{"id":134340,"date":"2026-08-26T15:47:10","date_gmt":"2026-08-26T10:17:10","guid":{"rendered":"https:\/\/www.guvi.in\/blog\/?p=134340"},"modified":"2026-11-19T18:21:57","modified_gmt":"2026-11-19T12:51:57","slug":"batch-normalization-explained","status":"publish","type":"post","link":"https:\/\/www.guvi.in\/blog\/batch-normalization-explained\/","title":{"rendered":"What Is Batch Normalization? Explained Simply"},"content":{"rendered":"\n<p>Batch normalization is a layer that adjusts the values flowing through a neural network so they stay on a similar scale. It does this by using the mean and variance of each mini-batch.<\/p>\n\n\n\n<p>Deep networks often train slowly because the inputs to each layer keep shifting as the earlier weights change. Batch normalization keeps those inputs steady, so the network learns faster and with fewer problems.<\/p>\n\n\n\n<p><\/p>\n\n\n\n<h2 class=\"wp-block-heading\"><strong>TL;DR Summary<\/strong><\/h2>\n\n\n\n<ul>\n<li>Batch normalization is a layer that keeps each layer\u2019s inputs on a steady scale, which makes deep networks train faster and more reliably.<\/li>\n\n\n\n<li>It standardizes values with the mean and variance of each mini-batch, then restores flexibility through the learnable parameters \u03b3 and \u03b2.<\/li>\n\n\n\n<li>Evaluation mode is essential, since inference relies on stored running statistics instead of the current batch.<\/li>\n\n\n\n<li>Very small batches make its statistics noisy, so accuracy can drop when the batch size is limited.<\/li>\n\n\n\n<li>Layer norm suits sequence models, and group norm suits image tasks with small batches.<\/li>\n<\/ul>\n\n\n\n<p><\/p>\n\n\n\n<p><\/p>\n\n\n\n<h2 class=\"wp-block-heading\"><strong>What Is Batch Normalization?<\/strong><\/h2>\n\n\n\n<p><a href=\"https:\/\/www.guvi.in\/hub\/deep-learning-tutorial\/batch-normalization-and-dropout\/?utm_source=blog&amp;utm_medium=hyperlink&amp;utm_campaign=batch-normalization-explained\" target=\"_blank\" rel=\"noreferrer noopener\">Batch normalization<\/a> is a layer that keeps the numbers flowing through a neural network within a steady range. For each small group of training examples, called a mini-batch, it centers the values around zero and evens out their spread. This lets the network learn faster and with fewer training problems.<\/p>\n\n\n\n<p>It processes each feature separately, or each channel in convolutional layers.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>Formula and Intuition<\/strong><\/h3>\n\n\n\n<p>For a mini-batch of size m, the layer computes the following:<\/p>\n\n\n\n<pre class=\"wp-block-code\"><code>mean:        \u03bc = (1\/m) * \u03a3 x_i\nvariance:    \u03c3\u00b2 = (1\/m) * \u03a3 (x_i - \u03bc)\u00b2\nnormalize:   x\u0302_i = (x_i - \u03bc) \/ sqrt(\u03c3\u00b2 + \u03b5)\noutput:      y_i = \u03b3 * x\u0302_i + \u03b2<\/code><\/pre>\n\n\n\n<p>Each symbol means the following:<\/p>\n\n\n\n<ul>\n<li><strong>x_i:<\/strong> the value of one feature for the i-th sample in the batch<\/li>\n\n\n\n<li><strong>\u03bc and \u03c3\u00b2:<\/strong> the mean and variance of that feature across the batch<\/li>\n\n\n\n<li><strong>\u03b5:<\/strong> a small constant, such as 1e-5, that prevents division by zero<\/li>\n\n\n\n<li><strong>\u03b3 and \u03b2:<\/strong> learnable parameters that set the final scale and offset<\/li>\n<\/ul>\n\n\n\n<p>Normalization gives each feature a mean of 0 and a variance of 1, so the values entering a layer stay on a steady scale from one batch to the next. <\/p>\n\n\n\n<p>The layer no longer has to adjust to large swings, allowing higher learning rates and more consistent gradient updates. The parameters \u03b3 and \u03b2 then let the network restore any scale or offset that suits the task.