{
  "source_file": "0930",
  "source_platform": "Medium",
  "author": "John Adeojo",
  "title_listed": "Ludwig - A \"Friendlier\" Deep Learning Framework",
  "keywords": "",
  "domain": "机器学习",
  "content_type": "深度技术解析",
  "is_paid_flag": "否",
  "url": "https://towardsdatascience.com/ludwig-a-friendlier-deep-learning-framework-946ee3d3b24/",
  "publish_date": "2023-06-27",
  "corpus_stage": "",
  "notes": "",
  "id": "e6b19855f20f",
  "collected_at": "2026-10-02T07:17:15",
  "title": "Ludwig - A \"Friendlier\" Deep Learning Framework | Towards Data Science",
  "body_markdown": "![Revisit consent button](../images/body-df29beacd7496220cc94.svg)\nWe value your privacy\nWe use cookies to enhance your browsing experience, serve personalised ads or content, and analyse our traffic. By clicking \"Accept All\", you consent to our use of cookies.\nCustomise Reject All Accept All\nCustomise Consent Preferences ![](../images/body-2159bb67f3de8ad870eb.svg)\nWe use cookies to help you navigate efficiently and perform certain functions. You will find detailed information about all cookies under each consent category below.\nThe cookies that are categorised as \"Necessary\" are stored on your browser as they are essential for enabling the basic functionalities of the site. ... Show more\nNecessaryAlways Active\nNecessary cookies are required to enable the basic features of this site, such as providing secure log-in or adjusting your consent preferences. These cookies do not store any personally identifiable data.\n  * Cookie\nBCTempID\n  * Duration\n10 minutes\n  * Description\nNo description available.\n\n\n  * Cookie\n__cf_bm\n  * Duration\n1 hour\n  * Description\nThis cookie, set by Cloudflare, is used to support Cloudflare Bot Management. \n\n\n  * Cookie\nAWSALBCORS\n  * Duration\n7 days\n  * Description\nAmazon Web Services set this cookie for load balancing.\n\n\n  * Cookie\n_cfuvid\n  * Duration\nsession\n  * Description\nCloudflare sets this cookie to track users across sessions to optimize user experience by maintaining session consistency and providing personalized services\n\n\n  * Cookie\nli_gc\n  * Duration\n6 months\n  * Description\nLinkedin set this cookie for storing visitor's consent regarding using cookies for non-essential purposes.\n\n\n  * Cookie\n__hssrc\n  * Duration\nsession\n  * Description\nThis cookie is set by Hubspot whenever it changes the session cookie. The __hssrc cookie set to 1 indicates that the user has restarted the browser, and if the cookie does not exist, it is assumed to be a new session.\n\n\n  * Cookie\n__hssc\n  * Duration\n1 hour\n  * Description\nHubSpot sets this cookie to keep track of sessions and to determine if HubSpot should increment the session number and timestamps in the __hstc cookie.\n\n\n  * Cookie\nwpEmojiSettingsSupports\n  * Duration\nsession\n  * Description\nWordPress sets this cookie when a user interacts with emojis on a WordPress site. It helps determine if the user's browser can display emojis properly.\n\n\n  * Cookie\nBCSessionID\n  * Duration\n1 year 1 month 4 days\n  * Description\nBlueconic sets this cookie as a unique identifier for the BlueConic profile.\n\n\n  * Cookie\n_octo\n  * Duration\n1 year\n  * Description\nNo description available.\n\n\n  * Cookie\nlogged_in\n  * Duration\n1 year\n  * Description\nNo description available.\n\n\n  * Cookie\n__Secure-YEC\n  * Duration\npast\n  * Description\nYouTube sets this cookie to stores the user's video player preferences using embedded YouTube video\n\n\n  * Cookie\n__eoi\n  * Duration\n6 months\n  * Description\nDescription is currently not available.\n\n\n  * Cookie\nAWSALBTGCORS\n  * Duration\n7 days\n  * Description\nNo description available.\n\n\n  * Cookie\nlogin-status-p\n  * Duration\npast\n  * Description\nDescription is currently not available.\n\n\n  * Cookie\nAWSALBTG\n  * Duration\n7 days\n  * Description\nNo description available.\n\n\n  * Cookie\ncsrf_token\n  * Duration\nsession\n  * Description\nNo description available.\n\n\n  * Cookie\ntoken_v2\n  * Duration\n1 day\n  * Description\nDescription is currently not available.\n\n\n  * Cookie\nD\n  * Duration\n1 year\n  * Description\nDescription is currently not available.\n\n\n  * Cookie\nPHPSESSID\n  * Duration\nsession\n  * Description\nThis cookie is native to PHP applications. The cookie stores and identifies a user's unique session ID to manage user sessions on the website. The cookie is a session cookie and will be deleted when all the browser windows are closed.