FEED Validator

for Atom and RSS and KML

It looks like this is a web page, not a feed. I looked for a feed associated with this page, but couldn't find one. Please enter the address of your feed to validate.

Source: https://stlplaces.com/blog/how-to-implement-batch-normalization-in-pytorch

<!doctype html>
<html data-n-head-ssr>
<head >
<meta data-n-head="ssr" charset="utf-8"><meta data-n-head="ssr" name="viewport" content="width=device-width, initial-scale=1"><meta data-n-head="ssr" property="twitter:title" content="How to Implement Batch Normalization In PyTorch?"><meta data-n-head="ssr" property="twitter:description" content="Looking to learn how to implement Batch Normalization in PyTorch effectively."><meta data-n-head="ssr" property="twitter:image" content="https://blogweb-static.fra1.cdn.digitaloceanspaces.com/images/d7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae/blog/465d418c-e01f-4744-92bf-7347d5442899/658324fe1a7a03cb8b15bf11.png"><meta data-n-head="ssr" property="twitter:card" content="summary_large_image"><meta data-n-head="ssr" property="og:title" content="How to Implement Batch Normalization In PyTorch?"><meta data-n-head="ssr" property="og:type" content="article"><meta data-n-head="ssr" property="og:url" content="https://stlplaces.com/blog/how-to-implement-batch-normalization-in-pytorch"><meta data-n-head="ssr" property="og:image" content="https://blogweb-static.fra1.cdn.digitaloceanspaces.com/images/d7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae/blog/465d418c-e01f-4744-92bf-7347d5442899/658324fe1a7a03cb8b15bf11.png"><meta data-n-head="ssr" property="og:description" content="Looking to learn how to implement Batch Normalization in PyTorch effectively."><meta data-n-head="ssr" property="og:site_name" content="St. Louis Blog"><meta data-n-head="ssr" name="twitter:label1" content="Reading time"><meta data-n-head="ssr" name="twitter:data1" content="12 minutes read"><meta data-n-head="ssr" data-hid="description" name="description" content="Looking to learn how to implement Batch Normalization in PyTorch effectively."><title>How to Implement Batch Normalization In PyTorch in 2024?</title><link data-n-head="ssr" rel="stylesheet" type="text/css" href="https://fonts.googleapis.com/css?family=Literata"><link data-n-head="ssr" rel="stylesheet" href="https://blogweb-static.fra1.cdn.digitaloceanspaces.com/css/d7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae/custom.css"><link data-n-head="ssr" rel="icon" type="image/png" href="https://blogweb-static.fra1.cdn.digitaloceanspaces.com/images/d7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae/favicon.png"><link data-n-head="ssr" data-hid="i18n-can" rel="canonical" href="https://stlplaces.com/blog/how-to-implement-batch-normalization-in-pytorch"><style data-n-head="vuetify" type="text/css" id="vuetify-theme-stylesheet" nonce="undefined">.v-application a { color: #993366; }
.v-application .primary {
background-color: #993366 !important;
border-color: #993366 !important;
}
.v-application .primary--text {
color: #993366 !important;
caret-color: #993366 !important;
}
.v-application .primary.lighten-5 {
background-color: #ffbaeb !important;
border-color: #ffbaeb !important;
}
.v-application .primary--text.text--lighten-5 {
color: #ffbaeb !important;
caret-color: #ffbaeb !important;
}
.v-application .primary.lighten-4 {
background-color: #ff9ecf !important;
border-color: #ff9ecf !important;
}
.v-application .primary--text.text--lighten-4 {
color: #ff9ecf !important;
caret-color: #ff9ecf !important;
}
.v-application .primary.lighten-3 {
background-color: #f083b4 !important;
border-color: #f083b4 !important;
}
.v-application .primary--text.text--lighten-3 {
color: #f083b4 !important;
caret-color: #f083b4 !important;
}
.v-application .primary.lighten-2 {
background-color: #d26899 !important;
border-color: #d26899 !important;
}
.v-application .primary--text.text--lighten-2 {
color: #d26899 !important;
caret-color: #d26899 !important;
}
.v-application .primary.lighten-1 {
background-color: #b54e7f !important;
border-color: #b54e7f !important;
}
.v-application .primary--text.text--lighten-1 {
color: #b54e7f !important;
caret-color: #b54e7f !important;
}
.v-application .primary.darken-1 {
background-color: #7d154e !important;
border-color: #7d154e !important;
}
.v-application .primary--text.text--darken-1 {
color: #7d154e !important;
caret-color: #7d154e !important;
}
.v-application .primary.darken-2 {
background-color: #620037 !important;
border-color: #620037 !important;
}
.v-application .primary--text.text--darken-2 {
color: #620037 !important;
caret-color: #620037 !important;
}
.v-application .primary.darken-3 {
background-color: #470022 !important;
border-color: #470022 !important;
}
.v-application .primary--text.text--darken-3 {
color: #470022 !important;
caret-color: #470022 !important;
}
.v-application .primary.darken-4 {
background-color: #33000c !important;
border-color: #33000c !important;
}
.v-application .primary--text.text--darken-4 {
color: #33000c !important;
caret-color: #33000c !important;
}
.v-application .secondary {
background-color: #424242 !important;
border-color: #424242 !important;
}
.v-application .secondary--text {
color: #424242 !important;
caret-color: #424242 !important;
}
.v-application .secondary.lighten-5 {
background-color: #c1c1c1 !important;
border-color: #c1c1c1 !important;
}
.v-application .secondary--text.text--lighten-5 {
color: #c1c1c1 !important;
caret-color: #c1c1c1 !important;
}
.v-application .secondary.lighten-4 {
background-color: #a6a6a6 !important;
border-color: #a6a6a6 !important;
}
.v-application .secondary--text.text--lighten-4 {
color: #a6a6a6 !important;
caret-color: #a6a6a6 !important;
}
.v-application .secondary.lighten-3 {
background-color: #8b8b8b !important;
border-color: #8b8b8b !important;
}
.v-application .secondary--text.text--lighten-3 {
color: #8b8b8b !important;
caret-color: #8b8b8b !important;
}
.v-application .secondary.lighten-2 {
background-color: #727272 !important;
border-color: #727272 !important;
}
.v-application .secondary--text.text--lighten-2 {
color: #727272 !important;
caret-color: #727272 !important;
}
.v-application .secondary.lighten-1 {
background-color: #595959 !important;
border-color: #595959 !important;
}
.v-application .secondary--text.text--lighten-1 {
color: #595959 !important;
caret-color: #595959 !important;
}
.v-application .secondary.darken-1 {
background-color: #2c2c2c !important;
border-color: #2c2c2c !important;
}
.v-application .secondary--text.text--darken-1 {
color: #2c2c2c !important;
caret-color: #2c2c2c !important;
}
.v-application .secondary.darken-2 {
background-color: #171717 !important;
border-color: #171717 !important;
}
.v-application .secondary--text.text--darken-2 {
color: #171717 !important;
caret-color: #171717 !important;
}
.v-application .secondary.darken-3 {
background-color: #000000 !important;
border-color: #000000 !important;
}
.v-application .secondary--text.text--darken-3 {
color: #000000 !important;
caret-color: #000000 !important;
}
.v-application .secondary.darken-4 {
background-color: #000000 !important;
border-color: #000000 !important;
}
.v-application .secondary--text.text--darken-4 {
color: #000000 !important;
caret-color: #000000 !important;
}
.v-application .accent {
background-color: #993366 !important;
border-color: #993366 !important;
}
.v-application .accent--text {
color: #993366 !important;
caret-color: #993366 !important;
}
.v-application .accent.lighten-5 {
background-color: #ffbaeb !important;
border-color: #ffbaeb !important;
}
.v-application .accent--text.text--lighten-5 {
color: #ffbaeb !important;
caret-color: #ffbaeb !important;
}
.v-application .accent.lighten-4 {
background-color: #ff9ecf !important;
border-color: #ff9ecf !important;
}
.v-application .accent--text.text--lighten-4 {
color: #ff9ecf !important;
caret-color: #ff9ecf !important;
}
.v-application .accent.lighten-3 {
background-color: #f083b4 !important;
border-color: #f083b4 !important;
}
.v-application .accent--text.text--lighten-3 {
color: #f083b4 !important;
caret-color: #f083b4 !important;
}
.v-application .accent.lighten-2 {
background-color: #d26899 !important;
border-color: #d26899 !important;
}
.v-application .accent--text.text--lighten-2 {
color: #d26899 !important;
caret-color: #d26899 !important;
}
.v-application .accent.lighten-1 {
background-color: #b54e7f !important;
border-color: #b54e7f !important;
}
.v-application .accent--text.text--lighten-1 {
color: #b54e7f !important;
caret-color: #b54e7f !important;
}
.v-application .accent.darken-1 {
background-color: #7d154e !important;
border-color: #7d154e !important;
}
.v-application .accent--text.text--darken-1 {
color: #7d154e !important;
caret-color: #7d154e !important;
}
.v-application .accent.darken-2 {
background-color: #620037 !important;
border-color: #620037 !important;
}
.v-application .accent--text.text--darken-2 {
color: #620037 !important;
caret-color: #620037 !important;
}
.v-application .accent.darken-3 {
background-color: #470022 !important;
border-color: #470022 !important;
}
.v-application .accent--text.text--darken-3 {
color: #470022 !important;
caret-color: #470022 !important;
}
.v-application .accent.darken-4 {
background-color: #33000c !important;
border-color: #33000c !important;
}
.v-application .accent--text.text--darken-4 {
color: #33000c !important;
caret-color: #33000c !important;
}
.v-application .error {
background-color: #993366 !important;
border-color: #993366 !important;
}
.v-application .error--text {
color: #993366 !important;
caret-color: #993366 !important;
}
.v-application .error.lighten-5 {
background-color: #ffbaeb !important;
border-color: #ffbaeb !important;
}
.v-application .error--text.text--lighten-5 {
color: #ffbaeb !important;
caret-color: #ffbaeb !important;
}
.v-application .error.lighten-4 {
background-color: #ff9ecf !important;
border-color: #ff9ecf !important;
}
.v-application .error--text.text--lighten-4 {
color: #ff9ecf !important;
caret-color: #ff9ecf !important;
}
.v-application .error.lighten-3 {
background-color: #f083b4 !important;
border-color: #f083b4 !important;
}
.v-application .error--text.text--lighten-3 {
color: #f083b4 !important;
caret-color: #f083b4 !important;
}
.v-application .error.lighten-2 {
background-color: #d26899 !important;
border-color: #d26899 !important;
}
.v-application .error--text.text--lighten-2 {
color: #d26899 !important;
caret-color: #d26899 !important;
}
.v-application .error.lighten-1 {
background-color: #b54e7f !important;
border-color: #b54e7f !important;
}
.v-application .error--text.text--lighten-1 {
color: #b54e7f !important;
caret-color: #b54e7f !important;
}
.v-application .error.darken-1 {
background-color: #7d154e !important;
border-color: #7d154e !important;
}
.v-application .error--text.text--darken-1 {
color: #7d154e !important;
caret-color: #7d154e !important;
}
.v-application .error.darken-2 {
background-color: #620037 !important;
border-color: #620037 !important;
}
.v-application .error--text.text--darken-2 {
color: #620037 !important;
caret-color: #620037 !important;
}
.v-application .error.darken-3 {
background-color: #470022 !important;
border-color: #470022 !important;
}
.v-application .error--text.text--darken-3 {
color: #470022 !important;
caret-color: #470022 !important;
}
.v-application .error.darken-4 {
background-color: #33000c !important;
border-color: #33000c !important;
}
.v-application .error--text.text--darken-4 {
color: #33000c !important;
caret-color: #33000c !important;
}
.v-application .info {
background-color: #993366 !important;
border-color: #993366 !important;
}
.v-application .info--text {
color: #993366 !important;
caret-color: #993366 !important;
}
.v-application .info.lighten-5 {
background-color: #ffbaeb !important;
border-color: #ffbaeb !important;
}
.v-application .info--text.text--lighten-5 {
color: #ffbaeb !important;
caret-color: #ffbaeb !important;
}
.v-application .info.lighten-4 {
background-color: #ff9ecf !important;
border-color: #ff9ecf !important;
}
.v-application .info--text.text--lighten-4 {
color: #ff9ecf !important;
caret-color: #ff9ecf !important;
}
.v-application .info.lighten-3 {
background-color: #f083b4 !important;
border-color: #f083b4 !important;
}
.v-application .info--text.text--lighten-3 {
color: #f083b4 !important;
caret-color: #f083b4 !important;
}
.v-application .info.lighten-2 {
background-color: #d26899 !important;
border-color: #d26899 !important;
}
.v-application .info--text.text--lighten-2 {
color: #d26899 !important;
caret-color: #d26899 !important;
}
.v-application .info.lighten-1 {
background-color: #b54e7f !important;
border-color: #b54e7f !important;
}
.v-application .info--text.text--lighten-1 {
color: #b54e7f !important;
caret-color: #b54e7f !important;
}
.v-application .info.darken-1 {
background-color: #7d154e !important;
border-color: #7d154e !important;
}
.v-application .info--text.text--darken-1 {
color: #7d154e !important;
caret-color: #7d154e !important;
}
.v-application .info.darken-2 {
background-color: #620037 !important;
border-color: #620037 !important;
}
.v-application .info--text.text--darken-2 {
color: #620037 !important;
caret-color: #620037 !important;
}
.v-application .info.darken-3 {
background-color: #470022 !important;
border-color: #470022 !important;
}
.v-application .info--text.text--darken-3 {
color: #470022 !important;
caret-color: #470022 !important;
}
.v-application .info.darken-4 {
background-color: #33000c !important;
border-color: #33000c !important;
}
.v-application .info--text.text--darken-4 {
color: #33000c !important;
caret-color: #33000c !important;
}
.v-application .success {
background-color: #993366 !important;
border-color: #993366 !important;
}
.v-application .success--text {
color: #993366 !important;
caret-color: #993366 !important;
}
.v-application .success.lighten-5 {
background-color: #ffbaeb !important;
border-color: #ffbaeb !important;
}
.v-application .success--text.text--lighten-5 {
color: #ffbaeb !important;
caret-color: #ffbaeb !important;
}
.v-application .success.lighten-4 {
background-color: #ff9ecf !important;
border-color: #ff9ecf !important;
}
.v-application .success--text.text--lighten-4 {
color: #ff9ecf !important;
caret-color: #ff9ecf !important;
}
.v-application .success.lighten-3 {
background-color: #f083b4 !important;
border-color: #f083b4 !important;
}
.v-application .success--text.text--lighten-3 {
color: #f083b4 !important;
caret-color: #f083b4 !important;
}
.v-application .success.lighten-2 {
background-color: #d26899 !important;
border-color: #d26899 !important;
}
.v-application .success--text.text--lighten-2 {
color: #d26899 !important;
caret-color: #d26899 !important;
}
.v-application .success.lighten-1 {
background-color: #b54e7f !important;
border-color: #b54e7f !important;
}
.v-application .success--text.text--lighten-1 {
color: #b54e7f !important;
caret-color: #b54e7f !important;
}
.v-application .success.darken-1 {
background-color: #7d154e !important;
border-color: #7d154e !important;
}
.v-application .success--text.text--darken-1 {
color: #7d154e !important;
caret-color: #7d154e !important;
}
.v-application .success.darken-2 {
background-color: #620037 !important;
border-color: #620037 !important;
}
.v-application .success--text.text--darken-2 {
color: #620037 !important;
caret-color: #620037 !important;
}
.v-application .success.darken-3 {
background-color: #470022 !important;
border-color: #470022 !important;
}
.v-application .success--text.text--darken-3 {
color: #470022 !important;
caret-color: #470022 !important;
}
.v-application .success.darken-4 {
background-color: #33000c !important;
border-color: #33000c !important;
}
.v-application .success--text.text--darken-4 {
color: #33000c !important;
caret-color: #33000c !important;
}
.v-application .warning {
background-color: #993366 !important;
border-color: #993366 !important;
}
.v-application .warning--text {
color: #993366 !important;
caret-color: #993366 !important;
}
.v-application .warning.lighten-5 {
background-color: #ffbaeb !important;
border-color: #ffbaeb !important;
}
.v-application .warning--text.text--lighten-5 {
color: #ffbaeb !important;
caret-color: #ffbaeb !important;
}
.v-application .warning.lighten-4 {
background-color: #ff9ecf !important;
border-color: #ff9ecf !important;
}
.v-application .warning--text.text--lighten-4 {
color: #ff9ecf !important;
caret-color: #ff9ecf !important;
}
.v-application .warning.lighten-3 {
background-color: #f083b4 !important;
border-color: #f083b4 !important;
}
.v-application .warning--text.text--lighten-3 {
color: #f083b4 !important;
caret-color: #f083b4 !important;
}
.v-application .warning.lighten-2 {
background-color: #d26899 !important;
border-color: #d26899 !important;
}
.v-application .warning--text.text--lighten-2 {
color: #d26899 !important;
caret-color: #d26899 !important;
}
.v-application .warning.lighten-1 {
background-color: #b54e7f !important;
border-color: #b54e7f !important;
}
.v-application .warning--text.text--lighten-1 {
color: #b54e7f !important;
caret-color: #b54e7f !important;
}
.v-application .warning.darken-1 {
background-color: #7d154e !important;
border-color: #7d154e !important;
}
.v-application .warning--text.text--darken-1 {
color: #7d154e !important;
caret-color: #7d154e !important;
}
.v-application .warning.darken-2 {
background-color: #620037 !important;
border-color: #620037 !important;
}
.v-application .warning--text.text--darken-2 {
color: #620037 !important;
caret-color: #620037 !important;
}
.v-application .warning.darken-3 {
background-color: #470022 !important;
border-color: #470022 !important;
}
.v-application .warning--text.text--darken-3 {
color: #470022 !important;
caret-color: #470022 !important;
}
.v-application .warning.darken-4 {
background-color: #33000c !important;
border-color: #33000c !important;
}
.v-application .warning--text.text--darken-4 {
color: #33000c !important;
caret-color: #33000c !important;
}</style><link rel="preload" href="https://pub-420acf56315e422bbbdab07717bee8cd.r2.dev/assets/0.1/50d1395.js" as="script"><link rel="preload" href="https://pub-420acf56315e422bbbdab07717bee8cd.r2.dev/assets/0.1/498f8f7.js" as="script"><link rel="preload" href="https://pub-420acf56315e422bbbdab07717bee8cd.r2.dev/assets/0.1/css/8ce80c3.css" as="style"><link rel="preload" href="https://pub-420acf56315e422bbbdab07717bee8cd.r2.dev/assets/0.1/fffc2dc.js" as="script"><link rel="preload" href="https://pub-420acf56315e422bbbdab07717bee8cd.r2.dev/assets/0.1/css/3afb08d.css" as="style"><link rel="preload" href="https://pub-420acf56315e422bbbdab07717bee8cd.r2.dev/assets/0.1/25d50b7.js" as="script"><link rel="stylesheet" href="https://pub-420acf56315e422bbbdab07717bee8cd.r2.dev/assets/0.1/css/8ce80c3.css"><link rel="stylesheet" href="https://pub-420acf56315e422bbbdab07717bee8cd.r2.dev/assets/0.1/css/3afb08d.css">
</head>
<body >
<div data-server-rendered="true" id="__nuxt"><div id="__layout"><div data-app="true" id="app" class="v-application v-application--is-ltr theme--light" style="font-family:Literata;background-image:url(https://blogweb-static.fra1.cdn.digitaloceanspaces.com/images/d7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae/background/555.png);background-size:auto;"><div class="v-application--wrap"><div tabindex="0" aria-label="Loading" class="vld-overlay is-active is-full-page" style="z-index:999;display:none;"><div class="vld-background" style="background:#fff;opacity:0.5;backdrop-filter:blur(2px);"></div><div class="vld-icon"><svg viewBox="0 0 38 38" xmlns="http://www.w3.org/2000/svg" width="65" height="65" stroke="#1976d2"><g fill="none" fill-rule="evenodd"><g transform="translate(1 1)" stroke-width="2"><circle stroke-opacity=".25" cx="18" cy="18" r="18"></circle><path d="M36 18c0-9.94-8.06-18-18-18"><animateTransform attributeName="transform" type="rotate" from="0 18 18" to="360 18 18" dur="0.8s" repeatCount="indefinite"></animateTransform></path></g></g></svg></div></div>  <div class="container"><div class="row main-menu"> <div class="d-none d-lg-flex d-xl-flex d-md-flex justify-space-between align-center col col-12"><div class="col-lg-3 col-4"><a href="/" class="nuxt-link-active"><div class="logo"><div aria-label="St. Louis Blog" role="img" class="v-image v-responsive theme--light" style="max-height:100px;"><div class="v-image__image v-image__image--preload v-image__image--contain" style="background-image:;background-position:center center;"></div><div class="v-responsive__content"></div></div></div></a></div> <div class="col-lg-5 col-4"><div class="v-input theme--light v-text-field v-text-field--filled v-text-field--enclosed v-text-field--rounded v-select v-autocomplete"><div class="v-input__control"><div role="combobox" aria-haspopup="listbox" aria-expanded="false" aria-owns="list-37278994" class="v-input__slot"><div class="v-select__slot"><label for="input-37278994" class="v-label theme--light" style="left:0px;right:auto;position:absolute;">Search</label><input autocomplete="off" id="input-37278994" type="text"><div class="v-input__append-inner"><div class="v-input__icon v-input__icon--clear"><button disabled="disabled" type="button" aria-label="clear icon" class="v-icon notranslate v-icon--disabled v-icon--link mdi mdi-close theme--light"></button></div></div><div class="v-input__append-inner"><div class="v-input__icon v-input__icon--append"><i aria-hidden="true" class="v-icon notranslate mdi mdi-menu-down theme--light"></i></div></div><input type="hidden"></div><div class="v-menu"></div></div><div class="v-text-field__details"><div class="v-messages theme--light"><div class="v-messages__wrapper"></div></div></div></div></div></div> <div class="col-lg-4 col-4"><div class="mx-auto v-card v-sheet v-sheet--outlined theme--light" style="max-width:344px;"><div tabindex="-1" class="v-list-item v-list-item--three-line theme--light"><div class="v-list-item__content"><div class="overline mb-4">Account</div>  <div class="v-list-item__subtitle">Sign in to you account</div></div> <div class="v-avatar" style="height:80px;min-width:80px;width:80px;">  <span class="headline"><i aria-hidden="true" class="v-icon notranslate mdi mdi-notebook-outline theme--light" style="font-size:40px;"></i></span></div></div>  <div class="v-card__actions"> <a href="/login" class="v-btn v-btn--contained v-btn--rounded v-btn--router theme--light v-size--small success"><span class="v-btn__content">Sign In</span></a> <a href="/register" class="v-btn v-btn--contained v-btn--rounded v-btn--router theme--light v-size--small info"><span class="v-btn__content">Sign Up</span></a></div></div></div></div> <nav class="hidden-md-and-up v-navigation-drawer v-navigation-drawer--close v-navigation-drawer--fixed v-navigation-drawer--is-mobile v-navigation-drawer--temporary theme--light" style="height:100vh;top:0px;transform:translateX(-100%);width:256px;"><div class="v-navigation-drawer__content"><div class="v-list v-sheet theme--light"><a href="/" tabindex="0" router="" class="v-list-item v-list-item--link theme--light"><div class="v-list-item__content"><div class="v-list-item__title">Home</div></div></a>    <a href="/login" tabindex="0" router="" class="v-list-item v-list-item--link theme--light"><div class="v-list-item__content"><div class="v-list-item__title">Sign In</div></div></a> <a href="/register" tabindex="0" router="" class="v-list-item v-list-item--link theme--light"><div class="v-list-item__content"><div class="v-list-item__title">Sign Up</div></div></a> </div></div><div class="v-navigation-drawer__border"></div></nav> <header class="hidden-md-and-up v-sheet theme--light v-toolbar v-app-bar v-app-bar--fixed" style="height:56px;margin-top:0px;transform:translateY(0px);left:0px;right:0px;"><div class="v-toolbar__content" style="height:56px;"><div class="row align-center justify-space-between"><button type="button" class="v-app-bar__nav-icon v-btn v-btn--flat v-btn--icon v-btn--round theme--light v-size--default"><span class="v-btn__content"><i aria-hidden="true" class="v-icon notranslate mdi mdi-menu theme--light"></i></span></button> <div class="col-sm-3 col-md-3 col-4"><a href="/" class="nuxt-link-active"><div class="logo"><div aria-label="St. Louis Blog" role="img" class="v-image v-responsive theme--light"><div class="v-responsive__sizer" style="padding-bottom:100%;"></div><div class="v-image__image v-image__image--preload v-image__image--contain" style="background-image:;background-position:center center;"></div><div class="v-responsive__content"></div></div></div></a></div></div></div></header> <div class="hidden-md-and-up mt-12 pt-6 mb-0 pb-0 col col-12"><div class="v-input theme--light v-text-field v-text-field--filled v-text-field--enclosed v-text-field--rounded v-select v-autocomplete"><div class="v-input__control"><div role="combobox" aria-haspopup="listbox" aria-expanded="false" aria-owns="list-37279023" class="v-input__slot"><div class="v-select__slot"><label for="input-37279023" class="v-label theme--light" style="left:0px;right:auto;position:absolute;">Search</label><input autocomplete="off" id="input-37279023" type="text"><div class="v-input__append-inner"><div class="v-input__icon v-input__icon--clear"><button disabled="disabled" type="button" aria-label="clear icon" class="v-icon notranslate v-icon--disabled v-icon--link mdi mdi-close theme--light"></button></div></div><div class="v-input__append-inner"><div class="v-input__icon v-input__icon--append"><i aria-hidden="true" class="v-icon notranslate mdi mdi-menu-down theme--light"></i></div></div><input type="hidden"></div><div class="v-menu"></div></div><div class="v-text-field__details"><div class="v-messages theme--light"><div class="v-messages__wrapper"></div></div></div></div></div></div> </div> <div class="flex"><ul itemscope="itemscope" itemtype="http://schema.org/BreadcrumbList" class="v-breadcrumbs v-breadcrumbs--large theme--light"><li itemprop="itemListElement" itemscope="itemscope" itemtype="http://schema.org/ListItem"><a href="/" itemprop="item" class="v-breadcrumbs__item nuxt-link-active"><span itemprop="name">Home</span></a> <meta content="1" itemprop="position"></li><li class="v-breadcrumbs__divider">></li><li itemprop="itemListElement" itemscope="itemscope" itemtype="http://schema.org/ListItem"><a href="/blog/category/programming" itemprop="item" class="v-breadcrumbs__item"><span itemprop="name">Programming</span></a> <meta content="2" itemprop="position"></li><li class="v-breadcrumbs__divider">></li><li itemprop="itemListElement" itemscope="itemscope" itemtype="http://schema.org/ListItem"><a href="#" itemprop="item" class="v-breadcrumbs__item v-breadcrumbs__item--disabled"><span itemprop="name">How to Implement Batch Normalization In PyTorch?</span></a> <meta content="3" itemprop="position"></li></ul></div> <div class="flex mt-1"><div class="row"><div class="col-md-3 col-lg-3 col-12"><div><div class="v-card v-sheet theme--light"><header class="v-sheet theme--dark v-toolbar info" style="height:56px;"><div class="v-toolbar__content" style="height:56px;"><div class="v-toolbar__title">Category</div> <div class="spacer"></div> <button type="button" class="d-flex d-sm-flex d-md-none d-lg-none d-xl-none v-btn v-btn--flat v-btn--icon v-btn--round v-btn--text theme--dark v-size--default"><span class="v-btn__content"><i aria-hidden="true" class="v-icon notranslate mdi mdi-dots-vertical theme--dark"></i></span></button></div></header> <div style="max-height:500px;overflow-y:auto;display:none;"><div><a href="/blog/category/about-st-louis" tabindex="0" router="" class="v-list-item v-list-item--link theme--light"><div class="v-list-item__title">About St. Louis</div></a></div><div><a href="/blog/category/services" tabindex="0" router="" class="v-list-item v-list-item--link theme--light"><div class="v-list-item__title">Services</div></a></div><div><a href="/blog/category/products" tabindex="0" router="" class="v-list-item v-list-item--link theme--light"><div class="v-list-item__title">Products</div></a></div><div><a href="/blog/category/books" tabindex="0" router="" class="v-list-item v-list-item--link theme--light"><div class="v-list-item__title">Books</div></a></div><div><a href="/blog/category/fitness" tabindex="0" router="" class="v-list-item v-list-item--link theme--light"><div class="v-list-item__title">Fitness</div></a></div><div><a href="/blog/category/finance" tabindex="0" router="" class="v-list-item v-list-item--link theme--light"><div class="v-list-item__title">Finance</div></a></div><div><a href="/blog/category/programming" tabindex="0" router="" class="v-list-item v-list-item--link theme--light"><div class="v-list-item__title">Programming</div></a></div></div></div>  <div class="v-card v-sheet theme--light mt-4" data-v-04cff9ac><header class="v-sheet theme--dark v-toolbar info" style="height:56px;" data-v-04cff9ac><div class="v-toolbar__content" style="height:56px;"><div class="v-toolbar__title" data-v-04cff9ac>Discussion</div> <div class="spacer" data-v-04cff9ac></div> <button type="button" class="d-flex d-sm-flex d-md-none d-lg-none d-xl-none v-btn v-btn--flat v-btn--icon v-btn--round v-btn--text theme--dark v-size--default" data-v-04cff9ac><span class="v-btn__content"><i aria-hidden="true" class="v-icon notranslate mdi mdi-dots-vertical theme--dark" data-v-04cff9ac></i></span></button></div></header> <div class="v-card__text" style="max-height:500px;overflow-y:auto;display:none;" data-v-04cff9ac data-v-04cff9ac><div class="row thread-row" data-v-04cff9ac><div class="align-items-center d-none d-md-none d-lg-none d-xl-flex d-sm-flex col col-1" data-v-04cff9ac><i aria-hidden="true" class="v-icon notranslate mdi mdi-message-text-outline theme--light" data-v-04cff9ac></i></div> <div class="col-sm-9 col-md-7 col-lg-8 col-xl-8 col-8" data-v-04cff9ac><p data-v-04cff9ac><a href="https://forum.stlplaces.com/thread/which-state-is-better-to-move-in-florida-or-north" target="_blank" title="Which state is better to move in: Florida or North Carolina?" data-v-04cff9ac>
Which state is better to move in: Florida or North Carolina?
</a></p></div> <div class="d-flex justify-end col-sm-2 col-md-5 col-lg-4 col-xl-3 col-3" data-v-04cff9ac><span draggable="false" class="v-chip v-chip--label theme--light v-size--small info white--text" data-v-04cff9ac><span class="v-chip__content"><i aria-hidden="true" class="v-icon notranslate v-icon--left mdi mdi-chat-processing-outline theme--light" data-v-04cff9ac></i>
1
</span></span></div></div><div class="row thread-row" data-v-04cff9ac><div class="align-items-center d-none d-md-none d-lg-none d-xl-flex d-sm-flex col col-1" data-v-04cff9ac><i aria-hidden="true" class="v-icon notranslate mdi mdi-message-text-outline theme--light" data-v-04cff9ac></i></div> <div class="col-sm-9 col-md-7 col-lg-8 col-xl-8 col-8" data-v-04cff9ac><p data-v-04cff9ac><a href="https://forum.stlplaces.com/thread/what-state-is-better-tennessee-or-iowa" target="_blank" title="What state is better: Tennessee or Iowa?" data-v-04cff9ac>
What state is better: Tennessee or Iowa?
</a></p></div> <div class="d-flex justify-end col-sm-2 col-md-5 col-lg-4 col-xl-3 col-3" data-v-04cff9ac><span draggable="false" class="v-chip v-chip--label theme--light v-size--small info white--text" data-v-04cff9ac><span class="v-chip__content"><i aria-hidden="true" class="v-icon notranslate v-icon--left mdi mdi-chat-processing-outline theme--light" data-v-04cff9ac></i>
1
</span></span></div></div><div class="row thread-row" data-v-04cff9ac><div class="align-items-center d-none d-md-none d-lg-none d-xl-flex d-sm-flex col col-1" data-v-04cff9ac><i aria-hidden="true" class="v-icon notranslate mdi mdi-message-text-outline theme--light" data-v-04cff9ac></i></div> <div class="col-sm-9 col-md-7 col-lg-8 col-xl-8 col-8" data-v-04cff9ac><p data-v-04cff9ac><a href="https://forum.stlplaces.com/thread/which-state-is-better-to-move-in-ohio-or-arizona" target="_blank" title="Which state is better to move in: Ohio or Arizona?" data-v-04cff9ac>
Which state is better to move in: Ohio or Arizona?
</a></p></div> <div class="d-flex justify-end col-sm-2 col-md-5 col-lg-4 col-xl-3 col-3" data-v-04cff9ac><span draggable="false" class="v-chip v-chip--label theme--light v-size--small info white--text" data-v-04cff9ac><span class="v-chip__content"><i aria-hidden="true" class="v-icon notranslate v-icon--left mdi mdi-chat-processing-outline theme--light" data-v-04cff9ac></i>
1
</span></span></div></div><div class="row thread-row" data-v-04cff9ac><div class="align-items-center d-none d-md-none d-lg-none d-xl-flex d-sm-flex col col-1" data-v-04cff9ac><i aria-hidden="true" class="v-icon notranslate mdi mdi-message-text-outline theme--light" data-v-04cff9ac></i></div> <div class="col-sm-9 col-md-7 col-lg-8 col-xl-8 col-8" data-v-04cff9ac><p data-v-04cff9ac><a href="https://forum.stlplaces.com/thread/which-state-is-best-to-visit-illinois-or-north" target="_blank" title="Which state is best to visit: Illinois or North Carolina?" data-v-04cff9ac>
Which state is best to visit: Illinois or North Carolina?
</a></p></div> <div class="d-flex justify-end col-sm-2 col-md-5 col-lg-4 col-xl-3 col-3" data-v-04cff9ac><span draggable="false" class="v-chip v-chip--label theme--light v-size--small info white--text" data-v-04cff9ac><span class="v-chip__content"><i aria-hidden="true" class="v-icon notranslate v-icon--left mdi mdi-chat-processing-outline theme--light" data-v-04cff9ac></i>
1
</span></span></div></div><div class="row thread-row" data-v-04cff9ac><div class="align-items-center d-none d-md-none d-lg-none d-xl-flex d-sm-flex col col-1" data-v-04cff9ac><i aria-hidden="true" class="v-icon notranslate mdi mdi-message-text-outline theme--light" data-v-04cff9ac></i></div> <div class="col-sm-9 col-md-7 col-lg-8 col-xl-8 col-8" data-v-04cff9ac><p data-v-04cff9ac><a href="https://forum.stlplaces.com/thread/which-state-is-better-to-move-in-ohio-or-indiana" target="_blank" title="Which state is better to move in: Ohio or Indiana?" data-v-04cff9ac>
Which state is better to move in: Ohio or Indiana?
</a></p></div> <div class="d-flex justify-end col-sm-2 col-md-5 col-lg-4 col-xl-3 col-3" data-v-04cff9ac><span draggable="false" class="v-chip v-chip--label theme--light v-size--small info white--text" data-v-04cff9ac><span class="v-chip__content"><i aria-hidden="true" class="v-icon notranslate v-icon--left mdi mdi-chat-processing-outline theme--light" data-v-04cff9ac></i>
1
</span></span></div></div></div></div></div></div> <div class="col-md-9 col-lg-9 col-12">   <div itemprop="blogPost" itemscope="itemscope" itemtype="https://schema.org/BlogPosting" class="v-card v-sheet theme--light"><div class="v-card__title"><div class="row"><meta itemprop="author"> <meta itemprop="mainEntityOfPage" content="https://stlplaces.com/blog/how-to-implement-batch-normalization-in-pytorch"> <div itemprop="publisher" itemscope="itemscope" itemtype="https://schema.org/Organization" class="d-none"><meta itemprop="name" content="stlplaces.com"> <div itemprop="logo" itemscope="itemscope" itemtype="https://schema.org/ImageObject"><meta itemprop="url" content="https://blogweb-static.fra1.cdn.digitaloceanspaces.com/images/d7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae/logo/993366.png"></div></div> <div class="col-md-12 col-lg-9 col-12"><h1 itemprop="name headline" class="font-weight-bold">
How to Implement Batch Normalization In PyTorch?
