{
    "componentChunkName": "component---src-templates-note-note-js",
    "path": "/notes/SFT-versus-offline-learning",
    "result": {"data":{"mdx":{"body":"var _excluded = [\"components\"];\nfunction _extends() { _extends = Object.assign ? Object.assign.bind() : function (target) { for (var i = 1; i < arguments.length; i++) { var source = arguments[i]; for (var key in source) { if (Object.prototype.hasOwnProperty.call(source, key)) { target[key] = source[key]; } } } return target; }; return _extends.apply(this, arguments); }\nfunction _objectWithoutProperties(source, excluded) { if (source == null) return {}; var target = _objectWithoutPropertiesLoose(source, excluded); var key, i; if (Object.getOwnPropertySymbols) { var sourceSymbolKeys = Object.getOwnPropertySymbols(source); for (i = 0; i < sourceSymbolKeys.length; i++) { key = sourceSymbolKeys[i]; if (excluded.indexOf(key) >= 0) continue; if (!Object.prototype.propertyIsEnumerable.call(source, key)) continue; target[key] = source[key]; } } return target; }\nfunction _objectWithoutPropertiesLoose(source, excluded) { if (source == null) return {}; var target = {}; var sourceKeys = Object.keys(source); var key, i; for (i = 0; i < sourceKeys.length; i++) { key = sourceKeys[i]; if (excluded.indexOf(key) >= 0) continue; target[key] = source[key]; } return target; }\n/* @jsxRuntime classic */\n/* @jsx mdx */\n\nvar _frontmatter = {};\nvar layoutProps = {\n  _frontmatter: _frontmatter\n};\nvar MDXLayout = \"wrapper\";\nreturn function MDXContent(_ref) {\n  var components = _ref.components,\n    props = _objectWithoutProperties(_ref, _excluded);\n  return mdx(MDXLayout, _extends({}, layoutProps, props, {\n    components: components,\n    mdxType: \"MDXLayout\"\n  }), mdx(\"p\", null, \"Offline learning in machine learning refers to a training approach where a model is trained on a fixed dataset that is not updated during the learning process\", \"[1]\", \". This method, also known as batch learning, involves training the model using all available data at once, rather than incrementally updating it as new data becomes available\", \"[2][5]\", \".\"), mdx(\"p\", null, \"In offline learning:\"), mdx(\"ol\", null, mdx(\"li\", {\n    parentName: \"ol\"\n  }, \"The entire dataset is used for training at once\", \"[2]\", \".\"), mdx(\"li\", {\n    parentName: \"ol\"\n  }, \"The model's parameters are updated after processing the complete dataset\", \"[2]\", \".\"), mdx(\"li\", {\n    parentName: \"ol\"\n  }, \"Once trained, the model remains static until retrained on new data\", \"[5]\", \".\")), mdx(\"h2\", null, \"Relation to Supervised Fine-Tuning (SFT)\"), mdx(\"p\", null, \"Supervised Fine-Tuning (SFT) is a technique often used in transfer learning, particularly for large language models. It shares some similarities with offline learning:\"), mdx(\"ol\", null, mdx(\"li\", {\n    parentName: \"ol\"\n  }, mdx(\"p\", {\n    parentName: \"li\"\n  }, mdx(\"strong\", {\n    parentName: \"p\"\n  }, \"Fixed Dataset\"), \": Both SFT and offline learning typically use a predetermined dataset for training\", \"[1][5]\", \".\")), mdx(\"li\", {\n    parentName: \"ol\"\n  }, mdx(\"p\", {\n    parentName: \"li\"\n  }, mdx(\"strong\", {\n    parentName: \"p\"\n  }, \"Batch Processing\"), \": SFT, like offline learning, often processes data in batches to update the model's parameters\", \"[2]\", \".\")), mdx(\"li\", {\n    parentName: \"ol\"\n  }, mdx(\"p\", {\n    parentName: \"li\"\n  }, mdx(\"strong\", {\n    parentName: \"p\"\n  }, \"Model Stability\"), \": After training, both approaches result in a static model that doesn't automatically adapt to new data without retraining\", \"[5]\", \".