INGMUYLEANG/VoxPocket-v4

VoxPocket v4: small Khmer text-to-speech

VoxPocket is a 0.8B Khmer TTS model (44.1 kHz). It was fine-tuned from VoxCPM 1.5 on real Khmer speech and on synthetic speech distilled from VoxCPM2. It reads Khmer, Khmer mixed with English words, and plain English. It can clone a voice from a short clip and take a style or a voice description.

What is new in v4 (compared with v3):

• English → Khmer switching: more practice on sentences that go from English words back to Khmer, including doubled words (Live Preview ភ្លាមភ្លាម) and Khmer endings after English (Format ទៀតទេ). In v3 these could repeat or break apart.
• Styles on cloned voices: many more clip + style pairs, so a cloned voice can change its tone.
• More synthetic data: about 2× the VoxCPM2 speech of v3.

🗣️ Try it in your browser, no download needed: VoxPocket v4 demo. It has ready voices, styles and voice design.

Use

from voxcpm.core import VoxCPM
import soundfile as sf

m = VoxCPM.from_pretrained(hf_model_id="INGMUYLEANG/VoxPocket-v4", load_denoiser=False)
sr = m.tts_model.sample_rate  # 44100

# 1) clone a voice: a clean 4–9 s clip + exactly what is said in it
wav = m.generate(text="សួស្តីបងប្អូនទាំងអស់គ្នា អរគុណច្រើនសម្រាប់ការគាំទ្រ។",
                 prompt_wav_path="my_voice.wav", prompt_text="<words spoken in my_voice.wav>",
                 cfg_value=2.0, inference_timesteps=10)
sf.write("clone.wav", wav, sr)

# 2) clone + style: put the style in brackets before the text
wav = m.generate(text="(sad, low and slow tone)សួស្តីបងប្អូនទាំងអស់គ្នា។",
                 prompt_wav_path="my_voice.wav", prompt_text="<words spoken in my_voice.wav>")

# 3) design a voice, no clip: describe it in brackets
wav = m.generate(text="(A young woman, clear and bright voice, cheerful tone)សួស្តីបងប្អូនទាំងអស់គ្នា។")

Tips:

• Reference clips: keep them short, about 4–9 s, and clean, with the exact words that are spoken.
• Long text: split it into sentences.
• Styles: the phrases used in training work best, for example sad, low and slow tone, excited and energetic, whispering quietly, happy, smiling tone, calm, slow and steady, serious and clear news-reader tone.

Results

The 40-sentence test (the same one used for v2 and v3) is not finished yet and will be added here. For reference, v3 scored a median letter error of 7.6% (v2: 12.2%).

Training

• Starting point: VoxPocket v3, about 15,500 more steps.
• Data: about 670 h in the training list.
  • 237 h of real Khmer speech
  • 270 h of synthetic VoxCPM2 speech, all checked by a speech recognizer: Khmer, mixed, English, clone, style and designed voices
  • 5 h of English → Khmer switch sentences (used twice)
  • 79 h of clip + style pairs (used twice)
• The v3 synthetic set is published as INGMUYLEANG/voxpocket-khmer-tts-v3.

Limits

• Styles on cloned voices are still weak: the tone changes only a little. Designed voices (no clip) follow descriptions better.
• Rare words and names are sometimes swapped for common ones. English loanwords inside Khmer can sound Khmer-ish.
• English is basic. The model is made for Khmer first.
• Do not use it to impersonate real people or to make deceptive audio. Only clone voices you have permission to use.

Contact

Questions about training details or hardware: contact the author through the Hugging Face profile or GitHub.