{"id":67655,"date":"2026-02-26T20:04:14","date_gmt":"2026-02-26T14:34:14","guid":{"rendered":"https:\/\/www.nextias.com\/ca\/?p=67655"},"modified":"2026-02-26T20:32:37","modified_gmt":"2026-02-26T15:02:37","slug":"llm-training-indian-firms","status":"publish","type":"post","link":"https:\/\/www.nextias.com\/ca\/current-affairs\/26-02-2026\/llm-training-indian-firms","title":{"rendered":"Training of Large Language Models (LLMs) by Indian Firms"},"content":{"rendered":"\n<p><strong>Syllabus: GS3\/ Science and Technology<\/strong><\/p>\n\n\n\n<h2 class=\"wp-block-heading\"><strong>Context<\/strong><\/h2>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Bengaluru-based startup Sarvam AI unveiled two indigenous Large Language Models (LLMs), underscoring India\u2019s push for sovereign, multilingual, and compute-efficient AI amid global competition.<\/li>\n<\/ul>\n\n\n\n<h2 class=\"wp-block-heading\"><strong>Large Language Models (LLMs<\/strong><strong>)<\/strong><\/h2>\n\n\n\n<ul class=\"wp-block-list\">\n<li>A large language model (LLM) is a type of artificial intelligence (AI) algorithm that uses <strong>deep learning techniques <\/strong>and massively large data sets to understand, summarize, generate and predict new content.<\/li>\n\n\n\n<li>Deep learning involves the <strong>probabilistic analysis of unstructured data<\/strong>, which eventually enables the deep learning model to recognize distinctions between pieces of content without human intervention.<\/li>\n\n\n\n<li>It helps to <strong>understand how characters, words, and sentences<\/strong> function together.<\/li>\n<\/ul>\n\n\n\n<h2 class=\"wp-block-heading\"><strong>Indigenous LLM Ecosystem in India<\/strong><\/h2>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>Sarvam AI Models:<\/strong> Focus on efficiency, accuracy, and Indian language capabilities. Intended to be open-source, though broader public scrutiny is ongoing.<\/li>\n\n\n\n<li><strong>BharatGen,<\/strong> incubated at IIT Bombay, trained a multilingual <strong>17-billion-parameter <\/strong>model for sectors like education and healthcare.<\/li>\n\n\n\n<li><strong>Gnani.ai<\/strong> launched compact speech and text-to-speech models.<\/li>\n<\/ul>\n\n\n\n<h2 class=\"wp-block-heading\"><strong>How LLMs Are Trained<\/strong><strong>?<\/strong><\/h2>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>GPU Clusters:<\/strong> LLM training requires massive computational power using clusters of Graphics Processing Units (GPUs). Thousands of GPUs operate simultaneously for weeks or months.<\/li>\n\n\n\n<li><strong>Data as the Core Input:<\/strong> Training relies on enormous datasets, often scraped from the Internet.<\/li>\n\n\n\n<li><strong>Model Parameters: <\/strong>Parameters represent the internal weights through which models learn patterns. Sarvam AI trained models with 35 billion and 105 billion parameters.\n<ul class=\"wp-block-list\">\n<li><strong>Larger parameter <\/strong>counts improve capability but require more computation.<\/li>\n<\/ul>\n<\/li>\n<\/ul>\n\n\n\n<h2 class=\"wp-block-heading\"><strong>Key Training Methodologies Used<\/strong><\/h2>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>Data Curation:<\/strong> It focuses on <strong>collecting high-quality datasets<\/strong> in Indian languages.\n<ul class=\"wp-block-list\">\n<li>It includes government documents, literature, media, and synthetic data generation.<\/li>\n\n\n\n<li>It is critical for improving performance beyond English-centric AI systems.<\/li>\n<\/ul>\n<\/li>\n\n\n\n<li><strong>Pre-Training:<\/strong> The models learn <strong>general language patterns<\/strong> by predicting the next token in large unlabelled datasets.\n<ul class=\"wp-block-list\">\n<li>This stage builds foundational reasoning and grammar capabilities.<\/li>\n<\/ul>\n<\/li>\n\n\n\n<li><strong>Fine-Tuning: <\/strong>Models are adapted for specific tasks using curated datasets.\n<ul class=\"wp-block-list\">\n<li>Tools such as <strong>Hugging Face<\/strong> and <strong>LangChain<\/strong> support instruction tuning, classification, and domain adaptation.<\/li>\n<\/ul>\n<\/li>\n\n\n\n<li><strong>Alignment\/RLHF (Reinforcement Learning from Human Feedback): <\/strong>Human raters rank model outputs to teach it to be safer, more accurate, and better aligned with human intent, discouraging harmful or biased responses.<\/li>\n<\/ul>\n\n\n\n<h2 class=\"wp-block-heading\"><strong>Challenges in Training LLMs in India<\/strong><\/h2>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>Limited Indian Language Data: <\/strong>Scarcity of high-quality datasets in Indian languages reduces model performance.\n<ul class=\"wp-block-list\">\n<li>Many systems rely on translation into English before processing, increasing token usage and latency. Suboptimal native performance affects adoption among non-English users.