</h1></div> <div class="d-flex justify-end align-start col-md-12 col-lg-3 col-12"><div><span class="d-flex caption"><i aria-hidden="true" class="v-icon notranslate mdi mdi-clock-outline theme--light"></i> <time datetime="2024-12-01T00:00:00Z">
December 1, 2024 12:00 AM</time> <meta content="2023-12-20T17:31:44Z" itemprop="datePublished"> <meta content="2024-12-01T00:00:00Z" itemprop="dateModified"></span> <span class="d-flex caption justify-end">
12 minutes read
</span></div></div></div></div> <div class="col col-12"></div> <div class="v-card__text post-text ql-viewer"><div class="row"><div itemprop="image" itemscope="itemscope" class="text-center col col-12"><div aria-label="How to Implement Batch Normalization In PyTorch?" role="img" itemprop="url contentUrl" itemtype="https://schema.org/ImageObject" class="v-image v-responsive theme--light" style="max-height:300px;"><div class="v-image__image v-image__image--preload v-image__image--contain" style="background-image:;background-position:center center;"></div><div class="v-responsive__content"></div></div></div></div> <div itemprop="articleBody" class="row"><div class="col"><div class="run-code"><p>Batch normalization is a widely used technique for <a href="https://stlplaces.com/blog/how-to-handle-overfitting-in-tensorflow-models">improving the training of deep neural networks</a>. It normalizes the activations of each mini-batch by subtracting the mini-batch mean and dividing by the mini-batch standard deviation. This helps in reducing internal covariate shift by ensuring that the input to each layer is normalized.</p><p><br/></p><p>Implementing batch normalization in PyTorch is straightforward. Here are the steps:</p><ol><li>Import the necessary libraries:</li></ol><div style="color:#f8f8f2;background-color:#272822;">
<table style="border-spacing:0;padding:0;margin:0;border:0;"><tbody><tr><td style="vertical-align:top;padding:0;margin:0;border:0;">
<pre tabindex="0" style="color:#f8f8f2;background-color:#272822;"><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">1
</span><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">2
</span></pre></td>
<td style="vertical-align:top;padding:0;margin:0;border:0;;width:100%">
<pre tabindex="0" style="color:#f8f8f2;background-color:#272822;"><span style="display:flex;"><span>import torch
</span></span><span style="display:flex;"><span>import torch.nn as nn
</span></span></pre></td></tr></tbody></table>
</div>
<p><br/></p><ol><li>Define a custom neural network architecture:</li></ol><div style="color:#f8f8f2;background-color:#272822;">
<table style="border-spacing:0;padding:0;margin:0;border:0;"><tbody><tr><td style="vertical-align:top;padding:0;margin:0;border:0;">
<pre tabindex="0" style="color:#f8f8f2;background-color:#272822;"><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">1
</span><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">2
</span><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">3
</span><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">4
</span><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">5
</span><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">6
</span></pre></td>
<td style="vertical-align:top;padding:0;margin:0;border:0;;width:100%">
<pre tabindex="0" style="color:#f8f8f2;background-color:#272822;"><span style="display:flex;"><span>class Net(nn.Module):
</span></span><span style="display:flex;"><span> def __init__(self):
</span></span><span style="display:flex;"><span> super(Net, self).__init__()
</span></span><span style="display:flex;"><span> self.fc1 = nn.Linear(10, 20)
</span></span><span style="display:flex;"><span> self.bn1 = nn.BatchNorm1d(20) # Batch normalization layer
</span></span><span style="display:flex;"><span> self.fc2 = nn.Linear(20, 10)
</span></span></pre></td></tr></tbody></table>
</div>
<p><br/></p><ol><li>Override the forward method of the neural network:</li></ol><div style="color:#f8f8f2;background-color:#272822;">
<table style="border-spacing:0;padding:0;margin:0;border:0;"><tbody><tr><td style="vertical-align:top;padding:0;margin:0;border:0;">
<pre tabindex="0" style="color:#f8f8f2;background-color:#272822;"><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">1
</span><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">2
</span><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">3
</span><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">4
</span><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">5
</span><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">6
</span></pre></td>
<td style="vertical-align:top;padding:0;margin:0;border:0;;width:100%">
<pre tabindex="0" style="color:#f8f8f2;background-color:#272822;"><span style="display:flex;"><span> def forward(self, x):
</span></span><span style="display:flex;"><span> x = self.fc1(x)
</span></span><span style="display:flex;"><span> x = self.bn1(x)
</span></span><span style="display:flex;"><span> x = torch.relu(x)
</span></span><span style="display:flex;"><span> x = self.fc2(x)
</span></span><span style="display:flex;"><span> return x
</span></span></pre></td></tr></tbody></table>
</div>
<p><br/></p><ol><li>Create an instance of the network:</li></ol><div style="color:#f8f8f2;background-color:#272822;">
<table style="border-spacing:0;padding:0;margin:0;border:0;"><tbody><tr><td style="vertical-align:top;padding:0;margin:0;border:0;">
<pre tabindex="0" style="color:#f8f8f2;background-color:#272822;"><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">1
</span></pre></td>
<td style="vertical-align:top;padding:0;margin:0;border:0;;width:100%">
<pre tabindex="0" style="color:#f8f8f2;background-color:#272822;"><span style="display:flex;"><span>net = Net()
</span></span></pre></td></tr></tbody></table>
</div>
<p><br/></p><p>That's it! Now the network <code>net</code> includes a batch normalization layer (<code>self.bn1</code>) after the first <a href="https://topminisite.com/blog/how-to-restore-in-fully-connected-layer-using" target="_blank">fully connected layer</a> (<code>self.fc1</code>). During training, as the mini-batches pass through this network, the batch normalization layer will normalize the activations.</p><p><br/></p><p>Note: It is essential to ensure that the network is in training mode using <code>net.train()</code> before training and in evaluation mode using <code>net.eval()</code> during inference/testing.</p><p><br/></p><p>You can now use this network for training and inference in your PyTorch project, while enjoying the benefits of batch normalization.</p>
<div class="rating">
<h2>Best PyTorch Books of December 2024</h2>
<div class="row mt-2">
<div class="col-12">
<div class="v-card elevation-6">
<div class="v-card__text rating-text">
<div class="rating-counter">
<span class="v-badge">
<span class="v-badge__wrapper">
<span aria-atomic="true" aria-label="Позиция" class="v-badge__badge primary">
1
</span>
</span>
</span>
</div>
<div class="row">
<div class="col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center">
<div>
<img src="https://blogweb-static.fra1.cdn.digitaloceanspaces.com/images/d7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae/rating/41jkoc6owal-sl160.jpg" alt="PyTorch Recipes: A Problem-Solution Approach to Build, Train and Deploy Neural Network Models" />
<p class="text-center font-weight-bold text-h6">Rating is 5 out of 5</p>
<div class="stars" style="--rating: 5;" aria-label="Rating is 5 out of 5" ></div>
</div>
</div>
<div class="col-lg-6 col-md-8 col-sm-6 col-12">
<p class="font-weight-bold rating-name">PyTorch Recipes: A Problem-Solution Approach to Build, Train and Deploy Neural Network Models</p>
</div>
<div class="col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center">
<div class="text-center d-flex flex-column">
<a href="https://gosrc.cc/go/MDdl-1Y4g" target="_blank" rel="nofollow noopener" class="v-btn v-btn--rounded elevation-5 v-size--large success mb-2">
<span class="v-btn__content">Get Book Now</span>
</a>
</div>
</div>
</div>
</div>
</div>
</div>
<div class="col-12">
<div class="v-card elevation-6">
<div class="v-card__text rating-text">
<div class="rating-counter">
<span class="v-badge">
<span class="v-badge__wrapper">
<span aria-atomic="true" aria-label="Позиция" class="v-badge__badge primary">
2
</span>
</span>
</span>
</div>
<div class="row">
<div class="col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center">
<div>
<img src="https://blogweb-static.fra1.cdn.digitaloceanspaces.com/images/d7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae/rating/4198ez8dpbl-sl160.jpg" alt="Mastering PyTorch: Build powerful deep learning architectures using advanced PyTorch features, 2nd Edition" />
<p class="text-center font-weight-bold text-h6">Rating is 4.9 out of 5</p>
<div class="stars" style="--rating: 4.9;" aria-label="Rating is 4.9 out of 5" ></div>
</div>
</div>
<div class="col-lg-6 col-md-8 col-sm-6 col-12">
<p class="font-weight-bold rating-name">Mastering PyTorch: Build powerful deep learning architectures using advanced PyTorch features, 2nd Edition</p>
</div>
<div class="col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center">
<div class="text-center d-flex flex-column">
<a href="https://gosrc.cc/go/UZp_a1L4R" target="_blank" rel="nofollow noopener" class="v-btn v-btn--rounded elevation-5 v-size--large success mb-2">
<span class="v-btn__content">Get Book Now</span>
</a>
</div>
</div>
</div>
</div>
</div>
</div>
<div class="col-12">
<div class="v-card elevation-6">
<div class="v-card__text rating-text">
<div class="rating-counter">
<span class="v-badge">
<span class="v-badge__wrapper">
<span aria-atomic="true" aria-label="Позиция" class="v-badge__badge primary">
3
</span>
</span>
</span>
</div>
<div class="row">
<div class="col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center">
<div>
<img src="https://blogweb-static.fra1.cdn.digitaloceanspaces.com/images/d7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae/rating/513exc7qqjl-sl160.jpg" alt="Natural Language Processing with PyTorch: Build Intelligent Language Applications Using Deep Learning" />
<p class="text-center font-weight-bold text-h6">Rating is 4.8 out of 5</p>
<div class="stars" style="--rating: 4.8;" aria-label="Rating is 4.8 out of 5" ></div>
</div>
</div>
<div class="col-lg-6 col-md-8 col-sm-6 col-12">
<p class="font-weight-bold rating-name">Natural Language Processing with PyTorch: Build Intelligent Language Applications Using Deep Learning</p>
</div>
<div class="col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center">
<div class="text-center d-flex flex-column">
<a href="https://gosrc.cc/go/AQT_-1L4R" target="_blank" rel="nofollow noopener" class="v-btn v-btn--rounded elevation-5 v-size--large success mb-2">
<span class="v-btn__content">Get Book Now</span>
</a>
</div>
</div>
</div>
</div>
</div>
</div>
<div class="col-12">
<div class="v-card elevation-6">
<div class="v-card__text rating-text">
<div class="rating-counter">
<span class="v-badge">
<span class="v-badge__wrapper">
<span aria-atomic="true" aria-label="Позиция" class="v-badge__badge primary">
4
</span>
</span>
</span>
</div>
<div class="row">
<div class="col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center">
<div>
<img src="https://blogweb-static.fra1.cdn.digitaloceanspaces.com/images/d7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae/rating/518jzrflrvl-sl160.jpg" alt="Deep Learning for Coders with Fastai and PyTorch: AI Applications Without a PhD" />
<p class="text-center font-weight-bold text-h6">Rating is 4.7 out of 5</p>
<div class="stars" style="--rating: 4.7;" aria-label="Rating is 4.7 out of 5" ></div>
</div>
</div>
<div class="col-lg-6 col-md-8 col-sm-6 col-12">
<p class="font-weight-bold rating-name">Deep Learning for Coders with Fastai and PyTorch: AI Applications Without a PhD</p>
</div>
<div class="col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center">
<div class="text-center d-flex flex-column">
<a href="https://gosrc.cc/go/7JalaJYVR" target="_blank" rel="nofollow noopener" class="v-btn v-btn--rounded elevation-5 v-size--large success mb-2">
<span class="v-btn__content">Get Book Now</span>
</a>
</div>
</div>
</div>
</div>
</div>
</div>
<div class="col-12">
<div class="v-card elevation-6">
<div class="v-card__text rating-text">
<div class="rating-counter">
<span class="v-badge">
<span class="v-badge__wrapper">
<span aria-atomic="true" aria-label="Позиция" class="v-badge__badge primary">
5
</span>
</span>
</span>
</div>
<div class="row">
<div class="col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center">
<div>
<img src="https://blogweb-static.fra1.cdn.digitaloceanspaces.com/images/d7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae/rating/41o-uy9lzzl-sl160.jpg" alt="Machine Learning with PyTorch and Scikit-Learn: Develop machine learning and deep learning models with Python" />
<p class="text-center font-weight-bold text-h6">Rating is 4.6 out of 5</p>
<div class="stars" style="--rating: 4.6;" aria-label="Rating is 4.6 out of 5" ></div>
</div>
</div>
<div class="col-lg-6 col-md-8 col-sm-6 col-12">
<p class="font-weight-bold rating-name">Machine Learning with PyTorch and Scikit-Learn: Develop machine learning and deep learning models with Python</p>
</div>
<div class="col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center">
<div class="text-center d-flex flex-column">
<a href="https://gosrc.cc/go/GZs_a1Y4g" target="_blank" rel="nofollow noopener" class="v-btn v-btn--rounded elevation-5 v-size--large success mb-2">
<span class="v-btn__content">Get Book Now</span>
</a>
</div>
</div>
</div>
</div>
</div>
</div>
<div class="col-12">
<div class="v-card elevation-6">
<div class="v-card__text rating-text">
<div class="rating-counter">
<span class="v-badge">
<span class="v-badge__wrapper">
<span aria-atomic="true" aria-label="Позиция" class="v-badge__badge primary">
6
</span>
</span>
</span>
</div>
<div class="row">
<div class="col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center">
<div>
<img src="https://blogweb-static.fra1.cdn.digitaloceanspaces.com/images/d7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae/rating/411ruzisg3l-sl160.jpg" alt="Deep Learning with PyTorch: Build, train, and tune neural networks using Python tools" />
<p class="text-center font-weight-bold text-h6">Rating is 4.5 out of 5</p>
<div class="stars" style="--rating: 4.5;" aria-label="Rating is 4.5 out of 5" ></div>
</div>
</div>
<div class="col-lg-6 col-md-8 col-sm-6 col-12">
<p class="font-weight-bold rating-name">Deep Learning with PyTorch: Build, train, and tune neural networks using Python tools</p>
</div>
<div class="col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center">
<div class="text-center d-flex flex-column">
<a href="https://gosrc.cc/go/8lwlaJY4R" target="_blank" rel="nofollow noopener" class="v-btn v-btn--rounded elevation-5 v-size--large success mb-2">
<span class="v-btn__content">Get Book Now</span>
</a>
</div>
</div>
</div>
</div>
</div>
</div>
<div class="col-12">
<div class="v-card elevation-6">
<div class="v-card__text rating-text">
<div class="rating-counter">
<span class="v-badge">
<span class="v-badge__wrapper">
<span aria-atomic="true" aria-label="Позиция" class="v-badge__badge primary">
7
</span>
</span>
</span>
</div>
<div class="row">
<div class="col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center">
<div>
<img src="https://blogweb-static.fra1.cdn.digitaloceanspaces.com/images/d7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae/rating/41sz-tftqpl-sl160.jpg" alt="Programming PyTorch for Deep Learning: Creating and Deploying Deep Learning Applications" />
<p class="text-center font-weight-bold text-h6">Rating is 4.4 out of 5</p>
<div class="stars" style="--rating: 4.4;" aria-label="Rating is 4.4 out of 5" ></div>
</div>
</div>
<div class="col-lg-6 col-md-8 col-sm-6 col-12">
<p class="font-weight-bold rating-name">Programming PyTorch for Deep Learning: Creating and Deploying Deep Learning Applications</p>
</div>
<div class="col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center">
<div class="text-center d-flex flex-column">
<a href="https://gosrc.cc/go/1d9_-JY4R" target="_blank" rel="nofollow noopener" class="v-btn v-btn--rounded elevation-5 v-size--large success mb-2">
<span class="v-btn__content">Get Book Now</span>
</a>
</div>
</div>
</div>
</div>
</div>
</div>
<div class="col-12">
<div class="v-card elevation-6">
<div class="v-card__text rating-text">
<div class="rating-counter">
<span class="v-badge">
<span class="v-badge__wrapper">
<span aria-atomic="true" aria-label="Позиция" class="v-badge__badge primary">
8
</span>
</span>
</span>
</div>
<div class="row">
<div class="col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center">
<div>
<img src="https://blogweb-static.fra1.cdn.digitaloceanspaces.com/images/d7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae/rating/41xz0sfhxsl-sl160.jpg" alt="PyTorch Pocket Reference: Building and Deploying Deep Learning Models" />
<p class="text-center font-weight-bold text-h6">Rating is 4.3 out of 5</p>
<div class="stars" style="--rating: 4.3;" aria-label="Rating is 4.3 out of 5" ></div>
</div>
</div>
<div class="col-lg-6 col-md-8 col-sm-6 col-12">
<p class="font-weight-bold rating-name">PyTorch Pocket Reference: Building and Deploying Deep Learning Models</p>
</div>
<div class="col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center">
<div class="text-center d-flex flex-column">
<a href="https://gosrc.cc/go/she_-1YVg" target="_blank" rel="nofollow noopener" class="v-btn v-btn--rounded elevation-5 v-size--large success mb-2">
<span class="v-btn__content">Get Book Now</span>
</a>
</div>
</div>
</div>
</div>
</div>
</div>
<div class="col-12">
<div class="v-card elevation-6">
<div class="v-card__text rating-text">
<div class="rating-counter">
<span class="v-badge">
<span class="v-badge__wrapper">
<span aria-atomic="true" aria-label="Позиция" class="v-badge__badge primary">
9
</span>
</span>
</span>
</div>
<div class="row">
<div class="col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center">
<div>
<img src="https://blogweb-static.fra1.cdn.digitaloceanspaces.com/images/d7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae/rating/41ayiicwf2l-sl160.jpg" alt="Deep Learning with PyTorch Lightning: Swiftly build high-performance Artificial Intelligence (AI) models using Python" />
<p class="text-center font-weight-bold text-h6">Rating is 4.2 out of 5</p>
<div class="stars" style="--rating: 4.2;" aria-label="Rating is 4.2 out of 5" ></div>
</div>
</div>
<div class="col-lg-6 col-md-8 col-sm-6 col-12">
<p class="font-weight-bold rating-name">Deep Learning with PyTorch Lightning: Swiftly build high-performance Artificial Intelligence (AI) models using Python</p>
</div>
<div class="col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center">
<div class="text-center d-flex flex-column">
<a href="https://gosrc.cc/go/kmiX-1LVg" target="_blank" rel="nofollow noopener" class="v-btn v-btn--rounded elevation-5 v-size--large success mb-2">
<span class="v-btn__content">Get Book Now</span>
</a>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
<p><br/></p><h2>How to implement batch normalization in a custom PyTorch model?</h2><p>To implement batch normalization in a custom PyTorch model, you can follow these steps:</p><ol><li>Import the required modules:</li></ol><div style="color:#f8f8f2;background-color:#272822;">
<table style="border-spacing:0;padding:0;margin:0;border:0;"><tbody><tr><td style="vertical-align:top;padding:0;margin:0;border:0;">
<pre tabindex="0" style="color:#f8f8f2;background-color:#272822;"><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">1
</span><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">2
</span></pre></td>
<td style="vertical-align:top;padding:0;margin:0;border:0;;width:100%">
<pre tabindex="0" style="color:#f8f8f2;background-color:#272822;"><span style="display:flex;"><span>import torch
</span></span><span style="display:flex;"><span>import torch.nn as nn
</span></span></pre></td></tr></tbody></table>
</div>
<p><br/></p><script async="" src="https://pagead2.googlesyndication.com/pagead/js/adsbygoogle.js"></script>
<ins class="adsbygoogle" style="display:block" data-ad-format="fluid" data-ad-layout-key="-ef+6k-30-ac+ty" data-ad-client="ca-pub-4833888168110763" data-ad-slot="3267362137"></ins>
<script>
(adsbygoogle = window.adsbygoogle || []).push({});
</script><ol><li>Define a basic custom model class:</li></ol><div style="color:#f8f8f2;background-color:#272822;">
<table style="border-spacing:0;padding:0;margin:0;border:0;"><tbody><tr><td style="vertical-align:top;padding:0;margin:0;border:0;">
<pre tabindex="0" style="color:#f8f8f2;background-color:#272822;"><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">1
</span><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">2
</span><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">3
</span><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">4
</span><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">5
</span></pre></td>
<td style="vertical-align:top;padding:0;margin:0;border:0;;width:100%">
<pre tabindex="0" style="color:#f8f8f2;background-color:#272822;"><span style="display:flex;"><span>class CustomModel(nn.Module):
</span></span><span style="display:flex;"><span> def __init__(self):
</span></span><span style="display:flex;"><span> super(CustomModel, self).__init__()
</span></span><span style="display:flex;"><span> self.fc1 = nn.Linear(in_features, hidden_units)
</span></span><span style="display:flex;"><span> self.fc2 = nn.Linear(hidden_units, out_features)
</span></span></pre></td></tr></tbody></table>
</div>
<p><br/></p><p>Replace <code>in_features</code>, <code>hidden_units</code>, and <code>out_features</code> with appropriate values for your model architecture.</p><ol><li>Add batch normalization layers and their parameters to the model:</li></ol><div style="color:#f8f8f2;background-color:#272822;">
<table style="border-spacing:0;padding:0;margin:0;border:0;"><tbody><tr><td style="vertical-align:top;padding:0;margin:0;border:0;">
<pre tabindex="0" style="color:#f8f8f2;background-color:#272822;"><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">1
</span><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">2
</span></pre></td>
<td style="vertical-align:top;padding:0;margin:0;border:0;;width:100%">
<pre tabindex="0" style="color:#f8f8f2;background-color:#272822;"><span style="display:flex;"><span>self.bn1 = nn.BatchNorm1d(hidden_units)
</span></span><span style="display:flex;"><span>self.bn2 = nn.BatchNorm1d(out_features)
</span></span></pre></td></tr></tbody></table>
</div>
<p><br/></p><p>Adjust the parameter value based on your model architecture.</p><ol><li>Define the forward pass of the model:</li></ol><div style="color:#f8f8f2;background-color:#272822;">
<table style="border-spacing:0;padding:0;margin:0;border:0;"><tbody><tr><td style="vertical-align:top;padding:0;margin:0;border:0;">
<pre tabindex="0" style="color:#f8f8f2;background-color:#272822;"><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">1
</span><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">2
</span><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">3
</span><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">4
</span><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">5
</span><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">6
</span><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">7
</span></pre></td>
<td style="vertical-align:top;padding:0;margin:0;border:0;;width:100%">
<pre tabindex="0" style="color:#f8f8f2;background-color:#272822;"><span style="display:flex;"><span>def forward(self, x):
</span></span><span style="display:flex;"><span> x = self.fc1(x)
</span></span><span style="display:flex;"><span> x = self.bn1(x)
</span></span><span style="display:flex;"><span> x = nn.functional.relu(x)
</span></span><span style="display:flex;"><span> x = self.fc2(x)
</span></span><span style="display:flex;"><span> x = self.bn2(x)
</span></span><span style="display:flex;"><span> return x
</span></span></pre></td></tr></tbody></table>
</div>
<p><br/></p><p>This example assumes the ReLU activation function, but you can replace it with any activation function you prefer.</p><ol><li>Create an instance of the custom model:</li></ol><div style="color:#f8f8f2;background-color:#272822;">
<table style="border-spacing:0;padding:0;margin:0;border:0;"><tbody><tr><td style="vertical-align:top;padding:0;margin:0;border:0;">
<pre tabindex="0" style="color:#f8f8f2;background-color:#272822;"><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">1
</span></pre></td>
<td style="vertical-align:top;padding:0;margin:0;border:0;;width:100%">
<pre tabindex="0" style="color:#f8f8f2;background-color:#272822;"><span style="display:flex;"><span>model = CustomModel()
</span></span></pre></td></tr></tbody></table>
</div>
<p><br/></p><p>Now you have implemented batch normalization in your custom PyTorch model.</p><p><br/></p><h2>What are the advantages of using batch normalization in PyTorch?</h2><p>Batch normalization is a regularization technique that is widely used in deep learning models. When applied to PyTorch models, it provides several advantages:</p><ol><li><strong>Improved convergence</strong>: Batch normalization normalizes the input to each neuron across a mini-batch, which helps in stabilizing the learning process. This leads to faster convergence and reduces the number of epochs required for training.
</li><li><strong>Reduced overfitting</strong>: By normalizing the inputs, batch normalization reduces the dependence of each neuron on the other neurons in the network. This reduces the chances of overfitting and improves the generalization ability of the model.
</li><li><strong>Increased learning rate</strong>: Batch normalization reduces the internal covariate shift by maintaining zero mean and unit variance activations. This enables the use of higher learning rates during training, which can speed up the training process.
</li><li><strong>Better gradient flow</strong>: Normalizing the inputs using batch normalization helps in ensuring that the gradients flow smoothly and consistently during backpropagation. This helps combat the vanishing and exploding gradient problems, making it easier to train deep networks.
</li><li><strong>Robustness to different input distributions</strong>: Batch normalization makes the model less sensitive to the <a href="https://studentprojectcode.com/blog/how-to-simplify-units-with-different-scales-in" target="_blank">scale</a> and distribution of the input data. This allows the model to perform well even when faced with inputs that are significantly different from the training data.
</li><li><strong>Weight initialization flexibility</strong>: Batch normalization helps in reducing the dependence of the model's performance on the choice of weight initialization. It allows the use of simpler initialization methods like random or small weights, which can speed up the training process.
</li></ol><p><br/></p><p>Overall, batch normalization is a useful tool for improving the performance and stability of deep learning models in PyTorch, leading to faster convergence, better generalization, and increased robustness.</p><p><br/></p><h2>What is the effect of batch size on batch normalization in PyTorch?</h2><p>The batch size affects the batch normalization in PyTorch in the following way:</p><ol><li><strong>Statistics estimation</strong>: Batch normalization relies on estimating the mean and variance of the input data to normalize it. With a larger batch size, there is more data available for statistics estimation, leading to more accurate estimates of the mean and variance. This can result in improved normalization and consequently, better performance.
</li><li><strong>Noise reduction</strong>: Batch normalization introduces some noise to the statistics estimation process. With a larger batch size, the noise is averaged out more effectively, resulting in more stable estimates of mean and variance. This can lead to reduced overfitting and improved generalization.
</li><li><strong>Training dynamics</strong>: Smaller batch sizes tend to introduce more stochasticity and randomness in the training process, as each batch's statistics differ significantly. On the other hand, larger batch sizes provide more consistent statistics, which can affect the optimization process. This can result in different training dynamics, such as convergence speed and stability.
</li></ol><p><br/></p><script async="" src="https://pagead2.googlesyndication.com/pagead/js/adsbygoogle.js"></script>
<ins class="adsbygoogle" style="display:block" data-ad-format="fluid" data-ad-layout-key="-ef+6k-30-ac+ty" data-ad-client="ca-pub-4833888168110763" data-ad-slot="3267362137"></ins>
<script>
(adsbygoogle = window.adsbygoogle || []).push({});
</script><p>It's important to note that the choice of batch size is often a trade-off. Larger batch sizes require more memory, may limit parallelization, and increase computational requirements. However, they can offer better normalization and estimation, while smaller batch sizes may introduce more noise but can be computationally more efficient.</p><p><br/></p><h2>What are the requirements for using batch normalization in PyTorch?</h2><p>To use batch normalization in PyTorch, the following requirements should be met:</p><ol><li><strong>PyTorch should be installed on the system. You can install it using pip</strong>: pip install torch.
</li><li>Import the necessary modules:
</li></ol><div style="color:#f8f8f2;background-color:#272822;">
<table style="border-spacing:0;padding:0;margin:0;border:0;"><tbody><tr><td style="vertical-align:top;padding:0;margin:0;border:0;">
<pre tabindex="0" style="color:#f8f8f2;background-color:#272822;"><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">1
</span><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">2
</span></pre></td>
<td style="vertical-align:top;padding:0;margin:0;border:0;;width:100%">
<pre tabindex="0" style="color:#f8f8f2;background-color:#272822;"><span style="display:flex;"><span>import torch
</span></span><span style="display:flex;"><span>import torch.nn as nn
</span></span></pre></td></tr></tbody></table>
</div>
<p><br/></p><ol><li>Define your model architecture using the nn.Module class. Use the torch.nn.BatchNorm2d or torch.nn.BatchNorm1d layer (based on your input dimensions) for batch normalization.
</li><li>Use batch normalization layer after the convolutional or linear layer in your model architecture. For example:
</li></ol><div style="color:#f8f8f2;background-color:#272822;">
<table style="border-spacing:0;padding:0;margin:0;border:0;"><tbody><tr><td style="vertical-align:top;padding:0;margin:0;border:0;">
<pre tabindex="0" style="color:#f8f8f2;background-color:#272822;"><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">1
</span><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">2
</span><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">3
</span><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">4
</span><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">5
</span><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">6
</span><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">7
</span><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">8
</span></pre></td>
<td style="vertical-align:top;padding:0;margin:0;border:0;;width:100%">
<pre tabindex="0" style="color:#f8f8f2;background-color:#272822;"><span style="display:flex;"><span>class MyModel(nn.Module):
</span></span><span style="display:flex;"><span> def __init__(self):
</span></span><span style="display:flex;"><span> super(MyModel, self).__init__()
</span></span><span style="display:flex;"><span> self.conv1 = nn.Conv2d(3, 64, kernel_size=3)
</span></span><span style="display:flex;"><span> self.bn1 = nn.BatchNorm2d(64)
</span></span><span style="display:flex;"><span> self.fc1 = nn.Linear(64, 10)
</span></span><span style="display:flex;"><span> self.bn2 = nn.BatchNorm1d(10)
</span></span><span style="display:flex;"><span> ...
</span></span></pre></td></tr></tbody></table>
</div>
<p><br/></p><ol><li>During the forward pass, apply batch normalization to the input tensor. For example:</li></ol><div style="color:#f8f8f2;background-color:#272822;">
<table style="border-spacing:0;padding:0;margin:0;border:0;"><tbody><tr><td style="vertical-align:top;padding:0;margin:0;border:0;">
<pre tabindex="0" style="color:#f8f8f2;background-color:#272822;"><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">1
</span><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">2
</span><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">3
</span><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">4
</span><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">5
</span><span style="white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f">6
</span></pre></td>
<td style="vertical-align:top;padding:0;margin:0;border:0;;width:100%">
<pre tabindex="0" style="color:#f8f8f2;background-color:#272822;"><span style="display:flex;"><span>def forward(self, x):
</span></span><span style="display:flex;"><span> x = self.conv1(x)
</span></span><span style="display:flex;"><span> x = self.bn1(x)
</span></span><span style="display:flex;"><span> x = self.fc1(x)
</span></span><span style="display:flex;"><span> x = self.bn2(x)
</span></span><span style="display:flex;"><span> ...
</span></span></pre></td></tr></tbody></table>
</div>
<p><br/></p><p>Note: Batch normalization is typically used before the activation function, but the order can vary depending on your problem and experiment settings.</p><p><br/></p><h2>What is the impact of batch normalization on model generalization in PyTorch?</h2><p>Batch normalization has a significant impact on model generalization in PyTorch. It helps to improve the generalization capability of neural networks by reducing the internal covariate shift.</p><p><br/></p><p>Internal covariate shift refers to the change in the distribution of network activations due to the change in parameter values during training. This can slow down the training process and hinder the performance of the model.</p><p><br/></p><p>Batch normalization solves this problem by normalizing the output of each layer using the mean and variance of the mini-batch. By doing so, it reduces the effect of the internal covariate shift and makes the optimization process more stable. Batch normalization also introduces additional trainable parameters, which allow the network to adaptively scale and shift the normalized values.</p><p><br/></p><p>The normalization of inputs helps in the generalization of the model because it keeps the values within a reasonable range. It prevents extreme values from causing instability in the network, which can lead to overfitting. Additionally, batch normalization acts as a regularizer, reducing the need for other regularization techniques like dropout.</p><script async="" src="https://pagead2.googlesyndication.com/pagead/js/adsbygoogle.js"></script>
<ins class="adsbygoogle" style="display:block" data-ad-format="fluid" data-ad-layout-key="-ef+6k-30-ac+ty" data-ad-client="ca-pub-4833888168110763" data-ad-slot="3267362137"></ins>
<script>
(adsbygoogle = window.adsbygoogle || []).push({});
</script><p><br/></p><p>Overall, batch normalization in PyTorch improves the generalization ability of models by reducing internal covariate shift, making the training process more stable, and acting as a regularizer.</p></div></div></div></div>  <div class="text-center col"></div> <div class="justify-center icons d-flex col col-12" data-v-80f69840 data-v-80f69840><a href="https://www.facebook.com/sharer.php?src=sp&u=https%3A%2F%2Fstlplaces.com%2Fblog%2Fhow-to-implement-batch-normalization-in-pytorch&quote=How%20to%20Implement%20Batch%20Normalization%20In%20PyTorch%3F&hashtag=%23blogweb" rel="nofollow noopener" target="_blank" title="Facebook" data-v-80f69840><img src="https://blogweb-static.fra1.cdn.digitaloceanspaces.com/assets/images/icons/32/fb.png" width="32" height="32" alt="Facebook" data-v-80f69840></a>   <a href="https://twitter.com/intent/tweet?url=https%3A%2F%2Fstlplaces.com%2Fblog%2Fhow-to-implement-batch-normalization-in-pytorch&text=How%20to%20Implement%20Batch%20Normalization%20In%20PyTorch%3F&hashtags=blogweb" rel="nofollow noopener" target="_blank" title="Twitter" data-v-80f69840><img src="https://blogweb-static.fra1.cdn.digitaloceanspaces.com/assets/images/icons/32/twitter.png" width="32" height="32" alt="Twitter" data-v-80f69840></a> <a href="https://www.linkedin.com/sharing/share-offsite/?url=https%3A%2F%2Fstlplaces.com%2Fblog%2Fhow-to-implement-batch-normalization-in-pytorch" rel="nofollow noopener" target="_blank" title="LinkedIn" data-v-80f69840><img src="https://blogweb-static.fra1.cdn.digitaloceanspaces.com/assets/images/icons/32/linkedin.png" width="32" height="32" alt="LinkedIn" data-v-80f69840></a>  <a href="https://api.whatsapp.com/send?text=https%3A%2F%2Fstlplaces.com%2Fblog%2Fhow-to-implement-batch-normalization-in-pytorch" rel="nofollow noopener" target="_blank" title="Whatsapp" data-v-80f69840><img src="https://blogweb-static.fra1.cdn.digitaloceanspaces.com/assets/images/icons/32/whatsapp.png" width="32" height="32" alt="Whatsapp" data-v-80f69840></a> <a href="https://getpocket.com/save?url=https%3A%2F%2Fstlplaces.com%2Fblog%2Fhow-to-implement-batch-normalization-in-pytorch" rel="nofollow noopener" target="_blank" title="Pocket" data-v-80f69840><img src="https://blogweb-static.fra1.cdn.digitaloceanspaces.com/assets/images/icons/32/pocket.png" width="32" height="32" alt="Pocket" data-v-80f69840></a></div></div>   <div class="row mt-2"><div class="col col-12"><h2 class="display-1">Related Posts:</h2></div> <div class="col-sm-12 col-md-6 col-lg-4 col-12"><div class="mx-auto v-card v-sheet theme--light" style="max-width:400px;"><div class="v-image v-responsive align-end theme--light" style="height:200px;"><div class="v-image__image v-image__image--preload v-image__image--cover" style="background-image:;background-position:center center;"></div><div class="v-responsive__content"></div></div> <div class="v-card__title"><a href="/blog/how-to-implement-batch-normalization-in-tensorflow" itemprop="mainEntityOfPage url">
How to Implement Batch Normalization In TensorFlow?
</a></div> <div class="v-card__text text--primary">
Batch normalization is a technique used to improve the speed, stability, and performance of neural networks. It works by normalizing the output of the previous layer within each batch of training examples. This helps in mitigating the issue of internal covaria...
</div></div></div><div class="col-sm-12 col-md-6 col-lg-4 col-12"><div class="mx-auto v-card v-sheet theme--light" style="max-width:400px;"><div class="v-image v-responsive align-end theme--light" style="height:200px;"><div class="v-image__image v-image__image--preload v-image__image--cover" style="background-image:;background-position:center center;"></div><div class="v-responsive__content"></div></div> <div class="v-card__title"><a href="/blog/how-to-batch-images-with-arbitrary-sizes-in" itemprop="mainEntityOfPage url">
How to Batch Images With Arbitrary Sizes In Tensorflow?
</a></div> <div class="v-card__text text--primary">
To batch images with arbitrary sizes in TensorFlow, you can use the tf.image.resize_with_pad() function to resize the images to a specific size before batching them together. You can specify the target size for resizing the images and pad them if necessary to ...
</div></div></div><div class="col-sm-12 col-md-6 col-lg-4 col-12"><div class="mx-auto v-card v-sheet theme--light" style="max-width:400px;"><div class="v-image v-responsive align-end theme--light" style="height:200px;"><div class="v-image__image v-image__image--preload v-image__image--cover" style="background-image:;background-position:center center;"></div><div class="v-responsive__content"></div></div> <div class="v-card__title"><a href="/blog/how-to-do-batch-filling-in-pytorch" itemprop="mainEntityOfPage url">
How to Do Batch Filling In Pytorch?
</a></div> <div class="v-card__text text--primary">
Batch filling in PyTorch refers to the process of creating a batch of data from a given dataset. It involves splitting the dataset into smaller batches, which are then used for model training or inference.To perform batch filling in PyTorch, you can follow the...
