app.py

# app.py — InstantID × Beautiful Realistic Asians v7（ZeroGPU / ControlNetMediaPipeFace）
# 2025-06-22 版

##############################################################################
# 0. 旧 API → 新 API 互換パッチ（必ず diffusers import の前に置く）
##############################################################################
from huggingface_hub import hf_hub_download
import huggingface_hub as _hf_hub

# diffusers-0.27 は cached_download() を呼び出すため、HF-Hub ≥0.28 でも使えるように注入
if not hasattr(_hf_hub, "cached_download"):
    _hf_hub.cached_download = hf_hub_download  # :contentReference[oaicite:1]{index=1}

##############################################################################
# 1. 標準 & 外部ライブラリ
##############################################################################
import os, io, base64, subprocess, traceback
from pathlib import Path
from typing import Optional

import numpy as np
import torch
import gradio as gr
import spaces
from fastapi import FastAPI, UploadFile, File, Form, HTTPException
from PIL import Image

from diffusers import (
    StableDiffusionControlNetPipeline,
    ControlNetModel,
    DPMSolverMultistepScheduler,
    AutoencoderKL,
)
from diffusers.loaders import AttnProcsLayers
from insightface.app import FaceAnalysis
from basicsr.utils.download_util import load_file_from_url
from realesrgan import RealESRGANer

##############################################################################
# 2. キャッシュ & 永続パス
##############################################################################
PERSIST_BASE = Path("/data")
CACHE_ROOT = (
    PERSIST_BASE / "instantid_cache"
    if PERSIST_BASE.exists() and os.access(PERSIST_BASE, os.W_OK)
    else Path.home() / ".cache" / "instantid_cache"
)
MODELS_DIR = CACHE_ROOT / "models"
LORA_DIR   = CACHE_ROOT / "lora"
UPSCALE_DIR = CACHE_ROOT / "realesrgan"
for p in (MODELS_DIR, LORA_DIR, UPSCALE_DIR):
    p.mkdir(parents=True, exist_ok=True)

##############################################################################
# 3. モデル識別子 & ファイル名
##############################################################################
# すべて HF Hub 側にバイナリがあるため、curl ではなく hf_hub_download() を推奨
BRA_REPO    = "i0switch-assets/Beautiful_Realistic_Asians_v7"
BRA_FILE    = "beautiful_realistic_asians_v7_fp16.safetensors"

IP_REPO     = "h94/IP-Adapter"
IP_FILE_BIN = "ip-adapter-plus-face_sd15.bin"       # Git LFS バイナリ :contentReference[oaicite:2]{index=2}

IP_LORA_REPO = "h94/IP-Adapter-FaceID"
IP_FILE_LORA = "ip-adapter-faceid-plusv2_sd15_lora.safetensors"  # Git LFS バイナリ

CN_REPO   = "CrucibleAI/ControlNetMediaPipeFace"  # 公開・無認証で DL 可 :contentReference[oaicite:3]{index=3}
CN_FOLDER = "diffusion_sd15"                      # SD-1.5 用フォルダ :contentReference[oaicite:4]{index=4}

REALESRGAN_REPO = "aimagelab/realesrgan"
REALESRGAN_FILE = "RealESRGAN_x4plus.pth"

##############################################################################
# 4. ダウンローダ（HF Hub 優先）
##############################################################################
def dl_hf(repo: str, filename: str, subfolder: Optional[str] = None) -> Path:
    """HF Hub から大容量バイナリを安全に取得（Git LFS ポインタ問題を回避）"""
    return Path(
        hf_hub_download(
            repo_id=repo,
            filename=filename,
            subfolder=subfolder,
            cache_dir=str(MODELS_DIR),
        )
    )

def dl_http(url: str, dst: Path):
    """小さなファイルのみ curl で取得（retry 付）"""
    if dst.exists():
        return dst
    for _ in range(2):
        try:
            subprocess.check_call(["curl", "-L", "-o", str(dst), url])
            return dst
        except subprocess.CalledProcessError:
            pass
    load_file_from_url(url=url, model_dir=str(dst.parent), file_name=dst.name)
    return dst

