init repo

.gitattributes

@@ -0,0 +1,36 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tar filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz filter=lfs diff=lfs merge=lfs -text
+*.zip filter=lfs diff=lfs merge=lfs -text
+*.zst filter=lfs diff=lfs merge=lfs -text
+*tfevents* filter=lfs diff=lfs merge=lfs -text
+*.wav filter=lfs diff=lfs merge=lfs -text

.gitignore

@@ -0,0 +1,2 @@
+__pycache__
+.gradio

README.md

@@ -0,0 +1,12 @@
+---
+title: Vietnamese TTS 
+emoji: 🗣️
+colorFrom: red
+colorTo: red
+sdk: gradio
+sdk_version: 5.15.0
+app_file: app.py
+pinned: true
+---
+
+Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

app.py

@@ -0,0 +1,102 @@
+import spaces
+import gradio as gr
+from cached_path import cached_path
+import tempfile
+
+from f5_tts.model import DiT
+from f5_tts.infer.utils_infer import (
+    preprocess_ref_audio_text,
+    load_vocoder,
+    load_model,
+    infer_process,
+    save_spectrogram,
+)
+
+
+vocoder = load_vocoder()
+model = load_model(
+    DiT,
+    dict(dim=1024, depth=22, heads=16, ff_mult=2, text_dim=512, conv_layers=4),
+    ckpt_path=str(
+        cached_path("hf://whatvn/F5-TTS-vietnamese-tts/model.tensors")
+    ),
+    vocab_file=str(cached_path("hf://whatvn/F5-TTS-vietnamese-tts/vocab.txt")),
+)
+
+
+@spaces.GPU
+def infer(ref_audio_orig: str, gen_text: str, speed: float = 1.0):
+    if ref_audio_orig is None:
+        raise gr.Error("Reference audio is required.")
+
+    if gen_text is None or gen_text.strip() == "":
+        raise gr.Error("Text to generate is required.")
+
+    try:
+        ref_audio, ref_text = preprocess_ref_audio_text(ref_audio_orig, "")
+        final_wave, final_sample_rate, combined_spectrogram = infer_process(
+            ref_audio,
+            ref_text,
+            gen_text,
+            model,
+            vocoder,
+            cross_fade_duration=0.15,
+            nfe_step=32,
+            speed=speed,
+        )
+
+        with tempfile.NamedTemporaryFile(
+            suffix=".png", delete=False
+        ) as tmp_spectrogram:
+            spectrogram_path = tmp_spectrogram.name
+            save_spectrogram(combined_spectrogram, spectrogram_path)
+
+        return (final_sample_rate, final_wave), spectrogram_path
+    except Exception as e:
+        raise gr.Error(f"An error occurred during inference: {e}")
+
+
+iface = gr.Interface(
+    title="Vietnamese TTS",
+    description="Vietnamese TTS model trained with public data (around 200 hours Vietnamese voice) using [F5-TTS](https://github.com/SWivid/F5-TTS) model",
+    fn=infer,
+    inputs=[
+        gr.components.Audio(type="filepath", label="Reference Audio"),
+        gr.components.Textbox(label="Text to Generate", lines=3),
+        gr.components.Slider(
+            label="Speed",
+            minimum=0.3,
+            maximum=2.0,
+            value=1.0,
+            step=0.1,
+            info="Adjust the speed of the audio.",
+        ),
+    ],
+    outputs=[
+        gr.components.Audio(type="numpy", label="Synthesized Audio"),
+        gr.components.Image(type="filepath", label="Spectrogram"),
+    ],
+    submit_btn="Synthesize",
+    clear_btn=None,
+    flagging_mode="never",
+    examples=[
+        [
+            "examples/01.wav",
+            "Kiểm soát cảm xúc thực chất là một quá trình đánh giá lại bản thân, để tìm thấy tự do, thoát khỏi sự cuốn hút của chính bản ngã.",
+            0.8,
+        ],
+        [
+            "examples/02.wav",
+            "Ngoài ra, nội dung ở bên kênh đấy tôi sẽ cố gắng là không nói bậy nhá.",
+            1.0,
+        ],
+        [
+            "examples/01.wav",
+            "Cho tôi năm trăm triệu tôi sẽ gạch tên Pew và con tôi ra khỏi danh sách bạn bè, thực tế còn chịu tham gia một trận bốc xing để kết thúc tình nghĩa.",
+            0.8,
+        ],
+    ],
+)
+
+if __name__ == "__main__":
+    iface.queue().launch()

requirements.txt

@@ -0,0 +1,6 @@
+torch
+torchaudio
+soundfile
+transformers
+f5_tts @ git+https://github.com/SWivid/F5-TTS.git
+bitsandbytes>0.37.0