</div></div></div></div></div> </div></div> <div pages="" class="footer" data-v-293f1957>  <footer class="v-footer footer mt-4 v-sheet theme--light" data-v-293f1957> <div class="flex xs12 sm12 md6 lg4" data-v-293f1957><h4 class="title text-center" data-v-293f1957>Menu</h4> <div role="list" class="v-list no-background v-sheet theme--light" data-v-293f1957><div data-v-293f1957><a tabindex="0" href="/page/privacy-policy" target="_blank" router="" role="listitem" class="v-list-item v-list-item--link theme--light" data-v-293f1957><span class="mx-auto" data-v-293f1957>Privacy Policy</span></a></div><div data-v-293f1957><a tabindex="0" href="/page/terms-of-use" target="_blank" router="" role="listitem" class="v-list-item v-list-item--link theme--light" data-v-293f1957><span class="mx-auto" data-v-293f1957>Terms of Use</span></a></div></div></div> <div class="flex xs12 sm12 md6 lg4" data-v-293f1957><div class="about-section white-text mx-auto" data-v-293f1957><h4 class="title text-center" data-v-293f1957>About us</h4> <ul class="text-center" data-v-293f1957><li data-v-293f1957><a href="/contact" rel="nofollow" data-v-293f1957>Contact us</a></li></ul></div></div> <div class="flex text-center mt-2 xs12" data-v-293f1957><span data-v-293f1957>© 2024 Copyright: stlplaces.com
</span> <p class="mb-0" data-v-293f1957>
This Site is Created By
<a href="https://blogweb.me" target="_blank" data-v-293f1957>blogweb.me</a></p> <p data-v-293f1957>
Get Your Own
<a href="https://blogweb.me" target="_blank" data-v-293f1957>free blog</a></p></div></footer>  </div> </div> <div class="js-scripts"></div></div></div></div></div><script>window.__NUXT__=(function(a,b,c,d,e,f,g,h,i,j,k,l,m,n,o,p,q,r,s,t,u,v,w,x,y,z,A,B,C,D,E,F,G,H,I,J,K,L,M,N,O,P,Q,R,S,T,U,V,W,X,Y,Z,_,$,aa,ab,ac,ad,ae,af,ag,ah,ai,aj,ak,al,am,an,ao,ap,aq,ar,as,at,au,av,aw,ax,ay,az,aA,aB,aC,aD,aE,aF,aG,aH,aI,aJ,aK,aL,aM,aN,aO,aP,aQ,aR,aS,aT,aU,aV,aW,aX,aY,aZ,a_,a$,ba,bb,bc,bd,be,bf,bg,bh,bi,bj,bk,bl,bm,bn,bo,bp,bq,br,bs,bt,bu,bv,bw,bx,by,bz,bA,bB,bC,bD,bE,bF,bG,bH,bI,bJ,bK,bL,bM){return {layout:ar,data:[{url:as,relatedPosts:[{id:at,text:au,title:a,image:av,summary:aw,slug:ax},{id:ay,text:az,title:a,image:aA,summary:aB,slug:aC},{id:aD,text:aE,title:a,image:aF,summary:aG,slug:aH}],hasTags:[],isDesktop:void 0,comments:[],totalComments:b,blogSettings:{id:h,toc:e,toc_collapse:e,add_source:e,add_source_text:U,is_related_post:e,related_post:aI},blogSocialSettings:{id:79,size:b,facebook:d,vk:c,ok:c,linkedin:d,telegram:c,whatsapp:d,pocket:d,twitter:d,active:d},categories:[{id:1359,name:V,meta_title:V,meta_description:V,order:b,children:[],description:a,slug:"about-st-louis"},{id:1425,name:W,meta_title:W,meta_description:W,order:b,children:[],description:a,slug:"services"},{id:1426,name:X,meta_title:X,meta_description:X,order:b,children:[],description:a,slug:"products"},{id:1603,name:x,meta_title:x,meta_description:x,order:b,children:[],description:x,slug:"books"},{id:2552,name:aJ,meta_title:aJ,meta_description:a,order:b,children:[],description:a,slug:"fitness"},{id:2574,name:"Finance",meta_title:"Financial Insights: Expert Tips & Strategies",meta_description:"Unlock your financial success with expert tips, investment strategies, and practical advice on managing money. Stay informed and make informed decisions in the world of finance.",order:b,children:[],description:a,slug:"finance"},{id:aK,name:Y,meta_title:"Code Crunchers: Unleash Your Programming Skills",meta_description:"Discover the latest trends, tips, and tricks in programming. Enhance your coding skills with expert insights, tutorials, and practical examples. Dive into the world of programming here!",order:b,children:[],description:a,slug:aL}],post:{id:"658325005e1d703a99de76d5",post_id:110028,site:{Id:h,Name:k,Hash:aM,Description:y,PrivacyPolicy:"\u003Cp class=\"ql-align-center\"\u003E\u003Cstrong\u003EPrivacy Policy\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp class=\"ql-align-center\"\u003E\u003Cstrong\u003ELast Updated: May 9, 2020\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003E\u003Cstrong\u003EIntroduction\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003EThis privacy policy (“\u003Cstrong\u003EPolicy\u003C\u002Fstrong\u003E”) describes how \u003Cstrong\u003ESelf-employment in California\u003C\u002Fstrong\u003E (“\u003Cstrong\u003ECompany,\u003C\u002Fstrong\u003E” “\u003Cstrong\u003Ewe,\u003C\u002Fstrong\u003E” and “\u003Cstrong\u003Eour\u003C\u002Fstrong\u003E”) processes, collects, uses and shares personal data when using this website https:\u002F\u002Fblogweb.me (the “\u003Cstrong\u003EWebsite\u003C\u002Fstrong\u003E”). Please read the following information carefully to understand our practices regarding your personal data and how we will process data.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cstrong\u003EPurposes of Processing\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003E\u003Cstrong\u003EWhat is personal data?\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003EWe collect information about you in a range of forms, including personal data. As used in this Policy, “personal data” is as defined in the General Data Protection Regulation, this includes any information which, either alone or in combination with other information we process about you, identifies you as an individual, including, for example, your name, postal address, email address and telephone number.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cstrong\u003EWhy do we need your personal data?\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003EWe will only process your personal data in accordance with applicable data protection and privacy laws. We need certain personal data in order to provide you with access to the website. If you registered with us, you will have been asked to tick to agree to provide this information in order to access our services, purchase our products, or view our content. This consent provides us with the legal basis we require under applicable law to process your data. You maintain the right to withdraw such consent at any time. If you do not agree to our use of your personal data in line with this Policy, please do not use our website.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cstrong\u003ECollecting Your Personal Data\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003EWe collect information about you in the following ways:\u003C\u002Fp\u003E\u003Cp\u003E\u003Cstrong\u003EInformation You Give Us.\u003C\u002Fstrong\u003E This includes:\u003C\u002Fp\u003E\u003Cul\u003E\u003Cli\u003Ethe personal data you provide when you register to use our website, including your name, postal address, email address, telephone number, username, password and demographic information (such as your gender);\u003C\u002Fli\u003E\u003Cli\u003Ethe personal data that may be contained in any video, comment or other submission you upload or post to the website;\u003C\u002Fli\u003E\u003Cli\u003Ethe personal data you provide in connection with our rewards program and other promotions we run on the website;\u003C\u002Fli\u003E\u003Cli\u003Ethe personal data you provide when you report a problem with our website or when we provide you with customer support;\u003C\u002Fli\u003E\u003Cli\u003Ethe personal data you provide when you make a purchase thorough our website; and\u003C\u002Fli\u003E\u003Cli\u003Ethe personal data you provide when you correspond with us by phone, email or otherwise.\u003C\u002Fli\u003E\u003C\u002Ful\u003E\u003Cp\u003E\u003Cstrong\u003EInformation from Social Networking Websites.\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003EOur website includes interfaces that allow you to connect with social networking sites (each a “SNS”). If you connect to a SNS through our website, you authorize us to access, use and store the information that you agreed the SNS could provide to us based on your settings on that SNS. We will access, use and store that information in accordance with this Policy. You can revoke our access to the information you provide in this way at any time by amending the appropriate settings from within your account settings on the applicable SNS.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cstrong\u003EInformation Automatically Collected.\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003EWe automatically log information about you and your computer or mobile device when you access our website. For example, when visiting our website, we log your computer or mobile device operating system name and version, manufacturer and model, browser type, browser language, screen resolution, the website you visited before browsing to our website, pages you viewed, how long you spent on a page, access times and information about your use of and actions on our website. We collect this information about you using cookies.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cstrong\u003EAutomated Decision Making and Profiling.\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003EWe do not use your personal data for the purposes of automated decision-making. However, we may do so in order to fulfill obligations imposed by law, in which case we will inform you of any such processing and provide you with an opportunity to object.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cstrong\u003ECookies\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003E\u003Cstrong\u003EWhat are cookies?\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003EWe may collect information using “cookies.” Cookies are small data files stored on the hard drive of your computer or mobile device by a website. We may use both session cookies (which expire once you close your web browser) and persistent cookies (which stay on your computer or mobile device until you delete them) to provide you with a more personal and interactive experience on our website.\u003C\u002Fp\u003E\u003Cp\u003EWe use two broad categories of cookies: (1) first party cookies, served directly by us to your computer or mobile device, which are used only by us to recognize your computer or mobile device when it revisits our website; and (2) third party cookies, which are served by service providers on our website, and can be used by such service providers to recognize your computer or mobile device when it visits other websites.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u003E\u003C\u002Fp\u003E\u003Cp\u003E\u003Cstrong\u003ECookies we use\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003EOur website uses the following types of cookies for the purposes set out below:\u003C\u002Fp\u003E\u003Cp\u003E\u003Cstrong\u003EType of cookie\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003E\u003Cstrong\u003EPurpose\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003E\u003Cem\u003EEssential Cookies\u003C\u002Fem\u003E\u003C\u002Fp\u003E\u003Cp\u003EThese cookies are essential to provide you with services available through our website and to enable you to use some of its features. For example, they allow you to log in to secure areas of our website and help the content of the pages you request load quickly.\u003Cstrong\u003E \u003C\u002Fstrong\u003EWithout these cookies, the services that you have asked for cannot be provided, and we only use these cookies to provide you with those services.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cem\u003EFunctionality Cookies\u003C\u002Fem\u003E\u003C\u002Fp\u003E\u003Cp\u003EThese cookies allow our website to remember choices you make when you use our website, such as remembering your language preferences, remembering your login details and remembering the changes you make to other parts of our website which you can customize. The purpose of these cookies is to provide you with a more personal experience and to avoid you having to re-enter your preferences every time you visit our website.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cem\u003EAnalytics and Performance Cookies\u003C\u002Fem\u003E\u003C\u002Fp\u003E\u003Cp\u003EThese cookies are used to collect information about traffic to our website and how users use our website. The information gathered does not identify any individual visitor. It includes the number of visitors to our website, the websites that referred them to our website, the pages they visited on our website, what time of day they visited our website, whether they have visited our website before, and other similar information. We use this information to help operate our website more efficiently, to gather broad demographic information and to monitor the level of activity on our website.\u003C\u002Fp\u003E\u003Cp\u003EWe use Google Analytics for this purpose. Google Analytics uses its own cookies. It is only used to improve how our website works. You can find out more information about Google Analytics cookies here: \u003Ca href=\"https:\u002F\u002Fdevelopers.google.com\u002Fanalytics\u002Fresources\u002Fconcepts\u002FgaConceptsCookies\" rel=\"noopener noreferrer\" target=\"_blank\" style=\"color: rgb(0, 0, 255);\"\u003E\u003Cu\u003Ehttps:\u002F\u002Fdevelopers.google.com\u002Fanalytics\u002Fresources\u002Fconcepts\u002FgaConceptsCookies\u003C\u002Fu\u003E\u003C\u002Fa\u003E\u003C\u002Fp\u003E\u003Cp\u003EYou can find out more about how Google protects your data here: \u003Ca href=\"https:\u002F\u002Fpolicies.google.com\u002Fprivacy\" rel=\"noopener noreferrer\" target=\"_blank\" style=\"color: rgb(0, 0, 255);\"\u003E\u003Cu\u003Ehttps:\u002F\u002Fpolicies.google.com\u002Fprivacy\u003C\u002Fu\u003E\u003C\u002Fa\u003E.\u003C\u002Fp\u003E\u003Cp\u003EYou can prevent the use of Google Analytics relating to your use of our website by downloading and installing the browser plugin available via this link: \u003Ca href=\"http:\u002F\u002Ftools.google.com\u002Fdlpage\u002Fgaoptout?hl=en-GB\" rel=\"noopener noreferrer\" target=\"_blank\" style=\"color: rgb(0, 0, 255);\"\u003E\u003Cu\u003Ehttp:\u002F\u002Ftools.google.com\u002Fdlpage\u002Fgaoptout?hl=en-GB\u003C\u002Fu\u003E\u003C\u002Fa\u003E\u003C\u002Fp\u003E\u003Cp\u003E\u003Cem\u003ETargeted and advertising cookies\u003C\u002Fem\u003E\u003C\u002Fp\u003E\u003Cp\u003EThese cookies track your browsing habits to enable us to show advertising which is more likely to be of interest to you. These cookies use information about your browsing history to group you with other users who have similar interests. Based on that information, and with our permission, third party advertisers can place cookies to enable them to show adverts which we think will be relevant to your interests while you are on third party websites.\u003C\u002Fp\u003E\u003Cp\u003EYou can disable cookies which remember your browsing habits and target advertising at you by visiting \u003Ca href=\"http:\u002F\u002Fwww.youronlinechoices.com\u002Fuk\u002Fyour-ad-choices\" rel=\"noopener noreferrer\" target=\"_blank\" style=\"color: rgb(0, 0, 255);\"\u003E\u003Cu\u003Ehttp:\u002F\u002Fwww.youronlinechoices.com\u002Fuk\u002Fyour-ad-choices\u003C\u002Fu\u003E\u003C\u002Fa\u003E. If you choose to remove targeted or advertising cookies, you will still see adverts but they may not be relevant to you. Even if you do choose to remove cookies by the companies listed at the above link, not all companies that serve online behavioral advertising are included in this list, and so you may still receive some cookies and tailored adverts from companies that are not listed.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cem\u003ESocial Media Cookies\u003C\u002Fem\u003E\u003C\u002Fp\u003E\u003Cp\u003EThese cookies are used when you share information using a social media sharing button or “like” button on our website or you link your account or engage with our content on or through a social networking website such as Facebook, Twitter or Google+. The social network will record that you have done this.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cstrong\u003EDisabling cookies\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003EYou can typically remove or reject cookies via your browser settings. In order to do this, follow the instructions provided by your browser (usually located within the “settings,” “help” “tools” or “edit” facility). Many browsers are set to accept cookies until you change your settings.\u003C\u002Fp\u003E\u003Cp\u003EIf you do not accept our cookies, you may experience some inconvenience in your use of our website. For example, we may not be able to recognize your computer or mobile device and you may need to log in every time you visit our website.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cstrong\u003EAdvertising\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003EWe may use other companies to serve third-party advertisements when you visit and use the website. These companies may collect and use click stream information, browser type, time and date, subject of advertisements clicked or scrolled over during your visits to the website and other websites in order to provide advertisements about goods and services likely to be of interest to you. These companies typically use tracking technologies to collect this information. Other companies' use of their tracking technologies is subject to their own privacy policies.\u003C\u002Fp\u003E\u003Cul\u003E\u003Cli\u003E\u003Cstrong\u003EUsing Your Personal Data\u003C\u002Fstrong\u003E\u003C\u002Fli\u003E\u003C\u002Ful\u003E\u003Cp\u003E\u003Cbr\u003E\u003C\u002Fp\u003E\u003Cp\u003EWe may use your personal data as follows:\u003C\u002Fp\u003E\u003Cul\u003E\u003Cli\u003Eto operate, maintain, and improve our website, products, and services;\u003C\u002Fli\u003E\u003Cli\u003Eto manage your account, including to communicate with you regarding your account, if you have an account on our website;\u003C\u002Fli\u003E\u003Cli\u003Eto operate and administer our rewards program and other promotions you participate in on our website;\u003C\u002Fli\u003E\u003Cli\u003Eto respond to your comments and questions and to provide customer service;\u003C\u002Fli\u003E\u003Cli\u003Eto send information including technical notices, updates, security alerts, and support and administrative messages;\u003C\u002Fli\u003E\u003Cli\u003Ewith your consent, to send you marketing e-mails about upcoming promotions, and other news, including information about products and services offered by us and our affiliates. You may opt-out of receiving such information at any time: such marketing emails tell you how to “opt-out.” Please note, even if you opt out of receiving marketing emails, we may still send you non-marketing emails. Non-marketing emails include emails about your account with us (if you have one) and our business dealings with you;\u003C\u002Fli\u003E\u003Cli\u003Eto process payments you make via our website;\u003C\u002Fli\u003E\u003Cli\u003Eas we believe necessary or appropriate (a) to comply with applicable laws; (b) to comply with lawful requests and legal process, including to respond to requests from public and government authorities; (c) to enforce our Policy; and (d) to protect our rights, privacy, safety or property, and\u002For that of you or others;\u003C\u002Fli\u003E\u003Cli\u003Efor analysis and study services; and\u003C\u002Fli\u003E\u003Cli\u003Eas described in the “Sharing of your Personal Data” section below.\u003C\u002Fli\u003E\u003C\u002Ful\u003E\u003Cp\u003E\u003Cstrong\u003ESharing Your Personal Data\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003EWe may share your personal data as follows:\u003C\u002Fp\u003E\u003Cul\u003E\u003Cli\u003E\u003Cstrong\u003EThird Parties Designated by You.\u003C\u002Fstrong\u003E We may share your personal data with third parties where you have provided your consent to do so.\u003C\u002Fli\u003E\u003Cli\u003E\u003Cstrong\u003EOur\u003C\u002Fstrong\u003E \u003Cstrong\u003EThird Party Service Providers\u003C\u002Fstrong\u003E. We may share your personal data with our third party service providers who provide services such as data analysis, payment processing, information technology and related infrastructure provision, customer service, email delivery, auditing and other similar services.\u003C\u002Fli\u003E\u003Cli\u003E\u003Cstrong\u003EThird Party Sites\u003C\u002Fstrong\u003E\u003C\u002Fli\u003E\u003C\u002Ful\u003E\u003Cp\u003EOur website may contain links to third party websites and features.\u003Cstrong\u003E \u003C\u002Fstrong\u003EThis Policy does not cover the privacy practices of such third parties.\u003Cstrong\u003E \u003C\u002Fstrong\u003EThese third parties have their own privacy policies and we do not accept any responsibility or liability for their websites, features or policies. Please read their privacy policies before you submit any data to them.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cstrong\u003EUser Generated Content\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003EYou may share personal data with us when you submit user generated content to our website, including via our rewards program, forums, message boards and blogs on our website. Please note that any information you post or disclose on our website will become public information, and will be available to other users of our website and to the general public. We urge you to be very careful when deciding to disclose your personal data, or any other information, on our website. Such personal data and other information will not be private or confidential once it is published on our website.\u003C\u002Fp\u003E\u003Cp\u003EIf you provide feedback to us, we may use and disclose such feedback on our website, provided we do not associate such feedback with your personal data. If you have provided your consent to do so, we may post your first and last name along with your feedback on our website. We will collect any information contained in such feedback and will treat the personal data in it in accordance with this Policy.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cstrong\u003EInternational Data Transfer\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003EYour information, including personal data that we collect from you, may be transferred to, stored at and processed by us outside the country in which you reside, where data protection and privacy regulations may not offer the same level of protection as in other parts of the world. By accepting this Policy, you agree to this transfer, storing or processing. We will take all steps reasonably necessary to ensure that your data is treated securely and in accordance with this Policy.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cstrong\u003ESecurity\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003EWe seek to use reasonable organizational, technical and administrative measures to protect personal data within our organization. Unfortunately, no transmission or storage system can be guaranteed to be completely secure, and transmission of information via the Internet is not completely secure. If you have reason to believe that your interaction with us is no longer secure (for example, if you feel that the security of any account you might have with us has been compromised), please immediately notify us of the problem by contacting us.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cstrong\u003ERetention\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003EWe will only retain your personal data as long reasonably required for you to use the website until you close your account\u002Fcancel your subscription unless a longer retention period is required or permitted by law (for example for regulatory purposes).\u003C\u002Fp\u003E\u003Cp\u003E\u003Cstrong\u003EOur Policy on Children\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003EOur website is\u002Fare not directed to children under 16.\u003Cstrong\u003E \u003C\u002Fstrong\u003EIf a parent or guardian becomes aware that his or her child has provided us with information without their consent, he or she should contact us. We will delete such information from our files as soon as reasonably practicable.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cstrong\u003EYour Rights\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u003E\u003C\u002Fp\u003E\u003Cul\u003E\u003Cli\u003E\u003Cstrong\u003EOpt-out. \u003C\u002Fstrong\u003EYou may contact us anytime to opt-out of: (i) direct marketing communications; (ii) automated decision-making and\u002For profiling; (iii) our collection of sensitive personal data; (iv) any new processing of your personal data that we may carry out beyond the original purpose; or (v) the transfer of your personal data outside the EEA. Please note that your use of some of the website may be ineffective upon opt-out.\u003C\u002Fli\u003E\u003Cli\u003E\u003Cstrong\u003EAccess. \u003C\u002Fstrong\u003EYou may access the information we hold about you at any time via your profile\u002Faccount or by contacting us directly.\u003C\u002Fli\u003E\u003Cli\u003E\u003Cstrong\u003EAmend. \u003C\u002Fstrong\u003EYou can also contact us to update or correct any inaccuracies in your personal data.\u003C\u002Fli\u003E\u003Cli\u003E\u003Cstrong\u003EMove. \u003C\u002Fstrong\u003EYour personal data is portable – i.e. you to have the flexibility to move your data to other service providers as you wish.\u003C\u002Fli\u003E\u003Cli\u003E\u003Cstrong\u003EErase and forget. \u003C\u002Fstrong\u003EIn certain situations, for example when the information we hold about you is no longer relevant or is incorrect, you can request that we erase your data.\u003C\u002Fli\u003E\u003C\u002Ful\u003E\u003Cp\u003EIf you wish to exercise any of these rights, please contact us. In your request, please make clear: (i) \u003Cstrong\u003Ewhat\u003C\u002Fstrong\u003E personal data is concerned; and (ii) \u003Cstrong\u003Ewhich of the above rights\u003C\u002Fstrong\u003E you would like to enforce. For your protection, we may only implement requests with respect to the personal data associated with the particular email address that you use to send us your request, and we may need to verify your identity before implementing your request. We will try to comply with your request as soon as reasonably practicable and in any event, within one month of your request. Please note that we may need to retain certain information for recordkeeping purposes and\u002For to complete any transactions that you began prior to requesting such change or deletion.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cstrong\u003EComplaints\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003EWe are committed to resolve any complaints about our collection or use of your personal data. If you would like to make a complaint regarding this Policy or our practices in relation to your personal data, please contact us through the information listed on our website. We will reply to your complaint as soon as we can and in any event, within 30 days. We hope to resolve any complaint brought to our attention, however if you feel that your complaint has not been adequately resolved, you reserve the right to contact your local data protection supervisory authority\u003C\u002Fp\u003E\u003Cp\u003E\u003Cstrong\u003EContact Information\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003EWe welcome your comments or questions about this Policy. You may contact us in writing or through our website.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u003E\u003C\u002Fp\u003E",Terms:"\u003Cp\u003E\u003Cstrong\u003ETerms of Use\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003EEffective as of May 9, 2020.\u003C\u002Fp\u003E\u003Cp\u003EWelcome to the Self-employment (the \"Service\"). The following Terms of Use apply when you view or use the Service located at: https:\u002F\u002Fblogweb.me. Please review the following terms carefully. By accessing or using the Service, you signify your agreement to these Terms of Use. If you do not agree to these Terms of Use, you may not access or use the Service.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cstrong\u003EPRIVACY POLICY\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003EThe company respects the privacy of its Service users. Please refer to the Company's Privacy Policy which explains how we collect, use, and disclose information that pertains to your privacy. When you access or use the Service, you signify your agreement to this Privacy Policy.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cstrong\u003EREGISTRATION; RULES FOR USER CONDUCT AND USE OF THE SERVICE\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003EYou need to be at least 16 years old to register for and use the Service.\u003C\u002Fp\u003E\u003Cp\u003EIf you are a user who signs up for the Service, the company will create a personalized account, which includes a unique username and a password to access the Service and allow you to receive messages from the Company. You agree to notify us immediately of any unauthorized use of your password and\u002For account. The Company will not be responsible for any liabilities, losses, or damages arising out of the unauthorized use of your member name, password and\u002For account.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cstrong\u003EUSE RESTRICTIONS.\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003EYour permission to use the Site is conditioned upon the following Use Restrictions and Conduct Restrictions: You agree that you will not under any circumstances:\u003C\u002Fp\u003E\u003Cul\u003E\u003Cli\u003Epost any information that is abusive, threatening, obscene, defamatory, libelous, or racially, sexually, religiously, or otherwise objectionable and offensive;\u003C\u002Fli\u003E\u003Cli\u003Euse the service for any unlawful purpose or for the promotion of illegal activities;\u003C\u002Fli\u003E\u003Cli\u003Eattempt to, or harass, abuse or harm another person or group;\u003C\u002Fli\u003E\u003Cli\u003Euse another user's account without permission;\u003C\u002Fli\u003E\u003Cli\u003Eprovide false or inaccurate information when registering an account;\u003C\u002Fli\u003E\u003Cli\u003Einterfere or attempt to interfere with the proper functioning of the Service;\u003C\u002Fli\u003E\u003Cli\u003Emake any automated use of the system, or take any action that we deem to impose or to potentially impose an unreasonable or disproportionately large load on our servers or network infrastructure;\u003C\u002Fli\u003E\u003Cli\u003Ebypass any robot exclusion headers or other measures we take to restrict access to the Service or use any software, technology, or device to scrape, spider, or crawl the Service or harvest or manipulate data; or\u003C\u002Fli\u003E\u003Cli\u003Epublish or link to malicious content intended to damage or disrupt another user's browser or computer.\u003C\u002Fli\u003E\u003C\u002Ful\u003E\u003Cp\u003E\u003Cstrong\u003EPOSTING AND CONDUCT RESTRICTIONS.\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003EWhen you create your own personalized account, you may be able to provide (\"User Content\"). You are solely responsible for the User Content that you post, upload, link to or otherwise make available via the Service. You agree that we are only acting as a passive conduit for your online distribution and publication of your User Content. The Company, however, reserves the right to remove any User Content from the Service at its discretion.\u003C\u002Fp\u003E\u003Cp\u003EThe following rules pertain to User Content. By transmitting and submitting any User Content while using the Service, you agree as follows:\u003C\u002Fp\u003E\u003Cul\u003E\u003Cli\u003EYou are solely responsible for your account and the activity that occurs while signed in to or while using your account;\u003C\u002Fli\u003E\u003Cli\u003EYou will not post information that is malicious, false or inaccurate;\u003C\u002Fli\u003E\u003Cli\u003EYou will not submit content that is copyrighted or subject to third party proprietary rights, including privacy, publicity, trade secret, etc., unless you are the owner of such rights or have the appropriate permission from their rightful owner to specifically submit such content; and\u003C\u002Fli\u003E\u003Cli\u003EYou hereby affirm we have the right to determine whether any of your User Content submissions are appropriate and comply with these Terms of Service, remove any and\u002For all of your submissions, and terminate your account with or without prior notice.\u003C\u002Fli\u003E\u003C\u002Ful\u003E\u003Cp\u003EYou understand and agree that any liability, loss or damage that occurs as a result of the use of any User Content that you make available or access through your use of the Service is solely your responsibility. The Company is not responsible for any public display or misuse of your User Content. The Company does not, and cannot, pre-screen or monitor all User Content. However, at our discretion, we, or the technology we employ, may monitor and\u002For record your interactions with the Service.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cstrong\u003EONLINE CONTENT DISCLAIMER\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003EOpinions, advice, statements, offers, or other information or content made available through the Service, but not directly by the Company, are those of their respective authors, and should not necessarily be relied upon. Such authors are solely responsible for such content. The Company does not guarantee the accuracy, completeness, or usefulness of any information on the Service and neither does the Company adopt nor endorse, nor is the Company responsible for the accuracy or reliability of any opinion, advice, or statement made by parties other than the Company. The Company takes no responsibility and assumes no liability for any User Content that you or any other user or third party posts or sends over the Service. Under no circumstances will the Company be responsible for any loss or damage resulting from anyone's reliance on information or other content posted on the Service, or transmitted to users.\u003C\u002Fp\u003E\u003Cp\u003EThough the Company strives to enforce these Terms of Use, you may be exposed to User Content that is inaccurate or objectionable. The Company reserves the right, but has no obligation, to monitor the materials posted in the public areas of the service or to limit or deny a user's access to the Service or take other appropriate action if a user violates these Terms of Use or engages in any activity that violates the rights of any person or entity or which we deem unlawful, offensive, abusive, harmful or malicious. The Company shall have the right to remove any such material that in its sole opinion violates, or is alleged to violate, the law or this agreement or which might be offensive, or that might violate the rights, harm, or threaten the safety of users or others. Unauthorized use may result in criminal and\u002For civil prosecution under the law. If you become aware of misuse of our Service, please contact us at https:\u002F\u002Fblogweb.me.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cstrong\u003ELINKS TO OTHER SITES AND\u002FOR MATERIALS\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003EAs part of the Service, the Company may provide you with convenient links to third party web site(s) (\"Third Party Sites\") as well as content or items belonging to or originating from third parties (the\"Third Party Applications, Software or Content\"). These links are provided as a courtesy to Service subscribers. The Company has no control over Third Party Sites and Third Party Applications, Software or Content or the promotions, materials, information, goods or services available on these Third Party Sites or Third Party Applications, Software or Content. Such Third Party Sites and Third Party Applications, Software or Content are not investigated, monitored or checked for accuracy, appropriateness, or completeness by the Company, and the Company is not responsible for any Third Party Sites accessed through the Site or any Third Party Applications, Software or Content posted on, available through or installed from the Site, including the content, accuracy, offensiveness, opinions, reliability, privacy practices or other policies of or contained in the Third Party Sites or the Third Party Applications, Software or Content. Inclusion of, linking to or permitting the use or installation of any Third Party Site or any Third Party Applications, Software or Content does not imply approval or endorsement thereof by the Company. If you decide to leave the Site and access the Third Party Sites or to use or install any Third Party Applications, Software or Content, you do so at your own risk and you should be aware that our terms and policies no longer govern. You should review the applicable terms and policies, including privacy and data gathering practices, of any site to which you navigate from the Site or relating to any applications you use or install from the site.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cstrong\u003ECOPYRIGHT COMPLAINTS AND COPYRIGHT AGENT\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003E(a) Termination of Repeat Infringe Accounts. The Company respects the intellectual property rights of others and requests that the users do the same. The Company has adopted and implemented a policy that provides for the termination in appropriate circumstances of users of the Service who are repeat infringers The Company may terminate access for participants or users who are found repeatedly to provide or post protected third party content without necessary rights and permissions.\u003C\u002Fp\u003E\u003Cp\u003E(b) Take-Down Notices. If you are a copyright owner or an agent thereof and believe, in good faith, that any materials provided on the Service infringe upon your copyrights, you may submit a notification pursuant by sending the following information in writing to the Company's designated copyright agent at Self-employment:\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003EThe date of your notification;\u003C\u002Fli\u003E\u003Cli\u003EA Physical or electronic signature of a person authorized to act on behalf of the owner of an exclusive right that is allegedly infringed;\u003C\u002Fli\u003E\u003Cli\u003EA description of the copyrighted work claimed to have been infringed, or, if multiple copyrighted works at a single online site are recovered by a single notification, a representative list of such works at that site;\u003C\u002Fli\u003E\u003Cli\u003EA description of the material that is claimed to be infringing or to be the subject of infringing activity and information sufficient to enable us to locate such work;\u003C\u002Fli\u003E\u003Cli\u003EInformation reasonably sufficient to permit the service provider to contact you, such as an address, telephone number, and\u002For email address;\u003C\u002Fli\u003E\u003Cli\u003EA statement that you have a good faith belief that use of the material in the manner complained of is not authorized by the copyright owner, its agent, or the law; and\u003C\u002Fli\u003E\u003Cli\u003EA statement that the information in the notification is accurate, and under penalty of perjury, that you are authorized to act on behalf of the owner of an exclusive right that is allegedly infringed.