1

app_local.py

@@ -28,35 +28,27 @@ def local_move2gpu(x):
 
 
 # ---------------------------------------------------------------------------
-# Model loading (lazy, singleton) / 模型懒加载（单例）
+# Model loading (eager, at startup) / 启动时立即加载，常驻内存
 # ---------------------------------------------------------------------------
-_model = None
-_separator = None
-
-
-def _load_model_impl():
-    """Internal: load YingMusicSinger (no GPU decorator, called inside GPU context)."""
-    download_files(task="infer")
-    global _model
-    if _model is None:
-        from src.YingMusicSinger.infer.YingMusicSinger import YingMusicSinger
-        _model = YingMusicSinger.from_pretrained("ASLP-lab/YingMusic-Singer")
-        _model = local_move2gpu(_model)
-        _model.eval()
-    return _model
-
-
-def _load_separator_impl():
-    """Internal: load MelBandRoformer separator (no GPU decorator, called inside GPU context)."""
-    download_files(task="infer")
-    global _separator
-    if _separator is None:
-        from src.third_party.MusicSourceSeparationTraining.inference_api import Separator
-        _separator = Separator(
-            config_path="ckpts/config_vocals_mel_band_roformer_kj.yaml",
-            checkpoint_path="ckpts/MelBandRoformer.ckpt",
-        )
-    return _separator
+print("🔄 Downloading required files...")
+download_files(task="infer")
+
+print("🔄 Loading YingMusicSinger model...")
+from src.YingMusicSinger.infer.YingMusicSinger import YingMusicSinger
+_model = YingMusicSinger.from_pretrained("ASLP-lab/YingMusic-Singer")
+_model = local_move2gpu(_model)
+_model.eval()
+print("✅ YingMusicSinger model loaded.")
+
+print("🔄 Loading MelBandRoformer separator...")
+from src.third_party.MusicSourceSeparationTraining.inference_api import Separator
+_separator = Separator(
+    config_path="ckpts/config_vocals_mel_band_roformer_kj.yaml",
+    checkpoint_path="ckpts/MelBandRoformer.ckpt",
+    device=LOCAL_DEVICE
+)
+print("✅ MelBandRoformer separator loaded.")
+print("🎤 All models ready. Starting UI...")
 
 
 # ---------------------------------------------------------------------------
@@ -67,7 +59,7 @@ def _separate_vocals_impl(audio_path: str) -> tuple:
     Separate audio into vocals and accompaniment using MelBandRoformer.
     Must be called within an active GPU context.
     """
-    separator = _load_separator_impl()
+    separator = _separator
 
     wav, sr = torchaudio.load(audio_path)
     vocal_wav, inst_wav, out_sr = separator.separate(wav, sr)
@@ -118,10 +110,8 @@ def mix_vocal_and_accompaniment(
 
 # ---------------------------------------------------------------------------
 # Inference wrapper / 推理入口
-# Single @spaces.GPU scope covers ALL heavy work (separation + synthesis)
-# so models stay resident in GPU memory across steps within one call.
 # ---------------------------------------------------------------------------
-
+@spaces.GPU
 def synthesize(
     ref_audio,
     melody_audio,
@@ -135,16 +125,8 @@ def synthesize(
     cfg_strength,
     seed,
 ):
-    """
-    主合成流程 / Main synthesis pipeline.
-
-    1. (可选) 用 MelBandRoformer 分离参考音频和旋律音频的人声与伴奏
-    2. 送入 YingMusicSinger 合成
-    3. (可选) 将合成人声与旋律音频的伴奏混合
-    """
     import random
 
