Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

resolving issue 1. #2

Open
wants to merge 1 commit into
base: master
Choose a base branch
from
Open

resolving issue 1. #2

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
219 changes: 1 addition & 218 deletions main.ipynb
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,218 +1 @@
{
"cells": [
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import os, warnings\n",
"import numpy as np\n",
"import matplotlib.pyplot as plt\n",
"warnings.filterwarnings(\"ignore\")\n",
"import torch\n",
"import torch.nn as nn\n",
"from torch.utils.data import DataLoader,random_split,Dataset\n",
"import torchaudio\n",
"from torchaudio import transforms\n",
"from torch import Tensor\n",
"from torchaudio.datasets.utils import (\n",
" download_url,\n",
" extract_archive,\n",
" walk_files\n",
")\n",
"\n",
"from train_utils import *\n",
"from model import *\n",
"from dataloader import *"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"train_audio_path = './data1/SpeechCommands/speech_commands_v0.02/'\n",
"\n",
"labels_dict=os.listdir(train_audio_path)\n",
"\n",
"a = torchaudio.datasets.SPEECHCOMMANDS('./data1/' , url = 'speech_commands_v0.02', \n",
" folder_in_archive= 'SpeechCommands', download = True)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"filename = \"./data1/SpeechCommands/speech_commands_v0.02/backward/0165e0e8_nohash_0.wav\"\n",
"waveform, sample_rate = torchaudio.load(filename)\n",
"\n",
"print(\"Shape of waveform: {}\".format(waveform.size()))\n",
"print(\"Sample rate of waveform: {}\".format(sample_rate))\n",
"\n",
"plt.figure()\n",
"plt.plot(waveform.t().numpy())"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"plt.plot(a[0][0].t())\n",
"plt.show()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"count=0\n",
"wave = []\n",
"labels = []\n",
"for i in range(0,105829):\n",
" if a[i][0].shape == (1,16000):\n",
" wave.append(a[i][0])\n",
" labels.append(a[i][2])"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"specgram = torchaudio.transforms.MFCC()(wave[0])\n",
"\n",
"print(\"Shape of spectrogram: {}\".format(specgram.size()))\n",
"\n",
"plt.figure(figsize=(10,5))\n",
"plt.imshow(specgram[0,:,:].numpy())\n",
"plt.colorbar()\n",
"plt.show()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"specgram = torchaudio.transforms.MelSpectrogram()(wave[0])\n",
"mfcc = torchaudio.transforms.MFCC()(wave[0])\n",
"\n",
"\n",
"fig,ax = plt.subplots(1,2)\n",
"\n",
"ax[0].imshow(specgram[0,:,:].numpy())\n",
"ax[1].imshow(mfcc[0,:,:].numpy())"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"data_transform = 0\n",
"\n",
"if data_transform == 1:\n",
" print(\"MFCC Features classification\")\n",
" train_audio_transforms = nn.Sequential(\n",
" torchaudio.transforms.MFCC(log_mels=False)\n",
" )\n",
" net = NN2D(num_class=35)\n",
"elif data_transform == 2:\n",
" print(\"Mel Spectogram Features classification\")\n",
" train_audio_transforms = nn.Sequential(\n",
" torchaudio.transforms.MelSpectrogram()\n",
" )\n",
" net = NN2DMEL(num_class=35)\n",
"else:\n",
" train_audio_transforms = None\n",
" net = NN(num_class=35)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"dataset= SpeechDataLoader(wave,labels,labels_dict, train_audio_transforms)\n",
"\n",
"traindata, testdata = random_split(dataset, [round(len(dataset)*.8), round(len(dataset)*.2)])\n",
"\n",
"trainloader = torch.utils.data.DataLoader(traindata, batch_size=100, shuffle=True)\n",
"\n",
"testloader = torch.utils.data.DataLoader(testdata, batch_size=100, shuffle=True)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"device = torch.device('cuda:9' if torch.cuda.is_available() else 'cpu')\n",
"\n",
"net = net.to(device)\n",
"\n",
"criterion = nn.CrossEntropyLoss()\n",
"optimizer = torch.optim.Adam(net.parameters(),lr=0.001)\n",
"scheduler = torch.optim.lr_scheduler.OneCycleLR(optimizer, max_lr=0.001,\n",
" steps_per_epoch=int(len(trainloader)),\n",
" epochs=num_epochs,\n",
" anneal_strategy='linear') \n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"num_epochs=20\n",
"\n",
"for epoch in range(0, num_epochs):\n",
" \n",
" train(net,trainloader,optimizer,scheduler,criterion,epoch,device)\n",
" best_acc = test(net,testloader,optimizer,criterion,epoch,device)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"kernelspec": {
"display_name": "Python [conda env:pyTorch]",
"language": "python",
"name": "conda-env-pyTorch-py"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.7.7"
}
},
"nbformat": 4,
"nbformat_minor": 4
}