<\/p>\n\n\n\n<p><\/p>\n\n\n\n<p><strong><em>Curious how neural networks really learn? Batch normalization is just one piece of the puzzle. HCL GUVI\u2019s Intel &amp; IITM Pravartak Certified AI\/ML Course walks you through deep learning with PyTorch, step by step, with no prior coding experience needed. Explore the course and start building real AI skills!<\/em><\/strong><\/p>\n\n\n<div class=\"wp-block-guvi-cta-button\" style=\"justify-content:center;\"><a href=\"https:\/\/www.guvi.in\/mlp\/artificial-intelligence-and-machine-learning?utm_source=blog&#038;utm_medium=cta_button&#038;utm_campaign=batch-normalization-explained\" class=\"guvi-btn-enquire\" target=\"_blank\" rel=\"noopener noreferrer\">Enroll in HCL GUVI&#039;s AI\/ML Course<svg class=\"guvi-btn-chevron\" style=\"transform:rotate(0deg);\" width=\"16\" height=\"16\" viewBox=\"0 0 24 24\" fill=\"none\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" aria-hidden=\"true\"><path fill-rule=\"evenodd\" clip-rule=\"evenodd\" d=\"M9.39935 18.1007C9.0674 17.7687 9.0674 17.2305 9.39935 16.8986L13.6922 12.6057C13.7508 12.5471 13.7508 12.4521 13.6922 12.3935L9.39935 8.10065C9.0674 7.7687 9.0674 7.23051 9.39935 6.89857C9.7313 6.56662 10.2695 6.56662 10.6014 6.89857L14.8943 11.1915C15.6168 11.9139 15.6168 13.0853 14.8943 13.8078L10.6014 18.1007C10.2695 18.4326 9.7313 18.4326 9.39935 18.1007Z\" fill=\"currentColor\"\/><\/svg><\/a><\/div>\n\n\n<p><\/p>\n\n\n\n<h2 class=\"wp-block-heading\"><strong>How Batch Normalization Works<\/strong><\/h2>\n\n\n\n<p>During training, a batch normalization layer processes every mini-batch in the following five steps.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>Step 1: Find the Average of the Mini-Batch<\/strong><\/h3>\n\n\n\n<p>The layer adds up the values of one feature across all samples in the mini-batch and divides by the number of samples. This tells the layer where the center of those values is.<\/p>\n\n\n\n<pre class=\"wp-block-code\"><code>\u03bc = (1\/m) * \u03a3 x_i<\/code><\/pre>\n\n\n\n<ul>\n<li><strong>m:<\/strong> the number of samples in the mini-batch<\/li>\n\n\n\n<li><strong>x_i:<\/strong> the value of the feature for the i-th sample<\/li>\n\n\n\n<li><strong>\u03bc:<\/strong> the mean of the feature across the mini-batch<\/li>\n<\/ul>\n\n\n\n<p><\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>Step 2: Measure How Spread Out the Values Are<\/strong><\/h3>\n\n\n\n<p>The layer then checks how far each value sits from that average. The average of these squared distances shows how widely the values are scattered.<\/p>\n\n\n\n<pre class=\"wp-block-code\"><code>\u03c3\u00b2 = (1\/m) * \u03a3 (x_i - \u03bc)\u00b2<\/code><\/pre>\n\n\n\n<ul>\n<li><strong>\u03c3\u00b2:<\/strong> the variance of the feature across the mini-batch<\/li>\n<\/ul>\n\n\n\n<p><\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>Step 3: Normalize the Values<\/strong><\/h3>\n\n\n\n<p>Each value has the mean subtracted, which centers the data at zero. The result is divided by the standard deviation, which brings the spread to a common scale.