Switch to Gradio SDK

Dockerfile

@@ -1,19 +0,0 @@
-# Use a more complete base image
-FROM python:3.11
-
-WORKDIR /app
-
-# System deps - use a single RUN to avoid cache issues
-RUN apt-get update && apt-get install -y espeak-ng libespeak-ng1 && rm -rf /var/lib/apt/lists/*
-
-# Python deps
-COPY requirements.txt .
-RUN pip install --no-cache-dir -r requirements.txt
-
-# App
-COPY main.py .
-
-ENV PORT=7860
-EXPOSE 7860
-
-CMD ["uvicorn", "main:app", "--host", "0.0.0.0", "--port", "7860"]

Dockerfile.test

@@ -1,5 +0,0 @@
-FROM python:3.11-slim
-WORKDIR /app
-COPY test_build.py .
-RUN pip install fastapi uvicorn
-CMD ["uvicorn", "test_build:app", "--host", "0.0.0.0", "--port", "7860"]

README.md

@@ -3,14 +3,14 @@ title: Kokoro 82M TTS
 emoji: 🗣️
 colorFrom: blue
 colorTo: purple
-sdk: docker
+sdk: gradio
 app_port: 7860
 pinned: false
 ---
 
-# Kokoro 82M TTS API
+# Kokoro 82M TTS
 
-High-quality, fast text-to-speech on CPU using **Kokoro 82M** (Apache 2.0).
+High-quality text-to-speech on CPU using **Kokoro 82M** (Apache 2.0).
 
 ## Features
 
@@ -18,39 +18,7 @@ High-quality, fast text-to-speech on CPU using **Kokoro 82M** (Apache 2.0).
 - **Speed**: RTF ~0.64 on CPU Basic (1.6× faster than real-time)
 - **Voices**: 50+ voices (US/UK English, multiple speakers)
 - **License**: Apache 2.0 (commercial use OK)
-- **Simple REST API**: `POST /tts` → WAV audio
-
-## API Endpoints
-
-### POST /tts
-
-Synthesize speech from text.
-
-**Request:**
-```json
-{
-  "text": "Hello, this is Kokoro 82M!",
-  "voice": "af_heart",
-  "speed": 1.0,
-  "sample_rate": 24000
-}
-```
-
-**Parameters:**
-- `text` (required): Text to synthesize
-- `voice` (default: `af_heart`): Voice ID (see `/health` or `/` for full list)
-- `speed` (default: `1.0`): Speed multiplier (0.5–2.0)
-- `sample_rate` (default: `24000`): Output sample rate (Kokoro native is 24kHz)
-
-**Response:** WAV audio file (24kHz, 16-bit PCM)
-
-### GET /health
-
-Health check + loaded voices.
-
-### GET /
-
-API info + voice catalog organized by category.
+- **Interface**: Gradio web UI + REST API
 
 ## Voice Categories
 
@@ -62,47 +30,28 @@ API info + voice catalog organized by category.
 | `bm_` | UK Male | `bm_george`, `bm_lewis` |
 | Others | Multilingual/Other | `jf_alpha` (JA), `zf_xiaobei` (ZH), etc. |
 
-## Usage Example (curl)
-
-```bash
-curl -X POST "https://your-username-kokoro-tts.hf.space/tts" \
-  -H "Content-Type: application/json" \
-  -d '{"text": "你好，这是 Kokoro 82M 语音合成测试", "voice": "af_heart"}' \
-  -o output.wav
-```
-
-## Python Client
-
-```python
-import requests
+## REST API (also available)
 