{"cells":[{"cell_type":"code","execution_count":null,"metadata":{"id":"Ah8VNBm5o93H"},"outputs":[],"source":["import os, warnings\n","import numpy as np\n","import matplotlib.pyplot as plt\n","warnings.filterwarnings(\"ignore\")\n","import torch\n","import torch.nn as nn\n","from torch.utils.data import DataLoader,random_split,Dataset\n","import torchaudio\n","from torchaudio import transforms\n","from torch import Tensor\n","from torchaudio.datasets.utils import (\n"," download_url,\n"," extract_archive,\n"," walk_files\n",")\n","\n","from train_utils import *\n","from model import *\n","from dataloader import *"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"fTcTT8A-o93P"},"outputs":[],"source":["train_audio_path = './data1/SpeechCommands/speech_commands_v0.02/'\n","\n","labels_dict=os.listdir(train_audio_path)\n","\n","a = torchaudio.datasets.SPEECHCOMMANDS('./data1/' , url = 'speech_commands_v0.02', \n"," folder_in_archive= 'SpeechCommands', download = True)"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"BuftEzk2o93Q"},"outputs":[],"source":["filename = \"./data1/SpeechCommands/speech_commands_v0.02/backward/0165e0e8_nohash_0.wav\"\n","waveform, sample_rate = torchaudio.load(filename)\n","\n","print(\"Shape of waveform: {}\".format(waveform.size()))\n","print(\"Sample rate of waveform: {}\".format(sample_rate))\n","\n","plt.figure()\n","plt.plot(waveform.t().numpy())"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"w9Ya04G0o93R"},"outputs":[],"source":["plt.plot(a[0][0].t())\n","plt.show()"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"ZR7Rlw6No93S"},"outputs":[],"source":["count=0\n","wave = []\n","labels = []\n","for i in range(0,105829):\n"," if a[i][0].shape == (1,16000):\n"," wave.append(a[i][0])\n"," labels.append(a[i][2])"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"VN6kcU3mo93U"},"outputs":[],"source":["specgram = torchaudio.transforms.MFCC()(wave[0])\n","\n","print(\"Shape of spectrogram: {}\".format(specgram.size()))\n","\n","plt.figure(figsize=(10,5))\n","plt.imshow(specgram[0,:,:].numpy())\n","plt.colorbar()\n","plt.show()"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"4aardBrqo93V"},"outputs":[],"source":["specgram = torchaudio.transforms.MelSpectrogram()(wave[0])\n","mfcc = torchaudio.transforms.MFCC()(wave[0])\n","\n","\n","fig,ax = plt.subplots(1,2)\n","\n","ax[0].imshow(specgram[0,:,:].numpy())\n","ax[1].imshow(mfcc[0,:,:].numpy())"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"yNhxe6vEo93W"},"outputs":[],"source":["data_transform = 0\n","\n","if data_transform == 1:\n"," print(\"MFCC Features classification\")\n"," train_audio_transforms = nn.Sequential(\n"," torchaudio.transforms.MFCC(log_mels=False)\n"," )\n"," net = NN2D(num_class=35)\n","elif data_transform == 2:\n"," print(\"Mel Spectogram Features classification\")\n"," train_audio_transforms = nn.Sequential(\n"," torchaudio.transforms.MelSpectrogram()\n"," )\n"," net = NN2DMEL(num_class=35)\n","else:\n"," train_audio_transforms = None\n"," net = NN(num_class=35)"]},{"cell_type":"code","source":["labels_dict=list(set(labels))"],"metadata":{"id":"CpmLFQPWpFlV"},"execution_count":null,"outputs":[]},{"cell_type":"code","execution_count":null,"metadata":{"id":"dmvFlZ-Ro93X"},"outputs":[],"source":["dataset= SpeechDataLoader(wave,labels,labels_dict, train_audio_transforms)\n","\n","traindata, testdata = random_split(dataset, [round(len(dataset)*.8), round(len(dataset)*.2)])\n","\n","trainloader = torch.utils.data.DataLoader(traindata, batch_size=100, shuffle=True)\n","\n","testloader = torch.utils.data.DataLoader(testdata, batch_size=100, shuffle=True)"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"2YefK0_4o93Y"},"outputs":[],"source":["device = torch.device('cuda:9' if torch.cuda.is_available() else 'cpu')\n","\n","net = net.to(device)\n","\n","criterion = nn.CrossEntropyLoss()\n","optimizer = torch.optim.Adam(net.parameters(),lr=0.001)\n","scheduler = torch.optim.lr_scheduler.OneCycleLR(optimizer, max_lr=0.001,\n"," steps_per_epoch=int(len(trainloader)),\n"," epochs=num_epochs,\n"," anneal_strategy='linear') \n"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"ePBxi5pno93Z"},"outputs":[],"source":["num_epochs=20\n","\n","for epoch in range(0, num_epochs):\n"," \n"," train(net,trainloader,optimizer,scheduler,criterion,epoch,device)\n"," best_acc = test(net,testloader,optimizer,criterion,epoch,device)"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"uF8eqS7Po93Z"},"outputs":[],"source":[]}],"metadata":{"kernelspec":{"display_name":"Python [conda env:pyTorch]","language":"python","name":"conda-env-pyTorch-py"},"language_info":{"codemirror_mode":{"name":"ipython","version":3},"file_extension":".py","mimetype":"text/x-python","name":"python","nbconvert_exporter":"python","pygments_lexer":"ipython3","version":"3.7.7"},"colab":{"provenance":[]}},"nbformat":4,"nbformat_minor":0}