Upload folder using huggingface_hub

tools/apply_preset.py

@@ -0,0 +1,119 @@
+#!/usr/bin/env python
+# -*- coding: utf-8 -*-
+"""
+配置预设应用工具
+快速切换不同的音质优化配置
+"""
+import json
+import shutil
+from pathlib import Path
+
+PRESETS_DIR = Path("configs/presets")
+CONFIG_FILE = Path("configs/config.json")
+BACKUP_FILE = Path("configs/config.backup.json")
+
+PRESETS = {
+    "1": "balanced.json",
+    "2": "clarity_priority.json",
+    "3": "timbre_priority.json"
+}
+
+def load_json(path: Path) -> dict:
+    """加载JSON文件"""
+    with open(path, "r", encoding="utf-8") as f:
+        return json.load(f)
+
+def save_json(path: Path, data: dict):
+    """保存JSON文件"""
+    with open(path, "w", encoding="utf-8") as f:
+        json.dump(data, f, indent=2, ensure_ascii=False)
+
+def backup_config():
+    """备份当前配置"""
+    if CONFIG_FILE.exists():
+        shutil.copy(CONFIG_FILE, BACKUP_FILE)
+        print(f"✓ 已备份当前配置到: {BACKUP_FILE}")
+
+def restore_config():
+    """恢复备份配置"""
+    if BACKUP_FILE.exists():
+        shutil.copy(BACKUP_FILE, CONFIG_FILE)
+        print(f"✓ 已恢复配置从: {BACKUP_FILE}")
+    else:
+        print("✗ 未找到备份文件")
+
+def apply_preset(preset_name: str):
+    """应用预设配置"""
+    preset_path = PRESETS_DIR / preset_name
+
+    if not preset_path.exists():
+        print(f"✗ 预设文件不存在: {preset_path}")
+        return
+
+    # 备份当前配置
+    backup_config()
+
+    # 加载预设和当前配置
+    preset = load_json(preset_path)
+    config = load_json(CONFIG_FILE)
+
+    # 合并配置 (只更新 cover 部分)
+    if "cover" in preset:
+        config["cover"].update(preset["cover"])
+
+    # 保存
+    save_json(CONFIG_FILE, config)
+
+    print(f"\n✓ 已应用预设: {preset.get('name', preset_name)}")
+    print(f"  描述: {preset.get('description', 'N/A')}")
+    print(f"\n主要参数:")
+    print(f"  - index_rate: {config['cover']['index_rate']}")
+    print(f"  - protect: {config['cover']['protect']}")
+    print(f"  - rms_mix_rate: {config['cover']['rms_mix_rate']}")
+    print(f"  - filter_radius: {config['cover']['filter_radius']}")
+    print(f"  - f0_stabilize: {config['cover']['f0_stabilize']}")
+
+def show_menu():
+    """显示菜单"""
+    print("\n" + "="*60)
+    print("RVC 音质优化配置预设")
+    print("="*60)
+    print("\n可用预设:")
+    print("  1. 平衡型配置 (推荐)")
+    print("     - 适合大多数歌曲")
+    print("     - 在音色转换和清晰度之间平衡")
+    print()
+    print("  2. 清晰度优先配置")
+    print("     - 减少伪影和失真")
+    print("     - 适合复杂歌曲和高音多的情况")
+    print("     - 保留更多源音频特征")
+    print()
+    print("  3. 音色优先配置")
+    print("     - 彻底的音色转换")
+    print("     - 适合音色特征明显的角色")
+    print("     - 可能有轻微口齿模糊")
+    print()
+    print("  r. 恢复备份配置")
+    print("  q. 退出")
+    print("="*60)
+
+def main():
+    """主函数"""
+    while True:
+        show_menu()
+        choice = input("\n请选择 (1-3/r/q): ").strip().lower()
+
+        if choice == "q":
+            print("\n再见!")
+            break
+        elif choice == "r":
+            restore_config()
+        elif choice in PRESETS:
+            apply_preset(PRESETS[choice])
+        else:
+            print("\n✗ 无效选择，请重试")
+
+        input("\n按回车继续...")
+
+if __name__ == "__main__":
+    main()

tools/character_models.py

@@ -0,0 +1,1792 @@
+# -*- coding: utf-8 -*-
+"""
+角色模型管理 - 从 HuggingFace 下载 RVC 角色模型
+"""
+import os
+import json
+import re
+import zipfile
+import shutil
+from pathlib import Path
+from typing import Optional, List, Dict, Callable
+
+try:
+    from huggingface_hub import hf_hub_download, list_repo_files
+    HF_AVAILABLE = True
+except ImportError:
+    HF_AVAILABLE = False
+
+
+# HuggingFace 仓库配置
+HF_REPO_ID = "trioskosmos/rvc_models"
+_VERSION_NOTE_CACHE: Dict[str, Optional[str]] = {}
+_VERSION_NOTE_CACHE_LOADED = False
+
+
+def _get_hf_token() -> Optional[str]:
+    """获取 HuggingFace Token（支持 HF_TOKEN / HUGGINGFACE_HUB_TOKEN / HUGGINGFACE_TOKEN）"""
+    return (
+        os.environ.get("HF_TOKEN")
+        or os.environ.get("HUGGINGFACE_HUB_TOKEN")
+        or os.environ.get("HUGGINGFACE_TOKEN")
+    )
+
+# 作品归类（用于 UI 分类筛选）
+SERIES_ALIASES = {
+    "Love Live!": "Love Live!",
+    "Love Live! Sunshine!!": "Love Live!",
+    "Love Live! Superstar!!": "Love Live!",
+    "Love Live! 虹咲学园": "Love Live!",
+    "Love Live! 虹咲学園": "Love Live!",
+    "Love Live! 莲之空女学院学园偶像俱乐部": "Love Live!",
+    "Love Live! Sunshine!! / 幻日夜羽": "Love Live!",
+    "Hololive Japan": "Hololive",
+    "Hololive English": "Hololive",
+    "Hololive Indonesia": "Hololive",
+    "崩坏：星穹铁道": "崩坏系列",
+    "崩坏3rd": "崩坏系列",
+    "偶像大师 灰姑娘女孩": "偶像大师",
+}
+
+
+def normalize_series(source: str) -> str:
+    """将来源归类到系列"""
+    if not source:
+        return "未知"
+    for key, series in SERIES_ALIASES.items():
+        if source.startswith(key):
+            return series
+    return source
+
+
+def _get_display_name(info: Dict, fallback: str) -> str:
+    """拼接中文名 / 英文名 / 日文名用于展示"""
+    zh_name = info.get("zh_name") or info.get("description") or fallback
+    en_name = info.get("en_name")
+    jp_name = info.get("jp_name")
+    parts = [zh_name]
+    if en_name and en_name != zh_name:
+        parts.append(en_name)
+    if jp_name and jp_name != zh_name and jp_name != en_name:
+        parts.append(jp_name)
+    display = " / ".join(parts)
+    variant = info.get("variant")
+    if variant:
+        display = f"{display} - {variant}"
+    variant_note = _get_version_note(fallback, info)
+    if variant_note:
+        display = f"{display} ({variant_note})"
+    return display
+
+
+def _find_index_file(pth_file: Path) -> Optional[Path]:
+    """尝试找到对应的索引文件"""
+    candidate = pth_file.with_suffix(".index")
+    if candidate.exists():
+        return candidate
+
+    index_files = list(pth_file.parent.glob("*.index"))
+    if not index_files:
+        return None
+
+    for idx in pth_file.parent.glob("*.index"):
+        if idx.stem.lower() == pth_file.stem.lower():
+            return idx
+
+    if len(index_files) == 1:
+        return index_files[0]
+
+    def _normalize_name(text: str) -> str:
+        return re.sub(r"[^a-z0-9]+", "", text.lower())
+
+    def _tokenize_name(text: str) -> List[str]:
+        return [token for token in re.split(r"[^a-z0-9]+", text.lower()) if len(token) >= 2]
+
+    model_norm = _normalize_name(pth_file.stem)
+    model_tokens = set(_tokenize_name(pth_file.stem))
+
+    best_match = None
+    best_score = -1
+    for idx in index_files:
+        idx_norm = _normalize_name(idx.stem)
+        idx_tokens = set(_tokenize_name(idx.stem))
+        score = 0
+        if idx_norm == model_norm:
+            score += 1000
+        if model_norm and (model_norm in idx_norm or idx_norm in model_norm):
+            score += 300
+        shared_tokens = len(model_tokens & idx_tokens)
+        score += shared_tokens * 40
+        if "added" in idx.stem.lower():
+            score += 10
+        if score > best_score:
+            best_score = score
+            best_match = idx
+
+    if best_match is not None and best_score > 0:
+        return best_match
+    return None
+
+
+def _safe_print(message: str):
+    """避免控制台编码问题导致的输出异常"""
+    try:
+        print(message)
+    except UnicodeEncodeError:
+        print(message.encode("utf-8", "backslashreplace").decode("utf-8"))
+
+
+def _load_version_note_cache():
+    global _VERSION_NOTE_CACHE_LOADED
+    if _VERSION_NOTE_CACHE_LOADED:
+        return
+    _VERSION_NOTE_CACHE_LOADED = True
+    try:
+        cache_path = get_character_models_dir() / "_version_notes.json"
+        if cache_path.exists():
+            with open(cache_path, "r", encoding="utf-8") as f:
+                data = json.load(f)
+                if isinstance(data, dict):
+                    _VERSION_NOTE_CACHE.update(data)
+    except Exception:
+        pass
+
+
+def _save_version_note_cache():
+    try:
+        cache_path = get_character_models_dir() / "_version_notes.json"
+        cache_path.parent.mkdir(parents=True, exist_ok=True)
+        with open(cache_path, "w", encoding="utf-8") as f:
+            json.dump(_VERSION_NOTE_CACHE, f, ensure_ascii=False, indent=2)
+    except Exception:
+        pass
+
+
+def _normalize_note(note: str) -> Optional[str]:
+    if not note:
+        return None
+    # 取首行，移除链接与多余空白
+    line = note.strip().splitlines()[0].strip()
+    if not line:
+        return None
+    line = re.sub(r"https?://\S+", "", line).strip()
+    line = re.sub(r"\s+", " ", line).strip()
+    if not line:
+        return None
+    if len(line) > 60:
+        return line[:57] + "..."
