{
"cells": [
{
"cell_type": "code",
"execution_count": 1,
"id": "8ccfdd2a",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"(2242, 5)"
]
},
"execution_count": 1,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"import pandas as pd\n",
"\n",
"df = pd.read_csv('malay-speech/UTTRANSINFO.txt', sep = '\\t')\n",
"df.shape"
]
},
{
"cell_type": "code",
"execution_count": 2,
"id": "b4a00d54",
"metadata": {},
"outputs": [],
"source": [
"import soundfile as sf\n",
"from glob import glob\n",
"import os"
]
},
{
"cell_type": "code",
"execution_count": 5,
"id": "a31241cf",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"0.0"
]
},
"execution_count": 5,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"timestamps = [i * 0.02 for i in range(1500 + 1)]\n",
"timestamps[0]"
]
},
{
"cell_type": "code",
"execution_count": 3,
"id": "448b3211",
"metadata": {},
"outputs": [],
"source": [
"!mkdir audio-whisper"
]
},
{
"cell_type": "code",
"execution_count": 14,
"id": "f5508d8d",
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"100%|██████████████████████████████████████████████████████████████████████████████████████████████████| 2242/2242 [00:04<00:00, 529.83it/s]\n"
]
}
],
"source": [
"from tqdm import tqdm\n",
"\n",
"data = []\n",
"for i in tqdm(range(len(df))):\n",
" row = df.iloc[i]\n",
" audio = os.path.join('malay-speech/WAV', row['SPEAKER_ID'], row['UTTRANS_ID'])\n",
" if not os.path.exists(audio):\n",
" continue\n",
" \n",
" y, sr = sf.read(audio)\n",
" length = len(y) / sr\n",
" \n",
" start = 0.0\n",
" end = min(timestamps, key=lambda t: abs(t - length))\n",
" t = row['PROMPT']\n",
" filtered = f\"<|{start:.2f}|> {t}<|{end:.2f}|>\"\n",
" text = f\"<|startoftranscript|><|ms|><|transcribe|>{filtered}<|endoftext|>\"\n",
" \n",
" new_f = os.path.join('audio-whisper', f'{i}.wav')\n",
" \n",
" data.append({\n",
" 'filename': new_f,\n",
" 'Y': text,\n",
" })\n",
" \n",
" sf.write(new_f, y, sr)"
]
},
{
"cell_type": "code",
"execution_count": 15,
"id": "70171283",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"2242"
]
},
"execution_count": 15,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"len(data)"
]
},
{
"cell_type": "code",
"execution_count": 16,
"id": "a8dcd8ae",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"{'filename': 'audio-whisper/0.wav',\n",
" 'Y': '<|startoftranscript|><|ms|><|transcribe|><|0.00|> Dia tidak tahu apa-apa .<|3.56|><|endoftext|>'}"
]
},
"execution_count": 16,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"data[0]"
]
},
{
"cell_type": "code",
"execution_count": 18,
"id": "3e494e41",
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"\n",
" \n",
" "
],
"text/plain": [
""
]
},
"execution_count": 18,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"import IPython.display as ipd\n",
"ipd.Audio('audio-whisper/2233.wav')"
]
},
{
"cell_type": "code",
"execution_count": 19,
"id": "31f5c2f7",
"metadata": {},
"outputs": [],
"source": [
"!zip -rq scripted-malay-daily-use-speech-corpus-whisper-format.zip audio-whisper"
]
},
{
"cell_type": "code",
"execution_count": 20,
"id": "81f25b63",
"metadata": {},
"outputs": [],
"source": [
"import json\n",
"\n",
"with open('scripted-malay-daily-use-speech-corpus-whisper-format.json', 'w') as fopen:\n",
" json.dump(data, fopen)"
]
},
{
"cell_type": "code",
"execution_count": 22,
"id": "ac25c536",
"metadata": {},
"outputs": [
{
"data": {
"application/vnd.jupyter.widget-view+json": {
"model_id": "44b8115898d640f8b0c4b1ff0fc5dedc",
"version_major": 2,
"version_minor": 0
},
"text/plain": [
"scripted-malay-daily-use-speech-corpus-whisper-format.zip: 0%| | 0.00/399M [00:00