Update README

.gitattributes

@@ -4,4 +4,5 @@
 *.npz filter=lfs diff=lfs merge=lfs -text
 *.bin filter=lfs diff=lfs merge=lfs -text
 *.wav filter=lfs diff=lfs merge=lfs -text
+*.mp3 filter=lfs diff=lfs merge=lfs -text
 banner.png filter=lfs diff=lfs merge=lfs -text

.gitignore

@@ -0,0 +1 @@
+*.DS_Store

README.md

@@ -96,8 +96,8 @@ model-index:
 
 ### Vietnamese Zero-Shot Text-to-Speech (TTS) with real-time streaming and voice cloning from seconds of audio. Fast, natural, and optimised for CPU inference.
 
-**The most accurate open Vietnamese TTS we know of — 13× fewer word errors than
-the next open model**, and it runs faster than real time on a laptop CPU.
+**The most accurate open Vietnamese TTS we know of — 4× fewer word errors than
+the next best model**, and it runs faster than real time on a laptop CPU.
 
 * 🎯 **Ultra-natural** — 2.91 UTMOS, ~0.5 MOS above every other open Vietnamese
   system, with near-zero dead air (0.029 s vs 0.23–0.53 s).
@@ -107,11 +107,30 @@ the next open model**, and it runs faster than real time on a laptop CPU.
 * ⚡ **Real-time on CPU, streaming** — ~2× faster than real time (RTF 0.5×),
   first audio chunk in ~70 ms. No GPU required.
 * 🇻🇳 **Built for Vietnamese** — tones, code-switched English, and a built-in
-  normalizer that reads `31/12/2025` and `1.250 tỷ` the way a person would.
+  normalizer that reads `31/12/2025` and `ZeroTTS` the way a person would.
 
 * Code, examples, browser demo: **https://github.com/zeroweight-ai/ZeroTTS**
 * Benchmark dataset: **https://huggingface.co/datasets/zeroweight-ai/ZeroBench-TTS**
 
+## Samples
+
+**Hội thoại hai giọng**
+
+<audio controls><source src="https://huggingface.co/zeroweight-ai/ZeroTTS/resolve/main/samples/conversation.mp3" type="audio/mpeg"></audio>
+
+**Kể chuyện, dạng dài**
+
+<audio controls><source src="https://huggingface.co/zeroweight-ai/ZeroTTS/resolve/main/samples/storytelling.mp3" type="audio/mpeg"></audio>
+
+**Bản tin, chèn tiếng Anh**
+
+<audio controls><source src="https://huggingface.co/zeroweight-ai/ZeroTTS/resolve/main/samples/news-code-switch.mp3" type="audio/mpeg"></audio>
+
+**Xuyên ngôn ngữ — giọng tham chiếu tiếng Việt, đọc tiếng Anh**
+
+<audio controls><source src="https://huggingface.co/zeroweight-ai/ZeroTTS/resolve/main/samples/cross-lingual-reference-vi.mp3" type="audio/mpeg"></audio>
+<audio controls><source src="https://huggingface.co/zeroweight-ai/ZeroTTS/resolve/main/samples/cross-lingual-english.mp3" type="audio/mpeg"></audio>
+
 ```python
 pip install zerotts
 ```
@@ -127,8 +146,33 @@ tts.save_audio(audio, "out.wav")
 Streaming, with first audio in roughly 70 ms:
 
 ```python
-for chunk in tts.synthesize_stream("Một đoạn văn bản dài hơn…", voice="maichi"):
-    play(chunk)   # (1, n) float32 at 48 kHz
+import queue
+
+import numpy as np
+import sounddevice as sd   # pip install sounddevice
+
+TEXT = ("Đây là chế độ phát trực tuyến. Âm thanh được tạo ra và phát ngay lập tức, "
+        "không cần chờ toàn bộ đoạn văn hoàn thành. Nhờ vậy, người nghe chỉ mất "
+        "khoảng 70 mili giây là đã nghe thấy câu đầu tiên, ngay cả khi mô "
+        "hình đang chạy trên CPU của một chiếc laptop bình thường.")
+
+pending, tail = queue.Queue(), np.zeros(0, dtype="float32")
+
+def feed(outdata, frames, _time, _status):
+    global tail
+    while len(tail) < frames and not pending.empty():
+        tail = np.concatenate([tail, pending.get_nowait()])
+    n = min(frames, len(tail))
+    outdata[:n, 0] = tail[:n]
+    outdata[n:] = 0
+    tail = tail[n:]
+
+with sd.OutputStream(samplerate=tts.sample_rate, channels=1,
+                     dtype="float32", callback=feed):
+    for chunk in tts.synthesize_stream(TEXT, voice="maichi"):
+        pending.put(chunk.reshape(-1))   # chunk is (1, n) float32 at 48 kHz
+    while not pending.empty() or len(tail):
+        sd.sleep(50)                     # let the buffer drain before closing
 ```
 