\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cp\u003E(c) Counter-Notices. If you believe that your User Content that has been removed from the Site is not infringing, or that you have the authorization from the copyright owner, the copyright owner's agent, or pursuant to the law, to post and use the content in your User Content, you may send a counter-notice containing the following information to our copyright agent using the contact information set forth above:\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003EYour physical or electronic signature;\u003C\u002Fli\u003E\u003Cli\u003EA description of the content that has been removed and the location at which the content appeared before it was removed;\u003C\u002Fli\u003E\u003Cli\u003EA statement that you have a good faith belief that the content was removed as a result of mistake or a misidentification of the content; and\u003C\u002Fli\u003E\u003Cli\u003EYour name, address, telephone number, and email address, a statement that you consent to the laws of California and a statement that you will accept service of process from the person who provided notification of the alleged infringement.\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cp\u003EIf a counter-notice is received by the Company copyright agent, the Company may send a copy of the counter-notice to the original complaining party informing such person that it may reinstate the removed content in 10 business days. Unless the copyright owner files an action seeking a court order against the content provider, member or user, the removed content may (in the Company's discretion) be reinstated on the Site in 10 to 14 business days or more after receipt of the counter-notice.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cstrong\u003ELICENSE GRANT\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003EBy posting any User Content via the Service, you expressly grant, and you represent and warrant that you have a right to grant, to the Company a royalty-free, sub licensable, transferable, perpetual, irrevocable, non-exclusive, worldwide license to use, reproduce, modify, publish, list information regarding, edit, translate, distribute, publicly perform, publicly display, and make derivative works of all such User Content and your name, voice, and\u002For likeness as contained in your User Content, if applicable, in whole or impart, and in any form, media or technology, whether now known or hereafter developed, for use in connection with the Service.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cstrong\u003EINTELLECTUAL PROPERTY\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003EYou acknowledge and agree that we and our licensors retain ownership of all intellectual property rights of any kind related to the Service, including applicable copyrights, trademarks and other proprietary rights. Other product and business names that are mentioned on the Service may be trademarks of their respective owners. We reserve all rights that are not expressly granted to you under this Agreement.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cstrong\u003EEMAIL MAY NOT BE USED TO PROVIDE NOTICE\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003ECommunications made through the Service's e-mail and messaging system, will not constitute legal notice to the Company or any of its officers, employees, agents or representatives in any situation where notice to the Company is required by contract or any law or regulation.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cstrong\u003EUSER CONSENT TO RECEIVE COMMUNICATIONS IN ELECTRONIC FORM\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003EFor contractual purposes, you (a) consent to receive communications from the Company in an electronic form via the email address you have submitted; and (b) agree that all Terms of Use, agreements, notices, disclosures, and other communications that the Company provides to you electronically satisfy any legal requirement that such communications would satisfy if it were in writing. The foregoing does not affect your non-waivable rights.\u003C\u002Fp\u003E\u003Cp\u003EWe may also use your email address, to send you other messages, including information about the Company and special offers. You may opt out of such email by changing your account settings or sending an email to Self-employment.\u003C\u002Fp\u003E\u003Cp\u003EOpting out may prevent you from receiving messages regarding the Company or Special Offers.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cstrong\u003EWARRANTY\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003ETHE SERVICE, IS PROVIDED \"AS IS,\" WITHOUT WARRANTY OF ANY KIND. WITHOUT LIMITING THE FOREGOING, THE COMPANY EXPRESSLY DISCLAIMS ALL WARRANTIES, WHETHER EXPRESS, IMPLIED OR STATUTORY, REGARDING THE SERVICE INCLUDING WITHOUT LIMITATION ANY WARRANTY OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, TITLE, SECURITY, ACCURACY AND NON-INFRINGEMENT. WITHOUT LIMITING THE FOREGOING, THE COMPANY MAKES NO WARRANTY OR REPRESENTATION THAT ACCESS TO OR OPERATION OF THE SERVICE WILL BE UNINTERRUPTED OR ERROR FREE. YOU ASSUME FULL RESPONSIBILITY AND RISK OF LOSS RESULTING FROM YOUR DOWNLOADING AND\u002FOR USE OF FILES, INFORMATION, CONTENT OR OTHER MATERIAL OBTAINED FROM THE SERVICE. SOME JURISDICTIONS LIMIT OR DO NOT PERMIT DISCLAIMERS OF WARRANTY, SO THIS PROVISION MAY NOT APPLY TO YOU.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cstrong\u003ELIMITATION OF DAMAGES; RELEASE\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003ETO THE EXTENT PERMITTED BY APPLICABLE LAW, IN NO EVENT SHALL THE COMPANY, ITS AFFILIATES, DIRECTORS, OR EMPLOYEES, OR ITS LICENSORS OR PARTNERS, BE LIABLE TO YOU FOR ANY LOSS OF PROFITS, USE, OR DATA, OR FOR ANY INCIDENTAL, INDIRECT, SPECIAL, CONSEQUENTIAL OR EXEMPLARY DAMAGES, HOWEVER ARISING, THAT RESULT FROM (A) THE USE, DISCLOSURE, OR DISPLAY OF YOUR USER CONTENT; (B) YOUR USE OR INABILITY TO USE THE SERVICE; (C) THE SERVICE GENERALLY OR THE SOFTWARE OR SYSTEMS THAT MAKE THE SERVICE AVAILABLE; OR (D) ANY OTHER INTERACTIONS WITH THE COMPANY OR ANY OTHER USER OF THE SERVICE, WHETHER BASED ON WARRANTY, CONTRACT, TORT (INCLUDING NEGLIGENCE) OR ANY OTHER LEGAL THEORY, AND WHETHER OR NOT THE COMPANY HAS BEEN INFORMED OF THE POSSIBILITY OF SUCH DAMAGE, AND EVEN IF A REMEDY SET FORTH HEREIN IS FOUND TO HAVE FAILED OF ITS ESSENTIAL PURPOSE. SOME JURISDICTIONS LIMIT OR DO NOT PERMIT DISCLAIMERS OF LIABILITY, SO THIS PROVISION MAY NOT APPLY TO YOU.\u003C\u002Fp\u003E\u003Cp\u003EIf you have a dispute with one or more users or a merchant of a product or service that you review using the Service, you release us (and our officers, directors, agents, subsidiaries, joint ventures and employees) from claims, demands and damages (actual and consequential) of every kind and nature, known and unknown, arising out of or in any way connected with such disputes.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cstrong\u003EMODIFICATION OF TERMS OF USE\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003EWe can amend these Terms of Use at any time and will update these Terms of Use in the event of any such amendments. It is your sole responsibility to check the Site from time to time to view any such changes in the Agreement. If you continue to use the Site, you signify your agreement to our revisions to these Terms of Use. However, we will notify you of material changes to the terms by posting a notice on our homepage and\u002For sending an email to the email address you provided to us upon registration. For this additional reason, you should keep your contact and profile information current. Any changes to these Terms or waiver of the Company's rights hereunder shall not be valid or effective except in a written agreement bearing the physical signature of an officer of the Company. No purported waiver or modification of this Agreement by the Company via telephonic or email communications shall be valid.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cstrong\u003EGENERAL TERMS\u003C\u002Fstrong\u003E\u003C\u002Fp\u003E\u003Cp\u003EIf any part of this Agreement is held invalid or unenforceable, that portion of the Agreement will be construed consistent with applicable law. The remaining portions will remain in full force and effect. Any failure on the part of the Company to enforce any provision of this Agreement will not be considered a waiver of our right to enforce such provision. Our rights under this Agreement will survive any termination of this Agreement.\u003C\u002Fp\u003E\u003Cp\u003EYou agree that any cause of action related to or arising out of your relationship with the Company must commence within ONE year after the cause of action accrues. Otherwise, such cause of action is permanently barred.\u003C\u002Fp\u003E\u003Cp\u003EThese Terms of Use and your use of the Site are governed by the laws of California, without regard to conflict of law provisions.\u003C\u002Fp\u003E\u003Cp\u003EThe Company may assign or delegate these Terms of Service and\u002For the Company's Privacy Policy, in whole or in part, to any person or entity at any time with or without your consent. You may not assign or delegate any rights or obligations under the Terms of Service or Privacy Policy without the Company's prior written consent, and any unauthorized assignment and delegation by you is void.\u003C\u002Fp\u003E\u003Cp\u003EYOU ACKNOWLEDGE THAT YOU HAVE READ THESE TERMS OF USE, UNDERSTAND THE TERMS OF USE, AND WILL BE BOUND BY THESE TERMS AND CONDITIONS. YOU FURTHER ACKNOWLEDGE THAT THESE TERMS OF USE TOGETHER WITH THE PRIVACY POLICY AT https:\u002F\u002Fblogweb.me REPRESENT THE COMPLETE AND EXCLUSIVE STATEMENT OF THE AGREEMENT BETWEEN US AND THAT IT SUPERSEDES ANY PROPOSAL OR PRIOR AGREEMENT ORAL OR WRITTEN, AND ANY OTHER COMMUNICATIONS BETWEEN US RELATING TO THE SUBJECT MATTER OF THIS AGREEMENT.\u003C\u002Fp\u003E\u003Cp class=\"ql-align-right\"\u003E\u003Cbr\u003E\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u003E\u003C\u002Fp\u003E",Domain:aN,Plan:e,PlanExpired:"2100-01-01T00:00:00Z",Port:aO,Active:e,Rating:b,CountVoted:b,Trusted:c,CreatedIp:a,Subject:{Id:44,Name:a,Slug:a,Icon:a,MetaTitle:a,MetaDescription:a,Locale:g,Site:g,Created:f,Updated:f},Settings:{Id:h,Title:k,Logo:aP,Locale:aQ,RobotsTxt:aR,FooterCode:aS,Description:y,Activation:aT,ScrollablePagination:b,AddWatermark:b,AddWatermarkPosition:b,LayoutSettings:{Id:h,Name:a,IsDark:b,BackgroundFull:b,PageTransition:a,CodeTheme:a,Background:a,BackgroundColor:a,TextColor:a,TextFontFamily:a,PrimaryColor:a,SecondaryColor:a,AccentColor:a,InfoColor:a,SuccessColor:a,ErrorColor:a,WarningColor:a,Created:f,Updated:f},ForumSettings:g,BlogSettings:{Id:h,Toc:b,TocCollapse:b,AddSource:b,AddSourceText:a,IsRelatedPost:b,RelatedPost:b,Created:f,Updated:f},MailSettings:{Id:33,Host:a,Email:a,FromName:a,User:a,Password:a,Encryption:a,Port:b,Created:f,Updated:f},SocialSettings:g,SecuritySettings:{Id:h,ThreadLimit:b,ThreadLimitType:b,RegisterLimit:b,RegisterLimitType:b,PostLimit:b,CommentLimitType:b,CommentLimit:b,PostLimitType:b,MessagesBeforeAutoApproved:b,MarkUncertainMessages:c,SecurityQuestions:g,Created:f,Updated:f},Created:aU,Updated:"2023-07-07T06:16:04Z"},User:{Id:499,Username:a,FirstName:a,Avatar:a,LastName:a,Company:a,Email:a,ConfirmationToken:a,CreatedIp:a,RestoreToken:a,PasswordRequestedAt:f,Password:a,Active:b,Trusted:c,Banned:b,Notifications:b,Role:g,Site:g,LastLogin:f,Created:f,Updated:f},Category:g,Created:aU,Updated:"2023-07-06T22:21:41Z"},title:Z,summary:"Batch normalization is a widely used technique for improving the training of deep neural networks. It normalizes the activations of each mini-batch by subtracting the mini-batch mean and dividing by the mini-batch standard deviation. This helps in reducing internal covariate shift by ensuring that the input to each layer is normalized.Implementing batch normalization in PyTorch is straightforward. Here are the steps:Import the necessary libraries:\nimport torch\nimport torch.",content:"\u003Cp\u003EBatch normalization is a widely used technique for \u003Ca class=\"auto-link\" href=\"https:\u002F\u002Fstlplaces.com\u002Fblog\u002Fhow-to-handle-overfitting-in-tensorflow-models\"\u003Eimproving the training of deep neural networks\u003C\u002Fa\u003E. It normalizes the activations of each mini-batch by subtracting the mini-batch mean and dividing by the mini-batch standard deviation. This helps in reducing internal covariate shift by ensuring that the input to each layer is normalized.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EImplementing batch normalization in PyTorch is straightforward. Here are the steps:\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003EImport the necessary libraries:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cpre class=\"code-block ql-syntax\" id=\"code-q2hyroj\"\u003Eimport torch\nimport torch.nn as nn\n\u003C\u002Fpre\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003EDefine a custom neural network architecture:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cpre class=\"code-block ql-syntax\" id=\"code-tz45vgq\"\u003Eclass Net(nn.Module):\n def __init__(self):\n super(Net, self).__init__()\n self.fc1 = nn.Linear(10, 20)\n self.bn1 = nn.BatchNorm1d(20) # Batch normalization layer\n self.fc2 = nn.Linear(20, 10)\n\u003C\u002Fpre\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003EOverride the forward method of the neural network:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cpre class=\"code-block ql-syntax\" id=\"code-ig8hxn3\"\u003E def forward(self, x):\n x = self.fc1(x)\n x = self.bn1(x)\n x = torch.relu(x)\n x = self.fc2(x)\n return x\n\u003C\u002Fpre\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003ECreate an instance of the network:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cpre class=\"code-block ql-syntax\" id=\"code-r682yo6\"\u003Enet = Net()\n\u003C\u002Fpre\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EThat's it! Now the network \u003Ccode\u003Enet\u003C\u002Fcode\u003E includes a batch normalization layer (\u003Ccode\u003Eself.bn1\u003C\u002Fcode\u003E) after the first \u003Ca href=\"https:\u002F\u002Ftopminisite.com\u002Fblog\u002Fhow-to-restore-in-fully-connected-layer-using\" class=\"auto-link\" target=\"_blank\"\u003Efully connected layer\u003C\u002Fa\u003E (\u003Ccode\u003Eself.fc1\u003C\u002Fcode\u003E). During training, as the mini-batches pass through this network, the batch normalization layer will normalize the activations.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003ENote: It is essential to ensure that the network is in training mode using \u003Ccode\u003Enet.train()\u003C\u002Fcode\u003E before training and in evaluation mode using \u003Ccode\u003Enet.eval()\u003C\u002Fcode\u003E during inference\u002Ftesting.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EYou can now use this network for training and inference in your PyTorch project, while enjoying the benefits of batch normalization.\u003C\u002Fp\u003E\n \u003Cdiv class=\"rating\"\u003E\n \u003Ch2\u003EBest PyTorch Books of December 2024\u003C\u002Fh2\u003E\n \u003Cdiv class=\"row mt-2\"\u003E\n \n \u003Cdiv class=\"col-12\"\u003E\n \u003Cdiv class=\"v-card elevation-6\"\u003E\n \u003Cdiv class=\"v-card__text rating-text\"\u003E\n \u003Cdiv class=\"rating-counter\"\u003E\n \u003Cspan class=\"v-badge\"\u003E\n \u003Cspan class=\"v-badge__wrapper\"\u003E\n \u003Cspan aria-atomic=\"true\" aria-label=\"Позиция\" class=\"v-badge__badge primary\"\u003E\n 1\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"row\"\u003E\n \u003Cdiv class=\"col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center\"\u003E\n \u003Cdiv\u003E\n \u003Cimg src=\"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com\u002Fimages\u002Fd7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae\u002Frating\u002F41jkoc6owal-sl160.jpg\" alt=\"PyTorch Recipes: A Problem-Solution Approach to Build, Train and Deploy Neural Network Models\" \u002F\u003E\n \u003Cp class=\"text-center font-weight-bold text-h6\"\u003ERating is 5 out of 5\u003C\u002Fp\u003E\n \u003Cdiv class=\"stars\" style=\"--rating: 5;\" aria-label=\"Rating is 5 out of 5\" \u003E\u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"col-lg-6 col-md-8 col-sm-6 col-12\"\u003E\n \u003Cp class=\"font-weight-bold rating-name\"\u003EPyTorch Recipes: A Problem-Solution Approach to Build, Train and Deploy Neural Network Models\u003C\u002Fp\u003E\n \n \n\n \n \n \u003C\u002Fdiv\u003E\n\n \u003Cdiv class=\"col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center\"\u003E\n \u003Cdiv class=\"text-center d-flex flex-column\"\u003E\n \n \u003Ca href=\"https:\u002F\u002Fgosrc.cc\u002Fgo\u002FMDdl-1Y4g\" target=\"_blank\" rel=\"nofollow noopener\" class=\"v-btn v-btn--rounded elevation-5 v-size--large success mb-2\"\u003E\n \u003Cspan class=\"v-btn__content\"\u003EGet Book Now\u003C\u002Fspan\u003E\n \u003C\u002Fa\u003E\n \n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \n \u003Cdiv class=\"col-12\"\u003E\n \u003Cdiv class=\"v-card elevation-6\"\u003E\n \u003Cdiv class=\"v-card__text rating-text\"\u003E\n \u003Cdiv class=\"rating-counter\"\u003E\n \u003Cspan class=\"v-badge\"\u003E\n \u003Cspan class=\"v-badge__wrapper\"\u003E\n \u003Cspan aria-atomic=\"true\" aria-label=\"Позиция\" class=\"v-badge__badge primary\"\u003E\n 2\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"row\"\u003E\n \u003Cdiv class=\"col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center\"\u003E\n \u003Cdiv\u003E\n \u003Cimg src=\"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com\u002Fimages\u002Fd7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae\u002Frating\u002F4198ez8dpbl-sl160.jpg\" alt=\"Mastering PyTorch: Build powerful deep learning architectures using advanced PyTorch features, 2nd Edition\" \u002F\u003E\n \u003Cp class=\"text-center font-weight-bold text-h6\"\u003ERating is 4.9 out of 5\u003C\u002Fp\u003E\n \u003Cdiv class=\"stars\" style=\"--rating: 4.9;\" aria-label=\"Rating is 4.9 out of 5\" \u003E\u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"col-lg-6 col-md-8 col-sm-6 col-12\"\u003E\n \u003Cp class=\"font-weight-bold rating-name\"\u003EMastering PyTorch: Build powerful deep learning architectures using advanced PyTorch features, 2nd Edition\u003C\u002Fp\u003E\n \n \n\n \n \n \u003C\u002Fdiv\u003E\n\n \u003Cdiv class=\"col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center\"\u003E\n \u003Cdiv class=\"text-center d-flex flex-column\"\u003E\n \n \u003Ca href=\"https:\u002F\u002Fgosrc.cc\u002Fgo\u002FUZp_a1L4R\" target=\"_blank\" rel=\"nofollow noopener\" class=\"v-btn v-btn--rounded elevation-5 v-size--large success mb-2\"\u003E\n \u003Cspan class=\"v-btn__content\"\u003EGet Book Now\u003C\u002Fspan\u003E\n \u003C\u002Fa\u003E\n \n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \n \u003Cdiv class=\"col-12\"\u003E\n \u003Cdiv class=\"v-card elevation-6\"\u003E\n \u003Cdiv class=\"v-card__text rating-text\"\u003E\n \u003Cdiv class=\"rating-counter\"\u003E\n \u003Cspan class=\"v-badge\"\u003E\n \u003Cspan class=\"v-badge__wrapper\"\u003E\n \u003Cspan aria-atomic=\"true\" aria-label=\"Позиция\" class=\"v-badge__badge primary\"\u003E\n 3\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"row\"\u003E\n \u003Cdiv class=\"col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center\"\u003E\n \u003Cdiv\u003E\n \u003Cimg src=\"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com\u002Fimages\u002Fd7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae\u002Frating\u002F513exc7qqjl-sl160.jpg\" alt=\"Natural Language Processing with PyTorch: Build Intelligent Language Applications Using Deep Learning\" \u002F\u003E\n \u003Cp class=\"text-center font-weight-bold text-h6\"\u003ERating is 4.8 out of 5\u003C\u002Fp\u003E\n \u003Cdiv class=\"stars\" style=\"--rating: 4.8;\" aria-label=\"Rating is 4.8 out of 5\" \u003E\u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"col-lg-6 col-md-8 col-sm-6 col-12\"\u003E\n \u003Cp class=\"font-weight-bold rating-name\"\u003ENatural Language Processing with PyTorch: Build Intelligent Language Applications Using Deep Learning\u003C\u002Fp\u003E\n \n \n\n \n \n \u003C\u002Fdiv\u003E\n\n \u003Cdiv class=\"col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center\"\u003E\n \u003Cdiv class=\"text-center d-flex flex-column\"\u003E\n \n \u003Ca href=\"https:\u002F\u002Fgosrc.cc\u002Fgo\u002FAQT_-1L4R\" target=\"_blank\" rel=\"nofollow noopener\" class=\"v-btn v-btn--rounded elevation-5 v-size--large success mb-2\"\u003E\n \u003Cspan class=\"v-btn__content\"\u003EGet Book Now\u003C\u002Fspan\u003E\n \u003C\u002Fa\u003E\n \n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \n \u003Cdiv class=\"col-12\"\u003E\n \u003Cdiv class=\"v-card elevation-6\"\u003E\n \u003Cdiv class=\"v-card__text rating-text\"\u003E\n \u003Cdiv class=\"rating-counter\"\u003E\n \u003Cspan class=\"v-badge\"\u003E\n \u003Cspan class=\"v-badge__wrapper\"\u003E\n \u003Cspan aria-atomic=\"true\" aria-label=\"Позиция\" class=\"v-badge__badge primary\"\u003E\n 4\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"row\"\u003E\n \u003Cdiv class=\"col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center\"\u003E\n \u003Cdiv\u003E\n \u003Cimg src=\"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com\u002Fimages\u002Fd7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae\u002Frating\u002F518jzrflrvl-sl160.jpg\" alt=\"Deep Learning for Coders with Fastai and PyTorch: AI Applications Without a PhD\" \u002F\u003E\n \u003Cp class=\"text-center font-weight-bold text-h6\"\u003ERating is 4.7 out of 5\u003C\u002Fp\u003E\n \u003Cdiv class=\"stars\" style=\"--rating: 4.7;\" aria-label=\"Rating is 4.7 out of 5\" \u003E\u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"col-lg-6 col-md-8 col-sm-6 col-12\"\u003E\n \u003Cp class=\"font-weight-bold rating-name\"\u003EDeep Learning for Coders with Fastai and PyTorch: AI Applications Without a PhD\u003C\u002Fp\u003E\n \n \n\n \n \n \u003C\u002Fdiv\u003E\n\n \u003Cdiv class=\"col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center\"\u003E\n \u003Cdiv class=\"text-center d-flex flex-column\"\u003E\n \n \u003Ca href=\"https:\u002F\u002Fgosrc.cc\u002Fgo\u002F7JalaJYVR\" target=\"_blank\" rel=\"nofollow noopener\" class=\"v-btn v-btn--rounded elevation-5 v-size--large success mb-2\"\u003E\n \u003Cspan class=\"v-btn__content\"\u003EGet Book Now\u003C\u002Fspan\u003E\n \u003C\u002Fa\u003E\n \n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \n \u003Cdiv class=\"col-12\"\u003E\n \u003Cdiv class=\"v-card elevation-6\"\u003E\n \u003Cdiv class=\"v-card__text rating-text\"\u003E\n \u003Cdiv class=\"rating-counter\"\u003E\n \u003Cspan class=\"v-badge\"\u003E\n \u003Cspan class=\"v-badge__wrapper\"\u003E\n \u003Cspan aria-atomic=\"true\" aria-label=\"Позиция\" class=\"v-badge__badge primary\"\u003E\n 5\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"row\"\u003E\n \u003Cdiv class=\"col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center\"\u003E\n \u003Cdiv\u003E\n \u003Cimg src=\"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com\u002Fimages\u002Fd7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae\u002Frating\u002F41o-uy9lzzl-sl160.jpg\" alt=\"Machine Learning with PyTorch and Scikit-Learn: Develop machine learning and deep learning models with Python\" \u002F\u003E\n \u003Cp class=\"text-center font-weight-bold text-h6\"\u003ERating is 4.6 out of 5\u003C\u002Fp\u003E\n \u003Cdiv class=\"stars\" style=\"--rating: 4.6;\" aria-label=\"Rating is 4.6 out of 5\" \u003E\u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"col-lg-6 col-md-8 col-sm-6 col-12\"\u003E\n \u003Cp class=\"font-weight-bold rating-name\"\u003EMachine Learning with PyTorch and Scikit-Learn: Develop machine learning and deep learning models with Python\u003C\u002Fp\u003E\n \n \n\n \n \n \u003C\u002Fdiv\u003E\n\n \u003Cdiv class=\"col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center\"\u003E\n \u003Cdiv class=\"text-center d-flex flex-column\"\u003E\n \n \u003Ca href=\"https:\u002F\u002Fgosrc.cc\u002Fgo\u002FGZs_a1Y4g\" target=\"_blank\" rel=\"nofollow noopener\" class=\"v-btn v-btn--rounded elevation-5 v-size--large success mb-2\"\u003E\n \u003Cspan class=\"v-btn__content\"\u003EGet Book Now\u003C\u002Fspan\u003E\n \u003C\u002Fa\u003E\n \n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \n \u003Cdiv class=\"col-12\"\u003E\n \u003Cdiv class=\"v-card elevation-6\"\u003E\n \u003Cdiv class=\"v-card__text rating-text\"\u003E\n \u003Cdiv class=\"rating-counter\"\u003E\n \u003Cspan class=\"v-badge\"\u003E\n \u003Cspan class=\"v-badge__wrapper\"\u003E\n \u003Cspan aria-atomic=\"true\" aria-label=\"Позиция\" class=\"v-badge__badge primary\"\u003E\n 6\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"row\"\u003E\n \u003Cdiv class=\"col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center\"\u003E\n \u003Cdiv\u003E\n \u003Cimg src=\"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com\u002Fimages\u002Fd7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae\u002Frating\u002F411ruzisg3l-sl160.jpg\" alt=\"Deep Learning with PyTorch: Build, train, and tune neural networks using Python tools\" \u002F\u003E\n \u003Cp class=\"text-center font-weight-bold text-h6\"\u003ERating is 4.5 out of 5\u003C\u002Fp\u003E\n \u003Cdiv class=\"stars\" style=\"--rating: 4.5;\" aria-label=\"Rating is 4.5 out of 5\" \u003E\u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"col-lg-6 col-md-8 col-sm-6 col-12\"\u003E\n \u003Cp class=\"font-weight-bold rating-name\"\u003EDeep Learning with PyTorch: Build, train, and tune neural networks using Python tools\u003C\u002Fp\u003E\n \n \n\n \n \n \u003C\u002Fdiv\u003E\n\n \u003Cdiv class=\"col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center\"\u003E\n \u003Cdiv class=\"text-center d-flex flex-column\"\u003E\n \n \u003Ca href=\"https:\u002F\u002Fgosrc.cc\u002Fgo\u002F8lwlaJY4R\" target=\"_blank\" rel=\"nofollow noopener\" class=\"v-btn v-btn--rounded elevation-5 v-size--large success mb-2\"\u003E\n \u003Cspan class=\"v-btn__content\"\u003EGet Book Now\u003C\u002Fspan\u003E\n \u003C\u002Fa\u003E\n \n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \n \u003Cdiv class=\"col-12\"\u003E\n \u003Cdiv class=\"v-card elevation-6\"\u003E\n \u003Cdiv class=\"v-card__text rating-text\"\u003E\n \u003Cdiv class=\"rating-counter\"\u003E\n \u003Cspan class=\"v-badge\"\u003E\n \u003Cspan class=\"v-badge__wrapper\"\u003E\n \u003Cspan aria-atomic=\"true\" aria-label=\"Позиция\" class=\"v-badge__badge primary\"\u003E\n 7\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"row\"\u003E\n \u003Cdiv class=\"col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center\"\u003E\n \u003Cdiv\u003E\n \u003Cimg src=\"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com\u002Fimages\u002Fd7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae\u002Frating\u002F41sz-tftqpl-sl160.jpg\" alt=\"Programming PyTorch for Deep Learning: Creating and Deploying Deep Learning Applications\" \u002F\u003E\n \u003Cp class=\"text-center font-weight-bold text-h6\"\u003ERating is 4.4 out of 5\u003C\u002Fp\u003E\n \u003Cdiv class=\"stars\" style=\"--rating: 4.4;\" aria-label=\"Rating is 4.4 out of 5\" \u003E\u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"col-lg-6 col-md-8 col-sm-6 col-12\"\u003E\n \u003Cp class=\"font-weight-bold rating-name\"\u003EProgramming PyTorch for Deep Learning: Creating and Deploying Deep Learning Applications\u003C\u002Fp\u003E\n \n \n\n \n \n \u003C\u002Fdiv\u003E\n\n \u003Cdiv class=\"col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center\"\u003E\n \u003Cdiv class=\"text-center d-flex flex-column\"\u003E\n \n \u003Ca href=\"https:\u002F\u002Fgosrc.cc\u002Fgo\u002F1d9_-JY4R\" target=\"_blank\" rel=\"nofollow noopener\" class=\"v-btn v-btn--rounded elevation-5 v-size--large success mb-2\"\u003E\n \u003Cspan class=\"v-btn__content\"\u003EGet Book Now\u003C\u002Fspan\u003E\n \u003C\u002Fa\u003E\n \n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \n \u003Cdiv class=\"col-12\"\u003E\n \u003Cdiv class=\"v-card elevation-6\"\u003E\n \u003Cdiv class=\"v-card__text rating-text\"\u003E\n \u003Cdiv class=\"rating-counter\"\u003E\n \u003Cspan class=\"v-badge\"\u003E\n \u003Cspan class=\"v-badge__wrapper\"\u003E\n \u003Cspan aria-atomic=\"true\" aria-label=\"Позиция\" class=\"v-badge__badge primary\"\u003E\n 8\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"row\"\u003E\n \u003Cdiv class=\"col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center\"\u003E\n \u003Cdiv\u003E\n \u003Cimg src=\"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com\u002Fimages\u002Fd7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae\u002Frating\u002F41xz0sfhxsl-sl160.jpg\" alt=\"PyTorch Pocket Reference: Building and Deploying Deep Learning Models\" \u002F\u003E\n \u003Cp class=\"text-center font-weight-bold text-h6\"\u003ERating is 4.3 out of 5\u003C\u002Fp\u003E\n \u003Cdiv class=\"stars\" style=\"--rating: 4.3;\" aria-label=\"Rating is 4.3 out of 5\" \u003E\u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"col-lg-6 col-md-8 col-sm-6 col-12\"\u003E\n \u003Cp class=\"font-weight-bold rating-name\"\u003EPyTorch Pocket Reference: Building and Deploying Deep Learning Models\u003C\u002Fp\u003E\n \n \n\n \n \n \u003C\u002Fdiv\u003E\n\n \u003Cdiv class=\"col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center\"\u003E\n \u003Cdiv class=\"text-center d-flex flex-column\"\u003E\n \n \u003Ca href=\"https:\u002F\u002Fgosrc.cc\u002Fgo\u002Fshe_-1YVg\" target=\"_blank\" rel=\"nofollow noopener\" class=\"v-btn v-btn--rounded elevation-5 v-size--large success mb-2\"\u003E\n \u003Cspan class=\"v-btn__content\"\u003EGet Book Now\u003C\u002Fspan\u003E\n \u003C\u002Fa\u003E\n \n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \n \u003Cdiv class=\"col-12\"\u003E\n \u003Cdiv class=\"v-card elevation-6\"\u003E\n \u003Cdiv class=\"v-card__text rating-text\"\u003E\n \u003Cdiv class=\"rating-counter\"\u003E\n \u003Cspan class=\"v-badge\"\u003E\n \u003Cspan class=\"v-badge__wrapper\"\u003E\n \u003Cspan aria-atomic=\"true\" aria-label=\"Позиция\" class=\"v-badge__badge primary\"\u003E\n 9\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"row\"\u003E\n \u003Cdiv class=\"col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center\"\u003E\n \u003Cdiv\u003E\n \u003Cimg src=\"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com\u002Fimages\u002Fd7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae\u002Frating\u002F41ayiicwf2l-sl160.jpg\" alt=\"Deep Learning with PyTorch Lightning: Swiftly build high-performance Artificial Intelligence (AI) models using Python\" \u002F\u003E\n \u003Cp class=\"text-center font-weight-bold text-h6\"\u003ERating is 4.2 out of 5\u003C\u002Fp\u003E\n \u003Cdiv class=\"stars\" style=\"--rating: 4.2;\" aria-label=\"Rating is 4.2 out of 5\" \u003E\u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"col-lg-6 col-md-8 col-sm-6 col-12\"\u003E\n \u003Cp class=\"font-weight-bold rating-name\"\u003EDeep Learning with PyTorch Lightning: Swiftly build high-performance Artificial Intelligence (AI) models using Python\u003C\u002Fp\u003E\n \n \n\n \n \n \u003C\u002Fdiv\u003E\n\n \u003Cdiv class=\"col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center\"\u003E\n \u003Cdiv class=\"text-center d-flex flex-column\"\u003E\n \n \u003Ca href=\"https:\u002F\u002Fgosrc.cc\u002Fgo\u002FkmiX-1LVg\" target=\"_blank\" rel=\"nofollow noopener\" class=\"v-btn v-btn--rounded elevation-5 v-size--large success mb-2\"\u003E\n \u003Cspan class=\"v-btn__content\"\u003EGet Book Now\u003C\u002Fspan\u003E\n \u003C\u002Fa\u003E\n \n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Ch2\u003EHow to implement batch normalization in a custom PyTorch model?\u003C\u002Fh2\u003E\u003Cp\u003ETo implement batch normalization in a custom PyTorch model, you can follow these steps:\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003EImport the required modules:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cpre class=\"code-block ql-syntax\" id=\"code-saa7pjg\"\u003Eimport torch\nimport torch.nn as nn\n\u003C\u002Fpre\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003EDefine a basic custom model class:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cpre class=\"code-block ql-syntax\" id=\"code-0qbn79f\"\u003Eclass CustomModel(nn.Module):\n def __init__(self):\n super(CustomModel, self).__init__()\n self.fc1 = nn.Linear(in_features, hidden_units)\n self.fc2 = nn.Linear(hidden_units, out_features)\n\u003C\u002Fpre\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EReplace \u003Ccode\u003Ein_features\u003C\u002Fcode\u003E, \u003Ccode\u003Ehidden_units\u003C\u002Fcode\u003E, and \u003Ccode\u003Eout_features\u003C\u002Fcode\u003E with appropriate values for your model architecture.\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003EAdd batch normalization layers and their parameters to the model:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cpre class=\"code-block ql-syntax\" id=\"code-l2fykbv\"\u003Eself.bn1 = nn.BatchNorm1d(hidden_units)\nself.bn2 = nn.BatchNorm1d(out_features)\n\u003C\u002Fpre\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EAdjust the parameter value based on your model architecture.\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003EDefine the forward pass of the model:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cpre class=\"code-block ql-syntax\" id=\"code-xy0rs7x\"\u003Edef forward(self, x):\n x = self.fc1(x)\n x = self.bn1(x)\n x = nn.functional.relu(x)\n x = self.fc2(x)\n x = self.bn2(x)\n return x\n\u003C\u002Fpre\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EThis example assumes the ReLU activation function, but you can replace it with any activation function you prefer.\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003ECreate an instance of the custom model:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cpre class=\"code-block ql-syntax\" id=\"code-8f8y751\"\u003Emodel = CustomModel()\n\u003C\u002Fpre\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003ENow you have implemented batch normalization in your custom PyTorch model.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Ch2\u003EWhat are the advantages of using batch normalization in PyTorch?\u003C\u002Fh2\u003E\u003Cp\u003EBatch normalization is a regularization technique that is widely used in deep learning models. When applied to PyTorch models, it provides several advantages:\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003E\u003Cstrong\u003EImproved convergence\u003C\u002Fstrong\u003E: Batch normalization normalizes the input to each neuron across a mini-batch, which helps in stabilizing the learning process. This leads to faster convergence and reduces the number of epochs required for training.\n\u003C\u002Fli\u003E\u003Cli\u003E\u003Cstrong\u003EReduced overfitting\u003C\u002Fstrong\u003E: By normalizing the inputs, batch normalization reduces the dependence of each neuron on the other neurons in the network. This reduces the chances of overfitting and improves the generalization ability of the model.\n\u003C\u002Fli\u003E\u003Cli\u003E\u003Cstrong\u003EIncreased learning rate\u003C\u002Fstrong\u003E: Batch normalization reduces the internal covariate shift by maintaining zero mean and unit variance activations. This enables the use of higher learning rates during training, which can speed up the training process.\n\u003C\u002Fli\u003E\u003Cli\u003E\u003Cstrong\u003EBetter gradient flow\u003C\u002Fstrong\u003E: Normalizing the inputs using batch normalization helps in ensuring that the gradients flow smoothly and consistently during backpropagation. This helps combat the vanishing and exploding gradient problems, making it easier to train deep networks.\n\u003C\u002Fli\u003E\u003Cli\u003E\u003Cstrong\u003ERobustness to different input distributions\u003C\u002Fstrong\u003E: Batch normalization makes the model less sensitive to the \u003Ca href=\"https:\u002F\u002Fstudentprojectcode.com\u002Fblog\u002Fhow-to-simplify-units-with-different-scales-in\" class=\"auto-link\" target=\"_blank\"\u003Escale\u003C\u002Fa\u003E and distribution of the input data. This allows the model to perform well even when faced with inputs that are significantly different from the training data.\n\u003C\u002Fli\u003E\u003Cli\u003E\u003Cstrong\u003EWeight initialization flexibility\u003C\u002Fstrong\u003E: Batch normalization helps in reducing the dependence of the model's performance on the choice of weight initialization. It allows the use of simpler initialization methods like random or small weights, which can speed up the training process.\n\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EOverall, batch normalization is a useful tool for improving the performance and stability of deep learning models in PyTorch, leading to faster convergence, better generalization, and increased robustness.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Ch2\u003EWhat is the effect of batch size on batch normalization in PyTorch?\u003C\u002Fh2\u003E\u003Cp\u003EThe batch size affects the batch normalization in PyTorch in the following way:\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003E\u003Cstrong\u003EStatistics estimation\u003C\u002Fstrong\u003E: Batch normalization relies on estimating the mean and variance of the input data to normalize it. With a larger batch size, there is more data available for statistics estimation, leading to more accurate estimates of the mean and variance. This can result in improved normalization and consequently, better performance.\n\u003C\u002Fli\u003E\u003Cli\u003E\u003Cstrong\u003ENoise reduction\u003C\u002Fstrong\u003E: Batch normalization introduces some noise to the statistics estimation process. With a larger batch size, the noise is averaged out more effectively, resulting in more stable estimates of mean and variance. This can lead to reduced overfitting and improved generalization.\n\u003C\u002Fli\u003E\u003Cli\u003E\u003Cstrong\u003ETraining dynamics\u003C\u002Fstrong\u003E: Smaller batch sizes tend to introduce more stochasticity and randomness in the training process, as each batch's statistics differ significantly. On the other hand, larger batch sizes provide more consistent statistics, which can affect the optimization process. This can result in different training dynamics, such as convergence speed and stability.