<\/p>\n\n\n\n<pre class=\"wp-block-code\"><code>x\u0302_i = (x_i - \u03bc) \/ sqrt(\u03c3\u00b2 + \u03b5)<\/code><\/pre>\n\n\n\n<ul>\n<li><strong>x\u0302_i:<\/strong> the normalized value, with a mean near 0 and a variance near 1<\/li>\n\n\n\n<li><strong>\u03b5:<\/strong> a tiny constant (1e-5 by default in PyTorch) that prevents division by zero when the variance is very small<\/li>\n<\/ul>\n\n\n\n<p><\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>Step 4: Scale and Shift the Result<\/strong><\/h3>\n\n\n\n<p>Forcing every feature to the same scale can limit what the network can express. The layer therefore stretches and moves the normalized values using two parameters that the network learns during training.<\/p>\n\n\n\n<pre class=\"wp-block-code\"><code>y_i = \u03b3 * x\u0302_i + \u03b2<\/code><\/pre>\n\n\n\n<ul>\n<li><strong>y_i:<\/strong> the final output passed to the next layer<\/li>\n\n\n\n<li><strong>\u03b3:<\/strong> the learnable scale, which stretches or shrinks the values<\/li>\n\n\n\n<li><strong>\u03b2:<\/strong> the learnable shift, which moves the values up or down<\/li>\n<\/ul>\n\n\n\n<p><\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>Step 5: Update the Running Averages<\/strong><\/h3>\n\n\n\n<p>The layer keeps a running mean and a running variance that are refreshed after every mini-batch. At inference time, these stored values replace the mini-batch statistics, so a prediction does not depend on which other samples are in the batch.<\/p>\n\n\n\n<pre class=\"wp-block-code\"><code>running_mean = (1 - momentum) * running_mean + momentum * \u03bc\nrunning_var  = (1 - momentum) * running_var  + momentum * \u03c3\u00b2<\/code><\/pre>\n\n\n\n<ul>\n<li><strong>running_mean and running_var:<\/strong> the long-term estimates of the mean and variance, used at inference<\/li>\n\n\n\n<li><strong>momentum:<\/strong> controls how quickly the running values follow new batches (0.1 by default in PyTorch)<\/li>\n<\/ul>\n\n\n\n<p>PyTorch updates the running variance with an unbiased estimate, dividing by m &#8211; 1 instead of m. The difference is negligible at normal batch sizes.<\/p>\n\n\n\n<p><\/p>\n\n\n\n<div style=\"background-color: #099f4e; border: 3px solid #110053; border-radius: 12px; padding: 18px 22px; color: #FFFFFF; font-size: 18px; font-family: Montserrat, Helvetica, sans-serif; line-height: 1.6; box-shadow: 0 4px 12px rgba(0, 0, 0, 0.15); max-width: 750px; margin: 22px auto;\">\n  <h4 style=\"margin-top: 0; font-size: 22px; font-weight: 700; color: #ffffff;\">\ud83d\udca1 Did You Know?<\/h4>\n  <p style=\"margin: 10px 0;\">\n    <strong style=\"color: #110053;\">Batch normalization<\/strong> was introduced in 2015 by <strong style=\"color: #110053;\">Sergey Ioffe<\/strong> and <strong style=\"color: #110053;\">Christian Szegedy<\/strong> at Google, and it quickly became a standard building block in deep learning models.\n  <\/p>\n<\/div>\n\n\n\n<p><\/p>\n\n\n\n<h2 class=\"wp-block-heading\"><strong>Why Use Batch Normalization?<\/strong><\/h2>\n\n\n\n<p>Batch normalization is used because it makes deep networks easier and faster to train.<\/p>\n\n\n\n<ul>\n<li><strong>Faster Training:<\/strong> Stable layer inputs allow higher learning rates, so the model reaches good accuracy in fewer epochs.<\/li>\n\n\n\n<li><strong>More Stable Gradients:<\/strong> Keeping activations on a consistent scale lowers the risk of vanishing or exploding gradients in deep networks.<\/li>\n\n\n\n<li><strong>Less Sensitivity to Initialization:<\/strong> Normalizing the inputs of each layer makes training less dependent on the starting weight values.<\/li>\n\n\n\n<li><strong>Mild Regularization:<\/strong> The statistics of each mini-batch add a small amount of noise, which can reduce overfitting.<\/li>\n\n\n\n<li><strong>Easier Hyperparameter Tuning:<\/strong> Training stays stable across a wider range of learning rates, so less trial and error is needed.