##############################################################################
# 5. グローバル変数（lazy-load）
##############################################################################
pipe: Optional[StableDiffusionControlNetPipeline] = None
face_analyser: Optional[FaceAnalysis] = None
upsampler: Optional[RealESRGANer] = None

##############################################################################
# 6. パイプライン初期化
##############################################################################
def initialize_pipelines():
    global pipe, face_analyser, upsampler
    if pipe is not None:
        return

    print("[INIT] Downloading model assets …")

    # 6-1 主要モデル
    bra_ckpt = dl_hf(BRA_REPO, BRA_FILE)
    ip_bin   = dl_hf(IP_REPO,  IP_FILE_BIN)
    ip_lora  = dl_hf(IP_LORA_REPO, IP_FILE_LORA)
    cn_model = ControlNetModel.from_pretrained(
        CN_REPO, subfolder=CN_FOLDER, torch_dtype=torch.float16, cache_dir=str(MODELS_DIR)
    )

    # 6-2 Diffusers パイプライン
    pipe_tmp = StableDiffusionControlNetPipeline.from_pretrained(
        "runwayml/stable-diffusion-v1-5",
        controlnet=cn_model,
        vae=AutoencoderKL.from_pretrained(
            "stabilityai/sd-vae-ft-mse", torch_dtype=torch.float16
        ),
        torch_dtype=torch.float16,
        cache_dir=str(MODELS_DIR),
        safety_checker=None,
    )
    pipe_tmp.scheduler = DPMSolverMultistepScheduler.from_pretrained(
        "runwayml/stable-diffusion-v1-5",
        subfolder="scheduler",
        cache_dir=str(MODELS_DIR),
    )

    # 6-3 IP-Adapter ロード（必須 3 引数） :contentReference[oaicite:5]{index=5}
    pipe_tmp.load_ip_adapter(
        str(ip_bin.parent),   # repo_or_path
        "",                   # subfolder（直下なので空文字）
        ip_bin.name           # weight_name
    )
    AttnProcsLayers(pipe_tmp.unet.attn_processors).load_lora_weights(
        ip_lora, adapter_name="ip_faceid", safe_load=True
    )
    pipe_tmp.set_adapters(["ip_faceid"], adapter_weights=[0.6])
    pipe_tmp.to("cuda")

    pipe = pipe_tmp

    # 6-4 InsightFace
    face_analyser = FaceAnalysis(
        name="buffalo_l", root=str(MODELS_DIR), providers=["CUDAExecutionProvider"]
    )
    face_analyser.prepare(ctx_id=0, det_size=(640, 640))

    # 6-5 Real-ESRGAN
    re_ckpt = dl_hf(REALESRGAN_REPO, REALESRGAN_FILE)
    upsampler = RealESRGANer(
        scale=4,
        model_path=str(re_ckpt),
        half=True,
        tile=512, tile_pad=10, pre_pad=0, gpu_id=0
    )

    print("[INIT] Pipelines ready.")

##############################################################################
# 7. プロンプトテンプレ
##############################################################################
BASE_PROMPT = (
    "(masterpiece:1.2), best quality, ultra-realistic, RAW photo, 8k, "
    "cinematic lighting, textured skin, "
)
NEG_PROMPT = (
    "verybadimagenegative_v1.3, ng_deepnegative_v1_75t, "
    "(worst quality:2), (low quality:2), lowres, blurry, bad anatomy, "
    "bad hands, extra digits, watermark, signature"
)

