{"id":441,"date":"2024-03-24T16:22:46","date_gmt":"2024-03-24T15:22:46","guid":{"rendered":"https:\/\/blog.zhaw.ch\/high-performance\/?p=441"},"modified":"2024-11-12T17:51:05","modified_gmt":"2024-11-12T16:51:05","slug":"train-analog-devices-max78002-directly-from-jupyter-notebook","status":"publish","type":"post","link":"https:\/\/blog.zhaw.ch\/high-performance\/2024\/03\/24\/train-analog-devices-max78002-directly-from-jupyter-notebook\/","title":{"rendered":"Train Analog Devices MAX78002 directly from Jupyter Notebook"},"content":{"rendered":"\n

For the past several months, we have been deep in the trenches with the\u00a0ai8x-training<\/a>\u00a0tool and the training of various Convolutional Neural Network (CNN) architectures tailored explicitly for the MAX78000 and MAX78002 devices. Nevertheless, the ai8x training<\/a> tool was more of a hindrance than a help.<\/p>\n\n\n\n

Among the myriad challenges we encountered, the inability to make real-time adjustments, fine-tune models while freezing or unfreezing specific layers, and transfer custom weights proved to be significant pain points. These restrictions prevented efficient and flexible development.<\/p>\n\n\n\n

But, we found a way to train the MAX78000 and MAX78002 devices right from your Jupyter notebook. With this approach, you’ll break free from the shackles of real-time debugging and fine-tuning woes, ensuring seamless interaction with your neural networks. Let’s dive into the nitty-gritty of this process and unlock the full potential of these devices.<\/p>\n\n\n\n

This guide<\/a> will walk you through\u00a0training MAX7800x directly from a Jupyter notebook<\/strong>. The best part? You won’t need the\u00a0ai8x-training<\/a>\u00a0tool for this process.<\/p>\n\n\n\n

The example provided at https:\/\/github.com\/InES-HPMM\/MAX7800x-Jupyter-training\/tree\/main<\/a> includes a simple classifier for the Fashion MNIST<\/a> dataset.<\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n

Training for only 7 epochs + 1 QAT epoch will lead to an accuracy of 99% on the test set! And the model is directly ready for deployment on the MAX78000\/MAX78002.<\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n

Enjoy!<\/strong><\/p>\n\n\n\n

<\/p>\n","protected":false},"excerpt":{"rendered":"