+    return line
+
+
+def _note_from_metadata(path: Path) -> Optional[str]:
+    try:
+        with open(path, "r", encoding="utf-8") as f:
+            data = json.load(f)
+    except Exception:
+        return None
+
+    parts = []
+    title = str(data.get("title") or "")
+    desc = str(data.get("description") or "")
+    type_val = str(data.get("type") or data.get("version") or "")
+    text = f"{title}\n{desc}"
+
+    epoch_match = re.search(r"(\d+)\s*epoch", text, re.IGNORECASE)
+    if epoch_match:
+        parts.append(f"{epoch_match.group(1)} epochs")
+
+    if type_val:
+        t = type_val.lower()
+        if t.startswith("v"):
+            parts.append(f"RVC {type_val}")
+        elif "rvc" in t:
+            parts.append(type_val)
+
+    if parts:
+        return " · ".join(dict.fromkeys(parts))
+
+    # fallback to title/description first line
+    return _normalize_note(desc) or _normalize_note(title)
+
+
+def _note_from_pth(path: Path) -> Optional[str]:
+    try:
+        import torch
+    except Exception:
+        return None
+
+    try:
+        obj = torch.load(path, map_location="cpu", weights_only=False)
+    except Exception:
+        return None
+
+    if not isinstance(obj, dict):
+        return None
+
+    parts = []
+    info = obj.get("info")
+    if isinstance(info, str):
+        match = re.search(r"(\d+)\s*epoch", info, re.IGNORECASE)
+        if match:
+            parts.append(f"{match.group(1)} epochs")
+
+    sr = obj.get("sr")
+    if sr:
+        if isinstance(sr, (int, float)):
+            sr_note = f"{int(sr/1000)}k" if sr >= 1000 else str(int(sr))
+        else:
+            sr_note = str(sr)
+        parts.append(sr_note)
+
+    if parts:
+        return " · ".join(dict.fromkeys(parts))
+    return None
+
+
+def _note_from_filename(name: str) -> Optional[str]:
+    if not name:
+        return None
+    lower = name.lower()
+    parts = []
+
+    if "rmvpe" in lower:
+        parts.append("RMVPE")
+    if "ov2" in lower or "ov2super" in lower:
+        parts.append("OV2")
+    if "pre-anime" in lower or "preanime" in lower:
+        parts.append("预TV")
+    # 训练轮次与步数
+    epoch_match = re.search(r"(?:^|[_-])e(\d{2,5})(?:[_-]|$)", lower)
+    if epoch_match:
+        parts.append(f"e{epoch_match.group(1)}")
+    step_match = re.search(r"(?:^|[_-])s(\d{3,7})(?:[_-]|$)", lower)
+    if step_match:
+        parts.append(f"s{step_match.group(1)}")
+
+    if parts:
+        return " · ".join(dict.fromkeys(parts))
+    return None
+
+
+def _get_version_note(name: str, info: Dict) -> Optional[str]:
+    _load_version_note_cache()
+    note = info.get("variant_note")
+    if note:
+        return _normalize_note(note)
+
+    cached = _VERSION_NOTE_CACHE.get(name)
+    if cached is not None:
+        return cached
+
+    # 1) 读取本地模型目录中的 metadata / info
+    char_dir = get_character_models_dir() / name
+    if char_dir.exists():
+        for candidate in ("metadata.json", "model_info.json", "info.json"):
+            path = char_dir / candidate
+            if path.exists():
+                note = _note_from_metadata(path)
+                if note:
+                    _VERSION_NOTE_CACHE[name] = note
+                    _save_version_note_cache()
+                    return note
+
+        # 2) 尝试从本地权重读取 info/版本
+        pth_files = sorted(char_dir.glob("*.pth"))
+        if pth_files:
+            note = _note_from_pth(pth_files[0]) or _note_from_filename(pth_files[0].name)
+            if note:
+                _VERSION_NOTE_CACHE[name] = note
+                _save_version_note_cache()
+                return note
+        index_files = sorted(char_dir.glob("*.index"))
+        if index_files:
+            note = _note_from_filename(index_files[0].name)
+            if note:
+                _VERSION_NOTE_CACHE[name] = note
+                _save_version_note_cache()
+                return note
+
+    # 3) 未下载时，从配置文件名解析
+    file_name = info.get("file") or ""
+    note = _note_from_filename(Path(str(file_name)).name)
+    if not note:
+        files = info.get("files") or []
+        if files:
+            note = _note_from_filename(Path(str(files[0])).name)
+
+    if not note and info.get("variant"):
+        note = "未提供版本说明"
+
+    _VERSION_NOTE_CACHE[name] = note
+    _save_version_note_cache()
+    return note
+
+
+def refresh_version_notes(force: bool = False) -> Dict[str, Optional[str]]:
+    """批量读取本地模型的版本说明，写入缓存文件"""
+    if force:
+        _VERSION_NOTE_CACHE.clear()
+        # 避免读取旧缓存文件
+        global _VERSION_NOTE_CACHE_LOADED
+        _VERSION_NOTE_CACHE_LOADED = True
+    notes = {}
+    for name in CHARACTER_MODELS.keys():
+        notes[name] = _get_version_note(name, CHARACTER_MODELS.get(name, {}))
+    return notes
+
+
+def _get_confirm_token(response) -> Optional[str]:
+    """获取 Google Drive 下载确认 token"""
+    import re
+    for key, value in response.cookies.items():
+        if key.startswith("download_warning"):
+            return value
+    try:
+        # 兼容通过页面确认下载的场景
+        match = re.search(r"confirm=([0-9A-Za-z_]+)", response.text)
+        if match:
+            return match.group(1)
+    except Exception:
+        pass
+    return None
+
+
+def _download_gdrive_file(file_id: str, dest_path: Path) -> bool:
+    """下载 Google Drive 文件（支持大文件确认）"""
+    import requests
+
+    dest_path.parent.mkdir(parents=True, exist_ok=True)
+    url = "https://drive.google.com/uc?export=download"
+    session = requests.Session()
+
+    response = session.get(url, params={"id": file_id}, stream=True, timeout=30)
+    token = _get_confirm_token(response)
+    if token:
+        response = session.get(
+            url, params={"id": file_id, "confirm": token}, stream=True, timeout=30
+        )
+
+    if response.status_code != 200:
+        print(f"  下载失败: HTTP {response.status_code}")
+        return False
+    content_type = response.headers.get("content-type", "")
+    if "text/html" in content_type:
+        print("  下载失败: 需要手动确认或无访问权限")
+        return False
+
+    total_size = int(response.headers.get("content-length", 0))
+    downloaded = 0
+    with open(dest_path, "wb") as f:
+        for chunk in response.iter_content(chunk_size=8192):
+            if chunk:
+                f.write(chunk)
+                downloaded += len(chunk)
+                # 简单进度输出
+                if total_size > 0:
+                    percent = downloaded * 100 / total_size
+                    if int(percent) % 10 == 0:
+                        print(f"  下载进度: {percent:.0f}%")
+
+    return dest_path.exists() and dest_path.stat().st_size > 0
+
+
+# 角色模型列表
+CHARACTER_MODELS = {
+    "Aimi": {
+        "file": "Aimi.zip",
+        "zh_name": "爱美",
+        "en_name": "Aimi",
+        "jp_name": "愛美",
+        "source": "原创角色"
+    },
+    "hanayo": {
+        "file": "kubotayurika.zip",
+        "zh_name": "小泉花阳",
+        "en_name": "Hanayo Koizumi",
+        "jp_name": "小泉花陽",
+        "source": "Love Live!"
+    },
+    "rin": {
+        "file": "rin.zip",
+        "zh_name": "星空凛",
+        "en_name": "Rin Hoshizora",
+        "jp_name": "星空凛",
+        "source": "Love Live!"
+    },
+    "umi": {
+        "file": "umi.zip",
+        "zh_name": "园田海未",
+        "en_name": "Umi Sonoda",
+        "jp_name": "園田海未",
+        "source": "Love Live!"
+    },
+    "nozomi": {
+        "file": "nozomi.zip",
+        "zh_name": "东条希",
+        "en_name": "Nozomi Tojo",
+        "jp_name": "東條希",
+        "source": "Love Live!"
+    },
+    "dia": {
+        "file": "dia.zip",
+        "zh_name": "黑泽黛雅",
+        "en_name": "Dia Kurosawa",
+        "jp_name": "黒澤ダイヤ",
+        "source": "Love Live! Sunshine!!"
+    },
+    "ayumu": {
+        "file": "ayumu.zip",
+        "zh_name": "上原步梦",
+        "en_name": "Ayumu Uehara",
+        "jp_name": "上原歩夢",
+        "source": "Love Live! 虹咲学园"
+    },
+    "chika": {
+        "file": "TakamiChika.zip",
+        "zh_name": "高海千歌",
+        "en_name": "Chika Takami",
+        "jp_name": "高海千歌",
+        "source": "Love Live! Sunshine!!",
+        "repo": "Icchan/LoveLive"
+    },
+    "riko": {
+        "file": "SakurauchiRiko.zip",
+        "zh_name": "樱内梨子",
+        "en_name": "Riko Sakurauchi",
+        "jp_name": "桜内梨子",
+        "source": "Love Live! Sunshine!!",
+        "repo": "Icchan/LoveLive"
+    },
+    "mari": {
+        "file": "Mari Ohara (Love Live! Sunshine!!) - Weights.gg Model.zip",
+        "zh_name": "小原鞠莉",
+        "en_name": "Mari Ohara",
+        "jp_name": "小原鞠莉",
+        "source": "Love Live! Sunshine!!",
+        "repo": "ChocoKat/Mari_Ohara"
+    },
+    "yohane": {
+        "file": "Yoshiko Tsushima.zip",
+        "zh_name": "津岛善子（夜羽）",
+        "en_name": "Yoshiko Tsushima (Yohane)",
+        "jp_name": "ヨハネ",
+        "source": "Love Live! Sunshine!! / 幻日夜羽",
+        "repo": "HarunaKasuga/YoshikoTsushima"
+    },
+    "nico": {
+        "url": "https://huggingface.co/Zurakichi/RVC/resolve/main/Models/Love%20Live/V2/%C2%B5%27s/NicoYazawa.zip",
+        "filename": "NicoYazawa.zip",
+        "zh_name": "矢泽妮可",
+        "en_name": "Nico Yazawa",
+        "jp_name": "矢澤にこ",
+        "source": "Love Live!",
+        "repo": "Zurakichi/RVC"
+    },
+    "kanata": {
+        "file": "Models/Love Live/V2/Nijigasaki/KanataKonoe.zip",
+        "zh_name": "近江彼方",
+        "en_name": "Kanata Konoe",
+        "jp_name": "近江彼方",
+        "source": "Love Live! 虹咲学园",
+        "repo": "Zurakichi/RVC"
+    },
+    "setsuna_yuki_nana": {
+        "file": "Models/Love Live/V2/Nijigasaki/SetsunaYuki.zip",
+        "zh_name": "优木雪菜",
+        "en_name": "Setsuna Yuki",
+        "jp_name": "優木せつ菜",
+        "variant": "Zurakichi v2",
+        "source": "Love Live! 虹咲学园",
+        "repo": "Zurakichi/RVC"
+    },
+    "hanamaru_zurakichi": {
+        "file": "Models/Love Live/V2/Aqours/HanamaruKunikida.zip",
+        "zh_name": "国木田花丸",
+        "en_name": "Hanamaru Kunikida",
+        "jp_name": "国木田花丸",
+        "variant": "Zurakichi v2",
+        "source": "Love Live! Sunshine!!",
+        "repo": "Zurakichi/RVC"
+    },
+    "kanan": {
+        "gdrive_id": "16dPSDGb3ciLsy1HtEXG2OSPhU7BuCty_",
+        "filename": "Kanan_Matsuura.zip",
+        "zh_name": "松浦果南",
+        "en_name": "Kanan Matsuura",
+        "jp_name": "松浦果南",
+        "source": "Love Live! Sunshine!!"
