599d8d11595e52 Mon Sep 17 00:00:00 2001
From: Mikei386 <44135113+Mikei386@users.noreply.github.com>
Date: Sun, 20 Sep 2026 23:23:57 +0200
Subject: [PATCH] Use validated Medium MTP2 and microbatch256 combination

---
 compose.yaml                    | 4 ++--
 config/profile-matrix.json      | 2 +-
 docs/STANDARD_PROFILE_MATRIX.md | 2 +-
 3 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/compose.yaml b/compose.yaml
index 7019b60..27a021b 100644
--- a/compose.yaml
+++ b/compose.yaml
@@ -206,7 +206,7 @@ services:
       - --batch-size
       - "${MEDIUM_BATCH_SIZE:-2048}"
       - --ubatch-size
-      - "${MEDIUM_UBATCH_SIZE:-512}"
+      - "${MEDIUM_UBATCH_SIZE:-256}"
       - --parallel
       - "${MEDIUM_PARALLEL_SLOTS:-1}"
       - --kv-unified
@@ -247,7 +247,7 @@ services:
       - --spec-type
       - draft-mtp
       - --spec-draft-n-max
-      - "3"
+      - "2"
       - --spec-draft-type-k
       - f16
       - --spec-draft-type-v
diff --git a/config/profile-matrix.json b/config/profile-matrix.json
index a39e2c3..9d0af93 100644
--- a/config/profile-matrix.json
+++ b/config/profile-matrix.json
@@ -24,7 +24,7 @@
       "model_family": "Qwen3.8-27B IQ4 XS Pure",
       "gpu_split": "85:15",
       "vision": true,
-      "mtp": 3,
+      "mtp": 2,
       "description": "Ausgewogenes Standardprofil für Alltag und lange agentische Aufgaben."
     },
     {
diff --git a/docs/STANDARD_PROFILE_MATRIX.md b/docs/STANDARD_PROFILE_MATRIX.md
index 0ed6dda..27f9630 100644
--- a/docs/STANDARD_PROFILE_MATRIX.md
+++ b/docs/STANDARD_PROFILE_MATRIX.md
@@ -7,7 +7,7 @@ Standardprofil: **medium** · globales Ausgabelimit: **8192 Token**
 | Profil | API-Alias | Gesamtkontext | Slots | Modell | GPU-Verteilung | Vision | MTP |
 |---|---|---:|---:|---|---|---|---:|
 | fast | `qwen-fast` | 76,800 | 1 | Qwen3.8-27B IQ4 Mix | 5080 only | ja | 2 |
-| medium | `qwen-medium` | 160,000 | 1 | Qwen3.8-27B IQ4 XS Pure | 85:15 | ja | 3 |
+| medium | `qwen-medium` | 160,000 | 1 | Qwen3.8-27B IQ4 XS Pure | 85:15 | ja | 2 |
 | large | `qwen-large` | 192,000 | 1 | Qwen3.8-27B IQ4 XS Pure | 86:14 | ja | 2 |
 | ultra | `qwen-ultra` | 262,144 | 1 | Qwen3.8-27B IQ4 XS Pure | 80:20 | nein | 2 |
 | uncensored | `qwen-uncensored` | 80,000 | 1 | Qwen3.8-27B Abliterated Q4_K_M | 90:10 | ja | 2 |