For the past several months, we have been deep in the trenches with the\u00a0ai8x-training\u00a0tool and the training of various Convolutional Neural Network (CNN) architectures tailored explicitly for the MAX78000 and MAX78002 devices. Nevertheless, the ai8x training tool was more of a hindrance than a help. Among the myriad challenges we encountered, the inability to make […]<\/p>\n","protected":false},"author":270,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"ngg_post_thumbnail":0,"footnotes":""},"categories":[1],"tags":[],"features":[],"class_list":["post-441","post","type-post","status-publish","format-standard","hentry","category-allgemein"],"yoast_head":"\nTrain Analog Devices MAX78002 directly from Jupyter Notebook - Embedded High Performance Multimedia Blog<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/blog.zhaw.ch\/high-performance\/2024\/03\/24\/train-analog-devices-max78002-directly-from-jupyter-notebook\/\" \/>\n<meta property=\"og:locale\" content=\"en_GB\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Train Analog Devices MAX78002 directly from Jupyter Notebook\" \/>\n<meta property=\"og:description\" content=\"For the past several months, we have been deep in the trenches with the\u00a0ai8x-training\u00a0tool and the training of various Convolutional Neural Network (CNN) architectures tailored explicitly for the MAX78000 and MAX78002 devices. Nevertheless, the ai8x training tool was more of a hindrance than a help. Among the myriad challenges we encountered, the inability to make […]\" \/>\n<meta property=\"og:url\" content=\"https:\/\/blog.zhaw.ch\/high-performance\/2024\/03\/24\/train-analog-devices-max78002-directly-from-jupyter-notebook\/\" \/>\n<meta property=\"og:site_name\" content=\"Embedded High Performance Multimedia Blog\" \/>\n<meta property=\"article:published_time\" content=\"2024-03-24T15:22:46+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2024-11-12T16:51:05+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/blog.zhaw.ch\/high-performance\/files\/2024\/03\/fashion_mnist-3.0.1.png\" \/>\n<meta name=\"author\" content=\"rosn\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"Written by\" \/>\n\t<meta name=\"twitter:data1\" content=\"rosn\" \/>\n\t<meta name=\"twitter:label2\" content=\"Estimated reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"2 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"Article\",\"@id\":\"https:\/\/blog.zhaw.ch\/high-performance\/2024\/03\/24\/train-analog-devices-max78002-directly-from-jupyter-notebook\/#article\",\"isPartOf\":{\"@id\":\"https:\/\/blog.zhaw.ch\/high-performance\/2024\/03\/24\/train-analog-devices-max78002-directly-from-jupyter-notebook\/\"},\"author\":{\"name\":\"rosn\",\"@id\":\"https:\/\/blog.zhaw.ch\/high-performance\/#\/schema\/person\/e1a77329f74257615afa71bc883106c9\"},\"headline\":\"Train Analog Devices MAX78002 directly from Jupyter Notebook\",\"datePublished\":\"2024-03-24T15:22:46+00:00\",\"dateModified\":\"2024-11-12T16:51:05+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\/\/blog.zhaw.ch\/high-performance\/2024\/03\/24\/train-analog-devices-max78002-directly-from-jupyter-notebook\/\"},\"wordCount\":233,\"commentCount\":0,\"image\":{\"@id\":\"https:\/\/blog.zhaw.ch\/high-performance\/2024\/03\/24\/train-analog-devices-max78002-directly-from-jupyter-notebook\/#primaryimage\"},\"thumbnailUrl\":\"https:\/\/blog.zhaw.ch\/high-performance\/files\/2024\/03\/fashion_mnist-3.0.1.png\",\"articleSection\":[\"Allgemein\"],\"inLanguage\":\"en-GB\",\"potentialAction\":[{\"@type\":\"CommentAction\",\"name\":\"Comment\",\"target\":[\"https:\/\/blog.zhaw.ch\/high-performance\/2024\/03\/24\/train-analog-devices-max78002-directly-from-jupyter-notebook\/#respond\"]}]},{\"@type\":\"WebPage\",\"@id\":\"https:\/\/blog.zhaw.ch\/high-performance\/2024\/03\/24\/train-analog-devices-max78002-directly-from-jupyter-notebook\/\",\"url\":\"https:\/\/blog.zhaw.ch\/high-performance\/2024\/03\/24\/train-analog-devices-max78002-directly-from-jupyter-notebook\/\",\"name\":\"Train Analog Devices MAX78002 directly from Jupyter Notebook - Embedded High Performance Multimedia