\n\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EIt's important to note that the choice of batch size is often a trade-off. Larger batch sizes require more memory, may limit parallelization, and increase computational requirements. However, they can offer better normalization and estimation, while smaller batch sizes may introduce more noise but can be computationally more efficient.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Ch2\u003EWhat are the requirements for using batch normalization in PyTorch?\u003C\u002Fh2\u003E\u003Cp\u003ETo use batch normalization in PyTorch, the following requirements should be met:\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003E\u003Cstrong\u003EPyTorch should be installed on the system. You can install it using pip\u003C\u002Fstrong\u003E: pip install torch.\n\u003C\u002Fli\u003E\u003Cli\u003EImport the necessary modules:\n\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cpre class=\"code-block ql-syntax\" id=\"code-xq2ctpg\"\u003Eimport torch\nimport torch.nn as nn\n\u003C\u002Fpre\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003EDefine your model architecture using the nn.Module class. Use the torch.nn.BatchNorm2d or torch.nn.BatchNorm1d layer (based on your input dimensions) for batch normalization.\n\u003C\u002Fli\u003E\u003Cli\u003EUse batch normalization layer after the convolutional or linear layer in your model architecture. For example:\n\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cpre class=\"code-block ql-syntax\" id=\"code-th2v391\"\u003Eclass MyModel(nn.Module):\n def __init__(self):\n super(MyModel, self).__init__()\n self.conv1 = nn.Conv2d(3, 64, kernel_size=3)\n self.bn1 = nn.BatchNorm2d(64)\n self.fc1 = nn.Linear(64, 10)\n self.bn2 = nn.BatchNorm1d(10)\n ...\n\u003C\u002Fpre\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003EDuring the forward pass, apply batch normalization to the input tensor. For example:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cpre class=\"code-block ql-syntax\" id=\"code-tz9yw27\"\u003Edef forward(self, x):\n x = self.conv1(x)\n x = self.bn1(x)\n x = self.fc1(x)\n x = self.bn2(x)\n ...\n\u003C\u002Fpre\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003ENote: Batch normalization is typically used before the activation function, but the order can vary depending on your problem and experiment settings.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Ch2\u003EWhat is the impact of batch normalization on model generalization in PyTorch?\u003C\u002Fh2\u003E\u003Cp\u003EBatch normalization has a significant impact on model generalization in PyTorch. It helps to improve the generalization capability of neural networks by reducing the internal covariate shift.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EInternal covariate shift refers to the change in the distribution of network activations due to the change in parameter values during training. This can slow down the training process and hinder the performance of the model.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EBatch normalization solves this problem by normalizing the output of each layer using the mean and variance of the mini-batch. By doing so, it reduces the effect of the internal covariate shift and makes the optimization process more stable. Batch normalization also introduces additional trainable parameters, which allow the network to adaptively scale and shift the normalized values.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EThe normalization of inputs helps in the generalization of the model because it keeps the values within a reasonable range. It prevents extreme values from causing instability in the network, which can lead to overfitting. Additionally, batch normalization acts as a regularizer, reducing the need for other regularization techniques like dropout.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EOverall, batch normalization in PyTorch improves the generalization ability of models by reducing internal covariate shift, making the training process more stable, and acting as a regularizer.\u003C\u002Fp\u003E",content_ad:"\u003Cp\u003EBatch normalization is a widely used technique for \u003Ca class=\"auto-link\" href=\"https:\u002F\u002Fstlplaces.com\u002Fblog\u002Fhow-to-handle-overfitting-in-tensorflow-models\"\u003Eimproving the training of deep neural networks\u003C\u002Fa\u003E. It normalizes the activations of each mini-batch by subtracting the mini-batch mean and dividing by the mini-batch standard deviation. This helps in reducing internal covariate shift by ensuring that the input to each layer is normalized.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EImplementing batch normalization in PyTorch is straightforward. Here are the steps:\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003EImport the necessary libraries:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cpre class=\"code-block ql-syntax\" id=\"code-q2hyroj\"\u003Eimport torch\nimport torch.nn as nn\n\u003C\u002Fpre\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003EDefine a custom neural network architecture:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cpre class=\"code-block ql-syntax\" id=\"code-tz45vgq\"\u003Eclass Net(nn.Module):\n def __init__(self):\n super(Net, self).__init__()\n self.fc1 = nn.Linear(10, 20)\n self.bn1 = nn.BatchNorm1d(20) # Batch normalization layer\n self.fc2 = nn.Linear(20, 10)\n\u003C\u002Fpre\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003EOverride the forward method of the neural network:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cpre class=\"code-block ql-syntax\" id=\"code-ig8hxn3\"\u003E def forward(self, x):\n x = self.fc1(x)\n x = self.bn1(x)\n x = torch.relu(x)\n x = self.fc2(x)\n return x\n\u003C\u002Fpre\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003ECreate an instance of the network:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cpre class=\"code-block ql-syntax\" id=\"code-r682yo6\"\u003Enet = Net()\n\u003C\u002Fpre\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EThat's it! Now the network \u003Ccode\u003Enet\u003C\u002Fcode\u003E includes a batch normalization layer (\u003Ccode\u003Eself.bn1\u003C\u002Fcode\u003E) after the first \u003Ca href=\"https:\u002F\u002Ftopminisite.com\u002Fblog\u002Fhow-to-restore-in-fully-connected-layer-using\" class=\"auto-link\" target=\"_blank\"\u003Efully connected layer\u003C\u002Fa\u003E (\u003Ccode\u003Eself.fc1\u003C\u002Fcode\u003E). During training, as the mini-batches pass through this network, the batch normalization layer will normalize the activations.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003ENote: It is essential to ensure that the network is in training mode using \u003Ccode\u003Enet.train()\u003C\u002Fcode\u003E before training and in evaluation mode using \u003Ccode\u003Enet.eval()\u003C\u002Fcode\u003E during inference\u002Ftesting.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EYou can now use this network for training and inference in your PyTorch project, while enjoying the benefits of batch normalization.\u003C\u002Fp\u003E\n \u003Cdiv class=\"rating\"\u003E\n \u003Ch2\u003EBest PyTorch Books of December 2024\u003C\u002Fh2\u003E\n \u003Cdiv class=\"row mt-2\"\u003E\n \n \u003Cdiv class=\"col-12\"\u003E\n \u003Cdiv class=\"v-card elevation-6\"\u003E\n \u003Cdiv class=\"v-card__text rating-text\"\u003E\n \u003Cdiv class=\"rating-counter\"\u003E\n \u003Cspan class=\"v-badge\"\u003E\n \u003Cspan class=\"v-badge__wrapper\"\u003E\n \u003Cspan aria-atomic=\"true\" aria-label=\"Позиция\" class=\"v-badge__badge primary\"\u003E\n 1\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"row\"\u003E\n \u003Cdiv class=\"col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center\"\u003E\n \u003Cdiv\u003E\n \u003Cimg src=\"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com\u002Fimages\u002Fd7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae\u002Frating\u002F41jkoc6owal-sl160.jpg\" alt=\"PyTorch Recipes: A Problem-Solution Approach to Build, Train and Deploy Neural Network Models\" \u002F\u003E\n \u003Cp class=\"text-center font-weight-bold text-h6\"\u003ERating is 5 out of 5\u003C\u002Fp\u003E\n \u003Cdiv class=\"stars\" style=\"--rating: 5;\" aria-label=\"Rating is 5 out of 5\" \u003E\u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"col-lg-6 col-md-8 col-sm-6 col-12\"\u003E\n \u003Cp class=\"font-weight-bold rating-name\"\u003EPyTorch Recipes: A Problem-Solution Approach to Build, Train and Deploy Neural Network Models\u003C\u002Fp\u003E\n \n \n\n \n \n \u003C\u002Fdiv\u003E\n\n \u003Cdiv class=\"col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center\"\u003E\n \u003Cdiv class=\"text-center d-flex flex-column\"\u003E\n \n \u003Ca href=\"https:\u002F\u002Fgosrc.cc\u002Fgo\u002FMDdl-1Y4g\" target=\"_blank\" rel=\"nofollow noopener\" class=\"v-btn v-btn--rounded elevation-5 v-size--large success mb-2\"\u003E\n \u003Cspan class=\"v-btn__content\"\u003EGet Book Now\u003C\u002Fspan\u003E\n \u003C\u002Fa\u003E\n \n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \n \u003Cdiv class=\"col-12\"\u003E\n \u003Cdiv class=\"v-card elevation-6\"\u003E\n \u003Cdiv class=\"v-card__text rating-text\"\u003E\n \u003Cdiv class=\"rating-counter\"\u003E\n \u003Cspan class=\"v-badge\"\u003E\n \u003Cspan class=\"v-badge__wrapper\"\u003E\n \u003Cspan aria-atomic=\"true\" aria-label=\"Позиция\" class=\"v-badge__badge primary\"\u003E\n 2\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"row\"\u003E\n \u003Cdiv class=\"col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center\"\u003E\n \u003Cdiv\u003E\n \u003Cimg src=\"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com\u002Fimages\u002Fd7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae\u002Frating\u002F4198ez8dpbl-sl160.jpg\" alt=\"Mastering PyTorch: Build powerful deep learning architectures using advanced PyTorch features, 2nd Edition\" \u002F\u003E\n \u003Cp class=\"text-center font-weight-bold text-h6\"\u003ERating is 4.9 out of 5\u003C\u002Fp\u003E\n \u003Cdiv class=\"stars\" style=\"--rating: 4.9;\" aria-label=\"Rating is 4.9 out of 5\" \u003E\u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"col-lg-6 col-md-8 col-sm-6 col-12\"\u003E\n \u003Cp class=\"font-weight-bold rating-name\"\u003EMastering PyTorch: Build powerful deep learning architectures using advanced PyTorch features, 2nd Edition\u003C\u002Fp\u003E\n \n \n\n \n \n \u003C\u002Fdiv\u003E\n\n \u003Cdiv class=\"col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center\"\u003E\n \u003Cdiv class=\"text-center d-flex flex-column\"\u003E\n \n \u003Ca href=\"https:\u002F\u002Fgosrc.cc\u002Fgo\u002FUZp_a1L4R\" target=\"_blank\" rel=\"nofollow noopener\" class=\"v-btn v-btn--rounded elevation-5 v-size--large success mb-2\"\u003E\n \u003Cspan class=\"v-btn__content\"\u003EGet Book Now\u003C\u002Fspan\u003E\n \u003C\u002Fa\u003E\n \n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \n \u003Cdiv class=\"col-12\"\u003E\n \u003Cdiv class=\"v-card elevation-6\"\u003E\n \u003Cdiv class=\"v-card__text rating-text\"\u003E\n \u003Cdiv class=\"rating-counter\"\u003E\n \u003Cspan class=\"v-badge\"\u003E\n \u003Cspan class=\"v-badge__wrapper\"\u003E\n \u003Cspan aria-atomic=\"true\" aria-label=\"Позиция\" class=\"v-badge__badge primary\"\u003E\n 3\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"row\"\u003E\n \u003Cdiv class=\"col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center\"\u003E\n \u003Cdiv\u003E\n \u003Cimg src=\"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com\u002Fimages\u002Fd7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae\u002Frating\u002F513exc7qqjl-sl160.jpg\" alt=\"Natural Language Processing with PyTorch: Build Intelligent Language Applications Using Deep Learning\" \u002F\u003E\n \u003Cp class=\"text-center font-weight-bold text-h6\"\u003ERating is 4.8 out of 5\u003C\u002Fp\u003E\n \u003Cdiv class=\"stars\" style=\"--rating: 4.8;\" aria-label=\"Rating is 4.8 out of 5\" \u003E\u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"col-lg-6 col-md-8 col-sm-6 col-12\"\u003E\n \u003Cp class=\"font-weight-bold rating-name\"\u003ENatural Language Processing with PyTorch: Build Intelligent Language Applications Using Deep Learning\u003C\u002Fp\u003E\n \n \n\n \n \n \u003C\u002Fdiv\u003E\n\n \u003Cdiv class=\"col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center\"\u003E\n \u003Cdiv class=\"text-center d-flex flex-column\"\u003E\n \n \u003Ca href=\"https:\u002F\u002Fgosrc.cc\u002Fgo\u002FAQT_-1L4R\" target=\"_blank\" rel=\"nofollow noopener\" class=\"v-btn v-btn--rounded elevation-5 v-size--large success mb-2\"\u003E\n \u003Cspan class=\"v-btn__content\"\u003EGet Book Now\u003C\u002Fspan\u003E\n \u003C\u002Fa\u003E\n \n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \n \u003Cdiv class=\"col-12\"\u003E\n \u003Cdiv class=\"v-card elevation-6\"\u003E\n \u003Cdiv class=\"v-card__text rating-text\"\u003E\n \u003Cdiv class=\"rating-counter\"\u003E\n \u003Cspan class=\"v-badge\"\u003E\n \u003Cspan class=\"v-badge__wrapper\"\u003E\n \u003Cspan aria-atomic=\"true\" aria-label=\"Позиция\" class=\"v-badge__badge primary\"\u003E\n 4\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"row\"\u003E\n \u003Cdiv class=\"col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center\"\u003E\n \u003Cdiv\u003E\n \u003Cimg src=\"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com\u002Fimages\u002Fd7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae\u002Frating\u002F518jzrflrvl-sl160.jpg\" alt=\"Deep Learning for Coders with Fastai and PyTorch: AI Applications Without a PhD\" \u002F\u003E\n \u003Cp class=\"text-center font-weight-bold text-h6\"\u003ERating is 4.7 out of 5\u003C\u002Fp\u003E\n \u003Cdiv class=\"stars\" style=\"--rating: 4.7;\" aria-label=\"Rating is 4.7 out of 5\" \u003E\u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"col-lg-6 col-md-8 col-sm-6 col-12\"\u003E\n \u003Cp class=\"font-weight-bold rating-name\"\u003EDeep Learning for Coders with Fastai and PyTorch: AI Applications Without a PhD\u003C\u002Fp\u003E\n \n \n\n \n \n \u003C\u002Fdiv\u003E\n\n \u003Cdiv class=\"col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center\"\u003E\n \u003Cdiv class=\"text-center d-flex flex-column\"\u003E\n \n \u003Ca href=\"https:\u002F\u002Fgosrc.cc\u002Fgo\u002F7JalaJYVR\" target=\"_blank\" rel=\"nofollow noopener\" class=\"v-btn v-btn--rounded elevation-5 v-size--large success mb-2\"\u003E\n \u003Cspan class=\"v-btn__content\"\u003EGet Book Now\u003C\u002Fspan\u003E\n \u003C\u002Fa\u003E\n \n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \n \u003Cdiv class=\"col-12\"\u003E\n \u003Cdiv class=\"v-card elevation-6\"\u003E\n \u003Cdiv class=\"v-card__text rating-text\"\u003E\n \u003Cdiv class=\"rating-counter\"\u003E\n \u003Cspan class=\"v-badge\"\u003E\n \u003Cspan class=\"v-badge__wrapper\"\u003E\n \u003Cspan aria-atomic=\"true\" aria-label=\"Позиция\" class=\"v-badge__badge primary\"\u003E\n 5\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"row\"\u003E\n \u003Cdiv class=\"col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center\"\u003E\n \u003Cdiv\u003E\n \u003Cimg src=\"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com\u002Fimages\u002Fd7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae\u002Frating\u002F41o-uy9lzzl-sl160.jpg\" alt=\"Machine Learning with PyTorch and Scikit-Learn: Develop machine learning and deep learning models with Python\" \u002F\u003E\n \u003Cp class=\"text-center font-weight-bold text-h6\"\u003ERating is 4.6 out of 5\u003C\u002Fp\u003E\n \u003Cdiv class=\"stars\" style=\"--rating: 4.6;\" aria-label=\"Rating is 4.6 out of 5\" \u003E\u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"col-lg-6 col-md-8 col-sm-6 col-12\"\u003E\n \u003Cp class=\"font-weight-bold rating-name\"\u003EMachine Learning with PyTorch and Scikit-Learn: Develop machine learning and deep learning models with Python\u003C\u002Fp\u003E\n \n \n\n \n \n \u003C\u002Fdiv\u003E\n\n \u003Cdiv class=\"col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center\"\u003E\n \u003Cdiv class=\"text-center d-flex flex-column\"\u003E\n \n \u003Ca href=\"https:\u002F\u002Fgosrc.cc\u002Fgo\u002FGZs_a1Y4g\" target=\"_blank\" rel=\"nofollow noopener\" class=\"v-btn v-btn--rounded elevation-5 v-size--large success mb-2\"\u003E\n \u003Cspan class=\"v-btn__content\"\u003EGet Book Now\u003C\u002Fspan\u003E\n \u003C\u002Fa\u003E\n \n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \n \u003Cdiv class=\"col-12\"\u003E\n \u003Cdiv class=\"v-card elevation-6\"\u003E\n \u003Cdiv class=\"v-card__text rating-text\"\u003E\n \u003Cdiv class=\"rating-counter\"\u003E\n \u003Cspan class=\"v-badge\"\u003E\n \u003Cspan class=\"v-badge__wrapper\"\u003E\n \u003Cspan aria-atomic=\"true\" aria-label=\"Позиция\" class=\"v-badge__badge primary\"\u003E\n 6\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"row\"\u003E\n \u003Cdiv class=\"col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center\"\u003E\n \u003Cdiv\u003E\n \u003Cimg src=\"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com\u002Fimages\u002Fd7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae\u002Frating\u002F411ruzisg3l-sl160.jpg\" alt=\"Deep Learning with PyTorch: Build, train, and tune neural networks using Python tools\" \u002F\u003E\n \u003Cp class=\"text-center font-weight-bold text-h6\"\u003ERating is 4.5 out of 5\u003C\u002Fp\u003E\n \u003Cdiv class=\"stars\" style=\"--rating: 4.5;\" aria-label=\"Rating is 4.5 out of 5\" \u003E\u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"col-lg-6 col-md-8 col-sm-6 col-12\"\u003E\n \u003Cp class=\"font-weight-bold rating-name\"\u003EDeep Learning with PyTorch: Build, train, and tune neural networks using Python tools\u003C\u002Fp\u003E\n \n \n\n \n \n \u003C\u002Fdiv\u003E\n\n \u003Cdiv class=\"col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center\"\u003E\n \u003Cdiv class=\"text-center d-flex flex-column\"\u003E\n \n \u003Ca href=\"https:\u002F\u002Fgosrc.cc\u002Fgo\u002F8lwlaJY4R\" target=\"_blank\" rel=\"nofollow noopener\" class=\"v-btn v-btn--rounded elevation-5 v-size--large success mb-2\"\u003E\n \u003Cspan class=\"v-btn__content\"\u003EGet Book Now\u003C\u002Fspan\u003E\n \u003C\u002Fa\u003E\n \n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \n \u003Cdiv class=\"col-12\"\u003E\n \u003Cdiv class=\"v-card elevation-6\"\u003E\n \u003Cdiv class=\"v-card__text rating-text\"\u003E\n \u003Cdiv class=\"rating-counter\"\u003E\n \u003Cspan class=\"v-badge\"\u003E\n \u003Cspan class=\"v-badge__wrapper\"\u003E\n \u003Cspan aria-atomic=\"true\" aria-label=\"Позиция\" class=\"v-badge__badge primary\"\u003E\n 7\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"row\"\u003E\n \u003Cdiv class=\"col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center\"\u003E\n \u003Cdiv\u003E\n \u003Cimg src=\"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com\u002Fimages\u002Fd7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae\u002Frating\u002F41sz-tftqpl-sl160.jpg\" alt=\"Programming PyTorch for Deep Learning: Creating and Deploying Deep Learning Applications\" \u002F\u003E\n \u003Cp class=\"text-center font-weight-bold text-h6\"\u003ERating is 4.4 out of 5\u003C\u002Fp\u003E\n \u003Cdiv class=\"stars\" style=\"--rating: 4.4;\" aria-label=\"Rating is 4.4 out of 5\" \u003E\u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"col-lg-6 col-md-8 col-sm-6 col-12\"\u003E\n \u003Cp class=\"font-weight-bold rating-name\"\u003EProgramming PyTorch for Deep Learning: Creating and Deploying Deep Learning Applications\u003C\u002Fp\u003E\n \n \n\n \n \n \u003C\u002Fdiv\u003E\n\n \u003Cdiv class=\"col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center\"\u003E\n \u003Cdiv class=\"text-center d-flex flex-column\"\u003E\n \n \u003Ca href=\"https:\u002F\u002Fgosrc.cc\u002Fgo\u002F1d9_-JY4R\" target=\"_blank\" rel=\"nofollow noopener\" class=\"v-btn v-btn--rounded elevation-5 v-size--large success mb-2\"\u003E\n \u003Cspan class=\"v-btn__content\"\u003EGet Book Now\u003C\u002Fspan\u003E\n \u003C\u002Fa\u003E\n \n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \n \u003Cdiv class=\"col-12\"\u003E\n \u003Cdiv class=\"v-card elevation-6\"\u003E\n \u003Cdiv class=\"v-card__text rating-text\"\u003E\n \u003Cdiv class=\"rating-counter\"\u003E\n \u003Cspan class=\"v-badge\"\u003E\n \u003Cspan class=\"v-badge__wrapper\"\u003E\n \u003Cspan aria-atomic=\"true\" aria-label=\"Позиция\" class=\"v-badge__badge primary\"\u003E\n 8\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"row\"\u003E\n \u003Cdiv class=\"col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center\"\u003E\n \u003Cdiv\u003E\n \u003Cimg src=\"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com\u002Fimages\u002Fd7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae\u002Frating\u002F41xz0sfhxsl-sl160.jpg\" alt=\"PyTorch Pocket Reference: Building and Deploying Deep Learning Models\" \u002F\u003E\n \u003Cp class=\"text-center font-weight-bold text-h6\"\u003ERating is 4.3 out of 5\u003C\u002Fp\u003E\n \u003Cdiv class=\"stars\" style=\"--rating: 4.3;\" aria-label=\"Rating is 4.3 out of 5\" \u003E\u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"col-lg-6 col-md-8 col-sm-6 col-12\"\u003E\n \u003Cp class=\"font-weight-bold rating-name\"\u003EPyTorch Pocket Reference: Building and Deploying Deep Learning Models\u003C\u002Fp\u003E\n \n \n\n \n \n \u003C\u002Fdiv\u003E\n\n \u003Cdiv class=\"col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center\"\u003E\n \u003Cdiv class=\"text-center d-flex flex-column\"\u003E\n \n \u003Ca href=\"https:\u002F\u002Fgosrc.cc\u002Fgo\u002Fshe_-1YVg\" target=\"_blank\" rel=\"nofollow noopener\" class=\"v-btn v-btn--rounded elevation-5 v-size--large success mb-2\"\u003E\n \u003Cspan class=\"v-btn__content\"\u003EGet Book Now\u003C\u002Fspan\u003E\n \u003C\u002Fa\u003E\n \n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \n \u003Cdiv class=\"col-12\"\u003E\n \u003Cdiv class=\"v-card elevation-6\"\u003E\n \u003Cdiv class=\"v-card__text rating-text\"\u003E\n \u003Cdiv class=\"rating-counter\"\u003E\n \u003Cspan class=\"v-badge\"\u003E\n \u003Cspan class=\"v-badge__wrapper\"\u003E\n \u003Cspan aria-atomic=\"true\" aria-label=\"Позиция\" class=\"v-badge__badge primary\"\u003E\n 9\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"row\"\u003E\n \u003Cdiv class=\"col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center\"\u003E\n \u003Cdiv\u003E\n \u003Cimg src=\"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com\u002Fimages\u002Fd7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae\u002Frating\u002F41ayiicwf2l-sl160.jpg\" alt=\"Deep Learning with PyTorch Lightning: Swiftly build high-performance Artificial Intelligence (AI) models using Python\" \u002F\u003E\n \u003Cp class=\"text-center font-weight-bold text-h6\"\u003ERating is 4.2 out of 5\u003C\u002Fp\u003E\n \u003Cdiv class=\"stars\" style=\"--rating: 4.2;\" aria-label=\"Rating is 4.2 out of 5\" \u003E\u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"col-lg-6 col-md-8 col-sm-6 col-12\"\u003E\n \u003Cp class=\"font-weight-bold rating-name\"\u003EDeep Learning with PyTorch Lightning: Swiftly build high-performance Artificial Intelligence (AI) models using Python\u003C\u002Fp\u003E\n \n \n\n \n \n \u003C\u002Fdiv\u003E\n\n \u003Cdiv class=\"col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center\"\u003E\n \u003Cdiv class=\"text-center d-flex flex-column\"\u003E\n \n \u003Ca href=\"https:\u002F\u002Fgosrc.cc\u002Fgo\u002FkmiX-1LVg\" target=\"_blank\" rel=\"nofollow noopener\" class=\"v-btn v-btn--rounded elevation-5 v-size--large success mb-2\"\u003E\n \u003Cspan class=\"v-btn__content\"\u003EGet Book Now\u003C\u002Fspan\u003E\n \u003C\u002Fa\u003E\n \n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Ch2\u003EHow to implement batch normalization in a custom PyTorch model?\u003C\u002Fh2\u003E\u003Cp\u003ETo implement batch normalization in a custom PyTorch model, you can follow these steps:\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003EImport the required modules:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cpre class=\"code-block ql-syntax\" id=\"code-saa7pjg\"\u003Eimport torch\nimport torch.nn as nn\n\u003C\u002Fpre\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cscript async=\"\" src=\"https:\u002F\u002Fpagead2.googlesyndication.com\u002Fpagead\u002Fjs\u002Fadsbygoogle.js\"\u003E\u003C\u002Fscript\u003E\n\u003Cins class=\"adsbygoogle\" style=\"display:block\" data-ad-format=\"fluid\" data-ad-layout-key=\"-ef+6k-30-ac+ty\" data-ad-client=\"ca-pub-4833888168110763\" data-ad-slot=\"3267362137\"\u003E\u003C\u002Fins\u003E\n\u003Cscript\u003E\n (adsbygoogle = window.adsbygoogle || []).push({});\n\u003C\u002Fscript\u003E\u003Col\u003E\u003Cli\u003EDefine a basic custom model class:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cpre class=\"code-block ql-syntax\" id=\"code-0qbn79f\"\u003Eclass CustomModel(nn.Module):\n def __init__(self):\n super(CustomModel, self).__init__()\n self.fc1 = nn.Linear(in_features, hidden_units)\n self.fc2 = nn.Linear(hidden_units, out_features)\n\u003C\u002Fpre\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EReplace \u003Ccode\u003Ein_features\u003C\u002Fcode\u003E, \u003Ccode\u003Ehidden_units\u003C\u002Fcode\u003E, and \u003Ccode\u003Eout_features\u003C\u002Fcode\u003E with appropriate values for your model architecture.\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003EAdd batch normalization layers and their parameters to the model:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cpre class=\"code-block ql-syntax\" id=\"code-l2fykbv\"\u003Eself.bn1 = nn.BatchNorm1d(hidden_units)\nself.bn2 = nn.BatchNorm1d(out_features)\n\u003C\u002Fpre\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EAdjust the parameter value based on your model architecture.\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003EDefine the forward pass of the model:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cpre class=\"code-block ql-syntax\" id=\"code-xy0rs7x\"\u003Edef forward(self, x):\n x = self.fc1(x)\n x = self.bn1(x)\n x = nn.functional.relu(x)\n x = self.fc2(x)\n x = self.bn2(x)\n return x\n\u003C\u002Fpre\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EThis example assumes the ReLU activation function, but you can replace it with any activation function you prefer.\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003ECreate an instance of the custom model:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cpre class=\"code-block ql-syntax\" id=\"code-8f8y751\"\u003Emodel = CustomModel()\n\u003C\u002Fpre\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003ENow you have implemented batch normalization in your custom PyTorch model.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Ch2\u003EWhat are the advantages of using batch normalization in PyTorch?\u003C\u002Fh2\u003E\u003Cp\u003EBatch normalization is a regularization technique that is widely used in deep learning models. When applied to PyTorch models, it provides several advantages:\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003E\u003Cstrong\u003EImproved convergence\u003C\u002Fstrong\u003E: Batch normalization normalizes the input to each neuron across a mini-batch, which helps in stabilizing the learning process. This leads to faster convergence and reduces the number of epochs required for training.\n\u003C\u002Fli\u003E\u003Cli\u003E\u003Cstrong\u003EReduced overfitting\u003C\u002Fstrong\u003E: By normalizing the inputs, batch normalization reduces the dependence of each neuron on the other neurons in the network. This reduces the chances of overfitting and improves the generalization ability of the model.\n\u003C\u002Fli\u003E\u003Cli\u003E\u003Cstrong\u003EIncreased learning rate\u003C\u002Fstrong\u003E: Batch normalization reduces the internal covariate shift by maintaining zero mean and unit variance activations. This enables the use of higher learning rates during training, which can speed up the training process.\n\u003C\u002Fli\u003E\u003Cli\u003E\u003Cstrong\u003EBetter gradient flow\u003C\u002Fstrong\u003E: Normalizing the inputs using batch normalization helps in ensuring that the gradients flow smoothly and consistently during backpropagation. This helps combat the vanishing and exploding gradient problems, making it easier to train deep networks.\n\u003C\u002Fli\u003E\u003Cli\u003E\u003Cstrong\u003ERobustness to different input distributions\u003C\u002Fstrong\u003E: Batch normalization makes the model less sensitive to the \u003Ca href=\"https:\u002F\u002Fstudentprojectcode.com\u002Fblog\u002Fhow-to-simplify-units-with-different-scales-in\" class=\"auto-link\" target=\"_blank\"\u003Escale\u003C\u002Fa\u003E and distribution of the input data. This allows the model to perform well even when faced with inputs that are significantly different from the training data.\n\u003C\u002Fli\u003E\u003Cli\u003E\u003Cstrong\u003EWeight initialization flexibility\u003C\u002Fstrong\u003E: Batch normalization helps in reducing the dependence of the model's performance on the choice of weight initialization. It allows the use of simpler initialization methods like random or small weights, which can speed up the training process.\n\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EOverall, batch normalization is a useful tool for improving the performance and stability of deep learning models in PyTorch, leading to faster convergence, better generalization, and increased robustness.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Ch2\u003EWhat is the effect of batch size on batch normalization in PyTorch?\u003C\u002Fh2\u003E\u003Cp\u003EThe batch size affects the batch normalization in PyTorch in the following way:\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003E\u003Cstrong\u003EStatistics estimation\u003C\u002Fstrong\u003E: Batch normalization relies on estimating the mean and variance of the input data to normalize it. With a larger batch size, there is more data available for statistics estimation, leading to more accurate estimates of the mean and variance. This can result in improved normalization and consequently, better performance.\n\u003C\u002Fli\u003E\u003Cli\u003E\u003Cstrong\u003ENoise reduction\u003C\u002Fstrong\u003E: Batch normalization introduces some noise to the statistics estimation process. With a larger batch size, the noise is averaged out more effectively, resulting in more stable estimates of mean and variance. This can lead to reduced overfitting and improved generalization.\n\u003C\u002Fli\u003E\u003Cli\u003E\u003Cstrong\u003ETraining dynamics\u003C\u002Fstrong\u003E: Smaller batch sizes tend to introduce more stochasticity and randomness in the training process, as each batch's statistics differ significantly. On the other hand, larger batch sizes provide more consistent statistics, which can affect the optimization process. This can result in different training dynamics, such as convergence speed and stability.\n\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cscript async=\"\" src=\"https:\u002F\u002Fpagead2.googlesyndication.com\u002Fpagead\u002Fjs\u002Fadsbygoogle.js\"\u003E\u003C\u002Fscript\u003E\n\u003Cins class=\"adsbygoogle\" style=\"display:block\" data-ad-format=\"fluid\" data-ad-layout-key=\"-ef+6k-30-ac+ty\" data-ad-client=\"ca-pub-4833888168110763\" data-ad-slot=\"3267362137\"\u003E\u003C\u002Fins\u003E\n\u003Cscript\u003E\n (adsbygoogle = window.adsbygoogle || []).push({});\n\u003C\u002Fscript\u003E\u003Cp\u003EIt's important to note that the choice of batch size is often a trade-off. Larger batch sizes require more memory, may limit parallelization, and increase computational requirements. However, they can offer better normalization and estimation, while smaller batch sizes may introduce more noise but can be computationally more efficient.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Ch2\u003EWhat are the requirements for using batch normalization in PyTorch?\u003C\u002Fh2\u003E\u003Cp\u003ETo use batch normalization in PyTorch, the following requirements should be met:\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003E\u003Cstrong\u003EPyTorch should be installed on the system. You can install it using pip\u003C\u002Fstrong\u003E: pip install torch.\n\u003C\u002Fli\u003E\u003Cli\u003EImport the necessary modules:\n\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cpre class=\"code-block ql-syntax\" id=\"code-xq2ctpg\"\u003Eimport torch\nimport torch.nn as nn\n\u003C\u002Fpre\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003EDefine your model architecture using the nn.Module class. Use the torch.nn.BatchNorm2d or torch.nn.BatchNorm1d layer (based on your input dimensions) for batch normalization.\n\u003C\u002Fli\u003E\u003Cli\u003EUse batch normalization layer after the convolutional or linear layer in your model architecture. For example:\n\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cpre class=\"code-block ql-syntax\" id=\"code-th2v391\"\u003Eclass MyModel(nn.Module):\n def __init__(self):\n super(MyModel, self).__init__()\n self.conv1 = nn.Conv2d(3, 64, kernel_size=3)\n self.bn1 = nn.BatchNorm2d(64)\n self.fc1 = nn.Linear(64, 10)\n self.bn2 = nn.BatchNorm1d(10)\n ...\n\u003C\u002Fpre\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003EDuring the forward pass, apply batch normalization to the input tensor. For example:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cpre class=\"code-block ql-syntax\" id=\"code-tz9yw27\"\u003Edef forward(self, x):\n x = self.conv1(x)\n x = self.bn1(x)\n x = self.fc1(x)\n x = self.bn2(x)\n ...\n\u003C\u002Fpre\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003ENote: Batch normalization is typically used before the activation function, but the order can vary depending on your problem and experiment settings.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Ch2\u003EWhat is the impact of batch normalization on model generalization in PyTorch?\u003C\u002Fh2\u003E\u003Cp\u003EBatch normalization has a significant impact on model generalization in PyTorch. It helps to improve the generalization capability of neural networks by reducing the internal covariate shift.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EInternal covariate shift refers to the change in the distribution of network activations due to the change in parameter values during training. This can slow down the training process and hinder the performance of the model.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EBatch normalization solves this problem by normalizing the output of each layer using the mean and variance of the mini-batch. By doing so, it reduces the effect of the internal covariate shift and makes the optimization process more stable. Batch normalization also introduces additional trainable parameters, which allow the network to adaptively scale and shift the normalized values.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EThe normalization of inputs helps in the generalization of the model because it keeps the values within a reasonable range. It prevents extreme values from causing instability in the network, which can lead to overfitting. Additionally, batch normalization acts as a regularizer, reducing the need for other regularization techniques like dropout.\u003C\u002Fp\u003E\u003Cscript async=\"\" src=\"https:\u002F\u002Fpagead2.googlesyndication.com\u002Fpagead\u002Fjs\u002Fadsbygoogle.js\"\u003E\u003C\u002Fscript\u003E\n\u003Cins class=\"adsbygoogle\" style=\"display:block\" data-ad-format=\"fluid\" data-ad-layout-key=\"-ef+6k-30-ac+ty\" data-ad-client=\"ca-pub-4833888168110763\" data-ad-slot=\"3267362137\"\u003E\u003C\u002Fins\u003E\n\u003Cscript\u003E\n (adsbygoogle = window.adsbygoogle || []).push({});\n\u003C\u002Fscript\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EOverall, batch normalization in PyTorch improves the generalization ability of models by reducing internal covariate shift, making the training process more stable, and acting as a regularizer.\u003C\u002Fp\u003E",formatted_content:"\u003Cp\u003EBatch normalization is a widely used technique for \u003Ca href=\"https:\u002F\u002Fstlplaces.com\u002Fblog\u002Fhow-to-handle-overfitting-in-tensorflow-models\"\u003Eimproving the training of deep neural networks\u003C\u002Fa\u003E. It normalizes the activations of each mini-batch by subtracting the mini-batch mean and dividing by the mini-batch standard deviation. This helps in reducing internal covariate shift by ensuring that the input to each layer is normalized.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EImplementing batch normalization in PyTorch is straightforward. Here are the steps:\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003EImport the necessary libraries:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cdiv style=\"color:#f8f8f2;background-color:#272822;\"\u003E\n\u003Ctable style=\"border-spacing:0;padding:0;margin:0;border:0;\"\u003E\u003Ctbody\u003E\u003Ctr\u003E\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E1\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E2\n\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\n\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;;width:100%\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003Eimport torch\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003Eimport torch.nn as nn\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\u003C\u002Ftr\u003E\u003C\u002Ftbody\u003E\u003C\u002Ftable\u003E\n\u003C\u002Fdiv\u003E\n\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003EDefine a custom neural network architecture:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cdiv style=\"color:#f8f8f2;background-color:#272822;\"\u003E\n\u003Ctable style=\"border-spacing:0;padding:0;margin:0;border:0;\"\u003E\u003Ctbody\u003E\u003Ctr\u003E\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E1\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E2\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E3\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E4\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E5\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E6\n\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\n\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;;width:100%\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003Eclass Net(nn.Module):\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E def __init__(self):\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E super(Net, self).