\n\n\n  * Cookie\nVISITOR_PRIVACY_METADATA\n  * Duration\n6 months\n  * Description\nYouTube sets this cookie to store the user's cookie consent state for the current domain. \n\n\n  * Cookie\ncookietest\n  * Duration\nsession\n  * Description\nThe cookietest cookie is typically used to determine whether the user's browser accepts cookies, essential for website functionality and user experience.\n\n\n  * Cookie\n__Host-airtable-session\n  * Duration\n1 year\n  * Description\nThis cookie is used to enable us to integrate the services of Airtable.\n\n\n  * Cookie\n__Host-airtable-session.sig\n  * Duration\n1 year\n  * Description\nThis cookie is used to enable us to integrate the services of Airtable.\n\n\n  * Cookie\nm\n  * Duration\n1 year 1 month 4 days\n  * Description\nStripe sets this cookie for fraud prevention purposes. It identifies the device used to access the website, allowing the website to be formatted accordingly.\n\n\n  * Cookie\nBIGipServer*\n  * Duration\nsession\n  * Description\nMarketo sets this cookie to collect information about the user's online activity and build a profile about their interests to provide advertisements relevant to the user. \n\n\n  * Cookie\n__cfruid\n  * Duration\nsession\n  * Description\nCloudflare sets this cookie to identify trusted web traffic.\n\n\n  * Cookie\n_GRECAPTCHA\n  * Duration\n6 months\n  * Description\nGoogle Recaptcha service sets this cookie to identify bots to protect the website against malicious spam attacks.\n\n\n  * Cookie\n__Secure-YNID\n  * Duration\n6 months\n  * Description\nGoogle cookie used to protect user security and prevent fraud, especially during the login process.\n\n\n  * Cookie\ncookieyes-consent\n  * Duration\n1 year\n  * Description\nCookieYes sets this cookie to remember users' consent preferences so that their preferences are respected on subsequent visits to this site. It does not collect or store any personal information about the site visitors.\n\n\nFunctional\nFunctional cookies help perform certain functionalities like sharing the content of the website on social media platforms, collecting feedback, and other third-party features.\n  * Cookie\nlidc\n  * Duration\n1 day\n  * Description\nLinkedIn sets the lidc cookie to facilitate data center selection.\n\n\n  * Cookie\nbrw\n  * Duration\n1 year\n  * Description\nNo description available.\n\n\n  * Cookie\nbrwConsent\n  * Duration\n5 minutes\n  * Description\nDescription is currently not available.\n\n\n  * Cookie\nWMF-Uniq\n  * Duration\n1 year\n  * Description\nDescription is currently not available.\n\n\n  * Cookie\nloom_anon_comment\n  * Duration\n1 year\n  * Description\nNo description available.\n\n\n  * Cookie\nloom_referral_video\n  * Duration\nsession\n  * Description\nDescription is currently not available.\n\n\n  * Cookie\nVISITOR_INFO1_LIVE\n  * Duration\n6 months\n  * Description\nA cookie set by YouTube to measure bandwidth that determines whether the user gets the new or old player interface.\n\n\n  * Cookie\nyt-remote-connected-devices\n  * Duration\nNever Expires\n  * Description\nYouTube sets this cookie to store the user's video preferences using embedded YouTube videos.\n\n\n  * Cookie\nytidb::LAST_RESULT_ENTRY_KEY\n  * Duration\nNever Expires\n  * Description\nThe cookie ytidb::LAST_RESULT_ENTRY_KEY is used by YouTube to store the last search result entry that was clicked by the user. This information is used to improve the user experience by providing more relevant search results in the future.\n\n\n  * Cookie\nyt-remote-device-id\n  * Duration\nNever Expires\n  * Description\nYouTube sets this cookie to store the user's video preferences using embedded YouTube videos.\n\n\n  * Cookie\nyt-remote-session-name\n  * Duration\nsession\n  * Description\nThe yt-remote-session-name cookie is used by YouTube to store the user's video player preferences using embedded YouTube video.\n\n\n  * Cookie\nyt-remote-fast-check-period\n  * Duration\nsession\n  * Description\nThe yt-remote-fast-check-period cookie is used by YouTube to store the user's video player preferences for embedded YouTube videos.\n\n\n  * Cookie\nyt-remote-session-app\n  * Duration\nsession\n  * Description\nThe yt-remote-session-app cookie is used by YouTube to store user preferences and information about the interface of the embedded YouTube video player.\n\n\n  * Cookie\nyt-remote-cast-available\n  * Duration\nsession\n  * Description\nThe yt-remote-cast-available cookie is used to store the user's preferences regarding whether casting is available on their YouTube video player.