Blog\",\"isPartOf\":{\"@id\":\"https:\/\/blog.zhaw.ch\/high-performance\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\/\/blog.zhaw.ch\/high-performance\/2024\/03\/24\/train-analog-devices-max78002-directly-from-jupyter-notebook\/#primaryimage\"},\"image\":{\"@id\":\"https:\/\/blog.zhaw.ch\/high-performance\/2024\/03\/24\/train-analog-devices-max78002-directly-from-jupyter-notebook\/#primaryimage\"},\"thumbnailUrl\":\"https:\/\/blog.zhaw.ch\/high-performance\/files\/2024\/03\/fashion_mnist-3.0.1.png\",\"datePublished\":\"2024-03-24T15:22:46+00:00\",\"dateModified\":\"2024-11-12T16:51:05+00:00\",\"author\":{\"@id\":\"https:\/\/blog.zhaw.ch\/high-performance\/#\/schema\/person\/e1a77329f74257615afa71bc883106c9\"},\"breadcrumb\":{\"@id\":\"https:\/\/blog.zhaw.ch\/high-performance\/2024\/03\/24\/train-analog-devices-max78002-directly-from-jupyter-notebook\/#breadcrumb\"},\"inLanguage\":\"en-GB\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/blog.zhaw.ch\/high-performance\/2024\/03\/24\/train-analog-devices-max78002-directly-from-jupyter-notebook\/\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"en-GB\",\"@id\":\"https:\/\/blog.zhaw.ch\/high-performance\/2024\/03\/24\/train-analog-devices-max78002-directly-from-jupyter-notebook\/#primaryimage\",\"url\":\"https:\/\/blog.zhaw.ch\/high-performance\/files\/2024\/03\/fashion_mnist-3.0.1.png\",\"contentUrl\":\"https:\/\/blog.zhaw.ch\/high-performance\/files\/2024\/03\/fashion_mnist-3.0.1.png\",\"width\":900,\"height\":900},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\/\/blog.zhaw.ch\/high-performance\/2024\/03\/24\/train-analog-devices-max78002-directly-from-jupyter-notebook\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Startseite\",\"item\":\"https:\/\/blog.zhaw.ch\/high-performance\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Train Analog Devices MAX78002 directly from Jupyter Notebook\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/blog.zhaw.ch\/high-performance\/#website\",\"url\":\"https:\/\/blog.zhaw.ch\/high-performance\/\",\"name\":\"Embedded High Performance Multimedia Blog\",\"description\":\"A Blog of the ZHAW Zurich University of Applied Sciences\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/blog.zhaw.ch\/high-performance\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"en-GB\"},{\"@type\":\"Person\",\"@id\":\"https:\/\/blog.zhaw.ch\/high-performance\/#\/schema\/person\/e1a77329f74257615afa71bc883106c9\",\"name\":\"rosn\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-GB\",\"@id\":\"https:\/\/secure.gravatar.com\/avatar\/eec853590b797a7cdaea9a010f9e4424a21362148edf1b8d8d51fc8c5bced337?s=96&d=mm&r=g\",\"url\":\"https:\/\/secure.gravatar.com\/avatar\/eec853590b797a7cdaea9a010f9e4424a21362148edf1b8d8d51fc8c5bced337?s=96&d=mm&r=g\",\"contentUrl\":\"https:\/\/secure.gravatar.com\/avatar\/eec853590b797a7cdaea9a010f9e4424a21362148edf1b8d8d51fc8c5bced337?s=96&d=mm&r=g\",\"caption\":\"rosn\"},\"url\":\"https:\/\/blog.zhaw.ch\/high-performance\/author\/rosn\/\"}]}<\/script>\n<!-- \/ Yoast SEO Premium plugin. -->","yoast_head_json":{"title":"Train Analog Devices MAX78002 directly from Jupyter Notebook - Embedded High Performance Multimedia Blog","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/blog.zhaw.ch\/high-performance\/2024\/03\/24\/train-analog-devices-max78002-directly-from-jupyter-notebook\/","og_locale":"en_GB","og_type":"article","og_title":"Train Analog Devices MAX78002 directly from Jupyter Notebook","og_description":"For the past several months, we have been deep in the trenches with the\u00a0ai8x-training\u00a0tool and the training of various Convolutional Neural Network (CNN) architectures tailored explicitly for the MAX78000 and MAX78002 devices. Nevertheless, the ai8x training tool was more of a hindrance than a help. Among the myriad challenges we encountered, the inability to make […]","og_url":"https:\/\/blog.zhaw.ch\/high-performance\/2024\/03\/24\/train-analog-devices-max78002-directly-from-jupyter-notebook\/","og_site_name":"Embedded High Performance Multimedia Blog","article_published_time":"2024-03-24T15:22:46+00:00","article_modified_time":"2024-11-12T16:51:05+00:00","og_image":[{"url":"https:\/\/blog.zhaw.ch\/high-performance\/files\/2024\/03\/fashion_mnist-3.0.1.png","type":"","width":"","height":""}],"author":"rosn","twitter_card":"summary_large_image","twitter_misc":{"Written by":"rosn","Estimated reading time":"2 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/blog.zhaw.ch\/high-performance\/2024\/03\/24\/train-analog-devices-max78002-directly-from-jupyter-notebook\/#article","isPartOf":{"@id":"https:\/\/blog.zhaw.ch\/high-performance\/2024\/03\/24\/train-analog-devices-max78002-directly-from-jupyter-notebook\/"},"author":{"name":"rosn","@id":"https:\/\/blog.zhaw.ch\/high-performance\/#\/schema\/person\/e1a77329f74257615afa71bc883106c9"},"headline":"Train Analog Devices MAX78002 directly from Jupyter