-    # ---- 输入校验 / Input validation ----------------------------------------
     if ref_audio is None:
         raise gr.Error("请上传参考音频 / Please upload Reference Audio")
     if melody_audio is None:
@@ -153,6 +135,8 @@ def synthesize(
         raise gr.Error("请输入参考音频对应的歌词 / Please enter Reference Text")
     if not target_text.strip():
         raise gr.Error("请输入目标合成歌词 / Please enter Target Text")
+    if mix_accompaniment_flag and not separate_vocals_flag:
+        raise gr.Error("「输出时混入伴奏」需要先开启「分离人声后过模型」/ 'Mix accompaniment into output' requires 'Separate vocals before synthesis' to be enabled first")
 
     ref_audio_path = ref_audio if isinstance(ref_audio, str) else ref_audio[0]
     melody_audio_path = (
@@ -163,7 +147,6 @@ def synthesize(
     if actual_seed < 0:
         actual_seed = random.randint(0, 2**31 - 1)
 
-    # ---- Step 1: 人声分离（合并在同一 GPU 上下文中）/ Vocal separation (same GPU context) ----
     melody_accomp_path = None
     actual_ref_path = ref_audio_path
     actual_melody_path = melody_audio_path
@@ -175,10 +158,7 @@ def synthesize(
         melody_vocals_path, melody_accomp_path = _separate_vocals_impl(melody_audio_path)
         actual_melody_path = melody_vocals_path
 
-    # ---- Step 2: 模型推理 / Model inference (same GPU context) ---------------
-    model = _load_model_impl()
-
-    audio_tensor, sr = model(
+    audio_tensor, sr = _model(
         ref_audio_path=actual_ref_path,
         melody_audio_path=actual_melody_path,
         ref_text=ref_text.strip(),
@@ -194,7 +174,6 @@ def synthesize(
     vocal_out_path = os.path.join(tempfile.mkdtemp(), "vocal_output.wav")
     torchaudio.save(vocal_out_path, audio_tensor.to("cpu"), sample_rate=sr)
 
-    # ---- Step 3: 混合伴奏 / Mix accompaniment (optional) ---------------------
     if (
         separate_vocals_flag
         and mix_accompaniment_flag
@@ -210,37 +189,64 @@ def synthesize(
 # Example presets / 预设示例
 # ---------------------------------------------------------------------------
 EXAMPLES_MELODY_CONTROL = [
-    # [ref_audio, melody_audio, ref_text, target_text, sep, mix, sil, t_shift, nfe, cfg, seed]
     [
-        "examples/melody_control/ref_01.wav",
-        "examples/melody_control/melody_01.wav",
-        "该体谅的不执着|如果那天我",
-        "好多天|看不完你",
-        True, False, 0.5, 0.5, 32, 3.0, -1,
+        "examples/hf_space/melody_control/melody_control_ZH_01_timbre.wav",
+        "examples/hf_space/melody_control/melody_control_ZH_01_melody.wav",
+        "人和人的沟通|有时候没有用",
+        "此刻记忆中的点滴啊|能否再次被珍藏",
+        True, True, 0.5, 0.5, 32, 3.0, -1,
     ],
     [
-        "examples/melody_control/ref_02.wav",
-        "examples/melody_control/melody_02.wav",
-        "月光下的身影|渐渐模糊",
-        "星光照亮前路|指引方向",
-        True, False, 0.5, 0.5, 32, 3.0, -1,
+        "examples/hf_space/melody_control/melody_control_EN_01_timbre.wav",
+        "examples/hf_space/melody_control/melody_control_EN_01_melody.wav",
+        "i don't know feel|but i wanna try",
+        "won't open the door|and say tomorrow",
+        True, True, 0.5, 0.5, 32, 3.0, -1,
     ],
+    [
+        "examples/hf_space/melody_control/melody_control_EN_02_timbre.wav",
+        "examples/hf_space/melody_control/melody_control_EN_02_melody.wav",
+        "and she'll never know your story like|i do",
+        "你将安然无恙|无人能再伤你",
+        False, False, 0.5, 0.5, 32, 3.0, -1,
+    ],
+    [
+        "examples/hf_space/melody_control/melody_control_ZH_02_timbre.wav",
+        "examples/hf_space/melody_control/melody_control_ZH_02_melody.wav",
+        "就让你|在别人怀里|快乐",
+        "Missing you in my mind|missing you in my heart",
+        False, False, 0.5, 0.5, 32, 3.0, -1,
+    ]
 ]
 