+    },
+    "yu_takasaki": {
+        "gdrive_id": "1xjIG_bsBzOTOwghGaLSMnO_vL2GgMPNw",
+        "filename": "Yu_Takasaki.zip",
+        "zh_name": "高咲侑",
+        "en_name": "Yu Takasaki",
+        "jp_name": "高咲侑",
+        "source": "Love Live! 虹咲学园"
+    },
+    "shizuku_osaka": {
+        "url": "https://mega.nz/file/UbZDEaRY#YnxExpDIJzh-rEDfEo2khTPAH1p6GZ5FzaMCfWdUQ34",
+        "zh_name": "樱坂雫",
+        "en_name": "Shizuku Osaka",
+        "jp_name": "桜坂しずく",
+        "source": "Love Live! 虹咲学园"
+    },
+    "sarah_kazuno": {
+        "file": "SarahKazuno2.zip",
+        "zh_name": "鹿角圣良",
+        "en_name": "Sarah Kazuno",
+        "jp_name": "鹿角聖良",
+        "variant": "v2",
+        "source": "Love Live! Sunshine!!",
+        "repo": "thebuddyadrian/RVC_Models"
+    },
+    "leah_kazuno": {
+        "file": "LeahKazuno2.zip",
+        "zh_name": "鹿角理亚",
+        "en_name": "Leah Kazuno",
+        "jp_name": "鹿角理亞",
+        "variant": "v2",
+        "source": "Love Live! Sunshine!!",
+        "repo": "thebuddyadrian/RVC_Models"
+    },
+    "eli": {
+        "file": "AyaseEli.zip",
+        "zh_name": "绚濑绘里",
+        "en_name": "Eli Ayase",
+        "jp_name": "絢瀬絵里",
+        "source": "Love Live!",
+        "repo": "Icchan/LoveLive"
+    },
+    "you": {
+        "file": "WatanabeYou.zip",
+        "zh_name": "渡边曜",
+        "en_name": "You Watanabe",
+        "jp_name": "渡辺曜",
+        "source": "Love Live! Sunshine!!",
+        "repo": "Icchan/LoveLive"
+    },
+    "honoka": {
+        "files": ["weights/Honoka.pth", "weights/honoka.index"],
+        "zh_name": "高坂穗乃果",
+        "en_name": "Honoka Kosaka",
+        "jp_name": "高坂穂乃果",
+        "source": "Love Live!",
+        "repo": "trioskosmos/rvc_models"
+    },
+    "kotori": {
+        "files": ["weights/Kotori.pth", "weights/kotori.index"],
+        "zh_name": "南小鸟",
+        "en_name": "Kotori Minami",
+        "jp_name": "南ことり",
+        "source": "Love Live!",
+        "repo": "trioskosmos/rvc_models"
+    },
+    "maki": {
+        "files": ["weights2/Maki.pth", "weights2/maki.index"],
+        "zh_name": "西木野真姬",
+        "en_name": "Maki Nishikino",
+        "jp_name": "西木野真姫",
+        "source": "Love Live!",
+        "repo": "trioskosmos/rvc_models"
+    },
+    "ruby": {
+        "files": ["weights2/Ruby.pth", "weights2/ruby.index"],
+        "zh_name": "黑泽露比",
+        "en_name": "Ruby Kurosawa",
+        "jp_name": "黒澤ルビィ",
+        "source": "Love Live! Sunshine!!",
+        "repo": "trioskosmos/rvc_models"
+    },
+    "kasumi": {
+        "files": ["weights/Kasumi.pth", "weights/kasumi.index"],
+        "zh_name": "中须霞",
+        "en_name": "Kasumi Nakasu",
+        "jp_name": "中須かすみ",
+        "source": "Love Live! 虹咲学园",
+        "repo": "trioskosmos/rvc_models"
+    },
+    "karin": {
+        "files": ["weights/Karin.pth", "weights/karin.index"],
+        "zh_name": "朝香果林",
+        "en_name": "Karin Asaka",
+        "jp_name": "朝香果林",
+        "source": "Love Live! 虹咲学园",
+        "repo": "trioskosmos/rvc_models"
+    },
+    "rina": {
+        "files": ["weights2/Rina.pth", "weights2/rina.index"],
+        "zh_name": "天王寺璃奈",
+        "en_name": "Rina Tennoji",
+        "jp_name": "天王寺璃奈",
+        "source": "Love Live! 虹咲学园",
+        "repo": "trioskosmos/rvc_models"
+    },
+    "lanzhu": {
+        "files": ["weights/Lanzhu.pth", "weights/lanzhu.index"],
+        "zh_name": "钟岚珠",
+        "en_name": "Lanzhu Zhong",
+        "jp_name": "鐘嵐珠",
+        "source": "Love Live! 虹咲学园",
+        "repo": "trioskosmos/rvc_models"
+    },
+    "keke": {
+        "files": ["weights/Keke.pth", "weights/keke.index"],
+        "zh_name": "唐可可",
+        "en_name": "Keke Tang",
+        "jp_name": "唐可可",
+        "source": "Love Live! Superstar!!",
+        "repo": "trioskosmos/rvc_models"
+    },
+    "ai_miyashita": {
+        "file": "AiMiyashitaV2.zip",
+        "zh_name": "宫下爱",
+        "en_name": "Ai Miyashita",
+        "jp_name": "宮下愛",
+        "source": "Love Live! 虹咲学园",
+        "repo": "0xMifune/LoveLive"
+    },
+    "emma_verde": {
+        "file": "EmmaVerdeV2.zip",
+        "zh_name": "艾玛·维尔德",
+        "en_name": "Emma Verde",
+        "jp_name": "エマ・ヴェルデ",
+        "source": "Love Live! 虹咲学园",
+        "repo": "0xMifune/LoveLive"
+    },
+    "shioriko_mifune": {
+        "file": "ShiorikoMifuneV2.zip",
+        "zh_name": "三船栞子",
+        "en_name": "Shioriko Mifune",
+        "jp_name": "三船栞子",
+        "source": "Love Live! 虹咲学园",
+        "repo": "0xMifune/LoveLive"
+    },
+    "chisato_arashi": {
+        "file": "ChisatoArashiV2.zip",
+        "zh_name": "岚千砂都",
+        "en_name": "Chisato Arashi",
+        "jp_name": "嵐千砂都",
+        "source": "Love Live! Superstar!!",
+        "repo": "0xMifune/LoveLive"
+    },
+    "ren_hazuki": {
+        "file": "RenHazukiV2.zip",
+        "zh_name": "叶月恋",
+        "en_name": "Ren Hazuki",
+        "jp_name": "葉月恋",
+        "source": "Love Live! Superstar!!",
+        "repo": "0xMifune/LoveLive"
+    },
+    "sumire_heanna": {
+        "file": "SumireHeannaV2.zip",
+        "zh_name": "平安名堇",
+        "en_name": "Sumire Heanna",
+        "jp_name": "平安名すみれ",
+        "source": "Love Live! Superstar!!",
+        "repo": "0xMifune/LoveLive"
+    },
+    "kinako_sakurakoji": {
+        "file": "SakurakojiKinakoV2.zip",
+        "zh_name": "樱小路希奈子",
+        "en_name": "Kinako Sakurakoji",
+        "jp_name": "桜小路きな子",
+        "source": "Love Live! Superstar!!",
+        "repo": "0xMifune/LoveLive"
+    },
+    "mei_yoneme": {
+        "file": "MeiYonemeV2.zip",
+        "zh_name": "米女芽衣",
+        "en_name": "Mei Yoneme",
+        "jp_name": "米女芽衣",
+        "source": "Love Live! Superstar!!",
+        "repo": "0xMifune/LoveLive"
+    },
+    "shiki_wakana": {
+        "file": "ShikiWakanaV2.zip",
+        "zh_name": "若菜四季",
+        "en_name": "Shiki Wakana",
+        "jp_name": "若菜四季",
+        "source": "Love Live! Superstar!!",
+        "repo": "0xMifune/LoveLive"
+    },
+    "natsumi_onitsuka": {
+        "file": "NatsumiOnitsukaV2.zip",
+        "zh_name": "鬼塚夏美",
+        "en_name": "Natsumi Onitsuka",
+        "jp_name": "鬼塚夏美",
+        "source": "Love Live! Superstar!!",
+        "repo": "0xMifune/LoveLive"
+    },
+    "sayaka_murano": {
+        "file": "SayakaMuranoV2.zip",
+        "zh_name": "村野纱香",
+        "en_name": "Sayaka Murano",
+        "jp_name": "村野さやか",
+        "source": "Love Live! 莲之空女学院学园偶像俱乐部",
+        "repo": "0xMifune/LoveLive"
+    },
+    "tsuzuri_yugiri": {
+        "file": "TsuzuriYugiri.zip",
+        "zh_name": "夕雾缀理",
+        "en_name": "Tsuzuri Yugiri",
+        "jp_name": "夕霧綴理",
+        "source": "Love Live! 莲之空女学院学园偶像俱乐部",
+        "repo": "0xMifune/LoveLive"
+    },
+    "hanamaru": {
+        "file": "Aqours_Hanamaru-Kunikida-RMVPE-Ov2.zip",
+        "zh_name": "国木田花丸",
+        "en_name": "Hanamaru Kunikida",
+        "jp_name": "国木田花丸",
+        "source": "Love Live! Sunshine!!",
+        "repo": "Swordsmagus/Love-Live-RVC"
+    },
+    "setsuna_yuki_og": {
+        "file": "Nijigasaki_Setsuna-Yuki-OG-RMVPE-Ov2.zip",
+        "zh_name": "优木雪菜",
+        "en_name": "Setsuna Yuki",
+        "jp_name": "優木せつ菜",
+        "variant": "OG",
+        "source": "Love Live! 虹咲学园",
+        "repo": "Swordsmagus/Love-Live-RVC"
+    },
+    "setsuna_yuki_coco": {
+        "file": "setsuna-yuki-coco-hayashi-ver-555epochs.zip",
+        "zh_name": "优木雪菜",
+        "en_name": "Setsuna Yuki",
+        "jp_name": "優木せつ菜",
+        "variant": "林鼓子版",
+        "source": "Love Live! 虹咲学园",
+        "repo": "Swordsmagus/Love-Live-RVC"
+    },
+    "hanayo_pre_anime_v0": {
+        "file": "Hanayo-Pre-Anime-v0.zip",
+        "zh_name": "小泉花阳",
+        "en_name": "Hanayo Koizumi",
+        "jp_name": "小泉花陽",
+        "variant": "预TV v0",
+        "source": "Love Live!",
+        "repo": "Swordsmagus/Love-Live-RVC"
+    },
+    "hanayo_pre_anime_v1": {
+        "file": "Hanayo-Pre-Anime-v1.zip",
+        "zh_name": "小泉花阳",
+        "en_name": "Hanayo Koizumi",
+        "jp_name": "小泉花陽",
+        "variant": "预TV v1",
+        "source": "Love Live!",
+        "repo": "Swordsmagus/Love-Live-RVC"
+    },
+    "kotori_pre_anime": {
+        "file": "Kotori-Pre-Anime.zip",
+        "zh_name": "南小鸟",
+        "en_name": "Kotori Minami",
+        "jp_name": "南ことり",