-url = "https://your-username-kokoro-tts.hf.space/tts"
-resp = requests.post(url, json={"text": "Hello world!", "voice": "am_michael"})
-with open("output.wav", "wb") as f:
-    f.write(resp.content)
+### POST /tts
+```json
+{"text": "Hello!", "voice": "af_heart", "speed": 1.0}
 ```
+Returns: WAV audio (24kHz, 16-bit PCM)
 
-## Deployment
-
-This Space uses Docker (`Dockerfile`) with:
-- Python 3.11 slim
-- `espeak-ng` for grapheme-to-phoneme
-- `kokoro-onnx` + `onnxruntime` for inference
-- Model files auto-downloaded on first run (~82 MB ONNX + ~30 MB voices bin)
+### GET /health
+Health check + loaded voices.
 
-**Startup time**: ~15-20 seconds (first run downloads models)
-**Concurrent requests**: Limited by CPU; recommend sequential or small queue
+### GET /
+API info + voice catalog.
 
 ## Model Info
 
 - **Architecture**: StyleTTS2-inspired, 82M parameters
-- **Training data**: High-quality English speech (Gigaspeech, etc.)
 - **Output**: 24kHz mono
-- **Source**: `hexgrad/Kokoro-82M` (PyTorch) → `onnx-community/Kokoro-82M-v1.0-ONNX`
+- **Source**: `hexgrad/Kokoro-82M` → `onnx-community/Kokoro-82M-v1.0-ONNX`
 - **ONNX release**: `thewh1teagle/kokoro-onnx` model files
 
 ## License
 
 - Model weights: Apache 2.0
 - Code: MIT
-# force rebuild

app.py

@@ -0,0 +1,125 @@
+import gradio as gr
+from kokoro_onnx import Kokoro
+import soundfile as sf
+import numpy as np
+from pathlib import Path
+import requests
+import tempfile
+import os
+
+# Model setup
+MODEL_DIR = Path("/tmp/kokoro_models")
+MODEL_DIR.mkdir(exist_ok=True)
+MODEL_PATH = MODEL_DIR / "kokoro-v1.0.onnx"
+VOICES_PATH = MODEL_DIR / "voices-v1.0.bin"
+MODEL_URL = "https://github.com/thewh1teagle/kokoro-onnx/releases/download/model-files-v1.0/kokoro-v1.0.onnx"
+VOICES_URL = "https://github.com/thewh1teagle/kokoro-onnx/releases/download/model-files-v1.0/voices-v1.0.bin"
+
+kokoro = None
+
+def download_file(url, dest):
+    if dest.exists():
+        return
+    print(f"Downloading {dest.name}...")
+    with requests.get(url, stream=True, timeout=120, allow_redirects=True) as r:
+        r.raise_for_status()
+        with open(dest, "wb") as f:
+            for chunk in r.iter_content(chunk_size=8192):
+                f.write(chunk)
+    print(f"✓ Downloaded {dest.name}")
+
+def get_kokoro():
+    global kokoro
+    if kokoro is not None:
+        return kokoro
+    download_file(MODEL_URL, MODEL_PATH)
+    download_file(VOICES_URL, VOICES_PATH)
+    kokoro = Kokoro(str(MODEL_PATH), str(VOICES_PATH))
+    print(f"✓ Kokoro loaded. Voices: {len(kokoro.get_voices())}")
+    return kokoro
+
+# Warm up
+try:
+    get_kokoro()
+except Exception as e:
+    print(f"Warmup failed: {e}")
+
+def tts(text, voice, speed):
+    if not text.strip():
+        return None, "请输入文本"
+    
+    k = get_kokoro()
+    voices = k.get_voices()
+    if voice not in voices:
+        return None, f"语音 '{voice}' 不可用"
+    
+    try:
+        audio, sr = k.create(text, voice=voice, speed=speed, lang="en-us")
+        # Save to temp file for Gradio
+        with tempfile.NamedTemporaryFile(suffix=".wav", delete=False) as f:
+            sf.write(f.name, audio, sr, format="WAV", subtype="PCM_16")
+            return f.name, "生成成功"
+    except Exception as e:
+        return None, f"生成失败: {e}"
+
+# Get voice list
+try:
+    voices_list = get_kokoro().get_voices()
+except:
+    voices_list = ["af_heart", "af_bella", "am_michael", "am_adam", "bf_emma", "bm_george"]
+
+# Categorize voices
+us_female = [v for v in voices_list if v.startswith("af_")]
+us_male = [v for v in voices_list if v.startswith("am_")]
+uk_female = [v for v in voices_list if v.startswith("bf_")]
+uk_male = [v for v in voices_list if v.startswith("bm_")]
+other = [v for v in voices_list if not v.startswith(("af_", "am_", "bf_", "bm_"))]
+
+voice_choices = us_female + us_male + uk_female + uk_male + other
+
+with gr.Blocks(title="Kokoro 82M TTS") as demo:
+    gr.Markdown("# 🗣️ Kokoro 82M TTS - High Quality CPU TTS")
+    gr.Markdown("**Model**: Kokoro 82M (Apache 2.0) | **Quality**: UTMOS MOS ~4.45 | **Speed**: RTF ~0.64 on CPU")
+    
+    with gr.Row():
+        with gr.Column(scale=2):
+            text = gr.Textbox(
+                label="Text to synthesize",
+                placeholder="Enter text here...",
+                lines=3,
+                value="Hello, this is Kokoro 82M text to speech!"
+            )
+            voice = gr.Dropdown(
+                label="Voice",
+                choices=voice_choices,
+                value="af_heart" if "af_heart" in voice_choices else voice_choices[0],
+                info=f"Total {len(voice_choices)} voices available"
+            )
+            speed = gr.Slider(0.5, 2.0, value=1.0, step=0.1, label="Speed")
+            btn = gr.Button("Generate", variant="primary")
+            status = gr.Textbox(label="Status", interactive=False)
+        
+        with gr.Column(scale=1):
+            audio_out = gr.Audio(label="Output", type="filepath")
+    
+    btn.click(tts, inputs=[text, voice, speed], outputs=[audio_out, status])
+    
+    with gr.Accordion("Voice Categories", open=False):
+        gr.Markdown(f"""
+        **US Female** ({len(us_female)}): {', '.join(us_female[:10])}{'...' if len(us_female) > 10 else ''}
+        **US Male** ({len(us_male)}): {', '.join(us_male[:10])}{'...' if len(us_male) > 10 else ''}
+        **UK Female** ({len(uk_female)}): {', '.join(uk_female[:10])}{'...' if len(uk_female) > 10 else ''}
+        **UK Male** ({len(uk_male)}): {', '.join(uk_male[:10])}{'...' if len(uk_male) > 10 else ''}
+        **Other** ({len(other)}): {', '.join(other[:10])}{'...' if len(other) > 10 else ''}
+        """)
+    
+    gr.Markdown("""
+    ---
+    **API**: Also available via REST:
+    - `POST /tts` - `{"text": "...", "voice": "af_heart", "speed": 1.0}` → WAV
+    - `GET /health` - Health check
+    - `GET /` - Voice catalog
+    """)
+
+if __name__ == "__main__":
+    demo.launch(server_name="0.0.0.0", server_port=7860)

requirements.txt

@@ -1,7 +1,5 @@
-fastapi==0.115.0
-uvicorn[standard]==0.32.0
+gradio==4.44.0
 kokoro-onnx==0.5.0
 onnxruntime==1.20.0
 soundfile==0.12.1
-pydantic==2.9.0
 requests==2.32.0

test_main.py

@@ -1,5 +0,0 @@
-from fastapi import FastAPI
-app = FastAPI()
-@app.get("/")
-def root():
-    return {"message": "test"}