Why the speaking does not pause at period or comma?

[thanhnew2001]

🐛 Bug

The speech does not include period or comma

To Reproduce

txt = f""" hello. how are you? """ print(f"Text: {txt}")

txt = preprocess_text(txt, text_mapper, hps, lang=LANG) stn_tst = text_mapper.get_text(txt, hps)

with torch.no_grad(): x_tst = stn_tst.unsqueeze(0).to(device) x_tst_lengths = torch.LongTensor([stn_tst.size(0)]).to(device)

start_time = time.time()
hyp = net_g.infer(
    x_tst, x_tst_lengths, noise_scale=0.667,
    noise_scale_w=0.9, length_scale=1.0
)[0][0, 0].cpu().float().numpy()
end_time = time.time()
inference_time = end_time - start_time

print(f"Generated audio (Inference Time: {inference_time:.2f} seconds)") Audio(hyp, rate=hps.data.sampling_rate)

Code sample

Expected behavior

Environment

• fairseq Version (e.g., 1.0 or main):
• PyTorch Version (e.g., 1.0)
• OS (e.g., Linux):
• How you installed fairseq (pip, source):
• Build command you used (if compiling from source):
• Python version:
• CUDA/cuDNN version:
• GPU models and configuration:
• Any other relevant information:

Additional context