__init__()\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E self.fc1 = nn.Linear(10, 20)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E self.bn1 = nn.BatchNorm1d(20) # Batch normalization layer\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E self.fc2 = nn.Linear(20, 10)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\u003C\u002Ftr\u003E\u003C\u002Ftbody\u003E\u003C\u002Ftable\u003E\n\u003C\u002Fdiv\u003E\n\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003EOverride the forward method of the neural network:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cdiv style=\"color:#f8f8f2;background-color:#272822;\"\u003E\n\u003Ctable style=\"border-spacing:0;padding:0;margin:0;border:0;\"\u003E\u003Ctbody\u003E\u003Ctr\u003E\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E1\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E2\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E3\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E4\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E5\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E6\n\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\n\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;;width:100%\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E def forward(self, x):\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E x = self.fc1(x)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E x = self.bn1(x)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E x = torch.relu(x)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E x = self.fc2(x)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E return x\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\u003C\u002Ftr\u003E\u003C\u002Ftbody\u003E\u003C\u002Ftable\u003E\n\u003C\u002Fdiv\u003E\n\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003ECreate an instance of the network:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cdiv style=\"color:#f8f8f2;background-color:#272822;\"\u003E\n\u003Ctable style=\"border-spacing:0;padding:0;margin:0;border:0;\"\u003E\u003Ctbody\u003E\u003Ctr\u003E\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E1\n\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\n\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;;width:100%\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003Enet = Net()\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\u003C\u002Ftr\u003E\u003C\u002Ftbody\u003E\u003C\u002Ftable\u003E\n\u003C\u002Fdiv\u003E\n\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EThat's it! Now the network \u003Ccode\u003Enet\u003C\u002Fcode\u003E includes a batch normalization layer (\u003Ccode\u003Eself.bn1\u003C\u002Fcode\u003E) after the first \u003Ca href=\"https:\u002F\u002Ftopminisite.com\u002Fblog\u002Fhow-to-restore-in-fully-connected-layer-using\" target=\"_blank\"\u003Efully connected layer\u003C\u002Fa\u003E (\u003Ccode\u003Eself.fc1\u003C\u002Fcode\u003E). During training, as the mini-batches pass through this network, the batch normalization layer will normalize the activations.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003ENote: It is essential to ensure that the network is in training mode using \u003Ccode\u003Enet.train()\u003C\u002Fcode\u003E before training and in evaluation mode using \u003Ccode\u003Enet.eval()\u003C\u002Fcode\u003E during inference\u002Ftesting.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EYou can now use this network for training and inference in your PyTorch project, while enjoying the benefits of batch normalization.\u003C\u002Fp\u003E\n \u003Cdiv class=\"rating\"\u003E\n \u003Ch2\u003EBest PyTorch Books of December 2024\u003C\u002Fh2\u003E\n \u003Cdiv class=\"row mt-2\"\u003E\n \n \u003Cdiv class=\"col-12\"\u003E\n \u003Cdiv class=\"v-card elevation-6\"\u003E\n \u003Cdiv class=\"v-card__text rating-text\"\u003E\n \u003Cdiv class=\"rating-counter\"\u003E\n \u003Cspan class=\"v-badge\"\u003E\n \u003Cspan class=\"v-badge__wrapper\"\u003E\n \u003Cspan aria-atomic=\"true\" aria-label=\"Позиция\" class=\"v-badge__badge primary\"\u003E\n 1\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"row\"\u003E\n \u003Cdiv class=\"col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center\"\u003E\n \u003Cdiv\u003E\n \u003Cimg src=\"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com\u002Fimages\u002Fd7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae\u002Frating\u002F41jkoc6owal-sl160.jpg\" alt=\"PyTorch Recipes: A Problem-Solution Approach to Build, Train and Deploy Neural Network Models\" \u002F\u003E\n \u003Cp class=\"text-center font-weight-bold text-h6\"\u003ERating is 5 out of 5\u003C\u002Fp\u003E\n \u003Cdiv class=\"stars\" style=\"--rating: 5;\" aria-label=\"Rating is 5 out of 5\" \u003E\u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"col-lg-6 col-md-8 col-sm-6 col-12\"\u003E\n \u003Cp class=\"font-weight-bold rating-name\"\u003EPyTorch Recipes: A Problem-Solution Approach to Build, Train and Deploy Neural Network Models\u003C\u002Fp\u003E\n \n \n\n \n \n \u003C\u002Fdiv\u003E\n\n \u003Cdiv class=\"col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center\"\u003E\n \u003Cdiv class=\"text-center d-flex flex-column\"\u003E\n \n \u003Ca href=\"https:\u002F\u002Fgosrc.cc\u002Fgo\u002FMDdl-1Y4g\" target=\"_blank\" rel=\"nofollow noopener\" class=\"v-btn v-btn--rounded elevation-5 v-size--large success mb-2\"\u003E\n \u003Cspan class=\"v-btn__content\"\u003EGet Book Now\u003C\u002Fspan\u003E\n \u003C\u002Fa\u003E\n \n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \n \u003Cdiv class=\"col-12\"\u003E\n \u003Cdiv class=\"v-card elevation-6\"\u003E\n \u003Cdiv class=\"v-card__text rating-text\"\u003E\n \u003Cdiv class=\"rating-counter\"\u003E\n \u003Cspan class=\"v-badge\"\u003E\n \u003Cspan class=\"v-badge__wrapper\"\u003E\n \u003Cspan aria-atomic=\"true\" aria-label=\"Позиция\" class=\"v-badge__badge primary\"\u003E\n 2\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"row\"\u003E\n \u003Cdiv class=\"col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center\"\u003E\n \u003Cdiv\u003E\n \u003Cimg src=\"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com\u002Fimages\u002Fd7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae\u002Frating\u002F4198ez8dpbl-sl160.jpg\" alt=\"Mastering PyTorch: Build powerful deep learning architectures using advanced PyTorch features, 2nd Edition\" \u002F\u003E\n \u003Cp class=\"text-center font-weight-bold text-h6\"\u003ERating is 4.9 out of 5\u003C\u002Fp\u003E\n \u003Cdiv class=\"stars\" style=\"--rating: 4.9;\" aria-label=\"Rating is 4.9 out of 5\" \u003E\u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"col-lg-6 col-md-8 col-sm-6 col-12\"\u003E\n \u003Cp class=\"font-weight-bold rating-name\"\u003EMastering PyTorch: Build powerful deep learning architectures using advanced PyTorch features, 2nd Edition\u003C\u002Fp\u003E\n \n \n\n \n \n \u003C\u002Fdiv\u003E\n\n \u003Cdiv class=\"col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center\"\u003E\n \u003Cdiv class=\"text-center d-flex flex-column\"\u003E\n \n \u003Ca href=\"https:\u002F\u002Fgosrc.cc\u002Fgo\u002FUZp_a1L4R\" target=\"_blank\" rel=\"nofollow noopener\" class=\"v-btn v-btn--rounded elevation-5 v-size--large success mb-2\"\u003E\n \u003Cspan class=\"v-btn__content\"\u003EGet Book Now\u003C\u002Fspan\u003E\n \u003C\u002Fa\u003E\n \n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \n \u003Cdiv class=\"col-12\"\u003E\n \u003Cdiv class=\"v-card elevation-6\"\u003E\n \u003Cdiv class=\"v-card__text rating-text\"\u003E\n \u003Cdiv class=\"rating-counter\"\u003E\n \u003Cspan class=\"v-badge\"\u003E\n \u003Cspan class=\"v-badge__wrapper\"\u003E\n \u003Cspan aria-atomic=\"true\" aria-label=\"Позиция\" class=\"v-badge__badge primary\"\u003E\n 3\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"row\"\u003E\n \u003Cdiv class=\"col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center\"\u003E\n \u003Cdiv\u003E\n \u003Cimg src=\"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com\u002Fimages\u002Fd7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae\u002Frating\u002F513exc7qqjl-sl160.jpg\" alt=\"Natural Language Processing with PyTorch: Build Intelligent Language Applications Using Deep Learning\" \u002F\u003E\n \u003Cp class=\"text-center font-weight-bold text-h6\"\u003ERating is 4.8 out of 5\u003C\u002Fp\u003E\n \u003Cdiv class=\"stars\" style=\"--rating: 4.8;\" aria-label=\"Rating is 4.8 out of 5\" \u003E\u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"col-lg-6 col-md-8 col-sm-6 col-12\"\u003E\n \u003Cp class=\"font-weight-bold rating-name\"\u003ENatural Language Processing with PyTorch: Build Intelligent Language Applications Using Deep Learning\u003C\u002Fp\u003E\n \n \n\n \n \n \u003C\u002Fdiv\u003E\n\n \u003Cdiv class=\"col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center\"\u003E\n \u003Cdiv class=\"text-center d-flex flex-column\"\u003E\n \n \u003Ca href=\"https:\u002F\u002Fgosrc.cc\u002Fgo\u002FAQT_-1L4R\" target=\"_blank\" rel=\"nofollow noopener\" class=\"v-btn v-btn--rounded elevation-5 v-size--large success mb-2\"\u003E\n \u003Cspan class=\"v-btn__content\"\u003EGet Book Now\u003C\u002Fspan\u003E\n \u003C\u002Fa\u003E\n \n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \n \u003Cdiv class=\"col-12\"\u003E\n \u003Cdiv class=\"v-card elevation-6\"\u003E\n \u003Cdiv class=\"v-card__text rating-text\"\u003E\n \u003Cdiv class=\"rating-counter\"\u003E\n \u003Cspan class=\"v-badge\"\u003E\n \u003Cspan class=\"v-badge__wrapper\"\u003E\n \u003Cspan aria-atomic=\"true\" aria-label=\"Позиция\" class=\"v-badge__badge primary\"\u003E\n 4\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"row\"\u003E\n \u003Cdiv class=\"col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center\"\u003E\n \u003Cdiv\u003E\n \u003Cimg src=\"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com\u002Fimages\u002Fd7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae\u002Frating\u002F518jzrflrvl-sl160.jpg\" alt=\"Deep Learning for Coders with Fastai and PyTorch: AI Applications Without a PhD\" \u002F\u003E\n \u003Cp class=\"text-center font-weight-bold text-h6\"\u003ERating is 4.7 out of 5\u003C\u002Fp\u003E\n \u003Cdiv class=\"stars\" style=\"--rating: 4.7;\" aria-label=\"Rating is 4.7 out of 5\" \u003E\u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"col-lg-6 col-md-8 col-sm-6 col-12\"\u003E\n \u003Cp class=\"font-weight-bold rating-name\"\u003EDeep Learning for Coders with Fastai and PyTorch: AI Applications Without a PhD\u003C\u002Fp\u003E\n \n \n\n \n \n \u003C\u002Fdiv\u003E\n\n \u003Cdiv class=\"col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center\"\u003E\n \u003Cdiv class=\"text-center d-flex flex-column\"\u003E\n \n \u003Ca href=\"https:\u002F\u002Fgosrc.cc\u002Fgo\u002F7JalaJYVR\" target=\"_blank\" rel=\"nofollow noopener\" class=\"v-btn v-btn--rounded elevation-5 v-size--large success mb-2\"\u003E\n \u003Cspan class=\"v-btn__content\"\u003EGet Book Now\u003C\u002Fspan\u003E\n \u003C\u002Fa\u003E\n \n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \n \u003Cdiv class=\"col-12\"\u003E\n \u003Cdiv class=\"v-card elevation-6\"\u003E\n \u003Cdiv class=\"v-card__text rating-text\"\u003E\n \u003Cdiv class=\"rating-counter\"\u003E\n \u003Cspan class=\"v-badge\"\u003E\n \u003Cspan class=\"v-badge__wrapper\"\u003E\n \u003Cspan aria-atomic=\"true\" aria-label=\"Позиция\" class=\"v-badge__badge primary\"\u003E\n 5\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"row\"\u003E\n \u003Cdiv class=\"col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center\"\u003E\n \u003Cdiv\u003E\n \u003Cimg src=\"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com\u002Fimages\u002Fd7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae\u002Frating\u002F41o-uy9lzzl-sl160.jpg\" alt=\"Machine Learning with PyTorch and Scikit-Learn: Develop machine learning and deep learning models with Python\" \u002F\u003E\n \u003Cp class=\"text-center font-weight-bold text-h6\"\u003ERating is 4.6 out of 5\u003C\u002Fp\u003E\n \u003Cdiv class=\"stars\" style=\"--rating: 4.6;\" aria-label=\"Rating is 4.6 out of 5\" \u003E\u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"col-lg-6 col-md-8 col-sm-6 col-12\"\u003E\n \u003Cp class=\"font-weight-bold rating-name\"\u003EMachine Learning with PyTorch and Scikit-Learn: Develop machine learning and deep learning models with Python\u003C\u002Fp\u003E\n \n \n\n \n \n \u003C\u002Fdiv\u003E\n\n \u003Cdiv class=\"col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center\"\u003E\n \u003Cdiv class=\"text-center d-flex flex-column\"\u003E\n \n \u003Ca href=\"https:\u002F\u002Fgosrc.cc\u002Fgo\u002FGZs_a1Y4g\" target=\"_blank\" rel=\"nofollow noopener\" class=\"v-btn v-btn--rounded elevation-5 v-size--large success mb-2\"\u003E\n \u003Cspan class=\"v-btn__content\"\u003EGet Book Now\u003C\u002Fspan\u003E\n \u003C\u002Fa\u003E\n \n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \n \u003Cdiv class=\"col-12\"\u003E\n \u003Cdiv class=\"v-card elevation-6\"\u003E\n \u003Cdiv class=\"v-card__text rating-text\"\u003E\n \u003Cdiv class=\"rating-counter\"\u003E\n \u003Cspan class=\"v-badge\"\u003E\n \u003Cspan class=\"v-badge__wrapper\"\u003E\n \u003Cspan aria-atomic=\"true\" aria-label=\"Позиция\" class=\"v-badge__badge primary\"\u003E\n 6\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"row\"\u003E\n \u003Cdiv class=\"col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center\"\u003E\n \u003Cdiv\u003E\n \u003Cimg src=\"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com\u002Fimages\u002Fd7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae\u002Frating\u002F411ruzisg3l-sl160.jpg\" alt=\"Deep Learning with PyTorch: Build, train, and tune neural networks using Python tools\" \u002F\u003E\n \u003Cp class=\"text-center font-weight-bold text-h6\"\u003ERating is 4.5 out of 5\u003C\u002Fp\u003E\n \u003Cdiv class=\"stars\" style=\"--rating: 4.5;\" aria-label=\"Rating is 4.5 out of 5\" \u003E\u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"col-lg-6 col-md-8 col-sm-6 col-12\"\u003E\n \u003Cp class=\"font-weight-bold rating-name\"\u003EDeep Learning with PyTorch: Build, train, and tune neural networks using Python tools\u003C\u002Fp\u003E\n \n \n\n \n \n \u003C\u002Fdiv\u003E\n\n \u003Cdiv class=\"col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center\"\u003E\n \u003Cdiv class=\"text-center d-flex flex-column\"\u003E\n \n \u003Ca href=\"https:\u002F\u002Fgosrc.cc\u002Fgo\u002F8lwlaJY4R\" target=\"_blank\" rel=\"nofollow noopener\" class=\"v-btn v-btn--rounded elevation-5 v-size--large success mb-2\"\u003E\n \u003Cspan class=\"v-btn__content\"\u003EGet Book Now\u003C\u002Fspan\u003E\n \u003C\u002Fa\u003E\n \n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \n \u003Cdiv class=\"col-12\"\u003E\n \u003Cdiv class=\"v-card elevation-6\"\u003E\n \u003Cdiv class=\"v-card__text rating-text\"\u003E\n \u003Cdiv class=\"rating-counter\"\u003E\n \u003Cspan class=\"v-badge\"\u003E\n \u003Cspan class=\"v-badge__wrapper\"\u003E\n \u003Cspan aria-atomic=\"true\" aria-label=\"Позиция\" class=\"v-badge__badge primary\"\u003E\n 7\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"row\"\u003E\n \u003Cdiv class=\"col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center\"\u003E\n \u003Cdiv\u003E\n \u003Cimg src=\"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com\u002Fimages\u002Fd7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae\u002Frating\u002F41sz-tftqpl-sl160.jpg\" alt=\"Programming PyTorch for Deep Learning: Creating and Deploying Deep Learning Applications\" \u002F\u003E\n \u003Cp class=\"text-center font-weight-bold text-h6\"\u003ERating is 4.4 out of 5\u003C\u002Fp\u003E\n \u003Cdiv class=\"stars\" style=\"--rating: 4.4;\" aria-label=\"Rating is 4.4 out of 5\" \u003E\u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"col-lg-6 col-md-8 col-sm-6 col-12\"\u003E\n \u003Cp class=\"font-weight-bold rating-name\"\u003EProgramming PyTorch for Deep Learning: Creating and Deploying Deep Learning Applications\u003C\u002Fp\u003E\n \n \n\n \n \n \u003C\u002Fdiv\u003E\n\n \u003Cdiv class=\"col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center\"\u003E\n \u003Cdiv class=\"text-center d-flex flex-column\"\u003E\n \n \u003Ca href=\"https:\u002F\u002Fgosrc.cc\u002Fgo\u002F1d9_-JY4R\" target=\"_blank\" rel=\"nofollow noopener\" class=\"v-btn v-btn--rounded elevation-5 v-size--large success mb-2\"\u003E\n \u003Cspan class=\"v-btn__content\"\u003EGet Book Now\u003C\u002Fspan\u003E\n \u003C\u002Fa\u003E\n \n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \n \u003Cdiv class=\"col-12\"\u003E\n \u003Cdiv class=\"v-card elevation-6\"\u003E\n \u003Cdiv class=\"v-card__text rating-text\"\u003E\n \u003Cdiv class=\"rating-counter\"\u003E\n \u003Cspan class=\"v-badge\"\u003E\n \u003Cspan class=\"v-badge__wrapper\"\u003E\n \u003Cspan aria-atomic=\"true\" aria-label=\"Позиция\" class=\"v-badge__badge primary\"\u003E\n 8\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"row\"\u003E\n \u003Cdiv class=\"col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center\"\u003E\n \u003Cdiv\u003E\n \u003Cimg src=\"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com\u002Fimages\u002Fd7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae\u002Frating\u002F41xz0sfhxsl-sl160.jpg\" alt=\"PyTorch Pocket Reference: Building and Deploying Deep Learning Models\" \u002F\u003E\n \u003Cp class=\"text-center font-weight-bold text-h6\"\u003ERating is 4.3 out of 5\u003C\u002Fp\u003E\n \u003Cdiv class=\"stars\" style=\"--rating: 4.3;\" aria-label=\"Rating is 4.3 out of 5\" \u003E\u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"col-lg-6 col-md-8 col-sm-6 col-12\"\u003E\n \u003Cp class=\"font-weight-bold rating-name\"\u003EPyTorch Pocket Reference: Building and Deploying Deep Learning Models\u003C\u002Fp\u003E\n \n \n\n \n \n \u003C\u002Fdiv\u003E\n\n \u003Cdiv class=\"col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center\"\u003E\n \u003Cdiv class=\"text-center d-flex flex-column\"\u003E\n \n \u003Ca href=\"https:\u002F\u002Fgosrc.cc\u002Fgo\u002Fshe_-1YVg\" target=\"_blank\" rel=\"nofollow noopener\" class=\"v-btn v-btn--rounded elevation-5 v-size--large success mb-2\"\u003E\n \u003Cspan class=\"v-btn__content\"\u003EGet Book Now\u003C\u002Fspan\u003E\n \u003C\u002Fa\u003E\n \n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \n \u003Cdiv class=\"col-12\"\u003E\n \u003Cdiv class=\"v-card elevation-6\"\u003E\n \u003Cdiv class=\"v-card__text rating-text\"\u003E\n \u003Cdiv class=\"rating-counter\"\u003E\n \u003Cspan class=\"v-badge\"\u003E\n \u003Cspan class=\"v-badge__wrapper\"\u003E\n \u003Cspan aria-atomic=\"true\" aria-label=\"Позиция\" class=\"v-badge__badge primary\"\u003E\n 9\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"row\"\u003E\n \u003Cdiv class=\"col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center\"\u003E\n \u003Cdiv\u003E\n \u003Cimg src=\"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com\u002Fimages\u002Fd7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae\u002Frating\u002F41ayiicwf2l-sl160.jpg\" alt=\"Deep Learning with PyTorch Lightning: Swiftly build high-performance Artificial Intelligence (AI) models using Python\" \u002F\u003E\n \u003Cp class=\"text-center font-weight-bold text-h6\"\u003ERating is 4.2 out of 5\u003C\u002Fp\u003E\n \u003Cdiv class=\"stars\" style=\"--rating: 4.2;\" aria-label=\"Rating is 4.2 out of 5\" \u003E\u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"col-lg-6 col-md-8 col-sm-6 col-12\"\u003E\n \u003Cp class=\"font-weight-bold rating-name\"\u003EDeep Learning with PyTorch Lightning: Swiftly build high-performance Artificial Intelligence (AI) models using Python\u003C\u002Fp\u003E\n \n \n\n \n \n \u003C\u002Fdiv\u003E\n\n \u003Cdiv class=\"col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center\"\u003E\n \u003Cdiv class=\"text-center d-flex flex-column\"\u003E\n \n \u003Ca href=\"https:\u002F\u002Fgosrc.cc\u002Fgo\u002FkmiX-1LVg\" target=\"_blank\" rel=\"nofollow noopener\" class=\"v-btn v-btn--rounded elevation-5 v-size--large success mb-2\"\u003E\n \u003Cspan class=\"v-btn__content\"\u003EGet Book Now\u003C\u002Fspan\u003E\n \u003C\u002Fa\u003E\n \n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Ch2\u003EHow to implement batch normalization in a custom PyTorch model?\u003C\u002Fh2\u003E\u003Cp\u003ETo implement batch normalization in a custom PyTorch model, you can follow these steps:\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003EImport the required modules:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cdiv style=\"color:#f8f8f2;background-color:#272822;\"\u003E\n\u003Ctable style=\"border-spacing:0;padding:0;margin:0;border:0;\"\u003E\u003Ctbody\u003E\u003Ctr\u003E\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E1\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E2\n\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\n\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;;width:100%\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003Eimport torch\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003Eimport torch.nn as nn\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\u003C\u002Ftr\u003E\u003C\u002Ftbody\u003E\u003C\u002Ftable\u003E\n\u003C\u002Fdiv\u003E\n\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003EDefine a basic custom model class:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cdiv style=\"color:#f8f8f2;background-color:#272822;\"\u003E\n\u003Ctable style=\"border-spacing:0;padding:0;margin:0;border:0;\"\u003E\u003Ctbody\u003E\u003Ctr\u003E\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E1\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E2\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E3\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E4\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E5\n\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\n\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;;width:100%\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003Eclass CustomModel(nn.Module):\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E def __init__(self):\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E super(CustomModel, self).__init__()\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E self.fc1 = nn.Linear(in_features, hidden_units)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E self.fc2 = nn.Linear(hidden_units, out_features)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\u003C\u002Ftr\u003E\u003C\u002Ftbody\u003E\u003C\u002Ftable\u003E\n\u003C\u002Fdiv\u003E\n\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EReplace \u003Ccode\u003Ein_features\u003C\u002Fcode\u003E, \u003Ccode\u003Ehidden_units\u003C\u002Fcode\u003E, and \u003Ccode\u003Eout_features\u003C\u002Fcode\u003E with appropriate values for your model architecture.\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003EAdd batch normalization layers and their parameters to the model:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cdiv style=\"color:#f8f8f2;background-color:#272822;\"\u003E\n\u003Ctable style=\"border-spacing:0;padding:0;margin:0;border:0;\"\u003E\u003Ctbody\u003E\u003Ctr\u003E\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E1\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E2\n\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\n\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;;width:100%\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003Eself.bn1 = nn.BatchNorm1d(hidden_units)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003Eself.bn2 = nn.BatchNorm1d(out_features)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\u003C\u002Ftr\u003E\u003C\u002Ftbody\u003E\u003C\u002Ftable\u003E\n\u003C\u002Fdiv\u003E\n\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EAdjust the parameter value based on your model architecture.\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003EDefine the forward pass of the model:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cdiv style=\"color:#f8f8f2;background-color:#272822;\"\u003E\n\u003Ctable style=\"border-spacing:0;padding:0;margin:0;border:0;\"\u003E\u003Ctbody\u003E\u003Ctr\u003E\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E1\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E2\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E3\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E4\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E5\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E6\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E7\n\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\n\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;;width:100%\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003Edef forward(self, x):\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E x = self.fc1(x)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E x = self.bn1(x)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E x = nn.functional.relu(x)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E x = self.fc2(x)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E x = self.bn2(x)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E return x\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\u003C\u002Ftr\u003E\u003C\u002Ftbody\u003E\u003C\u002Ftable\u003E\n\u003C\u002Fdiv\u003E\n\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EThis example assumes the ReLU activation function, but you can replace it with any activation function you prefer.\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003ECreate an instance of the custom model:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cdiv style=\"color:#f8f8f2;background-color:#272822;\"\u003E\n\u003Ctable style=\"border-spacing:0;padding:0;margin:0;border:0;\"\u003E\u003Ctbody\u003E\u003Ctr\u003E\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E1\n\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\n\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;;width:100%\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003Emodel = CustomModel()\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\u003C\u002Ftr\u003E\u003C\u002Ftbody\u003E\u003C\u002Ftable\u003E\n\u003C\u002Fdiv\u003E\n\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003ENow you have implemented batch normalization in your custom PyTorch model.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Ch2\u003EWhat are the advantages of using batch normalization in PyTorch?\u003C\u002Fh2\u003E\u003Cp\u003EBatch normalization is a regularization technique that is widely used in deep learning models. When applied to PyTorch models, it provides several advantages:\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003E\u003Cstrong\u003EImproved convergence\u003C\u002Fstrong\u003E: Batch normalization normalizes the input to each neuron across a mini-batch, which helps in stabilizing the learning process. This leads to faster convergence and reduces the number of epochs required for training.\n\u003C\u002Fli\u003E\u003Cli\u003E\u003Cstrong\u003EReduced overfitting\u003C\u002Fstrong\u003E: By normalizing the inputs, batch normalization reduces the dependence of each neuron on the other neurons in the network. This reduces the chances of overfitting and improves the generalization ability of the model.\n\u003C\u002Fli\u003E\u003Cli\u003E\u003Cstrong\u003EIncreased learning rate\u003C\u002Fstrong\u003E: Batch normalization reduces the internal covariate shift by maintaining zero mean and unit variance activations. This enables the use of higher learning rates during training, which can speed up the training process.\n\u003C\u002Fli\u003E\u003Cli\u003E\u003Cstrong\u003EBetter gradient flow\u003C\u002Fstrong\u003E: Normalizing the inputs using batch normalization helps in ensuring that the gradients flow smoothly and consistently during backpropagation. This helps combat the vanishing and exploding gradient problems, making it easier to train deep networks.\n\u003C\u002Fli\u003E\u003Cli\u003E\u003Cstrong\u003ERobustness to different input distributions\u003C\u002Fstrong\u003E: Batch normalization makes the model less sensitive to the \u003Ca href=\"https:\u002F\u002Fstudentprojectcode.com\u002Fblog\u002Fhow-to-simplify-units-with-different-scales-in\" target=\"_blank\"\u003Escale\u003C\u002Fa\u003E and distribution of the input data. This allows the model to perform well even when faced with inputs that are significantly different from the training data.\n\u003C\u002Fli\u003E\u003Cli\u003E\u003Cstrong\u003EWeight initialization flexibility\u003C\u002Fstrong\u003E: Batch normalization helps in reducing the dependence of the model's performance on the choice of weight initialization. It allows the use of simpler initialization methods like random or small weights, which can speed up the training process.\n\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EOverall, batch normalization is a useful tool for improving the performance and stability of deep learning models in PyTorch, leading to faster convergence, better generalization, and increased robustness.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Ch2\u003EWhat is the effect of batch size on batch normalization in PyTorch?\u003C\u002Fh2\u003E\u003Cp\u003EThe batch size affects the batch normalization in PyTorch in the following way:\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003E\u003Cstrong\u003EStatistics estimation\u003C\u002Fstrong\u003E: Batch normalization relies on estimating the mean and variance of the input data to normalize it. With a larger batch size, there is more data available for statistics estimation, leading to more accurate estimates of the mean and variance. This can result in improved normalization and consequently, better performance.\n\u003C\u002Fli\u003E\u003Cli\u003E\u003Cstrong\u003ENoise reduction\u003C\u002Fstrong\u003E: Batch normalization introduces some noise to the statistics estimation process. With a larger batch size, the noise is averaged out more effectively, resulting in more stable estimates of mean and variance. This can lead to reduced overfitting and improved generalization.\n\u003C\u002Fli\u003E\u003Cli\u003E\u003Cstrong\u003ETraining dynamics\u003C\u002Fstrong\u003E: Smaller batch sizes tend to introduce more stochasticity and randomness in the training process, as each batch's statistics differ significantly. On the other hand, larger batch sizes provide more consistent statistics, which can affect the optimization process. This can result in different training dynamics, such as convergence speed and stability.\n\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EIt's important to note that the choice of batch size is often a trade-off. Larger batch sizes require more memory, may limit parallelization, and increase computational requirements. However, they can offer better normalization and estimation, while smaller batch sizes may introduce more noise but can be computationally more efficient.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Ch2\u003EWhat are the requirements for using batch normalization in PyTorch?\u003C\u002Fh2\u003E\u003Cp\u003ETo use batch normalization in PyTorch, the following requirements should be met:\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003E\u003Cstrong\u003EPyTorch should be installed on the system. You can install it using pip\u003C\u002Fstrong\u003E: pip install torch.\n\u003C\u002Fli\u003E\u003Cli\u003EImport the necessary modules:\n\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cdiv style=\"color:#f8f8f2;background-color:#272822;\"\u003E\n\u003Ctable style=\"border-spacing:0;padding:0;margin:0;border:0;\"\u003E\u003Ctbody\u003E\u003Ctr\u003E\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E1\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E2\n\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\n\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;;width:100%\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003Eimport torch\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003Eimport torch.nn as nn\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\u003C\u002Ftr\u003E\u003C\u002Ftbody\u003E\u003C\u002Ftable\u003E\n\u003C\u002Fdiv\u003E\n\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003EDefine your model architecture using the nn.Module class. Use the torch.nn.BatchNorm2d or torch.nn.BatchNorm1d layer (based on your input dimensions) for batch normalization.\n\u003C\u002Fli\u003E\u003Cli\u003EUse batch normalization layer after the convolutional or linear layer in your model architecture. For example:\n\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cdiv style=\"color:#f8f8f2;background-color:#272822;\"\u003E\n\u003Ctable style=\"border-spacing:0;padding:0;margin:0;border:0;\"\u003E\u003Ctbody\u003E\u003Ctr\u003E\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E1\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E2\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E3\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E4\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E5\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E6\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E7\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E8\n\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\n\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;;width:100%\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003Eclass MyModel(nn.Module):\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E def __init__(self):\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E super(MyModel, self).__init__()\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E self.conv1 = nn.Conv2d(3, 64, kernel_size=3)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E self.bn1 = nn.BatchNorm2d(64)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E self.fc1 = nn.Linear(64, 10)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E self.bn2 = nn.BatchNorm1d(10)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E ...\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\u003C\u002Ftr\u003E\u003C\u002Ftbody\u003E\u003C\u002Ftable\u003E\n\u003C\u002Fdiv\u003E\n\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003EDuring the forward pass, apply batch normalization to the input tensor. For example:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cdiv style=\"color:#f8f8f2;background-color:#272822;\"\u003E\n\u003Ctable style=\"border-spacing:0;padding:0;margin:0;border:0;\"\u003E\u003Ctbody\u003E\u003Ctr\u003E\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E1\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E2\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E3\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E4\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E5\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E6\n\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\n\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;;width:100%\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003Edef forward(self, x):\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E x = self.conv1(x)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E x = self.bn1(x)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E x = self.fc1(x)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E x = self.bn2(x)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E ...\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\u003C\u002Ftr\u003E\u003C\u002Ftbody\u003E\u003C\u002Ftable\u003E\n\u003C\u002Fdiv\u003E\n\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003ENote: Batch normalization is typically used before the activation function, but the order can vary depending on your problem and experiment settings.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Ch2\u003EWhat is the impact of batch normalization on model generalization in PyTorch?\u003C\u002Fh2\u003E\u003Cp\u003EBatch normalization has a significant impact on model generalization in PyTorch. It helps to improve the generalization capability of neural networks by reducing the internal covariate shift.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EInternal covariate shift refers to the change in the distribution of network activations due to the change in parameter values during training. This can slow down the training process and hinder the performance of the model.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EBatch normalization solves this problem by normalizing the output of each layer using the mean and variance of the mini-batch. By doing so, it reduces the effect of the internal covariate shift and makes the optimization process more stable. Batch normalization also introduces additional trainable parameters, which allow the network to adaptively scale and shift the normalized values.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EThe normalization of inputs helps in the generalization of the model because it keeps the values within a reasonable range. It prevents extreme values from causing instability in the network, which can lead to overfitting. Additionally, batch normalization acts as a regularizer, reducing the need for other regularization techniques like dropout.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EOverall, batch normalization in PyTorch improves the generalization ability of models by reducing internal covariate shift, making the training process more stable, and acting as a regularizer.\u003C\u002Fp\u003E",formatted_content_ad:"\u003Cp\u003EBatch normalization is a widely used technique for \u003Ca href=\"https:\u002F\u002Fstlplaces.com\u002Fblog\u002Fhow-to-handle-overfitting-in-tensorflow-models\"\u003Eimproving the training of deep neural networks\u003C\u002Fa\u003E. It normalizes the activations of each mini-batch by subtracting the mini-batch mean and dividing by the mini-batch standard deviation. This helps in reducing internal covariate shift by ensuring that the input to each layer is normalized.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EImplementing batch normalization in PyTorch is straightforward. Here are the steps:\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003EImport the necessary libraries:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cdiv style=\"color:#f8f8f2;background-color:#272822;\"\u003E\n\u003Ctable style=\"border-spacing:0;padding:0;margin:0;border:0;\"\u003E\u003Ctbody\u003E\u003Ctr\u003E\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E1\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E2\n\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\n\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;;width:100%\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003Eimport torch\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003Eimport torch.nn as nn\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\u003C\u002Ftr\u003E\u003C\u002Ftbody\u003E\u003C\u002Ftable\u003E\n\u003C\u002Fdiv\u003E\n\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003EDefine a custom neural network architecture:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cdiv style=\"color:#f8f8f2;background-color:#272822;\"\u003E\n\u003Ctable style=\"border-spacing:0;padding:0;margin:0;border:0;\"\u003E\u003Ctbody\u003E\u003Ctr\u003E\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E1\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E2\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E3\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E4\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E5\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E6\n\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\n\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;;width:100%\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003Eclass Net(nn.Module):\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E def __init__(self):\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E super(Net, self).