+        "variant": "预TV",
+        "source": "Love Live!",
+        "repo": "Swordsmagus/Love-Live-RVC"
+    },
+    "kotori_pre_anime_v2": {
+        "file": "Kotori-Pre-Anime-v2.zip",
+        "zh_name": "南小鸟",
+        "en_name": "Kotori Minami",
+        "jp_name": "南ことり",
+        "variant": "预TV v2",
+        "source": "Love Live!",
+        "repo": "Swordsmagus/Love-Live-RVC"
+    },
+    "wien_margarete": {
+        "file": "Liella_Wien-Margarete-RMVPE-Ov2.zip",
+        "zh_name": "维恩·玛格丽特",
+        "en_name": "Wien Margarete",
+        "jp_name": "ウィーン・マルガレーテ",
+        "source": "Love Live! Superstar!!",
+        "repo": "Swordsmagus/Love-Live-RVC"
+    },
+    "mia_taylor": {
+        "file": "Mia-Taylor-SIFAS-Lines-Only.zip",
+        "zh_name": "米娅·泰勒",
+        "en_name": "Mia Taylor",
+        "jp_name": "ミア・テイラー",
+        "source": "Love Live! 虹咲学园",
+        "repo": "Swordsmagus/Love-Live-RVC"
+    },
+    "anju_yuki_arise": {
+        "file": "A-RISE_Anju-Yuki-RMVPE.zip",
+        "zh_name": "优木安朱",
+        "en_name": "Anju Yuki",
+        "jp_name": "優木あんじゅ",
+        "source": "Love Live!",
+        "repo": "Swordsmagus/Love-Live-RVC"
+    },
+    "erena_todo_arise": {
+        "file": "A-RISE_Erena-Todo-RMVPE.zip",
+        "zh_name": "统堂英玲奈",
+        "en_name": "Erena Todo",
+        "jp_name": "統堂英玲奈",
+        "source": "Love Live!",
+        "repo": "Swordsmagus/Love-Live-RVC"
+    },
+    "tsubasa_kira_arise": {
+        "file": "A-RISE_Tsubasa-Kira-RMVPE.zip",
+        "zh_name": "绮罗翼",
+        "en_name": "Tsubasa Kira",
+        "jp_name": "綺羅ツバサ",
+        "source": "Love Live!",
+        "repo": "Swordsmagus/Love-Live-RVC"
+    },
+    "kanon": {
+        "file": "ShibuyaKanonRVC.zip",
+        "zh_name": "涩谷香音",
+        "en_name": "Kanon Shibuya",
+        "jp_name": "渋谷かのん",
+        "source": "Love Live! Superstar!!",
+        "repo": "Phos252/RVCmodels"
+    },
+    "setsuna": {
+        "files": ["weights2/Setsuna.pth", "weights2/setsuna.index"],
+        "zh_name": "优木雪菜",
+        "en_name": "Setsuna Yuki",
+        "jp_name": "優木せつ菜",
+        "source": "Love Live! 虹咲学园",
+        "repo": "trioskosmos/rvc_models"
+    },
+    "setsuna_v2": {
+        "files": ["weights2/Setsuna2.pth", "weights2/setsuna2.index"],
+        "zh_name": "优木雪菜",
+        "en_name": "Setsuna Yuki",
+        "jp_name": "優木せつ菜",
+        "variant": "v2",
+        "source": "Love Live! 虹咲学园",
+        "repo": "trioskosmos/rvc_models"
+    },
+    "chika_v2": {
+        "files": ["weights/Chika2.pth", "weights/chika2.index"],
+        "zh_name": "高海千歌",
+        "en_name": "Chika Takami",
+        "jp_name": "高海千歌",
+        "variant": "v2",
+        "source": "Love Live! Sunshine!!",
+        "repo": "trioskosmos/rvc_models"
+    },
+    "dia_v2": {
+        "files": ["weights/Dia2.pth", "weights/dia2.index"],
+        "zh_name": "黑泽黛雅",
+        "en_name": "Dia Kurosawa",
+        "jp_name": "黒澤ダイヤ",
+        "variant": "v2",
+        "source": "Love Live! Sunshine!!",
+        "repo": "trioskosmos/rvc_models"
+    },
+    "hanamaru_v2": {
+        "files": ["weights/Hanamaru2.pth", "weights/hanamaru2.index"],
+        "zh_name": "国木田花丸",
+        "en_name": "Hanamaru Kunikida",
+        "jp_name": "国木田花丸",
+        "variant": "v2",
+        "source": "Love Live! Sunshine!!",
+        "repo": "trioskosmos/rvc_models"
+    },
+    "honoka_v2": {
+        "files": ["weights/Honoka2.pth", "weights/honoka2.index"],
+        "zh_name": "高坂穗乃果",
+        "en_name": "Honoka Kosaka",
+        "jp_name": "高坂穂乃果",
+        "variant": "v2",
+        "source": "Love Live!",
+        "repo": "trioskosmos/rvc_models"
+    },
+    "ayumu_v2": {
+        "files": ["weights/Ayumu2.pth", "weights/ayumu2.index"],
+        "zh_name": "上原步梦",
+        "en_name": "Ayumu Uehara",
+        "jp_name": "上原歩夢",
+        "variant": "v2",
+        "source": "Love Live! 虹咲学园",
+        "repo": "trioskosmos/rvc_models"
+    },
+    "ayumu_v3": {
+        "files": ["weights/Ayumu3.pth", "weights/ayumu3.index"],
+        "zh_name": "上原步梦",
+        "en_name": "Ayumu Uehara",
+        "jp_name": "上原歩夢",
+        "variant": "v3",
+        "source": "Love Live! 虹咲学园",
+        "repo": "trioskosmos/rvc_models"
+    },
+    "ayumu_v4": {
+        "files": ["weights/Ayumu4.pth", "weights/ayumu4.index"],
+        "zh_name": "上原步梦",
+        "en_name": "Ayumu Uehara",
+        "jp_name": "上原歩夢",
+        "variant": "v4",
+        "source": "Love Live! 虹咲学园",
+        "repo": "trioskosmos/rvc_models"
+    },
+    "karin_v2": {
+        "files": ["weights/Karin2.pth", "weights/karin2.index"],
+        "zh_name": "朝香果林",
+        "en_name": "Karin Asaka",
+        "jp_name": "朝香果林",
+        "variant": "v2",
+        "source": "Love Live! 虹咲学园",
+        "repo": "trioskosmos/rvc_models"
+    },
+    "keke_v2": {
+        "files": ["weights/Keke2.pth", "weights/keke2.index"],
+        "zh_name": "唐可可",
+        "en_name": "Keke Tang",
+        "jp_name": "唐可可",
+        "variant": "v2",
+        "source": "Love Live! Superstar!!",
+        "repo": "trioskosmos/rvc_models"
+    },
+    "riko_v2": {
+        "files": ["weights2/Riko2.pth", "weights2/riko2.index"],
+        "zh_name": "樱内梨子",
+        "en_name": "Riko Sakurauchi",
+        "jp_name": "桜内梨子",
+        "variant": "v2",
+        "source": "Love Live! Sunshine!!",
+        "repo": "trioskosmos/rvc_models"
+    },
+    "umi_v2": {
+        "files": ["weights2/Umi2.pth", "weights2/umi2.index"],
+        "zh_name": "园田海未",
+        "en_name": "Umi Sonoda",
+        "jp_name": "園田海未",
+        "variant": "v2",
+        "source": "Love Live!",
+        "repo": "trioskosmos/rvc_models"
+    },
+    "wien_v2": {
+        "files": ["weights2/Wien2.pth", "weights2/wien.index"],
+        "zh_name": "维恩·玛格丽特",
+        "en_name": "Wien Margarete",
+        "jp_name": "ウィーン・マルガレーテ",
+        "variant": "v2",
+        "source": "Love Live! Superstar!!",
+        "repo": "trioskosmos/rvc_models"
+    },
+    "yohane_trios": {
+        "files": ["weights2/Yohane.pth", "weights2/yohane.index"],
+        "zh_name": "津岛善子（夜羽）",
+        "en_name": "Yoshiko Tsushima (Yohane)",
+        "jp_name": "ヨハネ",
+        "variant": "trios",
+        "source": "Love Live! Sunshine!! / 幻日夜羽",
+        "repo": "trioskosmos/rvc_models"
+    },
+    "yoshiko_trios": {
+        "files": ["weights2/yoshiko.pth", "weights2/yoshiko.index"],
+        "zh_name": "津岛善子",
+        "en_name": "Yoshiko Tsushima",
+        "jp_name": "津島善子",
+        "variant": "trios",
+        "source": "Love Live! Sunshine!!",
+        "repo": "trioskosmos/rvc_models"
+    },
+    "yoshiko_v2": {
+        "files": ["weights2/Yoshiko2.pth", "weights2/yoshiko2.index"],
+        "zh_name": "津岛善子",
+        "en_name": "Yoshiko Tsushima",
+        "jp_name": "津島善子",
+        "variant": "v2",
+        "source": "Love Live! Sunshine!!",
+        "repo": "trioskosmos/rvc_models"
+    },
+    "tokai_teio": {
+        "file": "Tokai Teio (Uma Musume) - Weights Model.zip",
+        "zh_name": "东海帝皇",
+        "en_name": "Tokai Teio",
+        "jp_name": "トウカイテイオー",
+        "source": "赛马娘",
+        "repo": "kohaku12/RVC-MODELS"
+    },
+    "focalors": {
+        "file": "Focalors.zip",
+        "zh_name": "芙卡洛斯",
+        "en_name": "Focalors",
+        "jp_name": "フォカロルス",
+        "source": "原神",
+        "repo": "makiligon/RVC-Models"
+    },
+    "essex": {
+        "file": "Essex.zip",
+        "zh_name": "埃塞克斯",
+        "en_name": "Essex",
+        "jp_name": "エセックス",
+        "source": "碧蓝航线",
+        "repo": "makiligon/RVC-Models"
+    },
+    "ellie": {
+        "file": "ellie.zip",
+        "zh_name": "艾莉",
+        "en_name": "Ellie",
+        "jp_name": "エリー",
+        "source": "社区模型",
+        "repo": "megaaziib/my-rvc-models-collection"
+    },
+    "furina": {
+        "file": "Furina.zip",
+        "zh_name": "芙宁娜",
+        "en_name": "Furina",
+        "jp_name": "フリーナ",
+        "source": "原神",
+        "repo": "makiligon/RVC-Models"
+    },
+    "ayaka": {
+        "file": "Ayaka.zip",
+        "zh_name": "神里绫华",
+        "en_name": "Kamisato Ayaka",
+        "jp_name": "神里綾華",
+        "source": "原神",
+        "repo": "makiligon/RVC-Models"
+    },
+    "takane_shijou": {
+        "file": "TakaneShijou.zip",
+        "zh_name": "四条贵音",