\n\n\n  * Cookie\nyt-remote-cast-installed\n  * Duration\nsession\n  * Description\nThe yt-remote-cast-installed cookie is used to store the user's video player preferences using embedded YouTube video.\n\n\n  * Cookie\ncp_session\n  * Duration\n3 months\n  * Description\nCodepen sets this cookie for Help systems found in the website.\n\n\n  * Cookie\nloid\n  * Duration\n1 year 1 month 4 days\n  * Description\nThis cookie is set by the Reddit. The cookie enables the sharing of content from the website onto the social media platform.\n\n\nAnalytics\nAnalytical cookies are used to understand how visitors interact with the website. These cookies help provide information on metrics such as the number of visitors, bounce rate, traffic source, etc.\n  * Cookie\n__hstc\n  * Duration\n6 months\n  * Description\nHubspot set this main cookie for tracking visitors. It contains the domain, initial timestamp (first visit), last timestamp (last visit), current timestamp (this visit), and session number (increments for each subsequent session).\n\n\n  * Cookie\nhubspotutk\n  * Duration\n6 months\n  * Description\nHubSpot sets this cookie to keep track of the visitors to the website. This cookie is passed to HubSpot on form submission and used when deduplicating contacts.\n\n\n  * Cookie\n_ga\n  * Duration\n1 year 1 month 4 days\n  * Description\nGoogle Analytics sets this cookie to calculate visitor, session and campaign data and track site usage for the site's analytics report. The cookie stores information anonymously and assigns a randomly generated number to recognise unique visitors.\n\n\n  * Cookie\n_ga_*\n  * Duration\n1 year 1 month 4 days\n  * Description\nGoogle Analytics sets this cookie to store and count page views.\n\n\n  * Cookie\n__Host-psifi.analyticsTrace\n  * Duration\n6 hours\n  * Description\nDescription is currently not available.\n\n\n  * Cookie\n__Host-psifi.analyticsTraceV2\n  * Duration\n6 hours\n  * Description\nDescription is currently not available.\n\n\n  * Cookie\n_gh_sess\n  * Duration\nsession\n  * Description\nGitHub sets this cookie for temporary application and framework state between pages like what step the user is on in a multiple step form.\n\n\n  * Cookie\nYSC\n  * Duration\nsession\n  * Description\nYSC cookie is set by Youtube and is used to track the views of embedded videos on Youtube pages.\n\n\n  * Cookie\najs_anonymous_id\n  * Duration\n1 year\n  * Description\nThis cookie is set by Segment to count the number of people who visit a certain site by tracking if they have visited before.\n\n\n  * Cookie\nvuid\n  * Duration\n1 year 1 month 4 days\n  * Description\nVimeo installs this cookie to collect tracking information by setting a unique ID to embed videos on the website. \n\n\nPerformance\nPerformance cookies are used to understand and analyse the key performance indexes of the website which helps in delivering a better user experience for the visitors.\n  * Cookie\nAWSALB\n  * Duration\n7 days\n  * Description\nAWSALB is an application load balancer cookie set by Amazon Web Services to map the session to the target.\n\n\n  * Cookie\nacq\n  * Duration\npast\n  * Description\nDescription is currently not available.\n\n\n  * Cookie\nacq.sig\n  * Duration\npast\n  * Description\nDescription is currently not available.\n\n\n  * Cookie\nptc\n  * Duration\n2 years\n  * Description\nNo description available.\n\n\nAdvertisement\nAdvertisement cookies are used to provide visitors with customised advertisements based on the pages you visited previously and to analyse the effectiveness of the ad campaigns.\n  * Cookie\nmuc_ads\n  * Duration\n1 year 1 month 4 days\n  * Description\nTwitter sets this cookie to collect user behaviour and interaction data to optimize the website.\n\n\n  * Cookie\nguest_id_marketing\n  * Duration\n1 year 1 month 4 days\n  * Description\nTwitter sets this cookie to identify and track the website visitor.\n\n\n  * Cookie\nguest_id_ads\n  * Duration\n1 year 1 month 4 days\n  * Description\nTwitter sets this cookie to identify and track the website visitor.\n\n\n  * Cookie\npersonalization_id\n  * Duration\n1 year 1 month 4 days\n  * Description\nTwitter sets this cookie to integrate and share features for social media and also store information about how the user uses the website, for tracking and targeting.\n\n\n  * Cookie\nguest_id\n  * Duration\n1 year 1 month 4 days\n  * Description\nTwitter sets this cookie to identify and track the website visitor. It registers if a user is signed in to the Twitter platform and collects information about ad preferences.