__init__()\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E self.fc1 = nn.Linear(10, 20)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E self.bn1 = nn.BatchNorm1d(20) # Batch normalization layer\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E self.fc2 = nn.Linear(20, 10)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\u003C\u002Ftr\u003E\u003C\u002Ftbody\u003E\u003C\u002Ftable\u003E\n\u003C\u002Fdiv\u003E\n\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003EOverride the forward method of the neural network:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cdiv style=\"color:#f8f8f2;background-color:#272822;\"\u003E\n\u003Ctable style=\"border-spacing:0;padding:0;margin:0;border:0;\"\u003E\u003Ctbody\u003E\u003Ctr\u003E\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E1\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E2\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E3\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E4\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E5\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E6\n\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\n\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;;width:100%\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E def forward(self, x):\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E x = self.fc1(x)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E x = self.bn1(x)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E x = torch.relu(x)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E x = self.fc2(x)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E return x\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\u003C\u002Ftr\u003E\u003C\u002Ftbody\u003E\u003C\u002Ftable\u003E\n\u003C\u002Fdiv\u003E\n\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003ECreate an instance of the network:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cdiv style=\"color:#f8f8f2;background-color:#272822;\"\u003E\n\u003Ctable style=\"border-spacing:0;padding:0;margin:0;border:0;\"\u003E\u003Ctbody\u003E\u003Ctr\u003E\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E1\n\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\n\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;;width:100%\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003Enet = Net()\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\u003C\u002Ftr\u003E\u003C\u002Ftbody\u003E\u003C\u002Ftable\u003E\n\u003C\u002Fdiv\u003E\n\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EThat's it! Now the network \u003Ccode\u003Enet\u003C\u002Fcode\u003E includes a batch normalization layer (\u003Ccode\u003Eself.bn1\u003C\u002Fcode\u003E) after the first \u003Ca href=\"https:\u002F\u002Ftopminisite.com\u002Fblog\u002Fhow-to-restore-in-fully-connected-layer-using\" target=\"_blank\"\u003Efully connected layer\u003C\u002Fa\u003E (\u003Ccode\u003Eself.fc1\u003C\u002Fcode\u003E). During training, as the mini-batches pass through this network, the batch normalization layer will normalize the activations.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003ENote: It is essential to ensure that the network is in training mode using \u003Ccode\u003Enet.train()\u003C\u002Fcode\u003E before training and in evaluation mode using \u003Ccode\u003Enet.eval()\u003C\u002Fcode\u003E during inference\u002Ftesting.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EYou can now use this network for training and inference in your PyTorch project, while enjoying the benefits of batch normalization.\u003C\u002Fp\u003E\n \u003Cdiv class=\"rating\"\u003E\n \u003Ch2\u003EBest PyTorch Books of December 2024\u003C\u002Fh2\u003E\n \u003Cdiv class=\"row mt-2\"\u003E\n \n \u003Cdiv class=\"col-12\"\u003E\n \u003Cdiv class=\"v-card elevation-6\"\u003E\n \u003Cdiv class=\"v-card__text rating-text\"\u003E\n \u003Cdiv class=\"rating-counter\"\u003E\n \u003Cspan class=\"v-badge\"\u003E\n \u003Cspan class=\"v-badge__wrapper\"\u003E\n \u003Cspan aria-atomic=\"true\" aria-label=\"Позиция\" class=\"v-badge__badge primary\"\u003E\n 1\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"row\"\u003E\n \u003Cdiv class=\"col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center\"\u003E\n \u003Cdiv\u003E\n \u003Cimg src=\"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com\u002Fimages\u002Fd7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae\u002Frating\u002F41jkoc6owal-sl160.jpg\" alt=\"PyTorch Recipes: A Problem-Solution Approach to Build, Train and Deploy Neural Network Models\" \u002F\u003E\n \u003Cp class=\"text-center font-weight-bold text-h6\"\u003ERating is 5 out of 5\u003C\u002Fp\u003E\n \u003Cdiv class=\"stars\" style=\"--rating: 5;\" aria-label=\"Rating is 5 out of 5\" \u003E\u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"col-lg-6 col-md-8 col-sm-6 col-12\"\u003E\n \u003Cp class=\"font-weight-bold rating-name\"\u003EPyTorch Recipes: A Problem-Solution Approach to Build, Train and Deploy Neural Network Models\u003C\u002Fp\u003E\n \n \n\n \n \n \u003C\u002Fdiv\u003E\n\n \u003Cdiv class=\"col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center\"\u003E\n \u003Cdiv class=\"text-center d-flex flex-column\"\u003E\n \n \u003Ca href=\"https:\u002F\u002Fgosrc.cc\u002Fgo\u002FMDdl-1Y4g\" target=\"_blank\" rel=\"nofollow noopener\" class=\"v-btn v-btn--rounded elevation-5 v-size--large success mb-2\"\u003E\n \u003Cspan class=\"v-btn__content\"\u003EGet Book Now\u003C\u002Fspan\u003E\n \u003C\u002Fa\u003E\n \n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \n \u003Cdiv class=\"col-12\"\u003E\n \u003Cdiv class=\"v-card elevation-6\"\u003E\n \u003Cdiv class=\"v-card__text rating-text\"\u003E\n \u003Cdiv class=\"rating-counter\"\u003E\n \u003Cspan class=\"v-badge\"\u003E\n \u003Cspan class=\"v-badge__wrapper\"\u003E\n \u003Cspan aria-atomic=\"true\" aria-label=\"Позиция\" class=\"v-badge__badge primary\"\u003E\n 2\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"row\"\u003E\n \u003Cdiv class=\"col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center\"\u003E\n \u003Cdiv\u003E\n \u003Cimg src=\"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com\u002Fimages\u002Fd7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae\u002Frating\u002F4198ez8dpbl-sl160.jpg\" alt=\"Mastering PyTorch: Build powerful deep learning architectures using advanced PyTorch features, 2nd Edition\" \u002F\u003E\n \u003Cp class=\"text-center font-weight-bold text-h6\"\u003ERating is 4.9 out of 5\u003C\u002Fp\u003E\n \u003Cdiv class=\"stars\" style=\"--rating: 4.9;\" aria-label=\"Rating is 4.9 out of 5\" \u003E\u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"col-lg-6 col-md-8 col-sm-6 col-12\"\u003E\n \u003Cp class=\"font-weight-bold rating-name\"\u003EMastering PyTorch: Build powerful deep learning architectures using advanced PyTorch features, 2nd Edition\u003C\u002Fp\u003E\n \n \n\n \n \n \u003C\u002Fdiv\u003E\n\n \u003Cdiv class=\"col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center\"\u003E\n \u003Cdiv class=\"text-center d-flex flex-column\"\u003E\n \n \u003Ca href=\"https:\u002F\u002Fgosrc.cc\u002Fgo\u002FUZp_a1L4R\" target=\"_blank\" rel=\"nofollow noopener\" class=\"v-btn v-btn--rounded elevation-5 v-size--large success mb-2\"\u003E\n \u003Cspan class=\"v-btn__content\"\u003EGet Book Now\u003C\u002Fspan\u003E\n \u003C\u002Fa\u003E\n \n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \n \u003Cdiv class=\"col-12\"\u003E\n \u003Cdiv class=\"v-card elevation-6\"\u003E\n \u003Cdiv class=\"v-card__text rating-text\"\u003E\n \u003Cdiv class=\"rating-counter\"\u003E\n \u003Cspan class=\"v-badge\"\u003E\n \u003Cspan class=\"v-badge__wrapper\"\u003E\n \u003Cspan aria-atomic=\"true\" aria-label=\"Позиция\" class=\"v-badge__badge primary\"\u003E\n 3\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"row\"\u003E\n \u003Cdiv class=\"col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center\"\u003E\n \u003Cdiv\u003E\n \u003Cimg src=\"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com\u002Fimages\u002Fd7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae\u002Frating\u002F513exc7qqjl-sl160.jpg\" alt=\"Natural Language Processing with PyTorch: Build Intelligent Language Applications Using Deep Learning\" \u002F\u003E\n \u003Cp class=\"text-center font-weight-bold text-h6\"\u003ERating is 4.8 out of 5\u003C\u002Fp\u003E\n \u003Cdiv class=\"stars\" style=\"--rating: 4.8;\" aria-label=\"Rating is 4.8 out of 5\" \u003E\u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"col-lg-6 col-md-8 col-sm-6 col-12\"\u003E\n \u003Cp class=\"font-weight-bold rating-name\"\u003ENatural Language Processing with PyTorch: Build Intelligent Language Applications Using Deep Learning\u003C\u002Fp\u003E\n \n \n\n \n \n \u003C\u002Fdiv\u003E\n\n \u003Cdiv class=\"col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center\"\u003E\n \u003Cdiv class=\"text-center d-flex flex-column\"\u003E\n \n \u003Ca href=\"https:\u002F\u002Fgosrc.cc\u002Fgo\u002FAQT_-1L4R\" target=\"_blank\" rel=\"nofollow noopener\" class=\"v-btn v-btn--rounded elevation-5 v-size--large success mb-2\"\u003E\n \u003Cspan class=\"v-btn__content\"\u003EGet Book Now\u003C\u002Fspan\u003E\n \u003C\u002Fa\u003E\n \n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \n \u003Cdiv class=\"col-12\"\u003E\n \u003Cdiv class=\"v-card elevation-6\"\u003E\n \u003Cdiv class=\"v-card__text rating-text\"\u003E\n \u003Cdiv class=\"rating-counter\"\u003E\n \u003Cspan class=\"v-badge\"\u003E\n \u003Cspan class=\"v-badge__wrapper\"\u003E\n \u003Cspan aria-atomic=\"true\" aria-label=\"Позиция\" class=\"v-badge__badge primary\"\u003E\n 4\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"row\"\u003E\n \u003Cdiv class=\"col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center\"\u003E\n \u003Cdiv\u003E\n \u003Cimg src=\"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com\u002Fimages\u002Fd7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae\u002Frating\u002F518jzrflrvl-sl160.jpg\" alt=\"Deep Learning for Coders with Fastai and PyTorch: AI Applications Without a PhD\" \u002F\u003E\n \u003Cp class=\"text-center font-weight-bold text-h6\"\u003ERating is 4.7 out of 5\u003C\u002Fp\u003E\n \u003Cdiv class=\"stars\" style=\"--rating: 4.7;\" aria-label=\"Rating is 4.7 out of 5\" \u003E\u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"col-lg-6 col-md-8 col-sm-6 col-12\"\u003E\n \u003Cp class=\"font-weight-bold rating-name\"\u003EDeep Learning for Coders with Fastai and PyTorch: AI Applications Without a PhD\u003C\u002Fp\u003E\n \n \n\n \n \n \u003C\u002Fdiv\u003E\n\n \u003Cdiv class=\"col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center\"\u003E\n \u003Cdiv class=\"text-center d-flex flex-column\"\u003E\n \n \u003Ca href=\"https:\u002F\u002Fgosrc.cc\u002Fgo\u002F7JalaJYVR\" target=\"_blank\" rel=\"nofollow noopener\" class=\"v-btn v-btn--rounded elevation-5 v-size--large success mb-2\"\u003E\n \u003Cspan class=\"v-btn__content\"\u003EGet Book Now\u003C\u002Fspan\u003E\n \u003C\u002Fa\u003E\n \n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \n \u003Cdiv class=\"col-12\"\u003E\n \u003Cdiv class=\"v-card elevation-6\"\u003E\n \u003Cdiv class=\"v-card__text rating-text\"\u003E\n \u003Cdiv class=\"rating-counter\"\u003E\n \u003Cspan class=\"v-badge\"\u003E\n \u003Cspan class=\"v-badge__wrapper\"\u003E\n \u003Cspan aria-atomic=\"true\" aria-label=\"Позиция\" class=\"v-badge__badge primary\"\u003E\n 5\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"row\"\u003E\n \u003Cdiv class=\"col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center\"\u003E\n \u003Cdiv\u003E\n \u003Cimg src=\"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com\u002Fimages\u002Fd7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae\u002Frating\u002F41o-uy9lzzl-sl160.jpg\" alt=\"Machine Learning with PyTorch and Scikit-Learn: Develop machine learning and deep learning models with Python\" \u002F\u003E\n \u003Cp class=\"text-center font-weight-bold text-h6\"\u003ERating is 4.6 out of 5\u003C\u002Fp\u003E\n \u003Cdiv class=\"stars\" style=\"--rating: 4.6;\" aria-label=\"Rating is 4.6 out of 5\" \u003E\u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"col-lg-6 col-md-8 col-sm-6 col-12\"\u003E\n \u003Cp class=\"font-weight-bold rating-name\"\u003EMachine Learning with PyTorch and Scikit-Learn: Develop machine learning and deep learning models with Python\u003C\u002Fp\u003E\n \n \n\n \n \n \u003C\u002Fdiv\u003E\n\n \u003Cdiv class=\"col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center\"\u003E\n \u003Cdiv class=\"text-center d-flex flex-column\"\u003E\n \n \u003Ca href=\"https:\u002F\u002Fgosrc.cc\u002Fgo\u002FGZs_a1Y4g\" target=\"_blank\" rel=\"nofollow noopener\" class=\"v-btn v-btn--rounded elevation-5 v-size--large success mb-2\"\u003E\n \u003Cspan class=\"v-btn__content\"\u003EGet Book Now\u003C\u002Fspan\u003E\n \u003C\u002Fa\u003E\n \n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \n \u003Cdiv class=\"col-12\"\u003E\n \u003Cdiv class=\"v-card elevation-6\"\u003E\n \u003Cdiv class=\"v-card__text rating-text\"\u003E\n \u003Cdiv class=\"rating-counter\"\u003E\n \u003Cspan class=\"v-badge\"\u003E\n \u003Cspan class=\"v-badge__wrapper\"\u003E\n \u003Cspan aria-atomic=\"true\" aria-label=\"Позиция\" class=\"v-badge__badge primary\"\u003E\n 6\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"row\"\u003E\n \u003Cdiv class=\"col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center\"\u003E\n \u003Cdiv\u003E\n \u003Cimg src=\"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com\u002Fimages\u002Fd7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae\u002Frating\u002F411ruzisg3l-sl160.jpg\" alt=\"Deep Learning with PyTorch: Build, train, and tune neural networks using Python tools\" \u002F\u003E\n \u003Cp class=\"text-center font-weight-bold text-h6\"\u003ERating is 4.5 out of 5\u003C\u002Fp\u003E\n \u003Cdiv class=\"stars\" style=\"--rating: 4.5;\" aria-label=\"Rating is 4.5 out of 5\" \u003E\u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"col-lg-6 col-md-8 col-sm-6 col-12\"\u003E\n \u003Cp class=\"font-weight-bold rating-name\"\u003EDeep Learning with PyTorch: Build, train, and tune neural networks using Python tools\u003C\u002Fp\u003E\n \n \n\n \n \n \u003C\u002Fdiv\u003E\n\n \u003Cdiv class=\"col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center\"\u003E\n \u003Cdiv class=\"text-center d-flex flex-column\"\u003E\n \n \u003Ca href=\"https:\u002F\u002Fgosrc.cc\u002Fgo\u002F8lwlaJY4R\" target=\"_blank\" rel=\"nofollow noopener\" class=\"v-btn v-btn--rounded elevation-5 v-size--large success mb-2\"\u003E\n \u003Cspan class=\"v-btn__content\"\u003EGet Book Now\u003C\u002Fspan\u003E\n \u003C\u002Fa\u003E\n \n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \n \u003Cdiv class=\"col-12\"\u003E\n \u003Cdiv class=\"v-card elevation-6\"\u003E\n \u003Cdiv class=\"v-card__text rating-text\"\u003E\n \u003Cdiv class=\"rating-counter\"\u003E\n \u003Cspan class=\"v-badge\"\u003E\n \u003Cspan class=\"v-badge__wrapper\"\u003E\n \u003Cspan aria-atomic=\"true\" aria-label=\"Позиция\" class=\"v-badge__badge primary\"\u003E\n 7\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"row\"\u003E\n \u003Cdiv class=\"col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center\"\u003E\n \u003Cdiv\u003E\n \u003Cimg src=\"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com\u002Fimages\u002Fd7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae\u002Frating\u002F41sz-tftqpl-sl160.jpg\" alt=\"Programming PyTorch for Deep Learning: Creating and Deploying Deep Learning Applications\" \u002F\u003E\n \u003Cp class=\"text-center font-weight-bold text-h6\"\u003ERating is 4.4 out of 5\u003C\u002Fp\u003E\n \u003Cdiv class=\"stars\" style=\"--rating: 4.4;\" aria-label=\"Rating is 4.4 out of 5\" \u003E\u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"col-lg-6 col-md-8 col-sm-6 col-12\"\u003E\n \u003Cp class=\"font-weight-bold rating-name\"\u003EProgramming PyTorch for Deep Learning: Creating and Deploying Deep Learning Applications\u003C\u002Fp\u003E\n \n \n\n \n \n \u003C\u002Fdiv\u003E\n\n \u003Cdiv class=\"col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center\"\u003E\n \u003Cdiv class=\"text-center d-flex flex-column\"\u003E\n \n \u003Ca href=\"https:\u002F\u002Fgosrc.cc\u002Fgo\u002F1d9_-JY4R\" target=\"_blank\" rel=\"nofollow noopener\" class=\"v-btn v-btn--rounded elevation-5 v-size--large success mb-2\"\u003E\n \u003Cspan class=\"v-btn__content\"\u003EGet Book Now\u003C\u002Fspan\u003E\n \u003C\u002Fa\u003E\n \n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \n \u003Cdiv class=\"col-12\"\u003E\n \u003Cdiv class=\"v-card elevation-6\"\u003E\n \u003Cdiv class=\"v-card__text rating-text\"\u003E\n \u003Cdiv class=\"rating-counter\"\u003E\n \u003Cspan class=\"v-badge\"\u003E\n \u003Cspan class=\"v-badge__wrapper\"\u003E\n \u003Cspan aria-atomic=\"true\" aria-label=\"Позиция\" class=\"v-badge__badge primary\"\u003E\n 8\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"row\"\u003E\n \u003Cdiv class=\"col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center\"\u003E\n \u003Cdiv\u003E\n \u003Cimg src=\"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com\u002Fimages\u002Fd7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae\u002Frating\u002F41xz0sfhxsl-sl160.jpg\" alt=\"PyTorch Pocket Reference: Building and Deploying Deep Learning Models\" \u002F\u003E\n \u003Cp class=\"text-center font-weight-bold text-h6\"\u003ERating is 4.3 out of 5\u003C\u002Fp\u003E\n \u003Cdiv class=\"stars\" style=\"--rating: 4.3;\" aria-label=\"Rating is 4.3 out of 5\" \u003E\u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"col-lg-6 col-md-8 col-sm-6 col-12\"\u003E\n \u003Cp class=\"font-weight-bold rating-name\"\u003EPyTorch Pocket Reference: Building and Deploying Deep Learning Models\u003C\u002Fp\u003E\n \n \n\n \n \n \u003C\u002Fdiv\u003E\n\n \u003Cdiv class=\"col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center\"\u003E\n \u003Cdiv class=\"text-center d-flex flex-column\"\u003E\n \n \u003Ca href=\"https:\u002F\u002Fgosrc.cc\u002Fgo\u002Fshe_-1YVg\" target=\"_blank\" rel=\"nofollow noopener\" class=\"v-btn v-btn--rounded elevation-5 v-size--large success mb-2\"\u003E\n \u003Cspan class=\"v-btn__content\"\u003EGet Book Now\u003C\u002Fspan\u003E\n \u003C\u002Fa\u003E\n \n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \n \u003Cdiv class=\"col-12\"\u003E\n \u003Cdiv class=\"v-card elevation-6\"\u003E\n \u003Cdiv class=\"v-card__text rating-text\"\u003E\n \u003Cdiv class=\"rating-counter\"\u003E\n \u003Cspan class=\"v-badge\"\u003E\n \u003Cspan class=\"v-badge__wrapper\"\u003E\n \u003Cspan aria-atomic=\"true\" aria-label=\"Позиция\" class=\"v-badge__badge primary\"\u003E\n 9\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fspan\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"row\"\u003E\n \u003Cdiv class=\"col-lg-3 col-md-4 col-sm-6 col-12 d-flex justify-center align-center\"\u003E\n \u003Cdiv\u003E\n \u003Cimg src=\"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com\u002Fimages\u002Fd7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae\u002Frating\u002F41ayiicwf2l-sl160.jpg\" alt=\"Deep Learning with PyTorch Lightning: Swiftly build high-performance Artificial Intelligence (AI) models using Python\" \u002F\u003E\n \u003Cp class=\"text-center font-weight-bold text-h6\"\u003ERating is 4.2 out of 5\u003C\u002Fp\u003E\n \u003Cdiv class=\"stars\" style=\"--rating: 4.2;\" aria-label=\"Rating is 4.2 out of 5\" \u003E\u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003Cdiv class=\"col-lg-6 col-md-8 col-sm-6 col-12\"\u003E\n \u003Cp class=\"font-weight-bold rating-name\"\u003EDeep Learning with PyTorch Lightning: Swiftly build high-performance Artificial Intelligence (AI) models using Python\u003C\u002Fp\u003E\n \n \n\n \n \n \u003C\u002Fdiv\u003E\n\n \u003Cdiv class=\"col-lg-3 col-md-12 col-12 d-flex align-center justify-lg-end justify-center\"\u003E\n \u003Cdiv class=\"text-center d-flex flex-column\"\u003E\n \n \u003Ca href=\"https:\u002F\u002Fgosrc.cc\u002Fgo\u002FkmiX-1LVg\" target=\"_blank\" rel=\"nofollow noopener\" class=\"v-btn v-btn--rounded elevation-5 v-size--large success mb-2\"\u003E\n \u003Cspan class=\"v-btn__content\"\u003EGet Book Now\u003C\u002Fspan\u003E\n \u003C\u002Fa\u003E\n \n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n \n \u003C\u002Fdiv\u003E\n \u003C\u002Fdiv\u003E\n\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Ch2\u003EHow to implement batch normalization in a custom PyTorch model?\u003C\u002Fh2\u003E\u003Cp\u003ETo implement batch normalization in a custom PyTorch model, you can follow these steps:\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003EImport the required modules:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cdiv style=\"color:#f8f8f2;background-color:#272822;\"\u003E\n\u003Ctable style=\"border-spacing:0;padding:0;margin:0;border:0;\"\u003E\u003Ctbody\u003E\u003Ctr\u003E\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E1\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E2\n\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\n\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;;width:100%\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003Eimport torch\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003Eimport torch.nn as nn\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\u003C\u002Ftr\u003E\u003C\u002Ftbody\u003E\u003C\u002Ftable\u003E\n\u003C\u002Fdiv\u003E\n\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cscript async=\"\" src=\"https:\u002F\u002Fpagead2.googlesyndication.com\u002Fpagead\u002Fjs\u002Fadsbygoogle.js\"\u003E\u003C\u002Fscript\u003E\n\u003Cins class=\"adsbygoogle\" style=\"display:block\" data-ad-format=\"fluid\" data-ad-layout-key=\"-ef+6k-30-ac+ty\" data-ad-client=\"ca-pub-4833888168110763\" data-ad-slot=\"3267362137\"\u003E\u003C\u002Fins\u003E\n\u003Cscript\u003E\n (adsbygoogle = window.adsbygoogle || []).push({});\n\u003C\u002Fscript\u003E\u003Col\u003E\u003Cli\u003EDefine a basic custom model class:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cdiv style=\"color:#f8f8f2;background-color:#272822;\"\u003E\n\u003Ctable style=\"border-spacing:0;padding:0;margin:0;border:0;\"\u003E\u003Ctbody\u003E\u003Ctr\u003E\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E1\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E2\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E3\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E4\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E5\n\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\n\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;;width:100%\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003Eclass CustomModel(nn.Module):\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E def __init__(self):\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E super(CustomModel, self).__init__()\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E self.fc1 = nn.Linear(in_features, hidden_units)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E self.fc2 = nn.Linear(hidden_units, out_features)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\u003C\u002Ftr\u003E\u003C\u002Ftbody\u003E\u003C\u002Ftable\u003E\n\u003C\u002Fdiv\u003E\n\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EReplace \u003Ccode\u003Ein_features\u003C\u002Fcode\u003E, \u003Ccode\u003Ehidden_units\u003C\u002Fcode\u003E, and \u003Ccode\u003Eout_features\u003C\u002Fcode\u003E with appropriate values for your model architecture.\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003EAdd batch normalization layers and their parameters to the model:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cdiv style=\"color:#f8f8f2;background-color:#272822;\"\u003E\n\u003Ctable style=\"border-spacing:0;padding:0;margin:0;border:0;\"\u003E\u003Ctbody\u003E\u003Ctr\u003E\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E1\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E2\n\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\n\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;;width:100%\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003Eself.bn1 = nn.BatchNorm1d(hidden_units)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003Eself.bn2 = nn.BatchNorm1d(out_features)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\u003C\u002Ftr\u003E\u003C\u002Ftbody\u003E\u003C\u002Ftable\u003E\n\u003C\u002Fdiv\u003E\n\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EAdjust the parameter value based on your model architecture.\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003EDefine the forward pass of the model:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cdiv style=\"color:#f8f8f2;background-color:#272822;\"\u003E\n\u003Ctable style=\"border-spacing:0;padding:0;margin:0;border:0;\"\u003E\u003Ctbody\u003E\u003Ctr\u003E\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E1\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E2\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E3\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E4\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E5\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E6\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E7\n\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\n\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;;width:100%\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003Edef forward(self, x):\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E x = self.fc1(x)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E x = self.bn1(x)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E x = nn.functional.relu(x)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E x = self.fc2(x)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E x = self.bn2(x)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E return x\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\u003C\u002Ftr\u003E\u003C\u002Ftbody\u003E\u003C\u002Ftable\u003E\n\u003C\u002Fdiv\u003E\n\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EThis example assumes the ReLU activation function, but you can replace it with any activation function you prefer.\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003ECreate an instance of the custom model:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cdiv style=\"color:#f8f8f2;background-color:#272822;\"\u003E\n\u003Ctable style=\"border-spacing:0;padding:0;margin:0;border:0;\"\u003E\u003Ctbody\u003E\u003Ctr\u003E\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E1\n\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\n\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;;width:100%\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003Emodel = CustomModel()\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\u003C\u002Ftr\u003E\u003C\u002Ftbody\u003E\u003C\u002Ftable\u003E\n\u003C\u002Fdiv\u003E\n\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003ENow you have implemented batch normalization in your custom PyTorch model.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Ch2\u003EWhat are the advantages of using batch normalization in PyTorch?\u003C\u002Fh2\u003E\u003Cp\u003EBatch normalization is a regularization technique that is widely used in deep learning models. When applied to PyTorch models, it provides several advantages:\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003E\u003Cstrong\u003EImproved convergence\u003C\u002Fstrong\u003E: Batch normalization normalizes the input to each neuron across a mini-batch, which helps in stabilizing the learning process. This leads to faster convergence and reduces the number of epochs required for training.\n\u003C\u002Fli\u003E\u003Cli\u003E\u003Cstrong\u003EReduced overfitting\u003C\u002Fstrong\u003E: By normalizing the inputs, batch normalization reduces the dependence of each neuron on the other neurons in the network. This reduces the chances of overfitting and improves the generalization ability of the model.\n\u003C\u002Fli\u003E\u003Cli\u003E\u003Cstrong\u003EIncreased learning rate\u003C\u002Fstrong\u003E: Batch normalization reduces the internal covariate shift by maintaining zero mean and unit variance activations. This enables the use of higher learning rates during training, which can speed up the training process.\n\u003C\u002Fli\u003E\u003Cli\u003E\u003Cstrong\u003EBetter gradient flow\u003C\u002Fstrong\u003E: Normalizing the inputs using batch normalization helps in ensuring that the gradients flow smoothly and consistently during backpropagation. This helps combat the vanishing and exploding gradient problems, making it easier to train deep networks.\n\u003C\u002Fli\u003E\u003Cli\u003E\u003Cstrong\u003ERobustness to different input distributions\u003C\u002Fstrong\u003E: Batch normalization makes the model less sensitive to the \u003Ca href=\"https:\u002F\u002Fstudentprojectcode.com\u002Fblog\u002Fhow-to-simplify-units-with-different-scales-in\" target=\"_blank\"\u003Escale\u003C\u002Fa\u003E and distribution of the input data. This allows the model to perform well even when faced with inputs that are significantly different from the training data.\n\u003C\u002Fli\u003E\u003Cli\u003E\u003Cstrong\u003EWeight initialization flexibility\u003C\u002Fstrong\u003E: Batch normalization helps in reducing the dependence of the model's performance on the choice of weight initialization. It allows the use of simpler initialization methods like random or small weights, which can speed up the training process.\n\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EOverall, batch normalization is a useful tool for improving the performance and stability of deep learning models in PyTorch, leading to faster convergence, better generalization, and increased robustness.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Ch2\u003EWhat is the effect of batch size on batch normalization in PyTorch?\u003C\u002Fh2\u003E\u003Cp\u003EThe batch size affects the batch normalization in PyTorch in the following way:\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003E\u003Cstrong\u003EStatistics estimation\u003C\u002Fstrong\u003E: Batch normalization relies on estimating the mean and variance of the input data to normalize it. With a larger batch size, there is more data available for statistics estimation, leading to more accurate estimates of the mean and variance. This can result in improved normalization and consequently, better performance.\n\u003C\u002Fli\u003E\u003Cli\u003E\u003Cstrong\u003ENoise reduction\u003C\u002Fstrong\u003E: Batch normalization introduces some noise to the statistics estimation process. With a larger batch size, the noise is averaged out more effectively, resulting in more stable estimates of mean and variance. This can lead to reduced overfitting and improved generalization.\n\u003C\u002Fli\u003E\u003Cli\u003E\u003Cstrong\u003ETraining dynamics\u003C\u002Fstrong\u003E: Smaller batch sizes tend to introduce more stochasticity and randomness in the training process, as each batch's statistics differ significantly. On the other hand, larger batch sizes provide more consistent statistics, which can affect the optimization process. This can result in different training dynamics, such as convergence speed and stability.\n\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cscript async=\"\" src=\"https:\u002F\u002Fpagead2.googlesyndication.com\u002Fpagead\u002Fjs\u002Fadsbygoogle.js\"\u003E\u003C\u002Fscript\u003E\n\u003Cins class=\"adsbygoogle\" style=\"display:block\" data-ad-format=\"fluid\" data-ad-layout-key=\"-ef+6k-30-ac+ty\" data-ad-client=\"ca-pub-4833888168110763\" data-ad-slot=\"3267362137\"\u003E\u003C\u002Fins\u003E\n\u003Cscript\u003E\n (adsbygoogle = window.adsbygoogle || []).push({});\n\u003C\u002Fscript\u003E\u003Cp\u003EIt's important to note that the choice of batch size is often a trade-off. Larger batch sizes require more memory, may limit parallelization, and increase computational requirements. However, they can offer better normalization and estimation, while smaller batch sizes may introduce more noise but can be computationally more efficient.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Ch2\u003EWhat are the requirements for using batch normalization in PyTorch?\u003C\u002Fh2\u003E\u003Cp\u003ETo use batch normalization in PyTorch, the following requirements should be met:\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003E\u003Cstrong\u003EPyTorch should be installed on the system. You can install it using pip\u003C\u002Fstrong\u003E: pip install torch.\n\u003C\u002Fli\u003E\u003Cli\u003EImport the necessary modules:\n\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cdiv style=\"color:#f8f8f2;background-color:#272822;\"\u003E\n\u003Ctable style=\"border-spacing:0;padding:0;margin:0;border:0;\"\u003E\u003Ctbody\u003E\u003Ctr\u003E\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E1\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E2\n\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\n\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;;width:100%\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003Eimport torch\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003Eimport torch.nn as nn\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\u003C\u002Ftr\u003E\u003C\u002Ftbody\u003E\u003C\u002Ftable\u003E\n\u003C\u002Fdiv\u003E\n\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003EDefine your model architecture using the nn.Module class. Use the torch.nn.BatchNorm2d or torch.nn.BatchNorm1d layer (based on your input dimensions) for batch normalization.\n\u003C\u002Fli\u003E\u003Cli\u003EUse batch normalization layer after the convolutional or linear layer in your model architecture. For example:\n\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cdiv style=\"color:#f8f8f2;background-color:#272822;\"\u003E\n\u003Ctable style=\"border-spacing:0;padding:0;margin:0;border:0;\"\u003E\u003Ctbody\u003E\u003Ctr\u003E\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E1\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E2\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E3\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E4\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E5\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E6\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E7\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E8\n\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\n\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;;width:100%\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003Eclass MyModel(nn.Module):\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E def __init__(self):\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E super(MyModel, self).__init__()\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E self.conv1 = nn.Conv2d(3, 64, kernel_size=3)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E self.bn1 = nn.BatchNorm2d(64)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E self.fc1 = nn.Linear(64, 10)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E self.bn2 = nn.BatchNorm1d(10)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E ...\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\u003C\u002Ftr\u003E\u003C\u002Ftbody\u003E\u003C\u002Ftable\u003E\n\u003C\u002Fdiv\u003E\n\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Col\u003E\u003Cli\u003EDuring the forward pass, apply batch normalization to the input tensor. For example:\u003C\u002Fli\u003E\u003C\u002Fol\u003E\u003Cdiv style=\"color:#f8f8f2;background-color:#272822;\"\u003E\n\u003Ctable style=\"border-spacing:0;padding:0;margin:0;border:0;\"\u003E\u003Ctbody\u003E\u003Ctr\u003E\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E1\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E2\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E3\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E4\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E5\n\u003C\u002Fspan\u003E\u003Cspan style=\"white-space:pre;user-select:none;margin-right:0.4em;padding:0 0.4em 0 0.4em;color:#7f7f7f\"\u003E6\n\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\n\u003Ctd style=\"vertical-align:top;padding:0;margin:0;border:0;;width:100%\"\u003E\n\u003Cpre tabindex=\"0\" style=\"color:#f8f8f2;background-color:#272822;\"\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003Edef forward(self, x):\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E x = self.conv1(x)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E x = self.bn1(x)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E x = self.fc1(x)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E x = self.bn2(x)\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003Cspan style=\"display:flex;\"\u003E\u003Cspan\u003E ...\n\u003C\u002Fspan\u003E\u003C\u002Fspan\u003E\u003C\u002Fpre\u003E\u003C\u002Ftd\u003E\u003C\u002Ftr\u003E\u003C\u002Ftbody\u003E\u003C\u002Ftable\u003E\n\u003C\u002Fdiv\u003E\n\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003ENote: Batch normalization is typically used before the activation function, but the order can vary depending on your problem and experiment settings.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Ch2\u003EWhat is the impact of batch normalization on model generalization in PyTorch?\u003C\u002Fh2\u003E\u003Cp\u003EBatch normalization has a significant impact on model generalization in PyTorch. It helps to improve the generalization capability of neural networks by reducing the internal covariate shift.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EInternal covariate shift refers to the change in the distribution of network activations due to the change in parameter values during training. This can slow down the training process and hinder the performance of the model.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EBatch normalization solves this problem by normalizing the output of each layer using the mean and variance of the mini-batch. By doing so, it reduces the effect of the internal covariate shift and makes the optimization process more stable. Batch normalization also introduces additional trainable parameters, which allow the network to adaptively scale and shift the normalized values.