+        "en_name": "Takane Shijou",
+        "jp_name": "四条貴��",
+        "source": "偶像大师",
+        "repo": "makiligon/RVC-Models"
+    },
+    "kobo": {
+        "file": "kobo.zip",
+        "zh_name": "可波·卡娜埃露",
+        "en_name": "Kobo Kanaeru",
+        "jp_name": "こぼ・かなえる",
+        "source": "Hololive Indonesia",
+        "repo": "megaaziib/my-rvc-models-collection"
+    },
+    "kaela": {
+        "file": "kaela.zip",
+        "zh_name": "卡埃拉·科瓦尔斯基亚",
+        "en_name": "Kaela Kovalskia",
+        "jp_name": "カエラ・コヴァルスキア",
+        "source": "Hololive Indonesia",
+        "repo": "megaaziib/my-rvc-models-collection"
+    },
+    "pekora": {
+        "file": "pekora.zip",
+        "zh_name": "兔田佩克拉",
+        "en_name": "Usada Pekora",
+        "jp_name": "兎田ぺこら",
+        "source": "Hololive Japan",
+        "repo": "megaaziib/my-rvc-models-collection"
+    },
+    "kizuna_ai": {
+        "file": "kizuna-ai.zip",
+        "zh_name": "绊爱",
+        "en_name": "Kizuna AI",
+        "jp_name": "キズナアイ",
+        "source": "虚拟主播",
+        "repo": "megaaziib/my-rvc-models-collection"
+    },
+    "fuwawa": {
+        "file": "fuwawa.zip",
+        "zh_name": "软软·阿比斯加德",
+        "en_name": "Fuwawa Abyssgard",
+        "jp_name": "フワワ・アビスガード",
+        "source": "Hololive English",
+        "repo": "megaaziib/my-rvc-models-collection"
+    },
+    "mococo": {
+        "file": "mococo.zip",
+        "zh_name": "茸茸·阿比斯加德",
+        "en_name": "Mococo Abyssgard",
+        "jp_name": "モココ・アビスガード",
+        "source": "Hololive English",
+        "repo": "megaaziib/my-rvc-models-collection"
+    },
+    "vo_furina_kr": {
+        "file": "vo_furina_kr.zip",
+        "zh_name": "芙宁娜",
+        "en_name": "Furina",
+        "jp_name": "フリーナ",
+        "variant": "韩语",
+        "lang": "韩文",
+        "source": "原神",
+        "repo": "jarari/RVC-v2"
+    },
+    "silverwolf_kr": {
+        "file": "silverwolf_kr.zip",
+        "zh_name": "银狼",
+        "en_name": "Silver Wolf",
+        "jp_name": "銀狼",
+        "variant": "韩语",
+        "lang": "韩文",
+        "source": "崩坏：星穹铁道",
+        "repo": "jarari/RVC-v2"
+    },
+    "sparkle_e40": {
+        "file": "sparkle_e40.zip",
+        "zh_name": "花火",
+        "en_name": "Sparkle",
+        "jp_name": "花火",
+        "variant": "日语",
+        "source": "崩坏：星穹铁道",
+        "repo": "kohaku12/RVC-MODELS"
+    },
+    "ranko": {
+        "file": "ranko.zip",
+        "zh_name": "神崎兰子",
+        "en_name": "Ranko Kanzaki",
+        "jp_name": "神崎蘭子",
+        "source": "偶像大师 灰姑娘女孩"
+    },
+    "yumemiriamu": {
+        "file": "yumemiriamu.zip",
+        "zh_name": "梦见莉亚梦",
+        "en_name": "Riamu Yumemi",
+        "jp_name": "夢見りあむ",
+        "source": "偶像大师 灰姑娘女孩"
+    },
+    # ===== VOCALOID =====
+    "hatsune_miku": {
+        "file": "infamous_miku_v2.zip",
+        "zh_name": "初音未来",
+        "en_name": "Hatsune Miku",
+        "jp_name": "初音ミク",
+        "source": "VOCALOID",
+        "repo": "javinfamous/infamous_miku_v2"
+    },
+    # ===== 原神 (更多角色) =====
+    "nahida": {
+        "file": "Nahida JP (VA_ Yukari Tamura) - Weights.gg Model.zip",
+        "zh_name": "纳西妲",
+        "en_name": "Nahida",
+        "jp_name": "ナヒーダ",
+        "source": "原神",
+        "repo": "kohaku12/RVC-MODELS"
+    },
+    "nilou": {
+        "file": "Nilou%20JP%20(Kanemoto%20Hisako)%20-%20Weights.gg%20Model.zip",
+        "zh_name": "妮露",
+        "en_name": "Nilou",
+        "jp_name": "ニィロウ",
+        "source": "原神",
+        "repo": "kohaku12/RVC-MODELS"
+    },
+    # ===== 崩坏：星穹铁道 (更多角色) =====
+    "herta_hsr": {
+        "file": "Herta JP (VA_ Haruka Yamazaki) (Honkai_ Star Rail) - Weights.gg Model.zip",
+        "zh_name": "黑塔",
+        "en_name": "Herta",
+        "jp_name": "ヘルタ",
+        "source": "崩坏：星穹铁道",
+        "repo": "kohaku12/RVC-MODELS"
+    },
+    "the_herta": {
+        "file": "The Herta -Honkai Star Rail - Weights Model.zip",
+        "zh_name": "大黑塔",
+        "en_name": "The Herta",
+        "jp_name": "マダム・ヘルタ",
+        "source": "崩坏：星穹铁道",
+        "repo": "kohaku12/RVC-MODELS"
+    },
+    "firefly_hsr": {
+        "file": "Firefly _ Honkai_ Star Rail - Weights.gg Model.zip",
+        "zh_name": "流萤",
+        "en_name": "Firefly",
+        "jp_name": "ホタル",
+        "source": "崩坏：星穹铁道",
+        "repo": "kohaku12/RVC-MODELS"
+    },
+    "tingyun_hsr": {
+        "file": "Tingyun (Honkai_ Star Rail) (VA_ Yuuki Takada) - Weights.gg Model.zip",
+        "zh_name": "停云",
+        "en_name": "Tingyun",
+        "jp_name": "停雲",
+        "source": "崩坏：星穹铁道",
+        "repo": "kohaku12/RVC-MODELS"
+    },
+    "yunli_hsr": {
+        "file": "Yunli (Japanese Voice) Honkai Star Rail - Weights.gg Model.zip",
+        "zh_name": "云璃",
+        "en_name": "Yunli",
+        "jp_name": "雲璃",
+        "source": "崩坏：星穹铁道",
+        "repo": "kohaku12/RVC-MODELS"
+    },
+    "tribbie_hsr": {
+        "file": "Tribbie _ Honkai Star Rail JP (CV_ Hikaru Tono) - Weights Model.zip",
+        "zh_name": "缇宝",
+        "en_name": "Tribbie",
+        "jp_name": "トリビー",
+        "source": "崩坏：星穹铁道",
+        "repo": "kohaku12/RVC-MODELS"
+    },
+    "huohuo_hsr": {
+        "file": "huohuo2.zip",
+        "zh_name": "藿藿",
+        "en_name": "Huohuo",
+        "jp_name": "フォフォ",
+        "source": "崩坏：星穹铁道",
+        "repo": "kohaku12/RVC-MODELS"
+    },
+    "castorice_hsr": {
+        "file": "Castorice (Honkai_ Star Rail, Japanese - CV_ Chiwa Saito) [Preliminary] - Weights Model.zip",
+        "zh_name": "遐蝶",
+        "en_name": "Castorice",
+        "jp_name": "キャストリス",
+        "source": "崩坏：星穹铁道",
+        "repo": "kohaku12/RVC-MODELS"
+    },
+    # ===== 崩坏3rd =====
+    "seele_hi3": {
+        "file": "Seele (Honkai impact 3rd) - Weights.gg Model.zip",
+        "zh_name": "希儿",
+        "en_name": "Seele",
+        "jp_name": "ゼーレ",
+        "source": "崩坏3rd",
+        "repo": "kohaku12/RVC-MODELS"
+    },
+    "herrscher_ego_hi3": {
+        "file": "Herrscher of Human_ Ego (Honkai Impact 3rd) - Weights.gg Model.zip",
+        "zh_name": "真我·人之律者",
+        "en_name": "Herrscher of Human: Ego",
+        "jp_name": "真我・人の律者",
+        "source": "崩坏3rd",
+        "repo": "kohaku12/RVC-MODELS"
+    },
+    # ===== 绝区零 =====
+    "miyabi_zzz": {
+        "file": "hoshimi miyabi (Zenless Zone Zero)JP - Weights Model.zip",
+        "zh_name": "星见雅",
+        "en_name": "Hoshimi Miyabi",
+        "jp_name": "星見雅",
+        "source": "绝区零",
+        "repo": "kohaku12/RVC-MODELS"
+    },
+    # ===== Project SEKAI =====
+    "nene_kusanagi": {
+        "file": "Nene Kusanagi (Project Sekai) - Weights Model.zip",
+        "zh_name": "草薙宁宁",
+        "en_name": "Nene Kusanagi",
+        "jp_name": "草薙寧々",
+        "source": "Project SEKAI",
+        "repo": "kohaku12/RVC-MODELS"
+    },
+    # ===== Hololive (更多角色) =====
+    "miko_sakura": {
+        "file": "miko.zip",
+        "zh_name": "樱巫女",
+        "en_name": "Sakura Miko",
+        "jp_name": "さくらみこ",
+        "source": "Hololive Japan",
+        "repo": "megaaziib/my-rvc-models-collection"
+    },
+    "subaru_oozora": {
+        "file": "subaru.zip",
+        "zh_name": "大空昴",
+        "en_name": "Oozora Subaru",
+        "jp_name": "大空スバル",
+        "source": "Hololive Japan",
+        "repo": "megaaziib/my-rvc-models-collection"
+    },
+    "moona_hoshinova": {
+        "file": "moona.zip",
+        "zh_name": "穆娜・霍希诺瓦",
+        "en_name": "Moona Hoshinova",
+        "jp_name": "ムーナ・ホシノヴァ",
+        "source": "Hololive Indonesia",
+        "repo": "megaaziib/my-rvc-models-collection"
+    },
+    "risu_ayunda": {
+        "file": "risu.zip",
+        "zh_name": "阿云达·莉苏",
+        "en_name": "Ayunda Risu",
+        "jp_name": "アユンダ・リス",
+        "source": "Hololive Indonesia",
+        "repo": "megaaziib/my-rvc-models-collection"
+    },
+    "reine_pavolia": {
+        "file": "reine.zip",
+        "zh_name": "帕沃莉亚・蕾妮",
+        "en_name": "Pavolia Reine",
+        "jp_name": "パヴォリア・レイネ",
+        "source": "Hololive Indonesia",
+        "repo": "megaaziib/my-rvc-models-collection"
+    },
+    "zeta_vestia": {
+        "file": "zeta.zip",
+        "zh_name": "贝斯蒂亚・泽塔",
+        "en_name": "Vestia Zeta",
+        "jp_name": "ヴェスティア・ゼータ",
+        "source": "Hololive Indonesia",
+        "repo": "megaaziib/my-rvc-models-collection"