 ## Benchmarks
@@ -146,11 +190,13 @@ spoken out, from the benchmark's own curated reading.
 | **Dead air** (excess silence) ↓ | **0.029 s** | 0.386 s | 0.568 s | 0.215 s |
 | **RTF, CPU** ↓ | **0.50×** | 6.12× | 0.71× | 0.73× |
 | **Time to first audio, CPU** ↓ | **~70 ms** | ~34 s | ~6.1 s | ~5.1 s |
-| Size | **81 M**, CPU | 3.1 GB, GPU | 1.9 GB, GPU | 1.9 GB, GPU |
+| Size | **202 M** params, 0.86 GB fp32, CPU | 3.1 GB, GPU | 1.9 GB, GPU | 1.9 GB, GPU |
 
-**4× fewer word errors than the next-best system**, and the only one of the four
-that's actually fast on CPU — the others are sized and tuned for a GPU, and it
-shows.
+**4× fewer word errors than the next-best system**, and the fastest of the four
+on CPU. The gap is much wider in latency than in throughput: the two XTTS
+fine-tunes also beat real time (0.71×) but need seconds to emit their first
+sample, while OmniVoice is 6× *slower* than real time. All three are sized and
+tuned for a GPU, and it shows.
 
 Full comparison tables, per-subset breakdowns, and CPU speed methodology:
 **[docs/BENCHMARKS.md](https://github.com/zeroweight-ai/ZeroTTS/blob/main/docs/BENCHMARKS.md)**

voices/thanhtruc/preview.wav → samples/conversation.mp3

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:97e3120194132cd37ae79d6362aaa06cfc9bd971ec59e258be4f183bacbea5b6
-size 599084
+oid sha256:ac110a35c09efbdfd91354c53330df511d9eaf9faf21a378c94fc4f15f1efc34
+size 188588

voices/thanhtruc/voice.npz → samples/cross-lingual-english.mp3

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e4103e4f3aaee2396bc2a1c986e2db1a2ae62b227ed0f1bcc2f1a9e4ef4d8ed1
-size 31262
+oid sha256:3a318f85ed126ab59f17f22ac76552a42c5240f5c445270892630513a2db62c7
+size 263468

voices/thanhtruc/voice.bin → samples/cross-lingual-reference-vi.mp3

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9d443df8e986d5b08afe934c77764eaa76c326864407c2d6f96cbb09120acef5
-size 30720
+oid sha256:1b019bb424b186aaf26c721eb2a7ed68eeb3ca9cff8f3c3fe062151fbaaea003
+size 34220

samples/news-code-switch.mp3

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:be58f3701e6169d33ee23502b0b4674be269f03c6ae113035b671a3d20f9783a
+size 544172

samples/storytelling.mp3

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f0ab526814b46d0bd6c9fdf1026338dc84089cf7e7d9da5552d334ba972d7775
+size 971564

voices/hamy/preview.wav

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6aab12b5fa3d00b5e5dffe826b30fc7c7062c9e3da403d39831dab693d1f3f0d
-size 529964
+oid sha256:4bc33835023d7786bc75cb6c5c2681b1aec1ca3fb01300044b30d2161b29a171
+size 460844

voices/hamy/voice.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a385bc5fc73bc11ebc7cbde39c447ed1b476cb9970c293bd4e26ff66fd11119d
+oid sha256:e137f9544c20c491b3d139aeebb1d8549f8bd6f84f035a2df78830d3baec025d
 size 30720

voices/hamy/voice.npz

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b922963b868a0e403690194782b57f2a11aec55f59e0314566cb80bba71b332b
+oid sha256:a864c127386f74976154c840875f6ad615d74c377768bc6edcb75e3278f228b7
 size 31262

voices/index.json

@@ -159,26 +159,6 @@
         10,
         768
       ]
-    },
-    {
-      "name": "thanhtruc",
-      "display_name": "Thanh Trúc",
-      "language": "vi",
-      "gender": "nữ",
-      "description": "nữ, trẻ, dịu dàng, thủ thỉ, nhẹ",
-      "tags": [
-        "nữ",
-        "trẻ",
-        "dịu dàng",
-        "thủ thỉ",
-        "nhẹ"
-      ],
-      "n_voice_queries": 10,
-      "shape": [
-        1,
-        10,
-        768
-      ]
     }
   ]
-}
+}

voices/maichi/preview.wav

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a2b2c3fc41b652477139fc0b63bc35399e6d0d07b55470be4603e8024a64f4bf
-size 606764
+oid sha256:987b84f2287607d58d08b4979f0bf5387a8235bc2eff28263065de34481caa67
+size 491564

voices/thanhtruc/meta.json

@@ -1,15 +0,0 @@
-{
-  "name": "thanhtruc",
-  "display_name": "Thanh Trúc",
-  "language": "vi",
-  "gender": "nữ",
-  "description": "nữ, trẻ, dịu dàng, thủ thỉ, nhẹ",
-  "tags": [
-    "nữ",
-    "trẻ",
-    "dịu dàng",
-    "thủ thỉ",
-    "nhẹ"
-  ],
-  "n_voice_queries": 10
-}