<\/li>\n<\/ul>\n<\/li>\n\n\n\n<li><strong>High Capital Requirements:<\/strong> Training frontier models demands substantial financial investment. Startups often lack immediate commercial returns to justify such costs.<\/li>\n\n\n\n<li><strong>Infrastructure Constraints: <\/strong>Access to high-end computing facilities remains limited without government support.<\/li>\n<\/ul>\n\n\n\n<p><\/p>\n\n\n\n<div class=\"wp-block-group\"><div class=\"wp-block-group__inner-container is-layout-constrained wp-block-group-is-layout-constrained\">\n<div class=\"wp-block-group has-background\" style=\"background-color:#fff2cc\"><div class=\"wp-block-group__inner-container is-layout-constrained wp-block-group-is-layout-constrained\">\n<p><strong>IndiaAI Mission<\/strong><\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>The IndiaAI Mission is the <strong>flagship initiative<\/strong> to build a comprehensive, sovereign AI ecosystem for India.<\/li>\n\n\n\n<li>It focuses on developing high-performance computer infrastructure, indigenous foundational models, and safe, ethical AI, under the vision of &#8220;Making AI in India and Making AI Work for India&#8221;.<\/li>\n\n\n\n<li>India has achieved <strong>38,000 GPUs, <\/strong>providing affordable access to world-class AI resources.<\/li>\n\n\n\n<li><strong>A GPU or Graphics Processing Unit<\/strong> is a powerful computer chip that helps machines think faster, process images, run AI programs, and handle complex tasks more efficiently than a regular processor.<\/li>\n<\/ul>\n\n\n<div class=\"wp-block-image\">\n<figure class=\"aligncenter size-full is-resized\"><img data-dominant-color=\"e8ebf0\" data-has-transparency=\"false\" loading=\"lazy\" decoding=\"async\" width=\"611\" height=\"465\" src=\"https:\/\/wp-images.nextias.com\/cdn-cgi\/image\/format=auto\/ca\/uploads\/2026\/02\/image-140.png\" alt=\"LLMs\" class=\"not-transparent wp-image-67660\" style=\"--dominant-color: #e8ebf0; width:397px;height:auto\" srcset=\"https:\/\/wp-images.nextias.com\/cdn-cgi\/image\/format=auto\/ca\/uploads\/2026\/02\/image-140.png 611w, https:\/\/wp-images.nextias.com\/cdn-cgi\/image\/format=auto\/ca\/uploads\/2026\/02\/image-140-300x228.png 300w\" sizes=\"auto, (max-width: 611px) 100vw, 611px\" \/><\/figure>\n<\/div><\/div><\/div>\n\n\n\n<p><\/p>\n<\/div><\/div>\n\n\n\n<p><strong>Source: <\/strong><a href=\"https:\/\/www.thehindu.com\/sci-tech\/technology\/how-are-indian-firms-training-llms-explained\/article70676898.ece\" target=\"_blank\" rel=\"noopener\"><strong>TH<\/strong><\/a><\/p>\n","protected":false},"excerpt":{"rendered":"<p><strong> Context <\/strong><\/p>\n<li class=\"ms-5\"> Bengaluru-based startup Sarvam AI unveiled two indigenous Large Language Models (LLMs), underscoring India\u2019s push for sovereign, multilingual, and compute-efficient AI amid global competition. <\/li>\n<p><\/p>\n<p><strong> Large Language Models (LLMs) <\/strong><\/p>\n<li class=\"ms-5\"> A large language model (LLM) is a type of artificial intelligence (AI) algorithm that uses deep learning techniques and massively large data sets to understand, summarize, generate and predict new content. <\/li>\n<li class=\"ms-5\"> Deep learning involves the probabilistic analysis of unstructured data, which eventually enables the deep learning model to recognize distinctions between pieces of content without human intervention. <\/li>\n<li class=\"ms-5\"> It helps to understand how characters, words, and sentences function together. <\/li>\n<p><a href=\" https:\/\/www.nextias.com\/ca\/current-affairs\/26-02-2026\/llm-training-indian-firms \" class=\"btn btn-primary btn-sm float-end\">Read More<\/a><\/p>\n","protected":false},"author":4,"featured_media":0,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[21],"tags":[],"class_list":["post-67655","post","type-post","status-publish","format-standard","hentry","category-current-affairs"],"acf":[],"jetpack_featured_media_url":"","_links":{"self":[{"href":"https:\/\/www.nextias.com\/ca\/wp-json\/wp\/v2\/posts\/67655","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.nextias.com\/ca\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.nextias.com\/ca\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.nextias.com\/ca\/wp-json\/wp\/v2\/users\/4"}],"replies":[{"embeddable":true,"href":"https:\/\/www.nextias.com\/ca\/wp-json\/wp\/v2\/comments?post=67655"}],"version-history":[{"count":6,"href":"https:\/\/www.nextias.com\/ca\/wp-json\/wp\/v2\/posts\/67655\/revisions"}],"predecessor-version":[{"id":67670,"href":"https:\/\/www.nextias.com\/ca\/wp-json\/wp\/v2\/posts\/67655\/revisions\/67670"}],"wp:attachment":[{"href":"https:\/\/www.nextias.com\/ca\/wp-json\/wp\/v2\/media?parent=67655"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.nextias.com\/ca\/wp-json\/wp\/v2\/categories?post=67655"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.nextias.com\/ca\/wp-json\/wp\/v2\/tags?post=67655"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}