+    },
+    "anya_melfissa": {
+        "file": "anya.zip",
+        "zh_name": "安亚・梅尔菲莎",
+        "en_name": "Anya Melfissa",
+        "jp_name": "アーニャ・メルフィッサ",
+        "source": "Hololive Indonesia",
+        "repo": "megaaziib/my-rvc-models-collection"
+    },
+    "luna_himemori": {
+        "file": "luna.zip",
+        "zh_name": "姬森露娜",
+        "en_name": "Himemori Luna",
+        "jp_name": "姫森ルーナ",
+        "source": "Hololive Japan",
+        "repo": "megaaziib/my-rvc-models-collection"
+    },
+    # ===== 碧蓝航线 =====
+    "boothill": {
+        "file": "Boothill.zip",
+        "zh_name": "波提欧",
+        "en_name": "Boothill",
+        "jp_name": "ブートヒル",
+        "source": "崩坏：星穹铁道",
+        "repo": "makiligon/RVC-Models"
+    },
+    # ===== 明日方舟 =====
+    "shiroko_rosmontis": {
+        "file": "ShirokoRosmontis.zip",
+        "zh_name": "白子/迷迭香",
+        "en_name": "Shiroko / Rosmontis",
+        "jp_name": "シロコ / ロスモンティス",
+        "source": "蔚蓝档案 / 明日方舟",
+        "repo": "makiligon/RVC-Models"
+    },
+    # ===== 赛马娘 (更多角色) =====
+    "rice_shower": {
+        "file": "RiceShowerSinging.zip",
+        "zh_name": "米浴",
+        "en_name": "Rice Shower",
+        "jp_name": "ライスシャワー",
+        "source": "赛马娘",
+        "repo": "makiligon/RVC-Models"
+    },
+}
+
+
+def get_project_root() -> Path:
+    """获取项目根目录"""
+    return Path(__file__).parent.parent
+
+
+def get_character_models_dir() -> Path:
+    """获取角色模型目录"""
+    return get_project_root() / "assets" / "weights" / "characters"
+
+
+def list_available_characters() -> List[Dict]:
+    """
+    列出可用的角色模型
+
+    Returns:
+        list: 角色信息列表
+    """
+    result = []
+    for name, info in CHARACTER_MODELS.items():
+        source = info.get("source", "未知")
+        display = _get_display_name(info, name)
+        result.append({
+            "name": name,
+            "description": info.get("description", display),
+            "display": display,
+            "source": source,
+            "series": normalize_series(source),
+            "file": info.get("file"),
+            "files": info.get("files"),
+            "url": info.get("url"),
+            "repo": info.get("repo", HF_REPO_ID)
+        })
+    return result
+
+
+def list_downloaded_characters() -> List[Dict]:
+    """
+    列出已下载的角色模型
+
+    Returns:
+        list: 已下载的角色信息
+    """
+    models_dir = get_character_models_dir()
+    if not models_dir.exists():
+        return []
+
+    downloaded = []
+    seen = set()
+
+    # 递归搜索 .pth，优先使用顶层目录名作为角色名
+    for pth_file in models_dir.rglob("*.pth"):
+        rel_path = pth_file.relative_to(models_dir)
+        parts = rel_path.parts
+        if len(parts) == 1:
+            char_name = pth_file.stem
+        else:
+            char_name = parts[0]
+
+        if char_name in seen:
+            continue
+        seen.add(char_name)
+
+        info = CHARACTER_MODELS.get(char_name, {})
+        source = info.get("source", "未知")
+        display = _get_display_name(info, char_name)
+        index_file = _find_index_file(pth_file)
+        downloaded.append({
+            "name": char_name,
+            "description": info.get("description", display),
+            "display": display,
+            "source": source,
+            "series": normalize_series(source),
+            "model_path": str(pth_file),
+            "index_path": str(index_file) if index_file else None
+        })
+
+    return downloaded
+
+
+def get_character_model_path(name: str) -> Optional[Dict]:
+    """
+    获取角色模型路径
+
+    Args:
+        name: 角色名称
+
+    Returns:
+        dict: 包含 model_path 和 index_path 的字典，或 None
+    """
+    models_dir = get_character_models_dir()
+    char_dir = models_dir / name
+
+    # 1) 标准目录结构: characters/<name>/*.pth
+    if char_dir.exists():
+        pth_files = list(char_dir.glob("*.pth"))
+        if pth_files:
+            index_file = _find_index_file(pth_files[0])
+            return {
+                "model_path": str(pth_files[0]),
+                "index_path": str(index_file) if index_file else None
+            }
+
+    # 2) 允许直接放在 characters 目录
+    direct_pth = models_dir / f"{name}.pth"
+    if direct_pth.exists():
+        index_file = _find_index_file(direct_pth)
+        return {
+            "model_path": str(direct_pth),
+            "index_path": str(index_file) if index_file else None
+        }
+
+    # 3) 兜底：在 characters 目录递归查找同名模型
+    for pth_file in models_dir.rglob("*.pth"):
+        if pth_file.stem.lower() == name.lower():
+            index_file = _find_index_file(pth_file)
+            return {
+                "model_path": str(pth_file),
+                "index_path": str(index_file) if index_file else None
+            }
+
+    return None
+
+
+def download_character_model(
+    name: str,
+    progress_callback: Optional[Callable[[str, float], None]] = None
+) -> bool:
+    """
+    下载角色模型
+
+    Args:
+        name: 角色名称
+        progress_callback: 进度回调 (message, progress)
+
+    Returns:
+        bool: 是否成功
+    """
+    if name not in CHARACTER_MODELS:
+        _safe_print(f"未知角色: {name}")
+        return False
+
+    char_info = CHARACTER_MODELS[name]
+    repo_id = char_info.get("repo", HF_REPO_ID)
+    zip_file = char_info.get("file")
+    file_list = char_info.get("files")
+    direct_url = char_info.get("url")
+    gdrive_id = char_info.get("gdrive_id")
+    pattern = char_info.get("pattern")
+
+    # 检查是否已下载
+    char_dir = get_character_models_dir() / name
+    if char_dir.exists() and list(char_dir.glob("*.pth")):
+        _safe_print(f"角色模型已存在: {name}")
+        return True
+
+    if progress_callback:
+        progress_callback(f"正在下载 {name} 模型...", 0.1)
+
+    hf_token = _get_hf_token()
+
+    try:
+        # Google Drive 下载
+        if gdrive_id:
+            filename = char_info.get("filename") or f"{name}.zip"
+            temp_dir = get_project_root() / "temp" / "downloads"
+            temp_dir.mkdir(parents=True, exist_ok=True)
+            temp_path = temp_dir / filename
+
+            _safe_print(f"正在从 Google Drive 下载: {filename}")
+            if not _download_gdrive_file(gdrive_id, temp_path):
+                return False
+
+            char_dir.mkdir(parents=True, exist_ok=True)
+            if temp_path.suffix.lower() == ".zip":
+                if progress_callback:
+                    progress_callback(f"正在解压 {name} 模型...", 0.6)
+                with zipfile.ZipFile(temp_path, 'r') as zip_ref:
+                    zip_ref.extractall(char_dir)
+                _flatten_extracted_dir(char_dir)
+            else:
+                shutil.copy(str(temp_path), str(char_dir / temp_path.name))
+
+        # 直链下载（可用于非 HuggingFace 源）
+        elif direct_url:
+            if "mega.nz" in direct_url:
+                _safe_print("Mega 下载暂不支持，请手动下载并放入角色目录")
+                return False
+            from tools.download_models import download_file
+
+            filename = char_info.get("filename") or Path(direct_url).name
+            temp_dir = get_project_root() / "temp" / "downloads"
+            temp_dir.mkdir(parents=True, exist_ok=True)
+            temp_path = temp_dir / filename
+
+            _safe_print(f"正在下载: {direct_url}")
+            if not download_file(direct_url, temp_path, name):
+                return False
+
+            char_dir.mkdir(parents=True, exist_ok=True)
+            if temp_path.suffix.lower() == ".zip":
+                if progress_callback:
+                    progress_callback(f"正在解压 {name} 模型...", 0.6)
+                with zipfile.ZipFile(temp_path, 'r') as zip_ref:
+                    zip_ref.extractall(char_dir)
+                _flatten_extracted_dir(char_dir)
+            else:
+                shutil.copy(str(temp_path), str(char_dir / temp_path.name))
+
+        # HuggingFace 多文件下载
+        elif file_list:
+            if not HF_AVAILABLE:
+                raise ImportError("请安装 huggingface_hub: pip install huggingface_hub")
+
+            _safe_print(f"正在从 HuggingFace 下载: {repo_id} (files)")
+            char_dir.mkdir(parents=True, exist_ok=True)
+            total = len(file_list)
+            for idx, hf_file in enumerate(file_list, start=1):
+                if progress_callback and total > 0:
+                    progress_callback(
+                        f"正在下载 {name} 文件 {idx}/{total}...",
