add abilithy to generate audio from another audio

notebooks/test_model.ipynb

@@ -155,6 +155,7 @@
    },
    "outputs": [],
    "source": [
+    "start_steps = 500  #@param {type:\"slider\", min:0, max:1000, step:10}\n",
     "loop = AudioDiffusion.loop_it(audio, sample_rate, loops=1)\n",
     "for variation in range(12):\n",
     "    image2, (\n",
@@ -162,7 +163,7 @@
     "    ) = audio_diffusion.generate_spectrogram_and_audio_from_audio(\n",
     "        raw_audio=audio,\n",
     "        slice=0,\n",
-    "        start_step=500,\n",
+    "        start_step=start_steps,\n",
     "        steps=1000)\n",
     "    display(image2)\n",
     "    display(Audio(audio2, rate=sample_rate))\n",