Show examples using "torchaudio.save" instead of "scipy.io.wavfile.write" in README

Since torchaudio is already showcased in the examples for loading audio, it's more consistent to also showcase it for saving audio.

README.md

@@ -185,13 +185,13 @@ model = SeamlessM4Tv2Model.from_pretrained("facebook/seamless-m4t-v2-large")
 
 # from text
 text_inputs = processor(text = "Hello, my dog is cute", src_lang="eng", return_tensors="pt")
-audio_array_from_text = model.generate(**text_inputs, tgt_lang="rus")[0].cpu().numpy().squeeze()
+audio_tensor_from_text = model.generate(**text_inputs, tgt_lang="rus")[0]
 
 # from audio
 audio, orig_freq =  torchaudio.load("https://www2.cs.uic.edu/~i101/SoundFiles/preamble10.wav")
 audio =  torchaudio.functional.resample(audio, orig_freq=orig_freq, new_freq=16_000) # must be a 16 kHz waveform array
 audio_inputs = processor(audios=audio, return_tensors="pt")
-audio_array_from_audio = model.generate(**audio_inputs, tgt_lang="rus")[0].cpu().numpy().squeeze()
+audio_tensor_from_audio = model.generate(**audio_inputs, tgt_lang="rus")[0]
 ```
 
 3. Listen to the audio samples either in an ipynb notebook:
@@ -200,18 +200,18 @@ audio_array_from_audio = model.generate(**audio_inputs, tgt_lang="rus")[0].cpu()
 from IPython.display import Audio
 
 sample_rate = model.sampling_rate
-Audio(audio_array_from_text, rate=sample_rate)
-# Audio(audio_array_from_audio, rate=sample_rate)
+Audio(audio_tensor_from_text.cpu().numpy().squeeze(), rate=sample_rate)
+# Audio(audio_tensor_from_audio.cpu().numpy().squeeze(), rate=sample_rate)
 ```
 
-Or save them as a `.wav` file using a third-party library, e.g. `scipy`:
+Or save them as a `.wav` file:
 
 ```py
-import scipy
+import torchaudio
 
 sample_rate = model.sampling_rate
-scipy.io.wavfile.write("out_from_text.wav", rate=sample_rate, data=audio_array_from_text)
-# scipy.io.wavfile.write("out_from_audio.wav", rate=sample_rate, data=audio_array_from_audio)
+torchaudio.save(uri="out_from_text.wav", src=audio_tensor_from_text, sample_rate=sample_rate, channels_first=True)
+# torchaudio.save(uri="out_from_audio.wav", src=audio_tensor_from_audio, sample_rate=sample_rate, channels_first=True)
 ```
 For more details on using the SeamlessM4T model for inference using the 🤗 Transformers library, refer to the 
 **[SeamlessM4T v2 docs](https://huggingface.co/docs/transformers/main/en/model_doc/seamless_m4t_v2)** or to this **hands-on [Google Colab](https://colab.research.google.com/github/ylacombe/scripts_and_notebooks/blob/main/v2_seamless_m4t_hugging_face.ipynb).**