+                        0.1 + 0.8 * (idx / total)
+                    )
+                downloaded_path = hf_hub_download(
+                    repo_id=repo_id,
+                    filename=hf_file,
+                    cache_dir=str(get_project_root() / "temp" / "hf_cache"),
+                    token=hf_token
+                )
+                target_name = Path(hf_file).name
+                shutil.copy(str(downloaded_path), str(char_dir / target_name))
+
+        # HuggingFace: 根据 pattern 自动选择文件
+        elif pattern:
+            if not HF_AVAILABLE:
+                raise ImportError("请安装 huggingface_hub: pip install huggingface_hub")
+
+            _safe_print(f"正在从 HuggingFace 下载: {repo_id} (auto)")
+            files = list_repo_files(repo_id, token=hf_token)
+            if isinstance(pattern, str):
+                patterns = [pattern]
+            else:
+                patterns = list(pattern)
+
+            candidates = []
+            for f in files:
+                for p in patterns:
+                    if p in f or (p == ".zip" and f.lower().endswith(".zip")):
+                        candidates.append(f)
+                        break
+
+            if not candidates:
+                _safe_print(f"未找到匹配文件: {pattern}")
+                return False
+
+            # 优先 zip
+            zip_candidates = [c for c in candidates if c.lower().endswith(".zip")]
+            selected = zip_candidates[0] if zip_candidates else candidates[0]
+
+            if selected.lower().endswith(".zip"):
+                zip_file = selected
+                downloaded_path = hf_hub_download(
+                    repo_id=repo_id,
+                    filename=zip_file,
+                    cache_dir=str(get_project_root() / "temp" / "hf_cache"),
+                    token=hf_token
+                )
+                if progress_callback:
+                    progress_callback(f"正在解压 {name} 模型...", 0.6)
+                char_dir.mkdir(parents=True, exist_ok=True)
+                with zipfile.ZipFile(downloaded_path, 'r') as zip_ref:
+                    zip_ref.extractall(char_dir)
+                _flatten_extracted_dir(char_dir)
+            else:
+                # 多文件下载
+                file_list = candidates
+                char_dir.mkdir(parents=True, exist_ok=True)
+                total = len(file_list)
+                for idx, hf_file in enumerate(file_list, start=1):
+                    if progress_callback and total > 0:
+                        progress_callback(
+                            f"正在下载 {name} 文件 {idx}/{total}...",
+                            0.1 + 0.8 * (idx / total)
+                        )
+                    downloaded_path = hf_hub_download(
+                        repo_id=repo_id,
+                        filename=hf_file,
+                        cache_dir=str(get_project_root() / "temp" / "hf_cache"),
+                        token=hf_token
+                    )
+                    target_name = Path(hf_file).name
+                    shutil.copy(str(downloaded_path), str(char_dir / target_name))
+
+        # HuggingFace zip 下载
+        else:
+            if not HF_AVAILABLE:
+                raise ImportError("请安装 huggingface_hub: pip install huggingface_hub")
+            _safe_print(f"正在从 HuggingFace 下载: {repo_id}/{zip_file}")
+
+            downloaded_path = hf_hub_download(
+                repo_id=repo_id,
+                filename=zip_file,
+                cache_dir=str(get_project_root() / "temp" / "hf_cache"),
+                token=hf_token
+            )
+
+            if progress_callback:
+                progress_callback(f"正在解压 {name} 模型...", 0.6)
+
+            char_dir.mkdir(parents=True, exist_ok=True)
+
+            with zipfile.ZipFile(downloaded_path, 'r') as zip_ref:
+                zip_ref.extractall(char_dir)
+
+            _flatten_extracted_dir(char_dir)
+
+        if progress_callback:
+            progress_callback(f"{name} 模型下载完成", 1.0)
+
+        # 下载完成后更新版本说明缓存
+        _get_version_note(name, char_info)
+        _safe_print(f"角色模型已下载: {name}")
+        return True
+
+    except Exception as e:
+        _safe_print(f"下载失败: {e}")
+        if progress_callback:
+            progress_callback(f"下载失败: {e}", 0)
+        return False
+
+
+def _flatten_extracted_dir(char_dir: Path):
+    """
+    处理解压后可能的嵌套目录结构
+    """
+    # 检查是否有嵌套的单一目录
+    items = list(char_dir.iterdir())
+    if len(items) == 1 and items[0].is_dir():
+        nested_dir = items[0]
+        # 移动内容到上级目录
+        for item in nested_dir.iterdir():
+            shutil.move(str(item), str(char_dir / item.name))
+        nested_dir.rmdir()
+
+
+def delete_character_model(name: str) -> bool:
+    """
+    删除角色模型
+
+    Args:
+        name: 角色名称
+
+    Returns:
+        bool: 是否成功
+    """
+    char_dir = get_character_models_dir() / name
+    if char_dir.exists():
+        shutil.rmtree(char_dir)
+        print(f"已删除角色模型: {name}")
+        return True
+    return False
+
+
+def check_hf_available() -> bool:
+    """检查 huggingface_hub 是否可用"""
+    return HF_AVAILABLE
+
+
+def list_available_series() -> List[str]:
+    """
+    获取可用的作品/系列列表（去重）
+    """
+    series_set = set()
+    for info in CHARACTER_MODELS.values():
+        source = info.get("source", "未知")
+        series_set.add(normalize_series(source))
+    return sorted(series_set)
+
+
+def download_all_character_models(
+    series: Optional[str] = None,
+    progress_callback: Optional[Callable[[str, float], None]] = None
+) -> Dict[str, List[str]]:
+    """
+    批量下载角色模型
+
+    Args:
+        series: 仅下载指定系列（如 "Love Live!"），None 表示全部
+        progress_callback: 进度回调 (message, progress)
+
+    Returns:
+        dict: { "success": [...], "failed": [...] }
+    """
+    targets = list_available_characters()
+    if series and series != "全部":
+        targets = [c for c in targets if c.get("series") == series]
+
+    success = []
+    failed = []
+    total = max(len(targets), 1)
+
+    for idx, char in enumerate(targets, start=1):
+        if progress_callback:
+            progress_callback(
+                f"正在下载 {char['name']} ({idx}/{total})...",
+                idx / total
+            )
+        ok = download_character_model(char["name"])
+        if ok:
+            success.append(char["name"])
+        else:
+            failed.append(char["name"])
+
+    return {
+        "success": success,
+        "failed": failed
+    }
+
+
+def get_character_choices() -> List[str]:
+    """
+    获取角色选择列表 (用于 UI 下拉框)
+
+    Returns:
+        list: 角色名称列表
+    """
+    downloaded = list_downloaded_characters()
+    return [c["name"] for c in downloaded]

tools/download_models.py

@@ -0,0 +1,344 @@
+# -*- coding: utf-8 -*-
+"""
+模型下载工具 - 自动从 Hugging Face 下载所需模型
+"""
+import os
+import hashlib
+import requests
+from pathlib import Path
+from tqdm import tqdm
+from typing import Optional, Dict, List
+
+# 模型下载配置
+MODELS = {
+    "HP2_all_vocals.pth": {
+        "url": "https://huggingface.co/lj1995/VoiceConversionWebUI/resolve/main/uvr5_weights/HP2_all_vocals.pth",
+        "path": "assets/uvr5_weights/HP2_all_vocals.pth",
+        "size_mb": 140,
+        "description": "UVR5 HP2 vocal model"
+    },
+    "HP3_all_vocals.pth": {
+        "url": "https://huggingface.co/lj1995/VoiceConversionWebUI/resolve/main/uvr5_weights/HP3_all_vocals.pth",
+        "path": "assets/uvr5_weights/HP3_all_vocals.pth",
+        "size_mb": 140,
+        "description": "UVR5 HP3 vocal model"
+    },
+    "HP5_only_main_vocal.pth": {
+        "url": "https://huggingface.co/lj1995/VoiceConversionWebUI/resolve/main/uvr5_weights/HP5_only_main_vocal.pth",
+        "path": "assets/uvr5_weights/HP5_only_main_vocal.pth",
+        "size_mb": 140,
+        "description": "UVR5 HP5 main vocal model"
+    },
+    "VR-DeEchoAggressive.pth": {
+        "url": "https://huggingface.co/lj1995/VoiceConversionWebUI/resolve/main/uvr5_weights/VR-DeEchoAggressive.pth",
+        "path": "assets/uvr5_weights/VR-DeEchoAggressive.pth",
+        "size_mb": 130,
+        "description": "UVR5 de-echo aggressive"
+    },
+    "VR-DeEchoDeReverb.pth": {
+        "url": "https://huggingface.co/lj1995/VoiceConversionWebUI/resolve/main/uvr5_weights/VR-DeEchoDeReverb.pth",
+        "path": "assets/uvr5_weights/VR-DeEchoDeReverb.pth",
+        "size_mb": 130,
+        "description": "UVR5 de-echo + de-reverb"
+    },