<\/li>\n<\/ul>\n\n\n\n<p><strong>Also Read:<\/strong> <em><a href=\"https:\/\/www.guvi.in\/blog\/building-a-neural-network-using-pytorch\/\" target=\"_blank\" rel=\"noreferrer noopener\">Building a Neural Network Using PyTorch<\/a><\/em><\/p>\n\n\n\n<p><\/p>\n\n\n\n<h2 class=\"wp-block-heading\"><strong>Batch Normalization Code Example in PyTorch<\/strong><\/h2>\n\n\n\n<p>The example below applies <code><a href=\"https:\/\/docs.pytorch.org\/docs\/2.15\/generated\/torch.nn.BatchNorm1d.html\" target=\"_blank\" rel=\"noreferrer noopener nofollow\">BatchNorm1d<\/a><\/code> to a small batch of four samples with three features each.<\/p>\n\n\n\n<pre class=\"wp-block-code\"><code>import torch\nimport torch.nn as nn\n\nbn = nn.BatchNorm1d(3)\n\nx = torch.tensor(&#91;\n    &#91;1.0, 10.0, 100.0],\n    &#91;2.0, 20.0, 200.0],\n    &#91;3.0, 30.0, 300.0],\n    &#91;4.0, 40.0, 400.0],\n])\n\nbn.train()\ny = bn(x)\nprint(y.detach())<\/code><\/pre>\n\n\n\n<p><strong>Output:<\/strong><\/p>\n\n\n\n<pre class=\"wp-block-code\"><code>tensor(&#91;&#91;-1.3416, -1.3416, -1.3416],\n        &#91;-0.4472, -0.4472, -0.4472],\n        &#91; 0.4472,  0.4472,  0.4472],\n        &#91; 1.3416,  1.3416,  1.3416]])<\/code><\/pre>\n\n\n\n<ul>\n<li><code>nn.BatchNorm1d(3)<\/code> creates a layer for inputs with 3 features, and this number must match the input feature dimension.<\/li>\n\n\n\n<li>The input <code>x<\/code> holds 4 samples, and each feature column has a very different scale (1 to 4, 10 to 40, and 100 to 400).<\/li>\n\n\n\n<li>Calling <code>bn.train()<\/code> puts the layer in training mode, so it normalizes with the mean and variance of the current batch.<\/li>\n\n\n\n<li>The output <code>y<\/code> has identical values in all three columns, because each feature was normalized separately to a mean of 0 and a variance of 1.<\/li>\n\n\n\n<li>The call <code>y.detach()<\/code> removes gradient tracking, so only the values are printed.<\/li>\n\n\n\n<li>After training, calling <code>bn.eval()<\/code> makes the layer use its stored running statistics instead of the batch statistics.<\/li>\n<\/ul>\n\n\n\n<p><\/p>\n\n\n\n<p><\/p>\n\n\n\n<h2 class=\"wp-block-heading\"><strong>Placement and Inference Mode Mistakes<\/strong><\/h2>\n\n\n\n<p>Batch normalization usually goes after a linear or convolutional layer and before the <a href=\"https:\/\/www.guvi.in\/blog\/sigmoid-activation-function\/\" target=\"_blank\" rel=\"noreferrer noopener\">activation function<\/a>. Most errors come from using the wrong mode, batch size, or layer type.<\/p>\n\n\n\n<ul>\n<li><strong>Forgetting <code>model.eval()<\/code>:<\/strong> In training mode, predictions depend on the current batch. Switch to evaluation mode before inference so the stored running statistics are used.<\/li>\n\n\n\n<li><strong>Not calling <code>model.train()<\/code> again:<\/strong> After evaluation, switch back to training mode or the layer will stop using batch statistics.<\/li>\n\n\n\n<li><strong>Very small batches:<\/strong> Statistics from a few samples are noisy and lower accuracy. Use a larger batch, or switch to LayerNorm or GroupNorm.<\/li>\n\n\n\n<li><strong>Wrong layer type:<\/strong> Use <code>BatchNorm1d<\/code> for feature inputs and <code>BatchNorm2d<\/code> for image inputs. The <code>num_features<\/code> value must match the channel count, not the batch size.<\/li>\n\n\n\n<li><strong>Redundant bias:<\/strong> The shift parameter \u03b2 already acts as a bias, so set <code>bias=False<\/code> in the preceding layer.