{
"cells": [
{
"cell_type": "code",
"execution_count": 1,
"id": "b447e2c4",
"metadata": {},
"outputs": [],
"source": [
"import os\n",
"import sys\n",
"sys.path.insert(0, os.path.dirname(os.path.abspath(\"\")))"
]
},
{
"cell_type": "code",
"execution_count": 2,
"id": "c2fc0e7a",
"metadata": {},
"outputs": [],
"source": [
"import random\n",
"from src.mel import Mel\n",
"from PIL import ImageOps, Image\n",
"from IPython.display import Audio\n",
"from diffusers import DDPMPipeline\n",
"from datasets import load_from_disk"
]
},
{
"cell_type": "code",
"execution_count": 3,
"id": "a3d45c36",
"metadata": {},
"outputs": [],
"source": [
"mel = Mel(x_res=64, y_res=64, hop_length=1024)"
]
},
{
"cell_type": "markdown",
"id": "011fb5a1",
"metadata": {},
"source": [
"### Run model inference to generate Mel spectrogram"
]
},
{
"cell_type": "code",
"execution_count": 4,
"id": "b809fed5",
"metadata": {},
"outputs": [
{
"data": {
"application/vnd.jupyter.widget-view+json": {
"model_id": "5b2ca925758841d2accaac8229f84b53",
"version_major": 2,
"version_minor": 0
},
"text/plain": [
" 0%| | 0/1000 [00:00"
]
},
"execution_count": 6,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"image = Image.fromarray(images[0][0])\n",
"image"
]
},
{
"cell_type": "markdown",
"id": "7230c280",
"metadata": {},
"source": [
"### Transform Mel spectrogram to audio"
]
},
{
"cell_type": "code",
"execution_count": 5,
"id": "5f8a149d",
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"\n",
" \n",
" "
],
"text/plain": [
""
]
},
"execution_count": 5,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"audio = mel.image_to_audio(image)\n",
"Audio(data=audio, rate=mel.get_sample_rate())"
]
},
{
"cell_type": "markdown",
"id": "ef54cef3",
"metadata": {},
"source": [
"### Compare results with random sample from training set"
]
},
{
"cell_type": "code",
"execution_count": 7,
"id": "269ee816",
"metadata": {},
"outputs": [],
"source": [
"ds = load_from_disk('../data-64')"
]
},
{
"cell_type": "code",
"execution_count": 8,
"id": "af947751",
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
""
]
},
"execution_count": 8,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"image = random.choice(ds['train'])['image']\n",
"image"
]
},
{
"cell_type": "code",
"execution_count": 9,
"id": "492e2334",
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"\n",
" \n",
" "
],
"text/plain": [
""
]
},
"execution_count": 9,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"audio = mel.image_to_audio(image)\n",
"Audio(data=audio, rate=mel.get_sample_rate())"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "2ab3d0d2",
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"kernelspec": {
"display_name": "huggingface",
"language": "python",
"name": "huggingface"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.9.13"
},
"toc": {
"base_numbering": 1,
"nav_menu": {},
"number_sections": true,
"sideBar": true,
"skip_h1_title": false,
"title_cell": "Table of Contents",
"title_sidebar": "Contents",
"toc_cell": false,
"toc_position": {},
"toc_section_display": true,
"toc_window_display": false
}
},
"nbformat": 4,
"nbformat_minor": 5
}