##############################################################################
# 8. 生成コア（GPU アタッチ）
##############################################################################
@spaces.GPU(duration=60)  # ZeroGPU で 60 s まで実行可 :contentReference[oaicite:6]{index=6}
def generate_core(
    face_img: Image.Image,
    subject: str,
    add_prompt: str = "",
    add_neg: str = "",
    cfg: float = 7.5,
    ip_scale: float = 0.6,
    steps: int = 30,
    w: int = 768,
    h: int = 768,
    upscale: bool = False,
    up_factor: int = 4,
    progress: gr.Progress = gr.Progress(track_tqdm=True),
):
    try:
        if pipe is None:
            initialize_pipelines()

        if len(face_analyser.get(np.array(face_img))) == 0:
            raise ValueError("顔が検出できません。別の画像でお試しください。")

        pipe.set_adapters(["ip_faceid"], adapter_weights=[ip_scale])

        prompt   = BASE_PROMPT + subject + ", " + add_prompt
        negative = NEG_PROMPT  + ", " + add_neg

        result = pipe(
            prompt=prompt,
            negative_prompt=negative,
            num_inference_steps=int(steps),
            guidance_scale=float(cfg),
            image=face_img,
            control_image=None,
            width=int(w), height=int(h),
        ).images[0]

        if upscale:
            upsampler.scale = 4 if up_factor == 4 else 8
            result, _ = upsampler.enhance(np.array(result))
            result = Image.fromarray(result)

        return result

    except Exception as e:
        traceback.print_exc()
        raise e

##############################################################################
# 9. Gradio UI
##############################################################################
with gr.Blocks(title="InstantID × BRA v7 (ZeroGPU)") as demo:
    gr.Markdown("## InstantID × Beautiful Realistic Asians v7")
    with gr.Row():
        face_img = gr.Image(type="pil", label="Face ID", sources=["upload"])
        subject  = gr.Textbox(label="被写体説明（例: 30代日本人女性、黒髪セミロング）", interactive=True)
    add_prompt = gr.Textbox(label="追加プロンプト", interactive=True)
    add_neg    = gr.Textbox(label="追加ネガティブ", interactive=True)
    with gr.Row():
        cfg      = gr.Slider(1, 20, value=7.5, step=0.5, label="CFG Scale")
        ip_scale = gr.Slider(0.1, 1.0, value=0.6, step=0.05, label="IP-Adapter Weight")
    with gr.Row():
        steps = gr.Slider(10, 50, value=30, step=1, label="Steps")
        w     = gr.Slider(512, 1024, value=768, step=64, label="Width")
        h     = gr.Slider(512, 1024, value=768, step=64, label="Height")
    with gr.Row():
        upscale   = gr.Checkbox(label="Real-ESRGAN Upscale", value=False)
        up_factor = gr.Radio([4, 8], value=4, label="Upscale Factor")
    run_btn   = gr.Button("Generate")
    output_im = gr.Image(type="pil", label="Result")

    run_btn.click(
        fn=generate_core,
        inputs=[face_img, subject, add_prompt, add_neg,
                cfg, ip_scale, steps, w, h, upscale, up_factor],
        outputs=output_im, show_progress=True
    )

##############################################################################
# 10. FastAPI REST
##############################################################################
app = FastAPI()

@app.post("/api/generate")
async def api_generate(
    subject: str = Form(...),
    cfg: float = Form(7.5),
    steps: int = Form(30),
    ip_scale: float = Form(0.6),
    w: int = Form(768),
    h: int = Form(768),
    file: UploadFile = File(...),
):
    try:
        img = Image.open(io.BytesIO(await file.read())).convert("RGB")  # noqa
        res = generate_core(img, subject, "", "", cfg, ip_scale, steps, w, h, False, 4)
        buf = io.BytesIO(); res.save(buf, format="PNG")
        return {"image": "data:image/png;base64," + base64.b64encode(buf.getvalue()).decode()}
    except Exception as e:
        traceback.print_exc()
        raise HTTPException(status_code=500, detail=str(e))

##############################################################################
# 11. Launch（Gradio が自動で Uvicorn を起動）
##############################################################################
demo.queue(default_concurrency_limit=2).launch(share=False)  # :contentReference[oaicite:7]{index=7}