<\/li>\n<\/ul>\n\n\n\n<p><\/p>\n\n\n\n<p><strong><em>Debugging train and eval mode mistakes gets easier when you have experts beside you. In HCL GUVI\u2019s Intel &amp; IITM Pravartak Certified AI\/ML Course, 1:1 doubt sessions and 20+ hands-on projects turn confusing concepts into skills you can trust. Ready to stop guessing and start building with confidence!<\/em><\/strong><\/p>\n\n\n<div class=\"wp-block-guvi-cta-button\" style=\"justify-content:center;\"><a href=\"https:\/\/www.guvi.in\/mlp\/artificial-intelligence-and-machine-learning?utm_source=blog&#038;utm_medium=cta_button&#038;utm_campaign=batch-normalization-explained\" class=\"guvi-btn-enquire\" target=\"_blank\" rel=\"noopener noreferrer\">Enroll in HCL GUVI&#039;s AI\/ML Course<svg class=\"guvi-btn-chevron\" style=\"transform:rotate(0deg);\" width=\"16\" height=\"16\" viewBox=\"0 0 24 24\" fill=\"none\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" aria-hidden=\"true\"><path fill-rule=\"evenodd\" clip-rule=\"evenodd\" d=\"M9.39935 18.1007C9.0674 17.7687 9.0674 17.2305 9.39935 16.8986L13.6922 12.6057C13.7508 12.5471 13.7508 12.4521 13.6922 12.3935L9.39935 8.10065C9.0674 7.7687 9.0674 7.23051 9.39935 6.89857C9.7313 6.56662 10.2695 6.56662 10.6014 6.89857L14.8943 11.1915C15.6168 11.9139 15.6168 13.0853 14.8943 13.8078L10.6014 18.1007C10.2695 18.4326 9.7313 18.4326 9.39935 18.1007Z\" fill=\"currentColor\"\/><\/svg><\/a><\/div>\n\n\n<p><\/p>\n\n\n\n<h2 class=\"wp-block-heading\"><strong>Batch Norm vs Layer Norm vs Group Norm: Key Differences<\/strong><\/h2>\n\n\n\n<p>Batch norm, layer norm, and group norm all normalize activations, and they differ in which values they compute the mean and variance over.<\/p>\n\n\n\n<figure class=\"wp-block-table\"><table><thead><tr><th><strong>Feature<\/strong><\/th><th><strong>Batch Norm<\/strong><\/th><th><strong>Layer Norm<\/strong><\/th><th><strong>Group Norm<\/strong><\/th><\/tr><\/thead><tbody><tr><td><strong>Normalizes Over<\/strong><\/td><td>Each channel, across all samples in the batch and all spatial positions<\/td><td>All features of a single sample<\/td><td>Groups of channels within a single sample, plus spatial positions<\/td><\/tr><tr><td><strong>Depends on Batch Size<\/strong><\/td><td>Yes, statistics come from the mini-batch<\/td><td>No, statistics come from each sample alone<\/td><td>No, statistics come from each sample alone<\/td><\/tr><tr><td><strong>Training vs Inference<\/strong><\/td><td>Uses batch statistics in training and running statistics at inference<\/td><td>Computes the same way in both modes<\/td><td>Computes the same way in both modes<\/td><\/tr><tr><td><strong>Typical Use<\/strong><\/td><td>Convolutional networks for image tasks trained with moderate to large batches<\/td><td>Transformers and other sequence models in NLP<\/td><td>Image tasks with small batches, such as object detection and segmentation<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n<p><\/p>\n\n\n\n<h2 class=\"wp-block-heading\"><strong>Conclusion<\/strong><\/h2>\n\n\n\n<p>Batch normalization keeps the inputs of each layer on a steady scale, and it delivers its full benefit only when its training and inference behavior is handled correctly. Layer norm and group norm apply the same idea where small batches or sequence models make batch statistics unreliable. Together, they cover most normalization needs of modern architectures.<\/p>\n\n\n\n<p><\/p>\n\n\n\n<h2 class=\"wp-block-heading\"><strong>FAQs<\/strong><\/h2>\n\n\n<div id=\"rank-math-faq\" class=\"rank-math-block\">\n<div class=\"rank-math-list \">\n<div id=\"faq-question-1795091162192\" class=\"rank-math-list-item\">\n<h3 class=\"rank-math-question \">1. What is batch normalization in simple terms?