Spaces:
Runtime error
Runtime error
reproducible samples with seed
Browse files- notebooks/test_model.ipynb +20 -17
notebooks/test_model.ipynb
CHANGED
@@ -84,17 +84,9 @@
|
|
84 |
"metadata": {},
|
85 |
"outputs": [],
|
86 |
"source": [
|
87 |
-
"audio_diffusion = AudioDiffusion(model_id=model_id)"
|
88 |
-
|
89 |
-
|
90 |
-
{
|
91 |
-
"cell_type": "code",
|
92 |
-
"execution_count": null,
|
93 |
-
"id": "4dc17ac0",
|
94 |
-
"metadata": {},
|
95 |
-
"outputs": [],
|
96 |
-
"source": [
|
97 |
-
"mel = Mel(x_res=256, y_res=256)"
|
98 |
]
|
99 |
},
|
100 |
{
|
@@ -112,10 +104,13 @@
|
|
112 |
"metadata": {},
|
113 |
"outputs": [],
|
114 |
"source": [
|
115 |
-
"generator = torch.Generator()\n",
|
116 |
"for _ in range(10):\n",
|
117 |
-
"
|
118 |
-
"
|
|
|
|
|
|
|
|
|
119 |
" display(image)\n",
|
120 |
" display(Audio(audio, rate=sample_rate))\n",
|
121 |
" loop = AudioDiffusion.loop_it(audio, sample_rate)\n",
|
@@ -149,9 +144,10 @@
|
|
149 |
"outputs": [],
|
150 |
"source": [
|
151 |
"seed = 16183389798189209330 #@param {type:\"integer\"}\n",
|
|
|
152 |
"image, (sample_rate,\n",
|
153 |
-
" audio) = audio_diffusion.
|
154 |
-
" generator=
|
155 |
"display(image)\n",
|
156 |
"display(Audio(audio, rate=sample_rate))"
|
157 |
]
|
@@ -258,7 +254,6 @@
|
|
258 |
"overlap_samples = overlap_secs * mel.get_sample_rate()\n",
|
259 |
"slice_size = mel.x_res * mel.hop_length\n",
|
260 |
"stride = slice_size - overlap_samples\n",
|
261 |
-
"generator = torch.Generator()\n",
|
262 |
"seed = generator.seed()\n",
|
263 |
"print(f'Seed = {seed}')\n",
|
264 |
"track = np.array([])\n",
|
@@ -346,6 +341,14 @@
|
|
346 |
"audio = mel.image_to_audio(image)\n",
|
347 |
"Audio(data=audio, rate=mel.get_sample_rate())"
|
348 |
]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
349 |
}
|
350 |
],
|
351 |
"metadata": {
|
|
|
84 |
"metadata": {},
|
85 |
"outputs": [],
|
86 |
"source": [
|
87 |
+
"audio_diffusion = AudioDiffusion(model_id=model_id)\n",
|
88 |
+
"mel = Mel(x_res=256, y_res=256)\n",
|
89 |
+
"generator = torch.Generator()"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
90 |
]
|
91 |
},
|
92 |
{
|
|
|
104 |
"metadata": {},
|
105 |
"outputs": [],
|
106 |
"source": [
|
|
|
107 |
"for _ in range(10):\n",
|
108 |
+
" seed = generator.seed()\n",
|
109 |
+
" print(f'Seed = {seed}')\n",
|
110 |
+
" generator.manual_seed(seed)\n",
|
111 |
+
" image, (sample_rate,\n",
|
112 |
+
" audio) = audio_diffusion.generate_spectrogram_and_audio(\n",
|
113 |
+
" generator=generator)\n",
|
114 |
" display(image)\n",
|
115 |
" display(Audio(audio, rate=sample_rate))\n",
|
116 |
" loop = AudioDiffusion.loop_it(audio, sample_rate)\n",
|
|
|
144 |
"outputs": [],
|
145 |
"source": [
|
146 |
"seed = 16183389798189209330 #@param {type:\"integer\"}\n",
|
147 |
+
"generator.manual_seed(seed)\n",
|
148 |
"image, (sample_rate,\n",
|
149 |
+
" audio) = audio_diffusion.generate_spectrogram_and_audio(\n",
|
150 |
+
" generator=generator)\n",
|
151 |
"display(image)\n",
|
152 |
"display(Audio(audio, rate=sample_rate))"
|
153 |
]
|
|
|
254 |
"overlap_samples = overlap_secs * mel.get_sample_rate()\n",
|
255 |
"slice_size = mel.x_res * mel.hop_length\n",
|
256 |
"stride = slice_size - overlap_samples\n",
|
|
|
257 |
"seed = generator.seed()\n",
|
258 |
"print(f'Seed = {seed}')\n",
|
259 |
"track = np.array([])\n",
|
|
|
341 |
"audio = mel.image_to_audio(image)\n",
|
342 |
"Audio(data=audio, rate=mel.get_sample_rate())"
|
343 |
]
|
344 |
+
},
|
345 |
+
{
|
346 |
+
"cell_type": "code",
|
347 |
+
"execution_count": null,
|
348 |
+
"id": "4deb47f4",
|
349 |
+
"metadata": {},
|
350 |
+
"outputs": [],
|
351 |
+
"source": []
|
352 |
}
|
353 |
],
|
354 |
"metadata": {
|