chunked/nltk_chunking/tasks_text-to-speech/chunk_95.txt

inputs = processor(text=text, return_tensors="pt")

Create a spectrogram with your model: 

spectrogram = model.generate_speech(inputs["input_ids"], speaker_embeddings)

Visualize the spectrogram, if you'd like to: 

plt.figure()
plt.imshow(spectrogram.T)
plt.show()

Finally, use the vocoder to turn the spectrogram into sound.