Notebook","datePublished":"2024-03-24T15:22:46+00:00","dateModified":"2024-11-12T16:51:05+00:00","mainEntityOfPage":{"@id":"https:\/\/blog.zhaw.ch\/high-performance\/2024\/03\/24\/train-analog-devices-max78002-directly-from-jupyter-notebook\/"},"wordCount":233,"commentCount":0,"image":{"@id":"https:\/\/blog.zhaw.ch\/high-performance\/2024\/03\/24\/train-analog-devices-max78002-directly-from-jupyter-notebook\/#primaryimage"},"thumbnailUrl":"https:\/\/blog.zhaw.ch\/high-performance\/files\/2024\/03\/fashion_mnist-3.0.1.png","articleSection":["Allgemein"],"inLanguage":"en-GB","potentialAction":[{"@type":"CommentAction","name":"Comment","target":["https:\/\/blog.zhaw.ch\/high-performance\/2024\/03\/24\/train-analog-devices-max78002-directly-from-jupyter-notebook\/#respond"]}]},{"@type":"WebPage","@id":"https:\/\/blog.zhaw.ch\/high-performance\/2024\/03\/24\/train-analog-devices-max78002-directly-from-jupyter-notebook\/","url":"https:\/\/blog.zhaw.ch\/high-performance\/2024\/03\/24\/train-analog-devices-max78002-directly-from-jupyter-notebook\/","name":"Train Analog Devices MAX78002 directly from Jupyter Notebook - Embedded High Performance Multimedia Blog","isPartOf":{"@id":"https:\/\/blog.zhaw.ch\/high-performance\/#website"},"primaryImageOfPage":{"@id":"https:\/\/blog.zhaw.ch\/high-performance\/2024\/03\/24\/train-analog-devices-max78002-directly-from-jupyter-notebook\/#primaryimage"},"image":{"@id":"https:\/\/blog.zhaw.ch\/high-performance\/2024\/03\/24\/train-analog-devices-max78002-directly-from-jupyter-notebook\/#primaryimage"},"thumbnailUrl":"https:\/\/blog.zhaw.ch\/high-performance\/files\/2024\/03\/fashion_mnist-3.0.1.png","datePublished":"2024-03-24T15:22:46+00:00","dateModified":"2024-11-12T16:51:05+00:00","author":{"@id":"https:\/\/blog.zhaw.ch\/high-performance\/#\/schema\/person\/e1a77329f74257615afa71bc883106c9"},"breadcrumb":{"@id":"https:\/\/blog.zhaw.ch\/high-performance\/2024\/03\/24\/train-analog-devices-max78002-directly-from-jupyter-notebook\/#breadcrumb"},"inLanguage":"en-GB","potentialAction":[{"@type":"ReadAction","target":["https:\/\/blog.zhaw.ch\/high-performance\/2024\/03\/24\/train-analog-devices-max78002-directly-from-jupyter-notebook\/"]}]},{"@type":"ImageObject","inLanguage":"en-GB","@id":"https:\/\/blog.zhaw.ch\/high-performance\/2024\/03\/24\/train-analog-devices-max78002-directly-from-jupyter-notebook\/#primaryimage","url":"https:\/\/blog.zhaw.ch\/high-performance\/files\/2024\/03\/fashion_mnist-3.0.1.png","contentUrl":"https:\/\/blog.zhaw.ch\/high-performance\/files\/2024\/03\/fashion_mnist-3.0.1.png","width":900,"height":900},{"@type":"BreadcrumbList","@id":"https:\/\/blog.zhaw.ch\/high-performance\/2024\/03\/24\/train-analog-devices-max78002-directly-from-jupyter-notebook\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Startseite","item":"https:\/\/blog.zhaw.ch\/high-performance\/"},{"@type":"ListItem","position":2,"name":"Train Analog Devices MAX78002 directly from Jupyter Notebook"}]},{"@type":"WebSite","@id":"https:\/\/blog.zhaw.ch\/high-performance\/#website","url":"https:\/\/blog.zhaw.ch\/high-performance\/","name":"Embedded High Performance Multimedia Blog","description":"A Blog of the ZHAW Zurich University of Applied Sciences","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/blog.zhaw.ch\/high-performance\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-GB"},{"@type":"Person","@id":"https:\/\/blog.zhaw.ch\/high-performance\/#\/schema\/person\/e1a77329f74257615afa71bc883106c9","name":"rosn","image":{"@type":"ImageObject","inLanguage":"en-GB","@id":"https:\/\/secure.gravatar.com\/avatar\/eec853590b797a7cdaea9a010f9e4424a21362148edf1b8d8d51fc8c5bced337?s=96&d=mm&r=g","url":"https:\/\/secure.gravatar.com\/avatar\/eec853590b797a7cdaea9a010f9e4424a21362148edf1b8d8d51fc8c5bced337?s=96&d=mm&r=g","contentUrl":"https:\/\/secure.gravatar.com\/avatar\/eec853590b797a7cdaea9a010f9e4424a21362148edf1b8d8d51fc8c5bced337?s=96&d=mm&r=g","caption":"rosn"},"url":"https:\/\/blog.zhaw.ch\/high-performance\/author\/rosn\/"}]}},"_links":{"self":[{"href":"https:\/\/blog.zhaw.ch\/high-performance\/wp-json\/wp\/v2\/posts\/441","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/blog.zhaw.ch\/high-performance\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/blog.zhaw.ch\/high-performance\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/blog.zhaw.ch\/high-performance\/wp-json\/wp\/v2\/users\/270"}],"replies":[{"embeddable":true,"href":"https:\/\/blog.zhaw.ch\/high-performance\/wp-json\/wp\/v2\/comments?post=441"}],"version-history":[{"count":2,"href":"https:\/\/blog.zhaw.ch\/high-performance\/wp-json\/wp\/v2\/posts\/441\/revisions"}],"predecessor-version":[{"id":462,"href":"https:\/\/blog.zhaw.ch\/high-performance\/wp-json\/wp\/v2\/posts\/441\/revisions\/462"}],"wp:attachment":[{"href":"https:\/\/blog.zhaw.ch\/high-performance\/wp-json\/wp\/v2\/media?parent=441"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/blog.zhaw.ch\/high-performance\/wp-json\/wp\/v2\/categories?post=441"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/blog.zhaw.ch\/high-performance\/wp-json\/wp\/v2\/tags?post=441"},{"taxonomy":"features","embeddable":true,"href":"https:\/\/blog.zhaw.ch\/high-performance\/wp-json\/wp\/v2\/features?post=441"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}