 EXAMPLES_LYRIC_EDIT = [
     [
-        "examples/lyric_edit/ref_01.wav",
-        "examples/lyric_edit/melody_01.wav",
-        "该体谅的不执着|如果那天我",
-        "忘不掉的笑容|留在心里面",
-        True, False, 0.5, 0.5, 32, 3.0, -1,
+        "examples/hf_space/lyric_edit/SingEdit_ZH_01.wav",
+        "examples/hf_space/lyric_edit/SingEdit_ZH_01.wav",
+        "天青色等烟雨|而我在等你|炊烟袅袅升起",
+        "阳光中赏花香|花瓣在飘落|山间幽静致远",
+        True, True, 0.5, 0.5, 32, 3.0, -1,
+    ],
+    [
+        "examples/hf_space/lyric_edit/SingEdit_EN_01.wav",
+        "examples/hf_space/lyric_edit/SingEdit_EN_01.wav",
+        "can you tell my heart is speaking|my eyes will give you clues",
+        "can you spot the moon is grinning|my lips will show you hints",
+        True, True, 0.5, 0.5, 32, 3.0, -1,
     ],
     [
-        "examples/lyric_edit/ref_02.wav",
-        "examples/lyric_edit/melody_02.wav",
-        "夜深了还不睡|想着你的脸",
-        "春风又吹过来|带走我思念",
-        True, False, 0.5, 0.5, 32, 3.0, -1,
+        "examples/hf_space/lyric_edit/SingEdit_ZH_02.wav",
+        "examples/hf_space/lyric_edit/SingEdit_ZH_02.wav",
+        "歌声是翅膀|唱出了希望|所有的付出只因爱的力量|和你一样",
+        "火锅是梦想|煮出了欢畅|全部的辛劳全因肉的力量|与汤一样",
+        False, False, 0.5, 0.5, 32, 3.0, -1,
+    ],
+    [
+        "examples/hf_space/lyric_edit/SingEdit_EN_02.wav",
+        "examples/hf_space/lyric_edit/SingEdit_EN_02.wav",
+        "i can hear what you say|now i know|why know we can|make it",
+        "i can see where you go|but i say|why not we will|break it",
+        False, False, 0.5, 0.5, 32, 3.0, -1,
     ],
 ]
 
@@ -252,14 +258,16 @@ CUSTOM_CSS = """
 @import url('https://fonts.googleapis.com/css2?family=DM+Sans:ital,opsz,wght@0,9..40,300;0,9..40,500;0,9..40,700;1,9..40,400&family=Playfair+Display:wght@600;800&display=swap');
 
 :root {
-    --primary: #e85d04;
-    --primary-light: #f48c06;
+    --primary: #44ACFF;
+    --primary-light: #89D4FF;
+    --primary-warm: #FE9EC7;
+    --palette-yellow: #F9F6C4;
     --bg-dark: #0d1117;
     --surface: #161b22;
     --surface-light: #21262d;
     --text: #f0f6fc;
     --text-muted: #8b949e;
-    --accent-glow: rgba(232, 93, 4, 0.15);
+    --accent-glow: rgba(68, 172, 255, 0.10);
     --border: #30363d;
 }
 
@@ -269,7 +277,7 @@ CUSTOM_CSS = """
     margin: auto !important;
 }
 
-/* ---------- Badge links: no underline, no gap artifacts ---------- */
+/* ---------- Badge links ---------- */
 #app-header .badges a {
     text-decoration: none !important;
     display: inline-block;
@@ -286,7 +294,7 @@ CUSTOM_CSS = """
     line-height: 1.8;
 }
 
-/* ---------- Header / 头部 ---------- */
+/* ---------- Header ---------- */
 #app-header {
     text-align: center;
     padding: 1.8rem 1rem 0.5rem;
@@ -336,17 +344,29 @@ CUSTOM_CSS = """
     font-weight: 600;
 }
 