\u003C\u002Fp\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EThe normalization of inputs helps in the generalization of the model because it keeps the values within a reasonable range. It prevents extreme values from causing instability in the network, which can lead to overfitting. Additionally, batch normalization acts as a regularizer, reducing the need for other regularization techniques like dropout.\u003C\u002Fp\u003E\u003Cscript async=\"\" src=\"https:\u002F\u002Fpagead2.googlesyndication.com\u002Fpagead\u002Fjs\u002Fadsbygoogle.js\"\u003E\u003C\u002Fscript\u003E\n\u003Cins class=\"adsbygoogle\" style=\"display:block\" data-ad-format=\"fluid\" data-ad-layout-key=\"-ef+6k-30-ac+ty\" data-ad-client=\"ca-pub-4833888168110763\" data-ad-slot=\"3267362137\"\u003E\u003C\u002Fins\u003E\n\u003Cscript\u003E\n (adsbygoogle = window.adsbygoogle || []).push({});\n\u003C\u002Fscript\u003E\u003Cp\u003E\u003Cbr\u002F\u003E\u003C\u002Fp\u003E\u003Cp\u003EOverall, batch normalization in PyTorch improves the generalization ability of models by reducing internal covariate shift, making the training process more stable, and acting as a regularizer.\u003C\u002Fp\u003E",slug:"how-to-implement-batch-normalization-in-pytorch",image:"blog\u002F465d418c-e01f-4744-92bf-7347d5442899\u002F658324fe1a7a03cb8b15bf11.png",active:d,nofollow_links:c,hash_tags:["blogweb"],allow_comments:c,no_ad:c,update_daily:c,update_monthly:d,update_yearly:d,meta_title:"How to Implement Batch Normalization In PyTorch in 2024?",meta_description:aV,related_posts:[{id:at,text:au,title:a,image:av,summary:aw,slug:ax},{id:ay,text:az,title:a,image:aA,summary:aB,slug:aC},{id:aD,text:aE,title:a,image:aF,summary:aG,slug:aH},{id:154226,text:"How to Pass Parameters to A Batch File From Powershell?",title:a,image:"blog\u002Fd5909218-ed12-44a7-ae3c-e0eb3465deef\u002F6729074bfff57e469025a8fa.png",summary:"To pass parameters to a batch file from PowerShell, you can use the Start-Process cmdlet. You can pass the parameters as arguments to the batch file by specifying the -ArgumentList parameter followed by the parameters enclosed in quotes. For example, you can run the batch file example.bat with two parameters param1 and param2 as follows:\nStart-Process -FilePath "example.bat" -ArgumentList "param1", "param2"\nThis will execute the batch file example.",slug:"how-to-pass-parameters-to-a-batch-file-from"},{id:108964,text:"How to Save GPU Memory Usage In PyTorch?",title:a,image:"blog\u002F525ade24-90a9-4630-b1ff-35d422cd9125\u002F657b9c6081b5eea578801090.png",summary:"When working with PyTorch, it is essential to manage GPU memory efficiently to avoid out-of-memory errors and maximize the utilization of available resources. Here are some techniques to save GPU memory usage in PyTorch:Use smaller batch sizes: Reducing the batch size lowers the memory requirement for each mini-batch processed on the GPU. However, it may increase the training time due to more frequent parameter updates.",slug:"how-to-save-gpu-memory-usage-in-pytorch"},{id:108942,text:"How to \"Denormalize\" A Pytorch Tensor?",title:a,image:"blog\u002F694c2602-fc63-43a3-9fd2-ba4f10abdec0\u002F657b63990ab8e8d4bdcd0d68.png",summary:"Denormalizing a PyTorch tensor refers to the process of converting a normalized tensor back to its original scale or range. It involves reversing the normalization transformation that was applied to the tensor earlier.To denormalize a PyTorch tensor, you typically need two pieces of information: the mean and standard deviation used for normalization. These values are necessary to reverse the scaling process.",slug:"how-to-denormalize-a-pytorch-tensor"},{id:125346,text:"How to Set Batch Size When Inference With Tensorflow?",title:a,image:"blog\u002F677a4188-c769-493e-8477-4ab2fa972776\u002F6640bb05231af8264dae4b29.png",summary:"When performing inference with TensorFlow, setting the batch size can be important for optimizing the speed and efficiency of the process. The batch size refers to the number of samples that will be processed at once in each iteration. To set the batch size when performing inference with TensorFlow, you can typically specify it as a parameter when calling the prediction function.",slug:"how-to-set-batch-size-when-inference-with"},{id:109573,text:"How to Install PyTorch on My Machine?",title:a,image:"blog\u002F3a4e1b49-0223-42fa-8ed9-4fd7de9daead\u002F65800aed81b5eea57889c4eb.png",summary:"To install PyTorch on your machine, you need to follow these steps:Decide if you want to install PyTorch with or without CUDA support. If you have an NVIDIA GPU and want to utilize GPU acceleration, you will need to install PyTorch with CUDA.\nCheck if you have a compatible Python version installed. PyTorch supports Python 3.6 or above.\nOpen a terminal or command prompt on your machine.\nUse the package manager pip to install PyTorch.",slug:"how-to-install-pytorch-on-my-machine"},{id:118190,text:"How to Set Up Automatic Batch Processing on A Credit Card Machine?",title:a,image:"blog\u002F9fb10863-39ed-4f78-99db-b8fa8e9b12b3\u002F65d355667bb396817d9e413e.png",summary:"To set up automatic batch processing on a credit card machine, you will need to access the settings or configuration options on the machine. Look for an option that allows you to schedule automatic batch processing at a specific time each day.You may need to input the time and frequency at which you want the batches to be processed. Make sure to double check that the machine is connected to a reliable internet or phone line connection to ensure that the batches are processed successfully.",slug:"how-to-set-up-automatic-batch-processing-on-a"},{id:109596,text:"How to Create A Tensor In PyTorch?",title:a,image:"blog\u002Ffa736fa5-7557-4627-95fa-81255ba7b5b7\u002F65804383b27dd462faf907df.png",summary:"To create a tensor in PyTorch, you can follow the steps below:Import the PyTorch library: Begin by importing the PyTorch library using the import statement:\nimport torch\nCreate a tensor from a list or array: You can create a tensor by passing a Python list or an array to the torch.tensor() function. PyTorch will automatically infer the data type and shape of the tensor based on the inputs.\nmy_list = [1, 2, 3, 4, 5]\nmy_tensor = torch.",slug:"how-to-create-a-tensor-in-pytorch"}],category:{id:aK,name:Y,meta_title:a,meta_description:a,order:b,children:g,description:a,slug:aL},created:"2023-12-20T17:31:44Z",updated:"2024-12-01T00:00:00Z"}}],fetch:{},error:g,state:{loading:b,settings:{id:h,name:k,domain:aN,port:aO,plan:e,add_source:e,add_source_text:U,forum_active:c,footer_code:aS,scrollable_pagination:b,add_watermark:b,add_watermark_position:b,hash:aM,robots_txt:aR,locale:aQ,meta_title:k,modules:[{uuid:"52f05b96-2b7a-11eb-943e-6a24baf8d0e4",path:"amazon",name:"Amazon",active:d},{uuid:"39e96103-3de3-11eb-9b32-86f43b04e535",path:"tinysrc",name:"TinySRC",active:d},{uuid:"cc863ba7-13bd-11ed-a99e-8ebf5783113d",path:aW,name:"mywebforum.com",active:d},{uuid:"7671225a-2f09-11ee-9f18-9ac8ad3607b3",path:"openai",name:"OpenAI",active:d}],favicon_png:"\u002Ffavicon.png",favicon_ico:a,custom_css:".rating-text img{\n max-height: 150px !important;\n max-width: 190px !important;\n}\n\n.rating-text img{\n max-height: 150px !important;\n max-width: 190px !important;\n}\n\n.rating-text .row .d-flex \u003E div{\n text-align: center;\n}",meta_description:y,description:y,logo:aP,activation:aT},layout:{id:h,is_dark:b,name:ar,page_transition:"fadeUp",background:"background\u002F555.png",code_theme:"monokai",background_full:c,background_color:a,text_color:a,text_font_family:"Literata",primary_color:i,secondary_color:"#424242",accent_color:i,info_color:i,success_color:i,error_color:i,warning_color:i},menus:[{id:17,name:aX,position:_,link:"\u002Fpage\u002Fprivacy-policy",open_new_tab:d,order:b,no_follow:c},{id:18,name:"Terms of Use",position:_,link:"\u002Fpage\u002Fterms-of-use",open_new_tab:d,order:b,no_follow:c}],isFooterVisible:c,showAd:c,cdnUrl:"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com",metaOg:{title:Z,url:as,image:"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com\u002Fimages\u002Fd7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae\u002Fblog\u002F465d418c-e01f-4744-92bf-7347d5442899\u002F658324fe1a7a03cb8b15bf11.png",type:"article",description:aV,site_name:k},ad:[{id:8,name:"Own Domain",css_selector:a,position:e,one_time:c,show_every:b,code:"\u003Cdiv class=\"flex\"\u003E\n\u003Ca href=\"https:\u002F\u002Fgosrc.cc\u002Fgo\u002FoJqr0c6SR\" target=\"_blank\"\u003E\u003Cimg src=\"https:\u002F\u002Fblogweb-static.fra1.cdn.digitaloceanspaces.com\u002Fpromo\u002Fbanner.png\" style=\"max-height:200px; max-width:100%;\" \u002F\u003E\u003C\u002Fa\u003E\n\u003Cdiv\u003E",active:d,free:d},{id:112,name:"posttop",css_selector:a,position:aY,one_time:c,show_every:b,code:aZ,active:d,free:c},{id:113,name:"category",css_selector:a,position:aI,one_time:c,show_every:aY,code:aZ,active:d,free:c},{id:114,name:_,css_selector:a,position:2,one_time:c,show_every:b,code:a_,active:d,free:c},{id:115,name:"post",css_selector:"p",position:6,one_time:c,show_every:15,code:a_,active:d,free:c},{id:116,name:"menu",css_selector:a,position:a$,one_time:c,show_every:b,code:"\u003Cscript async src=\"https:\u002F\u002Fpagead2.googlesyndication.com\u002Fpagead\u002Fjs\u002Fadsbygoogle.js\"\u003E\u003C\u002Fscript\u003E\n\u003C!-- stlplaces3 --\u003E\n\u003Cins class=\"adsbygoogle\"\n style=\"display:block\"\n data-ad-client=\"ca-pub-4833888168110763\"\n data-ad-slot=\"8845076150\"\n data-ad-format=\"auto\"\n data-full-width-responsive=\"true\"\u003E\u003C\u002Fins\u003E\n\u003Cscript\u003E\n (adsbygoogle = window.adsbygoogle || []).push({});\n\u003C\u002Fscript\u003E",active:d,free:c}],forum:{domain:"forum.stlplaces.com",header:"Discussion",position:a$,active:d,threads:[{id:171090,name:"Which state is better to move in: Florida or North Carolina?",username:"jake",user_avatar:"avatar\u002Fjake\u002Favatar-94-s.jpg",role:l,message:ba,formatted_message:ba,category:m,category_slug:n,category_id:o,is_moderated:d,posts:e,likes:7,dislikes:b,slug:"which-state-is-better-to-move-in-florida-or-north",created:"2024-01-06T02:26:12Z",is_closed:c,is_pinned:c},{id:171003,name:"What state is better: Tennessee or Iowa?",username:"louie",user_avatar:"avatar\u002Flouie\u002Ffood007.jpg",role:l,message:bb,formatted_message:bb,category:m,category_slug:n,category_id:o,is_moderated:d,posts:e,likes:$,dislikes:b,slug:"what-state-is-better-tennessee-or-iowa",created:"2024-01-05T11:16:14Z",is_closed:c,is_pinned:c},{id:170966,name:"Which state is better to move in: Ohio or Arizona?",username:"dave",user_avatar:"avatar\u002Fdave\u002Fpozdravlenie-f-it-ru72.gif",role:l,message:bc,formatted_message:bc,category:m,category_slug:n,category_id:o,is_moderated:d,posts:e,likes:$,dislikes:b,slug:"which-state-is-better-to-move-in-ohio-or-arizona",created:"2024-01-05T08:14:08Z",is_closed:c,is_pinned:c},{id:170828,name:"Which state is best to visit: Illinois or North Carolina?",username:"chris",user_avatar:"avatar\u002Fchris\u002Ftest5.jpg",role:l,message:bd,formatted_message:bd,category:m,category_slug:n,category_id:o,is_moderated:d,posts:e,likes:13,dislikes:b,slug:"which-state-is-best-to-visit-illinois-or-north",created:"2024-01-04T01:54:11Z",is_closed:c,is_pinned:c},{id:170799,name:"Which state is better to move in: Ohio or Indiana?",username:"jaylin",user_avatar:"avatar\u002Fjaylin\u002F39-2.jpg",role:l,message:be,formatted_message:be,category:m,category_slug:n,category_id:o,is_moderated:d,posts:e,likes:$,dislikes:b,slug:"which-state-is-better-to-move-in-ohio-or-indiana",created:"2024-01-03T22:52:10Z",is_closed:c,is_pinned:c}],total:514},author:{},profileStats:{fetched:c,personalMessages:b},stats:{fetched:c,threads:b,comments:b,posts:b},breadcrumbs:[{text:aa,disabled:c,href:ab},{text:Y,disabled:c,href:"\u002Fblog\u002Fcategory\u002Fprogramming"},{text:Z,disabled:d}],i18n:{routeParams:{}},auth:{user:g,loggedIn:c,strategy:"local"}},serverRendered:d,routePath:"\u002Fblog\u002Fhow-to-implement-batch-normalization-in-pytorch",config:{_app:{basePath:ab,assetsPath:ab,cdnURL:"https:\u002F\u002Fpub-420acf56315e422bbbdab07717bee8cd.r2.dev\u002Fassets\u002F0.1\u002F"}},__i18n:{langs:{en:{common:{common_badge_blog:"blog",common_badge_page:"page",common_badge_post:aW,check_internet_connection:"You are offline! Check your internet connection.",score:"Score",open:"Open",got_it:"Got it",links_posted:bf,our_privacy:"See our Privacy Policy",cookie_text:"This site uses cookies to ensure you get the best experience. By clicking \"Got it\" or continuing to browse the website, you are agreeing to our use of cookies.",no_transition:"No Transition",mail_services:"Mail Services",not_available_this_plan:"Not available in this plan! Please upgrade your plan.",free_files:"The current plan has a limit 10 files",confirm_email_text:"We just sent an email to you with the confirmation code. Please put the confirmation code in field below.",main_top_ad:"Ad in all pages on the top",main_bottom_ad:"Ad in all pages on the bottom",blog_in_category_ad:"Ad in blog category",max_tag_reached:"Max Count of Tags Reached",last_threads:"Latest Threads",count_comments:"Blog Comments",blog_post_top_ad:"Ad in blog post on the top",blog_post_bottom_ad:"Ad in blog post on the bottom",blog_in_comment:"Ad in blog comment",blog_in_post_ad:"Ad in blog post",forum_in_thread_ad:"Ad in forum thread",forum_in_post_ad:"Ad in forum post",sliding_ad:"Sliding Ad on all pages",page_ad:"Ad on the page",sidebar_under_menu:"Left sidebar under menu",read_time:"read",site_is_not_ready:"You site is not ready yet",security_limit:"You can configure limit per IP to prevent any spam and bot messages. If not limit set it 0.",thanks_choosing_us:"Thank you for giving us the opportunity to serve you",register_site_first:"You need register your site first at",create_forum:"create a blog",create_forum_free:"free blog",create_forum_free_text:"Get Your Own",support_text:"support forum",post_subject:bg,how_to_create_text:"If you have any question about our platform - you can easily ask us. We are always ready to help you by contact form or by our",subcategory:"Subcategories",category:ac,main_page:"Main Page",banned:"Banned User",no_banned:"Not Banned User",new_message:"New",not_read:"Not Read",spam:"Spam",trusted:bh,no_comments:"No comments",no_trusted:"Common User",security_questions:"Registration Questions",security_questions_descr:"Registration Questions help you protect your system from bot. We will randomly show user question before registration and he can continue if answer right on this question.",nice_to_see_you:"Nice to see you again",records_not_found:"No records found",not_found:"There were no results for your search.",not_authorized:"Sign in to you account",account:"Account",last_message:" Last Reply ",from:" from ",thanks_confirming:"Thank you for using our service. Account successfully confirmed!",thanks_confirming_details:"You can use your credentials to login into system",mail_settings_valid:"Congratulations! Mail Settings verified successfully and ready to use",terms:"Terms & Conditions",privacy:aX,accept_terms:"I Agree to the Terms and Conditions",need_accept_terms:"You must agree to the Terms and Conditions in order to register",copyright_text:"This Site is Created By",security_messages:"Secure messages",na:"N\u002FA",active:bi,no_active:"Inactive",attention:"Attention!",count_topics:bj,need_moderation:"Approve message",no_data:"No data",no_files:"No files",no_records:"No records",no_messages:"No messages",count_messages:z,menu:p,search:A,action:"Action",pin_topic:"Pin topic",unpin_topic:"Unpin topic",existing_client:"Existing user? Sign In",close_topic:"Close topic",open_topic:"Open topic",rows_per_page:"Rows per page",top_menu:"Top menu",bottom_menu:"Bottom menu",waiting_approval:"Waiting approval",second:"Second",minute:"Minute",hour:"Hour",message:z,messages:z,topic_closed:"Topic closed",important_topic:"Important topic",threads:"Threads"},animation:{default_fade:"Default Fade",fade_up:"Fade Up",fade_right:"Fade Right",fade_down:"Fade Down",fade_left:"Fade Left",rotate:"Rotate",rotate_down_left:"Rotate Down Left",rotate_down_right:"Rotate Down Right",rotate_up_left:"Rotate Up Left",rotate_up_right:"Rotate Up Right",slide_down:"Slide Down",slide_left:"Slide Left",slide_right:"Slide Right",slide_up:"Slide Up",flip:"Flip",flip_x:"Flip X",flip_y:"Flip Y",zoom:"Zoom",zoom_down:"Zoom Down",zoom_left:"Zoom Left",zoom_right:"Zoom Right",zoom_up:"Zoom Up",light_speed:"Light Speed",bounce:"Bounce",bounce_down:"Bounce Down",bounce_left:"Bounce Left",bounce_right:"Bounce Right",bounce_up:"Bounce Up"},mail:{other:"Other",gmail:"Gmail",yandex:"Yandex",yahoo:"Yahoo",mail:"Mail.ru",rambler:"Rambler",bing:"Bing",outlook:"Outlook",aol:"AOL"},tabs:{settings_dashboard:bk,keywords_dashboard:"Keyword Dashboard",modules:q,inbox_pm:"Inbox",sent_pm:"Sent",moderator_main:"Statistic",moderator_threads:bl,links_settings:"Auto Linker",hash_tags_settings:ad,moderator_users:B,export_settings:r,moderator_posts:C,moderator_comments:D,blog_settings:E,privacy_settings:F,ad_settings:G,author_settings:bm,file_settings:s,my_profile_stat:"My Statistics",main_settings:bk,mail_settings:H,layout_settings:I,forum_settings:ae,member_settings:bn,page_settings:J,menu_settings:"Menu Settings",social_settings:K,security_settings:L,profile:"Profile",messages:z,redirect_settings:af,topics:bj},modals:{ask_ai:bo,new_pm:"Compose Message",delete_link:"Are you sure you want to delete this link?",delete_link_confirm:"Please confirm that you want to delete this link",delete_tag:"Are you sure you want to delete this hash tag?",delete_tag_confirm:"Please confirm that you want to delete this hash tag",delete_user:"Are you sure you want to delete this user?",delete_file:"Are you sure you want to delete this file?",delete_file_confirm:"Please confirm that you want to delete this file",delete_user_confirm:"Please confirm that you want to delete this user",delete_thread:"Are you sure you want to delete this thread?",delete_thread_confirm:"Please confirm that you want to delete this thread",delete_ad:"Are you sure you want to delete this Ad?",delete_ad_confirm:"Please confirm that you want to delete this Ad",delete_blog_content:"Are you sure you want to delete this post?",delete_blog_content_confirm:"Please confirm that you want to delete this post",delete_blog_comment:"Are you sure you want to delete this comment?",delete_blog_comment_confirm:"Please confirm that you want to delete this comment",delete_all_blog_comments_text:"Are you sure you want to delete all comments?",delete_all_blog_comments_text_confirm:"Please confirm that you want to delete all comments",delete_blog_category:bp,delete_blog_category_confirm:"Please confirm that you want to delete this category and related subcategories, posts.",delete_redirect:"Are you sure you want to delete this redirect?",delete_redirect_confirm:"Please confirm that you want to delete this redirect link.",delete_category:bp,delete_category_confirm:"Please confirm that you want to delete this category and related subcategories, threads, posts.",delete_menu:"Are you sure you want to delete this menu link?",delete_menu_confirm:"Please confirm that you want to delete this menu link.",delete_page:"Are you sure you want to delete this page?",delete_page_confirm:"Please confirm that you want to delete this page.",delete_message:"Are you sure you want to delete this message?",delete_message_confirm:"Please confirm that you want to delete this message.",delete_topic:"Are you sure you want to delete this topic?",delete_topic_confirm:"Please confirm that you want to delete this topic and related posts.",delete_all_topic_message:"Are you sure you want to delete all messages?",delete_all_topic_message_confirm:"Please confirm that you want to delete all messages",delete_all_topic_text:"Are you sure you want to delete all topics?",delete_all_topic_text_confirm:"Please confirm that you want to delete all topics"},errors:{page_not_found:"Page not found",other_error:"Error happened"},toasts:{email_sent:"Confirmation Code Sent to Email",vote_review:"We will review it and update rating",already_voted:"You have already voted",logout:"You successfully logged out",category_created:"Category successfully created!",category_updated:"Category successfully updated!",ad_created:"Ad successfully created!",ad_updated:"Ad successfully updated!",menu_created:"Menu link successfully created!",menu_updated:"Menu link successfully updated!",page_created:"Page successfully created!",page_updated:"Page successfully updated!",register_successfully:"You have successfully registered",confirm_email:"Please confirm your email address",comments_waiting_approval:"Comments waiting approval:",posts_waiting_approval:"Posts waiting approval:",users_waiting_approval:"Users waiting approval:",threads_waiting_approval:"Threads waiting approval:",favicon_updated:"Favicon successfully updated",background_updated:"Background successfully updated",select_logo:"Please select logo",select_favicon:"Please select favicon",logo_updated:"Logo successfully updated",profile_updated:"Profile successfully updated",avatar_updated:"Avatar successfully updated",select_avatar:"Please select image file",select_background:"Please select background image file",auth_success:"Logged in successfully",settings_success:"Settings successfully updated",terms_success:"Terms successfully updated",message_send:"Message has been sent",created:bq,updated:"Updated",password_changed:"Password changed successfully",record_created:bq,deleted:"Deleted",success:"Success",error:"Error happened."},titles:{link_settings:ag,link_keywords:ah,content_rating:M,modules:q,forum_home:ai,search:A,author:N,confirm_account:br,how_to_create:"How To Create Your Own Free Forum",category_settings:bs,edit_member:aj,export_settings:r,file_settings:s,blog_settings:E,ad_settings:G,blog_content:O,blog_category:t,blog:P,my_pm:Q,view_pm:ak,members:bt,redirects:af,menu:p,pages:J,layout_settings:I,social_settings:K,security_settings:L,terms_settings:F,mail_settings:H,admin_index:R,moderator_index:S,moderator_comments:D,moderator_threads:bu,moderator_users:B,moderator_posts:C,change_password:u,view_profile:"user profile",my_profile:al,register:bv,login:v,restore_password:w,contact_us:T},descriptions:{link_keywords:ah,link_settings:ag,content_rating:M,modules:q,author:N,forum_home:ai,confirm_account:br,how_to_create:"How To Create Your Own Free Blog. If you need a free modern blog - you can create it right now.",category_settings:bs,export_settings:r,edit_member:aj,moderator_index:S,file_settings:s,moderator_comments:D,moderator_threads:bu,moderator_users:B,moderator_posts:C,ad_settings:G,blog_settings:E,my_pm:Q,view_pm:ak,blog:P,blog_category:t,blog_content:O,members:bt,redirects:af,menu:p,terms_settings:F,layout_settings:I,pages:J,social_settings:K,security_settings:L,admin_index:R,mail_settings:H,change_password:u,view_profile:"user profile. Contacts, messages, topics.",my_profile:al,register:bv,login:v,restore_password:w,contact_us:"You can always send us message and we always ready to help. "},activations:{email:"Email confirmation required",manually:"Manually by admin\u002Fmoderator",automatically:"Automatically, after registration"},toc:{disabled:"No Active",top:"Top",bottom:"Bottom"},author:{rating:"Author Rating",author:N,rank_author:"Your Rate?",articles:"Articles",author_page:"About Author"},export:{yes:"Yes",no:"No",blog_category:t,blog:"Blog Posts",forum:ae,profile:"Profiles"},chudoshop:{name:j,commission:"Commission",category:ac,description:am,product_code:"Product Code",manufacturer:"Manufacturer",picture:an,price:ao},amazon:{price:ao,picture:an,thumbnail:"Thumbnail",features:"Features",name:j,product_url:"Product Url",product_group:"Product Group",brand:"Brand",model:bw,url:"Url",original_picture:"OriginalPicture"},gde_slon:{price:ao,vendor:"Vendor",model:bw,charge:"Charge",picture:an,name:j,description:am},watermark:{center:"Center",top_left:"Top-Left Corner",top_right:"Top-Right Corner",bottom_left:"Bottom-Left Corner",bottom_right:"Bottom-Right Corner"},forms:{url:"URL",links_posted:bf,keyword:"Keyword",max_links:"Max Links",nofollow_link:"Nofollow Link",max_links_per_article:"Max Links in One Article",frequency_between_links:"Characters between links",exclude_html_tags:"Exclude links in HTML tags(comma-separated)",hash_tag:"Hash Tag",meta_title:"Meta Title",meta_description:"Meta Description",ai_question:ap,openai_setting:"OpenAI Settings",generate_tags:"Auto Generate Tags",generate_image:"Generate Image",code_theme:"Code Editor Theme:",add_watermark:"Add Watermark to Images?",add_watermark_position:"Watermark Position",move_to_category:"Move Threads and Messages to Category",blog_move_to_category:"Move Content to Category",scrollable_pagination:"Scrollable pagination",category:"Category:",call_to_action_as_image:"Make it as image",chudoshop_setting:"Chudoshop Settings",header_text:"Block label",forum_domain:"Forum URL",forum_setting:bx,count_threads:"Count of Threads",tinysrc_setting:"TinySRC Settings",amazon_setting:"Amazon Settings",role:by,hash_tags:"Tags",partner_tag:"Partner Tag",confirm_your_email:"Confirm your email",access_key:"Access Key",secret_key:"Secret Key",toc_collapse:"Toc Collapsable",api_key:bz,call_to_action_text:"Call To Action Text",default_call_to_action_text:"Visit Site",gde_slon_lookup:"Gde Slon Lookup",amazon_lookup:"Amazon Lookup",chudoshop_lookup:"Chudoshop Lookup",unselect:"Unselect",select:"Select",gde_slon_query:bA,chudoshop_query:bA,gde_slon:"Gde Slon Settings",module_name:"Название модуля",add_source:"Add Source Whenever Copy Text?",add_source_text:"Text:",default_add_source_text:U,background_full:"Full Screen Image",notifications:"Notifications",to_username:"To",insert_code:"Code Block",page_transition:"Page Transition",review_link:"Review Link",rating:"Rating",benefits:"Benefits",disadvantages:"Disadvantages",author:N,cross_link:"Articles:",facebook:"Facebook",twitter:"Twitter",vkontakte:"VK",add_link:"Add Link",subject:"Subject",no_ad:"No Ad in this post",allow_comments:"Allow Comments",max_length:"Max XX characters",custom_css:"Custom CSS",security_question_question:ap,is_related_post:"Enable Related Posts",related_post:"Count of Related Posts",is_related_thread:"Enable Related Threads",related_thread:"Count of Related Threads",from_name:"From:",to_name:"To:",created_date:"Created at",status:"Status",export_data:"export data",export_data_deleted:"Unfortunately export deleted because we store it just in 3 days. If you need fresh data please create a new request.",export_data_deleted_short:"Unfortunately export deleted",export_settings:"Export Settings",security_question_answer:"Answer",messages_before_auto_approved:"How many approved messages for new clients",mark_uncertain_messages:"Moderate manually uncertain messages",thread_limit:"Thread limit",directory_name:"Directory name",nofollow:"Add for links rel=nofollow",blog_social_sizes:"Icon Size",notify_on_moderation:"Notify me about messages\u002Fthreads on moderation",directory:"Directory",summary:"Summary",banned:"Banned",robots_txt:"Robots.txt",toc:bg,one_time:"Show one time",show_every:"Show every",css_selector:"CSS selector",code:"Code",file:"File",question:ap,answer:"You answer",thread_limit_type:"Thread limit by",register_limit_type:"Register limit",trusted:bh,edit_profile:bB,post_limit:"Post limit",post_limit_type:"Post limit by",category_name:"Category name",category_description:"Category Description",blog_category:ac,text_font_family:"Font Family",add_in_menu:"Add In Menu",avatar:"Avatar",template:"Template",old_url:"Old URL",new_url:"New URL",image:bC,is_dark:"Dark Mode",logo:"Logo",topic_name:bD,open_new_tab:"Open in new tab",no_index:"Noindex page by google, bing, yahoo etc...",text_link:"Link text",select_page:"Select page",select_manually:"Manually",parent_category:"Parent Category",link:"Link",name:j,encryption:"Encryption",host:"Host",port:"Port",user:bE,page:"Page",position:"Position",type_background:"Background",no_background:"None",background_color:"Color",background_image:bC,activation:"Activate new members",password:"Password",new_password:"New password",confirm_password:"Confirm password",username:bE,firstname:"First name",favicon:"Favicon",favicon_png:"Favicon PNG",forum_name:j,active:bi,page_name:j,forum_description:am,footer_code:"Footer custom code (ex. google analytics code)",lastname:"Last name",update:bF,email:"Email",company:"Company",message:"Message",text:"Text"},social_size:{small:"Small",medium:"Medium",large:"Large"},buttons:{generate_meta_description:"Generate Meta Description",generate_meta_title:"Generate Meta Title",edit_keyword:"Edit Keyword",new_keyword:"Add Keyword",edit_hash_tag:"Edit Hash Tag",new_hash_tag:"Add Hash Tag",insert_into_content:"Add To Article",duplicate:"Duplicate",tinysrc_all:"Shortener All URLs",ask:"Ask",ask_ai:bo,become_partner:"Become a partner",capitalize:"Capitalize",tinysrc:"Shortener URL",chudoshop:"From Chudoshop",resend_code:"Resend Code",confirm_email:"Confirm Email",amazon:"From Amazon",gde_slon:"From Gde Slon",api_key:bz,settings:"Module Settings",copy:"Copy",new_content_rating:"New Rating",edit_content_rating:"Edit Rating",moderate:"Moderator",content_rating:M,add_redirect:"New Redirect",edit_redirect:"Edit Redirect",new_author:"Add Author",edit_author:"Edit Author",add_ad:bG,edit_ad:"Edit Ad",new_ad:bG,new_file:"New File",deactivate:"Deactivate",activate:"Activate",unbanned:"Unbanned",new_comment:"New Comment",show_more:"Show More",search_more:"Search More",delete_all_comments:"Delete All Comments",bann:bH,new_pm:"Compose",spam:"Mark As Spam",blog_social_share:"Social Share Buttons",new_export:bI,export_data:bI,new_directory:"New Directory",upload_file:"Upload File",upload_directory:"Create folder",edit_category:bJ,edit_thread:"Edit Thread",add_category:bK,add_menu_link:bL,blog_category:t,blog_content:O,edit_menu_link:"Edit link",new_menu:bL,validate_mail_settings:"Validate",new_page:"New Page",read:"Read More",new_blog_category:bK,edit_blog_category:bJ,new_blog_content:"New Post",edit_blog_content:"Edit Post",edit_page:"Edit Page",upload_avatar:"Upload avatar",upload_logo:"Upload Logo",upload_favicon:"Upload Favicon",upload_favicon_png:"Upload Favicon PNG",topic_name:bD,contact_us:T,home:aa,upload_background:"Upload background image",ban:bH,unban:"Unban",delete_all_topic:"Delete all topics",delete_all_messages:"Delete all messages",admin:"Admin Dashboard",profile:bM,signin:"Sign In",register:aq,change_password:u,role:by,logout:"Logout",reply:"Reply to this topic",reply_pm:"Reply",edit_post:"Edit post",new_post:"Add message",save:"Save",edit:"Edit",new_message:"New message",restore_password:w,remove:"Remove",login:v,close:"Close",add:"Add",update:bF,delete:"Delete",reset:"Reset Color Schema",send:"Send",clear:"Clear"},headers:{admin_hash_tags:ad,admin_keywords:ah,search:A,admin_modules:q,forum_home:ai,related_post:"Related Posts",comments:"Comments",related_thread:"Related Threads",edit_profile:bB,admin_members:"Users",export_settings:r,select_export:"Please select export settings",posts_on_moderation:"Posts on moderation",comments_on_moderation:"Comments on moderation",threads_on_moderation:"Threads on moderation",users_on_moderation:"Users on moderation",moderator_home:S,my_pm:Q,moderator_index:"Statistics",admin_home:R,blog:P,change_password:u,admin_redirects:"List of Redirect",admin_ad:"List of Ads",file_settings:s,view_profile:"User profile",color_settings:"Theme Color Settings",primary_color:"Primary Color",secondary_color:"Secondary Color",accent_color:"Accent Color",info_color:"Info Color",success_color:"Success Color",error_color:"Error Color",warning_color:"Warning Color",my_profile:al,register:aq,menu:p,follow_us:"Follow us",about_us:"About us",login:v,restore_password:w,contact_us:T},breadcrumbs:{hash_tag_settings:ad,link_settings:ag,content_rating:M,modules:q,author:bm,menu_home:p,export_settings:r,confirm_account:"Account confirmation",blog_home:P,search:A,forum_home:ae,pm:Q,view_pm:ak,blog_content:O,blog_settings:E,ad_settings:G,file_settings:s,privacy_settings:F,forum_settings:bx,mail_settings:H,layout_settings:I,redirects:"Redirects",blog_categories:t,edit_member:aj,members:bn,pages:J,social_settings:K,security_settings:L,admin_index:R,moderator_comments:D,moderator_threads:bl,moderator_posts:C,moderator_users:B,moderator_index:S,change_password:u,home:aa,view_profile:"User Profile",my_profile:bM,login:v,register:aq,contact_us:T,restore_password:w}}}}}}("",0,false,true,1,"0001-01-01T00:00:00Z",null,99,"#993366","Name","St. Louis Blog","Member","Saint Louis Community","saint-louis-community",69011,"Menu","Modules","Export","File Settings","Blog Category","Change password","Login","Restore password","Books","St. Louis MO places to see","Messages","Search","Moderate Users","Moderate Posts","Moderate Comments","Blog Settings","Privacy Settings","Ad Settings","Mail Settings","Layout Settings","Pages","Social Settings","Security Settings","Ratings","Author","Blog Content","Blog","Personal Messages","Admin dashboard","Moderator dashboard","Contact us","Read more at:","About St. Louis","Services","Products","Programming","How to Implement Batch Normalization In PyTorch?","bottom",14,"Home","\u002F","Category","Hash Tags","Forum","Redirect","Auto Linker Settings","Keywords List","Community Forum","Edit User Details","View Message","My Profile","Description","Picture","Price","Question","Sign Up","default","https:\u002F\u002Fstlplaces.com\u002Fblog\u002Fhow-to-implement-batch-normalization-in-pytorch",114546,"How to Implement Batch Normalization In TensorFlow?","blog\u002F1600d5af-18ea-4205-8fa6-4ffbde52eb24\u002F65b57622dc91f25ac074b50a.png","Batch normalization is a technique used to improve the speed, stability, and performance of neural networks. It works by normalizing the output of the previous layer within each batch of training examples. This helps in mitigating the issue of internal covariate shift, where the distribution of the input to each layer changes during training.To implement batch normalization in TensorFlow, follow these steps:Import the necessary TensorFlow libraries:\nimport tensorflow as tf\nfrom tensorflow.keras.","how-to-implement-batch-normalization-in-tensorflow",124619,"How to Batch Images With Arbitrary Sizes In Tensorflow?","blog\u002F2901b934-4a6b-474d-b8aa-94ff825d19d9\u002F663cbe2a5b80f0a99ffdf3ca.png","To batch images with arbitrary sizes in TensorFlow, you can use the tf.image.resize_with_pad() function to resize the images to a specific size before batching them together. You can specify the target size for resizing the images and pad them if necessary to fit the batch size. Once the images are resized and padded to the desired size, you can use the tf.data.Dataset.batch() function to batch the images together based on the batch size you want.","how-to-batch-images-with-arbitrary-sizes-in",108922,"How to Do Batch Filling In Pytorch?","blog\u002F1d4c2d95-99f0-459a-a7aa-5421cefb8c8d\u002F657b2b0481b5eea5787dfee9.png","Batch filling in PyTorch refers to the process of creating a batch of data from a given dataset. It involves splitting the dataset into smaller batches, which are then used for model training or inference.To perform batch filling in PyTorch, you can follow these steps:Load the dataset: Start by loading your dataset into memory. This could be a collection of images, texts, or any other data format.","how-to-do-batch-filling-in-pytorch",3,"Fitness",2583,"programming","d7c1b18a-4f7a-44c9-8ac5-1b45600ea4ae","stlplaces.com",80,"logo\u002F993366.png","en","User-agent: *\nDisallow: \u002Fsearch\nDisallow: \u002Fadmin\nDisallow: \u002Fprofile\nDisallow: \u002Flogin\nDisallow: \u002Fregister\n\nSitemap: https:\u002F\u002Fstlplaces.com\u002Fsitemap.xml","\u003C!-- Google tag (gtag.js) --\u003E\n\u003Cscript async src=\"https:\u002F\u002Fwww.googletagmanager.com\u002Fgtag\u002Fjs?id=G-YEX948KVXV\"\u003E\u003C\u002Fscript\u003E\n\u003Cscript\u003E\n window.dataLayer = window.dataLayer || [];\n function gtag(){dataLayer.push(arguments);}\n gtag('js', new Date());\n\n gtag('config', 'G-YEX948KVXV');\n\u003C\u002Fscript\u003E\n\n\u003Cscript data-ad-client=\"ca-pub-4833888168110763\" async src=\"https:\u002F\u002Fpagead2.googlesyndication.com\u002Fpagead\u002Fjs\u002Fadsbygoogle.js\"\u003E\u003C\u002Fscript\u003E","email","2020-12-10T08:12:31Z","Looking to learn how to implement Batch Normalization in PyTorch effectively.","forum","Privacy Policy",4,"\u003Cscript async src=\"https:\u002F\u002Fpagead2.googlesyndication.com\u002Fpagead\u002Fjs\u002Fadsbygoogle.js\"\u003E\u003C\u002Fscript\u003E\n\u003C!-- stlplaces --\u003E\n\u003Cins class=\"adsbygoogle\"\n style=\"display:block\"\n data-ad-client=\"ca-pub-4833888168110763\"\n data-ad-slot=\"4211055944\"\n data-ad-format=\"auto\"\n data-full-width-responsive=\"true\"\u003E\u003C\u002Fins\u003E\n\u003Cscript\u003E\n (adsbygoogle = window.adsbygoogle || []).push({});\n\u003C\u002Fscript\u003E","\u003Cscript async src=\"https:\u002F\u002Fpagead2.googlesyndication.com\u002Fpagead\u002Fjs\u002Fadsbygoogle.js\"\u003E\u003C\u002Fscript\u003E\n\u003Cins class=\"adsbygoogle\"\n style=\"display:block\"\n data-ad-format=\"fluid\"\n data-ad-layout-key=\"-ef+6k-30-ac+ty\"\n data-ad-client=\"ca-pub-4833888168110763\"\n data-ad-slot=\"3267362137\"\u003E\u003C\u002Fins\u003E\n\u003Cscript\u003E\n (adsbygoogle = window.adsbygoogle || []).push({});\n\u003C\u002Fscript\u003E",12,"\u003Cp\u003EWhich state is better to move in: Florida or North Carolina?\u003C\u002Fp\u003E","\u003Cp\u003EWhat state is better: Tennessee or Iowa?\u003C\u002Fp\u003E","\u003Cp\u003EWhich state is better to move in: Ohio or Arizona?\u003C\u002Fp\u003E","\u003Cp\u003EWhich state is best to visit: Illinois or North Carolina?\u003C\u002Fp\u003E","\u003Cp\u003EWhich state is better to move in: Ohio or Indiana?\u003C\u002Fp\u003E","Posted Links","Table of Contents","Trusted User","Active","Topics","General Settings","Moderate Threads","Authors","Members","Ask AI","Are you sure you want to delete this category?","Created","Your account was successfully confirmed","Forum Category Settings","List Users","Moderate Thread","New User","Model","Forum Settings","Role","Api Key","Query:","Edit Profile","Image","Title","Username","Update","New Ad","Ban","Export Data","Edit Category","New Category","Add a new menu link","My profile"));</script><script src="https://pub-420acf56315e422bbbdab07717bee8cd.r2.dev/assets/0.1/50d1395.js" defer></script><script src="https://pub-420acf56315e422bbbdab07717bee8cd.r2.dev/assets/0.1/498f8f7.js" defer></script><script src="https://pub-420acf56315e422bbbdab07717bee8cd.r2.dev/assets/0.1/fffc2dc.js" defer></script><script src="https://pub-420acf56315e422bbbdab07717bee8cd.r2.dev/assets/0.1/25d50b7.js" defer></script>
</body>
</html>

Home · About · News · Docs · Terms