<\/h3>\n<div class=\"rank-math-answer \">\n\n<p>Batch normalization is a layer that rescales the values inside a neural network so they stay on a steady scale, which helps the network train faster.<\/p>\n\n<\/div>\n<\/div>\n<div id=\"faq-question-1795091176074\" class=\"rank-math-list-item\">\n<h3 class=\"rank-math-question \">2. Why is batch normalization used in deep learning?<\/h3>\n<div class=\"rank-math-answer \">\n\n<p>It stabilizes layer inputs, which allows higher learning rates, steadier gradients, and faster convergence.<\/p>\n\n<\/div>\n<\/div>\n<div id=\"faq-question-1795091176714\" class=\"rank-math-list-item\">\n<h3 class=\"rank-math-question \">3. Where should batch normalization be placed in a neural network?<\/h3>\n<div class=\"rank-math-answer \">\n\n<p>It is usually placed after a linear or convolutional layer and before the activation function.<\/p>\n\n<\/div>\n<\/div>\n<div id=\"faq-question-1795091177645\" class=\"rank-math-list-item\">\n<h3 class=\"rank-math-question \">4. What is the difference between batch normalization and layer normalization?<\/h3>\n<div class=\"rank-math-answer \">\n\n<p>Batch normalization computes statistics across the mini-batch, while layer normalization computes them across the features of a single sample.<\/p>\n\n<\/div>\n<\/div>\n<div id=\"faq-question-1795091178556\" class=\"rank-math-list-item\">\n<h3 class=\"rank-math-question \">5. Does batch normalization work with small batch sizes?<\/h3>\n<div class=\"rank-math-answer \">\n\n<p>It works poorly because statistics from a few samples are noisy, so layer norm or group norm is the better choice.<\/p>\n\n<\/div>\n<\/div>\n<\/div>\n<\/div>","protected":false},"excerpt":{"rendered":"<p>Batch normalization is a layer that adjusts the values flowing through a neural network so they stay on a similar scale. It does this by using the mean and variance of each mini-batch. Deep networks often train slowly because the inputs to each layer keep shifting as the earlier weights change. Batch normalization keeps those [&hellip;]<\/p>\n","protected":false},"author":64,"featured_media":135764,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[933],"tags":[],"views":"463","authorinfo":{"name":"Abhishek Pati","url":"https:\/\/www.guvi.in\/blog\/author\/abhishek-pati\/"},"thumbnailURL":"https:\/\/www.guvi.in\/blog\/wp-content\/uploads\/2026\/08\/batch-normalization-explained-300x117.webp","_links":{"self":[{"href":"https:\/\/www.guvi.in\/blog\/wp-json\/wp\/v2\/posts\/134340"}],"collection":[{"href":"https:\/\/www.guvi.in\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.guvi.in\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.guvi.in\/blog\/wp-json\/wp\/v2\/users\/64"}],"replies":[{"embeddable":true,"href":"https:\/\/www.guvi.in\/blog\/wp-json\/wp\/v2\/comments?post=134340"}],"version-history":[{"count":14,"href":"https:\/\/www.guvi.in\/blog\/wp-json\/wp\/v2\/posts\/134340\/revisions"}],"predecessor-version":[{"id":149036,"href":"https:\/\/www.guvi.in\/blog\/wp-json\/wp\/v2\/posts\/134340\/revisions\/149036"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.guvi.in\/blog\/wp-json\/wp\/v2\/media\/135764"}],"wp:attachment":[{"href":"https:\/\/www.guvi.in\/blog\/wp-json\/wp\/v2\/media?parent=134340"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.guvi.in\/blog\/wp-json\/wp\/v2\/categories?post=134340"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.guvi.in\/blog\/wp-json\/wp\/v2\/tags?post=134340"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}