-/* ---------- Section labels / 分区标题 ---------- */
-.section-title {
-    font-family: 'DM Sans', sans-serif !important;
-    font-weight: 700 !important;
-    font-size: 1rem !important;
+/* ========== 2. Section labels: left accent bar ========== */
+.resume-container .section-title {
+    border: none;
+    outline: none;
+    box-shadow: none;
+    font-family: 'DM Sans', sans-serif;
+    font-weight: 700;
+    font-size: 1rem;
     letter-spacing: 0.06em;
     text-transform: uppercase;
-    color: var(--primary-light) !important;
-    border-bottom: 2px solid var(--primary);
-    padding-bottom: 6px;
-    margin-bottom: 12px !important;
+    color: var(--primary);
+    display: block;
+    padding: 3px 0 3px 10px;
+    margin-bottom: 14px;
+    border-left: 4px solid var(--primary-warm);
+    background: linear-gradient(90deg, rgb(254 158 199 / 8%) 0%, transparent 70%);
+    border-radius: 0 4px 4px 0;
+}
+
+.resume-container .section-title * {
+    border: inherit;
+    outline: none;
+    box-shadow: none;
 }
 
 /* ---------- Example tabs ---------- */
@@ -355,29 +375,41 @@ CUSTOM_CSS = """
     font-size: 0.95rem !important;
 }
 
-/* ---------- Run button / 合成按钮 ---------- */
+/* ========== 3. Run button — palette blue ========== */
 #run-btn {
-    background: linear-gradient(135deg, #e85d04, #dc2f02) !important;
+    background: linear-gradient(135deg, #44ACFF, #89D4FF) !important;
     border: none !important;
-    color: #fff !important;
+    color: #0a1628 !important;
     font-weight: 700 !important;
     font-size: 1.1rem !important;
     letter-spacing: 0.04em;
     padding: 12px 0 !important;
     border-radius: 10px !important;
     transition: transform 0.15s, box-shadow 0.25s !important;
-    box-shadow: 0 4px 20px rgba(232, 93, 4, 0.35) !important;
+    box-shadow: 0 4px 20px rgba(68, 172, 255, 0.35) !important;
 }
 #run-btn:hover {
     transform: translateY(-1px) !important;
-    box-shadow: 0 6px 28px rgba(232, 93, 4, 0.5) !important;
+    box-shadow: 0 6px 28px rgba(68, 172, 255, 0.5) !important;
 }
 
-/* ---------- Output audio / 输出音频 ---------- */
+/* ---------- Output audio ---------- */
 #output-audio {
-    border: 2px solid var(--primary) !important;
+    border: 2px solid #44ACFF !important;
     border-radius: 12px !important;
-    background: var(--accent-glow) !important;
+    background: rgba(68, 172, 255, 0.07) !important;
+}
+
+/* ========== 4. Vocal-sep info box: fix highlighted text colors ========== */
+.vocal-sep-info b {
+    color: #c9d1d9;
+    font-weight: 700;
+}
+b[data-ref] {
+    color: #FE9EC7 !important;
+}
+b[data-melody] {
+    color: #89D4FF !important;
 }
 """
 
@@ -472,7 +504,7 @@ def build_ui():
                     type="filepath",
                 )
             with gr.Column(scale=1):
-                gr.Markdown("#### ✏️ 歌词输入 / Lyrics", elem_classes="section-title")
+                gr.Markdown("#### ✏️ 歌词输入 / Lyrics")
                 ref_text = gr.Textbox(
                     label="参考音频歌词 / Reference Lyrics",
                     placeholder="例如 / e.g.：该体谅的不执着|如果那天我",
@@ -485,58 +517,17 @@ def build_ui():
                 )
 