\"))), mdx(\"p\", null, \"However, SFT differs from traditional offline learning in some key aspects:\"), mdx(\"ol\", null, mdx(\"li\", {\n    parentName: \"ol\"\n  }, mdx(\"p\", {\n    parentName: \"li\"\n  }, mdx(\"strong\", {\n    parentName: \"p\"\n  }, \"Pre-trained Base\"), \": SFT starts with a pre-trained model and fine-tunes it for specific tasks, while offline learning may train a model from scratch\", \"[1]\", \".\")), mdx(\"li\", {\n    parentName: \"ol\"\n  }, mdx(\"p\", {\n    parentName: \"li\"\n  }, mdx(\"strong\", {\n    parentName: \"p\"\n  }, \"Task Specificity\"), \": SFT is often used to adapt a general model to specific tasks or domains, whereas offline learning can be used for various types of models and tasks\", \"[2]\", \".\")), mdx(\"li\", {\n    parentName: \"ol\"\n  }, mdx(\"p\", {\n    parentName: \"li\"\n  }, mdx(\"strong\", {\n    parentName: \"p\"\n  }, \"Data Efficiency\"), \": SFT can often achieve good results with smaller datasets compared to training a model from scratch, making it more data-efficient in many cases\", \"[5]\", \".\"))), mdx(\"p\", null, \"In the context of AI development, offline learning and SFT are both valuable approaches, with SFT being particularly useful for adapting large, pre-trained models to specific applications while maintaining the benefits of offline learning's simplicity and stability.\"), mdx(\"p\", null, \"Citations:\\n\", \"[1]\", \" \", mdx(\"a\", {\n    parentName: \"p\",\n    \"href\": \"https://en.wikipedia.org/wiki/Offline_learning\"\n  }, \"https://en.wikipedia.org/wiki/Offline_learning\"), \"\\n\", \"[2]\", \" \", mdx(\"a\", {\n    parentName: \"p\",\n    \"href\": \"https://www.baeldung.com/cs/online-vs-offline-learning\"\n  }, \"https://www.baeldung.com/cs/online-vs-offline-learning\"), \"\\n\", \"[3]\", \" \", mdx(\"a\", {\n    parentName: \"p\",\n    \"href\": \"https://www.qwak.com/post/online-vs-offline-machine-learning-whats-the-difference\"\n  }, \"https://www.qwak.com/post/online-vs-offline-machine-learning-whats-the-difference\"), \"\\n\", \"[4]\", \" \", mdx(\"a\", {\n    parentName: \"p\",\n    \"href\": \"https://stats.stackexchange.com/questions/897/online-vs-offline-learning\"\n  }, \"https://stats.stackexchange.com/questions/897/online-vs-offline-learning\"), \"\\n\", \"[5]\", \" \", mdx(\"a\", {\n    parentName: \"p\",\n    \"href\": \"https://www.geeksforgeeks.org/batch-offline-learning-vs-online-learning-in-artificial-intelligence/\"\n  }, \"https://www.geeksforgeeks.org/batch-offline-learning-vs-online-learning-in-artificial-intelligence/\"), \"\\n\", \"[6]\", \" \", mdx(\"a\", {\n    parentName: \"p\",\n    \"href\": \"https://learn.microsoft.com/zh-tw/dotnet/machine-learning/mldotnet-api\"\n  }, \"https://learn.microsoft.com/zh-tw/dotnet/machine-learning/mldotnet-api\"), \"\\n\", \"[7]\", \" \", mdx(\"a\", {\n    parentName: \"p\",\n    \"href\": \"https://developers.google.com/machine-learning/glossary\"\n  }, \"https://developers.google.com/machine-learning/glossary\"), \"\\n\", \"[8]\", \" \", mdx(\"a\", {\n    parentName: \"p\",\n    \"href\": \"https://www.nettechindia.com/machine-learning-course\"\n  }, \"https://www.nettechindia.com/machine-learning-course\")));\n}\n;\nMDXContent.isMDXComponent = true;","inboundReferences":[]}},"pageContext":{"slug":"SFT-versus-offline-learning","node":{"childMdx":{"slug":"SFT-versus-offline-learning"}}}},
    "staticQueryHashes": []}