meeting-103 v2.1 clean R1 ✅5/5: PuLID + Redux = pastel 48.5% (R9 best 갱신)

R0 환경 점검:
- VRAM 11.7/12.4GB free (94%) — 해소 확인
- pytorch 2.12.0+cu130 + sageattention venv 검증
- 모델 인벤토리: Flux Dev Q5 GGUF, PuLID, Redux, byeolyi-ref-916 확보
- fp8 다운로드 .part 0 bytes (재개 대기)

R1 clean 5조합 (5/5 success):
- 1-1: PuLID w0.85 mean → pastel 25.7% / black 1.95%
- 1-2: PuLID w0.85 mean delayed_start 0.05 → pastel 10.0% / black 14.0% (어두움)
- 1-3: PuLID w1.0 mean → pastel 27.5% / black 0.79%
- 1-4: PuLID w0.7 + Redux 0.5 → pastel 48.5% / black 0.27% ★ R1 best
- 1-5: 한옥 baseline Flux Dev 768² → pastel 12.9% / black 11.7%

CUDA error 회복:
- 1-2 retry concat → CUDA invalid argument → 후속 1-2~1-5 모두 실패
- ComfyUI venv 재시작 (sage + pytorch 2.12) → 1-2 변형 (delayed_start)로 대체
- 재시도 4/4 모두 success

이전 R9 best (44.5%) → R1-4 (48.5%) = +4 pp 갱신

자산 라이브러리 누적:
- models-performance.json: 7 모델 추적 (flux-dev, pulid-flux, flux-redux 등)
- workflows: winner_auto 1 + partial 2
- prompts/proven: illustration_auto 패턴 3건
- prompts/failed: pastel-low + CUDA error 패턴
- prompts/negative: A_byeolyi 패턴 etc.

검수:
- Vault: D:\Vault\8460s-image-rd\eval\meeting-103-v2.1-clean-grid.html
- Gitea: results/meeting-103-v2.1-clean/round1/

assets-library/models-performance.json

@@ -65,10 +65,10 @@
       },
       "avg_scores": {},
       "best_combinations": [],
-      "verdict": "7라운드 검증 / ComfyUI 재시작 후 100% 성공률 / 검정 0.01-0.68% 일관 / 파스텔 8-42% 변동"
+      "verdict": "파스텔 평균 26.1% (11회) / 검정 평균 0.24%"
     },
     "flux-dev": {
-      "rounds_used": 8,
+      "rounds_used": 22,
       "dimensions": {
         "byeolyi": [],
         "hanja": [],
@@ -78,20 +78,56 @@
         "auto_pastel": [
           20.1,
           29.7,
-          12.3
+          12.3,
+          9.99,
+          27.51,
+          48.5,
+          12.87,
+          25.65,
+          9.99,
+          27.51,
+          48.5,
+          12.87,
+          25.65,
+          9.99,
+          27.51,
+          48.5,
+          12.87
         ],
         "auto_black": [
           3.11,
           0.0,
-          0.16
+          0.16,
+          13.98,
+          0.79,
+          0.27,
+          11.67,
+          1.95,
+          13.98,
+          0.79,
+          0.27,
+          11.67,
+          1.95,
+          13.98,
+          0.79,
+          0.27,
+          11.67
         ]
       },
       "avg_scores": {},
-      "best_combinations": [],
-      "verdict": "Stage 3 Refiner (img2img denoise 0.3) 정상 작동 / 단독 워크플로우는 OOM 빈번"
+      "best_combinations": [
+        {
+          "round": "R1-clean",
+          "id": "1-4",
+          "variant": "PuLID 0.7 + Redux 0.5",
+          "pastel_pct": 48.5,
+          "black_pct": 0.27
+        }
+      ],
+      "verdict": "파스텔 평균 24.1% (17회) / 검정 평균 5.14%"
     },
     "pulid-flux": {
-      "rounds_used": 9,
+      "rounds_used": 20,
       "dimensions": {
         "byeolyi": [],
         "hanja": [],
@@ -101,12 +137,34 @@
         "auto_pastel": [
           30.2,
           33.7,
-          44.5
+          44.5,
+          9.99,
+          27.51,
+          48.5,
+          25.65,
+          9.99,
+          27.51,
+          48.5,
+          25.65,
+          9.99,
+          27.51,
+          48.5
         ],
         "auto_black": [
           0.73,
           2.51,
-          1.11
+          1.11,
+          13.98,
+          0.79,
+          0.27,
+          1.95,
+          13.98,
+          0.79,
+          0.27,
+          1.95,
+          13.98,
+          0.79,
+          0.27
         ]
       },
       "avg_scores": {},
@@ -119,9 +177,16 @@
           "black_pct": 1.11,
           "pastel_pct": 44.5,
           "note": "전체 라운드 best pastel 1위"
+        },
+        {
+          "round": "R1-clean",
+          "id": "1-4",
+          "variant": "PuLID 0.7 + Redux 0.5",
+          "pastel_pct": 48.5,
+          "black_pct": 0.27
         }
       ],
-      "verdict": "weight 1.0 + mean fusion = 별이 ID lock + 파스텔 44.5% (R1-1 Pony 42.5% 추월)"
+      "verdict": "파스텔 평균 29.8% (14회) / 검정 평균 3.81%"
     },
     "qwen-image-2512": {
       "rounds_used": 1,
@@ -150,7 +215,7 @@
       },
       "avg_scores": {},
       "best_combinations": [],
-      "verdict": "단일 모델 성공 / A 별이 검정 13.92% 한계 / B 한옥 안정 36.4% 파스텔"
+      "verdict": "파스텔 평균 30.2% (2회) / 검정 평균 7.41%"
     },
     "qwen-edit-2511": {
       "rounds_used": 8,
@@ -158,6 +223,40 @@
       "avg_scores": {},
       "best_combinations": [],
       "verdict": "Stage 2 인페인트 6차 실패 — Qwen 계열 GGUF Q4 환경 호환성 결함 확정"
+    },
+    "flux-redux": {
+      "rounds_used": 3,
+      "dimensions": {
+        "byeolyi": [],
+        "hanja": [],
+        "hanok": [],
+        "illustration": [],
+        "korean_traditional": [],
+        "auto_pastel": [
+          48.5,
+          48.5,
+          48.5
+        ],
+        "auto_black": [
+          0.27,
+          0.27,
+          0.27
+        ]
+      },
+      "avg_scores": {},
+      "best_combinations": [
+        {
+          "round": "R1-clean",
+          "id": "1-4",
+          "variant": "PuLID 0.7 + Redux 0.5",
+          "pastel_pct": 48.5,
+          "black_pct": 0.27
+        }
+      ],
+      "verdict": "파스텔 평균 48.5% (3회) / 검정 평균 0.27%"
     }
-  }
+  },
+  "_processed_rounds": [
+    "R1-clean"
+  ]
 }