         # ================================================================
-        # ROW 2 – 预设示例 / Example Presets  ← before vocal separation
+        # ROW 2 – 伴奏分离
         # ================================================================
         gr.HTML("<hr style='border-color:#30363d; margin: 16px 0 12px;'>")
-        gr.Markdown("#### 🎵 预设示例 / Example Presets", elem_classes="section-title")
-        gr.Markdown(
-            "<small style='color:#8b949e;'>点击任意行自动填入上方输入区域 / Click any row to auto-fill the inputs above</small>"
-        )
-
-        # Hidden advanced-param components so gr.Examples can reference them
-        # (real sliders rendered inside the accordion below override these values)
-        with gr.Row(visible=False):
-            _sep_flag_ex   = gr.Checkbox(value=True)
-            _mix_flag_ex   = gr.Checkbox(value=False)
-            _sil_ex        = gr.Number(value=0.5)
-            _tshift_ex     = gr.Number(value=0.5)
-            _nfe_ex        = gr.Number(value=32)
-            _cfg_ex        = gr.Number(value=3.0)
-            _seed_ex       = gr.Number(value=-1, precision=0)
+        gr.Markdown("#### 🎚️ 伴奏分离 / Vocal Separation")
 
-        _example_inputs = [
-            ref_audio, melody_audio, ref_text, target_text,
-            _sep_flag_ex, _mix_flag_ex,
-            _sil_ex, _tshift_ex, _nfe_ex, _cfg_ex, _seed_ex,
-        ]
-
-        with gr.Tabs():
-            with gr.Tab("🎼 Melody Control"):
-                gr.Examples(
-                    examples=EXAMPLES_MELODY_CONTROL,
-                    inputs=_example_inputs,
-                    label="Melody Control Examples",
-                    examples_per_page=5,
-                )
-            with gr.Tab("✏️ Lyric Edit"):
-                gr.Examples(
-                    examples=EXAMPLES_LYRIC_EDIT,
-                    inputs=_example_inputs,
-                    label="Lyric Edit Examples",
-                    examples_per_page=5,
-                )
-
-        # ================================================================
-        # ROW 3 – 伴奏分离 / Vocal Separation
-        # ================================================================
-        gr.HTML("<hr style='border-color:#30363d; margin: 16px 0 12px;'>")
-        gr.Markdown("#### 🎚️ 伴奏分离 / Vocal Separation", elem_classes="section-title")
         gr.HTML("""
 <div style="font-size:0.85rem; color:#8b949e; line-height:1.75; margin: 0 0 12px; padding: 10px 16px;
             background: rgba(255,255,255,0.03); border-radius: 8px; border: 1px solid #21262d;">
   <ul style="margin:0; padding-left:1.2em; list-style: none;">
     <li style="margin-bottom:7px;">
-      💡 若输入的<b style="color:#c9d1d9;">参考音频</b>或<b style="color:#c9d1d9;">旋律音频</b>中含有伴奏或背景噪音，请开启「分离人声后过模型」—— 模型基于纯人声训练，混合音频会影响合成质量。<br>
+      💡 若输入的<b style="color:#FE9EC7; font-weight:700;">参考音频</b>或<b style="color:#89D4FF; font-weight:700;">旋律音频</b>中含有伴奏或背景噪音，请开启「分离人声后过模型」—— 模型基于纯人声训练，混合音频会影响合成质量。<br>
       <span style="color:#6e7681; font-size:0.82rem;">If either input contains accompaniment or background noise, enable <i>Separate vocals before synthesis</i> — the model is trained on clean vocals only and mixed audio degrades quality.</span>
     </li>
     <li style="margin-bottom:7px;">
@@ -544,7 +535,7 @@ def build_ui():
       <span style="color:#6e7681; font-size:0.82rem;">If both inputs are already clean vocals, skip separation — enabling it unnecessarily may introduce artifacts from the separation model.</span>
     </li>
     <li>
-      💡 若旋律音频含有伴奏，开启「分离人声后过模型」后，最终输出是否保留伴奏由「输出时混入伴奏」控制。<br>
+      💡 若<b style="color:#89D4FF; font-weight:700;">旋律音频</b>含有伴奏，开启「分离人声后过模型」后，最终输出是否保留伴奏由「输出时混入伴奏」控制。<br>
       <span style="color:#6e7681; font-size:0.82rem;">If the melody audio contains accompaniment and separation is enabled, use <i>Mix accompaniment into output</i> to decide whether to include it in the final result.</span>
     </li>
   </ul>
@@ -563,6 +554,9 @@ def build_ui():
                 info="将合成人声与分离出的伴奏混合作为最终输出（需先开启人声分离）/ Mix synthesised vocals with the separated accompaniment (requires separation enabled)",
             )
 