+    "VR-DeEchoNormal.pth": {
+        "url": "https://huggingface.co/lj1995/VoiceConversionWebUI/resolve/main/uvr5_weights/VR-DeEchoNormal.pth",
+        "path": "assets/uvr5_weights/VR-DeEchoNormal.pth",
+        "size_mb": 130,
+        "description": "UVR5 de-echo normal"
+    },
+    "onnx_dereverb_By_FoxJoy/vocals.onnx": {
+        "url": "https://huggingface.co/lj1995/VoiceConversionWebUI/resolve/main/uvr5_weights/onnx_dereverb_By_FoxJoy/vocals.onnx",
+        "path": "assets/uvr5_weights/onnx_dereverb_By_FoxJoy/vocals.onnx",
+        "size_mb": 50,
+        "description": "UVR5 ONNX dereverb"
+    },
+    "hubert_base.pt": {
+        "url": "https://huggingface.co/lj1995/VoiceConversionWebUI/resolve/main/hubert_base.pt",
+        "path": "assets/hubert/hubert_base.pt",
+        "size_mb": 189,
+        "description": "HuBERT 特征提取模型"
+    },
+    "rmvpe.pt": {
+        "url": "https://huggingface.co/lj1995/VoiceConversionWebUI/resolve/main/rmvpe.pt",
+        "path": "assets/rmvpe/rmvpe.pt",
+        "size_mb": 181,
+        "description": "RMVPE 音高提取模型"
+    },
+    "f0G48k.pth": {
+        "url": "https://huggingface.co/lj1995/VoiceConversionWebUI/resolve/main/pretrained_v2/f0G48k.pth",
+        "path": "assets/pretrained_v2/f0G48k.pth",
+        "size_mb": 55,
+        "description": "48kHz 生成器预训练权重"
+    },
+    "f0D48k.pth": {
+        "url": "https://huggingface.co/lj1995/VoiceConversionWebUI/resolve/main/pretrained_v2/f0D48k.pth",
+        "path": "assets/pretrained_v2/f0D48k.pth",
+        "size_mb": 55,
+        "description": "48kHz 判别器预训练权重"
+    },
+    "f0G40k.pth": {
+        "url": "https://huggingface.co/lj1995/VoiceConversionWebUI/resolve/main/pretrained_v2/f0G40k.pth",
+        "path": "assets/pretrained_v2/f0G40k.pth",
+        "size_mb": 55,
+        "description": "40kHz 生成器预训练权重"
+    },
+    "f0D40k.pth": {
+        "url": "https://huggingface.co/lj1995/VoiceConversionWebUI/resolve/main/pretrained_v2/f0D40k.pth",
+        "path": "assets/pretrained_v2/f0D40k.pth",
+        "size_mb": 55,
+        "description": "40kHz 判别器预训练权重"
+    }
+}
+
+# 必需模型列表
+REQUIRED_MODELS = ["hubert_base.pt", "rmvpe.pt", "HP2_all_vocals.pth"]
+
+# Mature DeEcho / DeReverb models downloaded separately
+MATURE_DEECHO_MODELS = [
+    "VR-DeEchoDeReverb.pth",
+    "onnx_dereverb_By_FoxJoy/vocals.onnx",
+    "VR-DeEchoNormal.pth",
+    "VR-DeEchoAggressive.pth",
+]
+
+
+def get_project_root() -> Path:
+    """获取项目根目录"""
+    return Path(__file__).parent.parent
+
+
+def download_file(url: str, dest_path: Path, desc: str = None) -> bool:
+    """
+    下载文件，支持断点续传和进度显示
+
+    Args:
+        url: 下载链接
+        dest_path: 目标路径
+        desc: 进度条描述
+
+    Returns:
+        bool: 下载是否成功
+    """
+    dest_path.parent.mkdir(parents=True, exist_ok=True)
+
+    # 检查已下载的部分
+    resume_pos = 0
+    if dest_path.exists():
+        resume_pos = dest_path.stat().st_size
+
+    headers = {}
+    if resume_pos > 0:
+        headers["Range"] = f"bytes={resume_pos}-"
+    if "huggingface.co" in url:
+        hf_token = (
+            os.environ.get("HF_TOKEN")
+            or os.environ.get("HUGGINGFACE_HUB_TOKEN")
+            or os.environ.get("HUGGINGFACE_TOKEN")
+        )
+        if hf_token:
+            headers["Authorization"] = f"Bearer {hf_token}"
+
+    try:
+        response = requests.get(url, headers=headers, stream=True, timeout=30)
+
+        # 检查是否支持断点续传
+        if response.status_code == 416:  # Range not satisfiable
+            print(f"  文件已完整下载: {dest_path.name}")
+            return True
+
+        if response.status_code not in [200, 206]:
+            print(f"  下载失败: HTTP {response.status_code}")
+            return False
+
+        # 获取文件总大小
+        total_size = int(response.headers.get("content-length", 0))
+        if response.status_code == 206:
+            total_size += resume_pos
+
+        # 下载模式
+        mode = "ab" if resume_pos > 0 else "wb"
+
+        with open(dest_path, mode) as f:
+            with tqdm(
+                total=total_size,
+                initial=resume_pos,
+                unit="B",
+                unit_scale=True,
+                desc=desc or dest_path.name
+            ) as pbar:
+                for chunk in response.iter_content(chunk_size=8192):
+                    if chunk:
+                        f.write(chunk)
+                        pbar.update(len(chunk))
+
+        return True
+
+    except requests.exceptions.RequestException as e:
+        print(f"  下载错误: {e}")
+        return False
+
+
+def check_model(name: str) -> bool:
+    """
+    检查模型是否已下载
+
+    Args:
+        name: 模型名称
+
+    Returns:
+        bool: 模型是否存在
+    """
+    if name not in MODELS:
+        return False
+
+    model_path = get_project_root() / MODELS[name]["path"]
+    return model_path.exists()
+
+
+def download_model(name: str) -> bool:
+    """
+    下载指定模型
+
+    Args:
+        name: 模型名称
+
+    Returns:
+        bool: 下载是否成功
+    """
+    if name not in MODELS:
+        print(f"未知模型: {name}")
+        return False
+
+    model_info = MODELS[name]
+    model_path = get_project_root() / model_info["path"]
+
+    if model_path.exists():
+        print(f"模型已存在: {name}")
+        return True
+
+    print(f"正在下载: {model_info['description']} ({model_info['size_mb']}MB)")
+    return download_file(model_info["url"], model_path, name)
+
+
+def download_required_models() -> bool:
+    """
+    下载所有必需模型
+
+    Returns:
+        bool: 是否全部下载成功
+    """
+    print("=" * 50)
+    print("检查必需模型...")
+    print("=" * 50)
+
+    success = True
+    for name in REQUIRED_MODELS:
+        if not check_model(name):
+            if not download_model(name):
+                success = False
+        else:
+            print(f"[OK] {name} 已存在")
+
+    return success
+
+
+def download_all_models() -> bool:
+    """
+    下载所有模型
+
+    Returns:
+        bool: 是否全部下载成功
+    """
+    print("=" * 50)
+    print("下载所有模型...")
+    print("=" * 50)
+
+    success = True
+    for name in MODELS:
+        if not check_model(name):
+            if not download_model(name):
+                success = False
+        else:
+            print(f"[OK] {name} 已存在")
+
+    return success
+
+
+def check_all_models() -> Dict[str, bool]:
+    """
+    检查所有模型状态
+
+    Returns:
+        dict: 模型名称 -> 是否存在
+    """
+    return {name: check_model(name) for name in MODELS}
+
+
+def get_available_mature_deecho_models() -> List[str]:
+    """Return locally available mature DeEcho / DeReverb models."""
+    return [name for name in MATURE_DEECHO_MODELS if check_model(name)]
+
+
+def get_preferred_mature_deecho_model() -> Optional[str]:
+    """Return the preferred learned DeEcho model by priority."""
+    available = set(get_available_mature_deecho_models())
+    for name in MATURE_DEECHO_MODELS:
+        if name in available:
+            return name
+    return None
+
+
+def download_mature_deecho_models() -> bool:
+    """Download mature DeEcho / DeReverb recommended models."""
+    print("=" * 50)
+    print("Downloading mature DeEcho / DeReverb models...")
+    print("=" * 50)
+
+    success = True
+    for name in MATURE_DEECHO_MODELS:
+        if not check_model(name):
+            if not download_model(name):
+                success = False
+        else:
+            print(f"[OK] {name} already exists")
+
+    return success
+
+
+def print_model_status():
+    """打印模型状态"""
+    print("=" * 50)
+    print("模型状态")
+    print("=" * 50)
+
+    status = check_all_models()
+    for name, exists in status.items():
+        info = MODELS[name]
+        mark = "OK" if exists else "MISSING"
+        print(f"  {mark} {name}")
+        print(f"      {info['description']}")
+        print(f"      大小: {info['size_mb']}MB")
+        if name in REQUIRED_MODELS:
+            print(f"      [必需]")
+        print()
+
+
+if __name__ == "__main__":
+    import argparse
+
+    parser = argparse.ArgumentParser(description="RVC 模型下载工具")
+    parser.add_argument("--check", action="store_true", help="检查模型状态")
+    parser.add_argument("--all", action="store_true", help="下载所有模型")
+    parser.add_argument("--model", type=str, help="下载指定模型")
+
+    args = parser.parse_args()
+
+    if args.check:
+        print_model_status()
+    elif args.model:
+        download_model(args.model)
+    elif args.all:
+        download_all_models()
+    else:
+        download_required_models()