6623ea4172
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/
37 lines
1.6 KiB
Markdown
37 lines
1.6 KiB
Markdown
# R0 — meeting-103 v2.1 clean 환경 점검
|
|
|
|
**일자**: 2026-05-19 23:00 KST
|
|
**목적**: VRAM 해소 확인 + 모델 인벤토리 + clean 폴더 구조 생성
|
|
|
|
## VRAM 상태 (해소 검증)
|
|
- RTX 4070 SUPER 12GB
|
|
- vram_free: 11.7 GB / 12.4 GB total (94% free)
|
|
- pytorch 2.12.0+cu130, sageattention OK
|
|
- comfyui 0.21.1 venv
|
|
|
|
## 모델 인벤토리 (gpu:/opt/ComfyUI/models)
|
|
- diffusion_models: Qwen-Edit-2511 Q4_K_S GGUF (12G), Qwen-2512 Q4_K_S GGUF (12G), Flux Fill Q4_K_S GGUF (6.4G), Flux Kontext Q4_K_S GGUF (6.4G), Qwen fp8 .part (0 bytes 미완료)
|
|
- unet: Flux Dev Q5_K_S GGUF (7.8G), Flux Schnell Q4_K_S GGUF (6.4G)
|
|
- checkpoints: SDXL base, Pony V6 XL, NoobAI-XL v1.0
|
|
- pulid: pulid_flux_v0.9.1.safetensors ✓
|
|
- style_models: flux1-redux-dev.safetensors ✓
|
|
- clip: t5xxl_fp8, clip_l, qwen_2.5_vl_7b_fp8 ✓
|
|
- clip_vision: sigclip_vision_patch14_384 ✓ (Redux 용)
|
|
- vae: ae (flux), qwen_image_vae ✓
|
|
- text_encoders: qwen_2.5_vl_7b_fp8 ✓
|
|
- controlnet: flux1-dev-controlnet-union-pro, flux-canny-controlnet-v3
|
|
- loras: 비어 있음 (byeolyi LoRA 미존재 → PuLID + byeolyi-ref-916.png 으로 별이 reference 제공)
|
|
|
|
## 별이 reference (PuLID input)
|
|
- /opt/ComfyUI/input/byeolyi-ref-768x1344.png ✓
|
|
- /opt/ComfyUI/input/byeolyi-ref-916.png ✓
|
|
- /opt/ComfyUI/input/byeolyi-default.png
|
|
|
|
## 디스크
|
|
- /opt: 593G/915G (276G free) — 충분
|
|
|
|
## 회피 검증 (failed-patterns)
|
|
- Qwen GGUF Q4 = 6회 실패 confirmed → fp8/bf16 우선 (fp8 다운로드 재개 대기)
|
|
- Qwen-family GGUF Q4 = 환경 비호환 confirmed
|
|
- PuLID concat/max = R2 실패 후 재시도 OK (VRAM 해소 후)
|