+        # ================================================================
+        # ROW 3 – 高级参数
+        # ================================================================
         with gr.Accordion("⚙️ 高级参数 / Advanced Parameters", open=False):
             with gr.Row():
                 nfe_step = gr.Slider(
@@ -592,7 +586,45 @@ def build_ui():
                 )
 
         # ================================================================
-        # ROW 5 – 合成按钮与输出 / Run & Output
+        # ROW 4 – 预设示例（放在所有真实控件定义之后）
+        # ================================================================
+        gr.HTML("<hr style='border-color:#30363d; margin: 16px 0 12px;'>")
+        gr.Markdown("#### 🎵 预设示例 / Example Presets")
+        gr.Markdown(
+            "<small style='color:#8b949e;'>点击任意行自动填入上方输入区域 / Click any row to auto-fill the inputs above</small>"
+        )
+        gr.HTML("""
+<p style="text-align:center; font-size:0.78rem; color:#484f58; margin: 2px 0 10px; line-height:1.7;">
+  示例中所用音频片段均仅供学术研究与演示目的，不用于任何商业用途。如有版权问题，请联系作者予以删除。<br>
+  Audio clips used in the examples are for academic research and demonstration purposes only, with no commercial use intended. If you believe any content infringes your copyright, please contact the authors for removal.
+</p>
+""")
+
+        # 所有真实控件均已定义，直接绑定
+        _example_inputs = [
+            ref_audio, melody_audio, ref_text, target_text,
+            separate_vocals_flag, mix_accompaniment_flag,
+            sil_len_to_end, t_shift, nfe_step, cfg_strength, seed,
+        ]
+
+        with gr.Tabs():
+            with gr.Tab("🎼 Melody Control"):
+                gr.Examples(
+                    examples=EXAMPLES_MELODY_CONTROL,
+                    inputs=_example_inputs,
+                    label="Melody Control Examples",
+                    examples_per_page=5,
+                )
+            with gr.Tab("✏️ Lyric Edit"):
+                gr.Examples(
+                    examples=EXAMPLES_LYRIC_EDIT,
+                    inputs=_example_inputs,
+                    label="Lyric Edit Examples",
+                    examples_per_page=5,
+                )
+
+        # ================================================================
+        # ROW 5 – 合成按钮与输出
         # ================================================================
         gr.HTML("<hr style='border-color:#30363d; margin: 12px 0;'>")
         run_btn = gr.Button("🎤  开始合成 / Start Synthesizing", elem_id="run-btn", size="lg")
@@ -603,36 +635,25 @@ def build_ui():
             elem_id="output-audio",
         )
 
-        # All inputs for the synthesize() call (uses real sliders, not example placeholders)
         _all_inputs = [
             ref_audio, melody_audio, ref_text, target_text,
             separate_vocals_flag, mix_accompaniment_flag,
             sil_len_to_end, t_shift, nfe_step, cfg_strength, seed,
         ]
 
-        # ================================================================
-        # Event wiring / 事件绑定
-        # ================================================================
-        separate_vocals_flag.change(
-            fn=lambda sep: gr.update(interactive=sep, value=False),
-            inputs=[separate_vocals_flag],
-            outputs=[mix_accompaniment_flag],
-        )
-
         run_btn.click(
             fn=synthesize,
             inputs=_all_inputs,
             outputs=output_audio,
         )
 
-        # ---- 页脚：免责声明 / Footer: disclaimer ----
         gr.HTML(DISCLAIMER_HTML)
 
     return demo
 
 
 # ---------------------------------------------------------------------------
-# Entry point / 启动入口
+# Entry point
 # ---------------------------------------------------------------------------
 if __name__ == "__main__":
     demo = build_ui()