\n\n\n  * Cookie\nbcookie\n  * Duration\n1 year\n  * Description\nLinkedIn sets this cookie from LinkedIn share buttons and ad tags to recognize browser IDs.\n\n\n  * Cookie\n__Secure-ROLLOUT_TOKEN\n  * Duration\n6 months\n  * Description\nYouTube sets this cookie to manage feature rollout and experimentation. It helps Google control which new features or interface changes are shown to users as part of testing and staged rollouts, ensuring consistent experience for a given user during an experiment.\n\n\n  * Cookie\nyt.innertube::nextId\n  * Duration\nNever Expires\n  * Description\nYouTube sets this cookie to register a unique ID to store data on what videos from YouTube the user has seen.\n\n\n  * Cookie\nyt.innertube::requests\n  * Duration\nNever Expires\n  * Description\nYouTube sets this cookie to register a unique ID to store data on what videos from YouTube the user has seen.\n\n\n  * Cookie\nsession_tracker\n  * Duration\nsession\n  * Description\nThis cookie is set by the Reddit. This cookie is used to identify trusted web traffic. It also helps in adverstising on the website.\n\n\n  * Cookie\nedgebucket\n  * Duration\nsession\n  * Description\nReddit sets this cookie to save the information about a log-on Reddit user, for the purpose of advertisement recommendations and updating the content.\n\n\n  * Cookie\ndid\n  * Duration\n1 year\n  * Description\nArbor sets this cookie to show targeted ads to site visitors.This cookie expires after 2 months or 1 year.\n\n\nUncategorised\nOther uncategorised cookies are those that are being analysed and have not been classified into a category as yet.\nNo cookies to display.\nReject All Save My Preferences Accept All\nToggle Search[](https://towardsdatascience.com/)Toggle Mobile Navigation\n[](https://towardsdatascience.com/)\nPublish AI, ML & data-science insights to a global community of data professionals.\nSign in[Submit an Article](https://write.insightmediagroup.io)\n[Latest](https://towardsdatascience.com/latest/)[Editor's Picks](https://towardsdatascience.com/tag/editors-pick/)[Newsletter](https://towardsdatascience.com/tag/the-variable/)[ShipAI](https://towardsdatascience.com/shipai/)\n* * *\n[Write For TDS](https://towardsdatascience.com/submissions/)\n[LinkedIn](https://www.linkedin.com/company/towards-data-science/)[X](https://x.com/TDataScience)Toggle Search\n[Data Science](https://towardsdatascience.com/category/data-science/)\n# Ludwig - A \"Friendlier\" Deep Learning Framework\nDeep Learning Made Easier with this low-code Declarative Framework\n[John Adeojo](https://towardsdatascience.com/author/johnadeojo/)\nJune 26, 202312 min read\n### Deep Learning Made Easy with this Low-Code, Declarative Framework\n![Image by Author: Generated with Midjourney](../images/000.png)\nImage by Author: Generated with Midjourney\nLearn this step by step with the interactive [AI and Data Scientist](https://roadmap.sh/ai-data-scientist), [Computer Science](https://roadmap.sh/computer-science) and [Machine Learning](https://roadmap.sh/machine-learning) roadmaps.\n## Background - Deep Learning, too Complex?\nI have tended to shy away from deep learning for industry use cases. Not due to a lack of interest, rather I find the popular deep learning frameworks cumbersome. I appreciate [Pytorch](https://pytorch.org/) and [TensorFlow](https://www.tensorflow.org/) are fantastic tools for research purposes, but the APIs are not the most user friendly. In situations where I need to deliver a quick proof of concept for a client, the last thing I want to be doing is tinkering with Pytorch tensors.\nWhile attending the AI summit in London, I stumbled across a team claiming to have a solution to my deep learning problem. They have been leveraging a different approach, which they described as \"the mid-point between TensorFlow and AutoML\", a framework called Ludwig.\n### What is Ludwig?\nDeveloped by Uber, [Ludwig](https://ludwig.ai/latest/) is an open-source framework for building deep learning models. It is declarative, which means that instead of constructing complex models layer-by-layer like you would in TensorFlow, you simply declare the structure of the model with a configuration file. This sounded too good to be true, so I had to see for myself. In the remainder of this article, I detail my experience with Ludwig through an example project I have taken from [Kaggle](https://www.kaggle.com/datasets/shelvigarg/sales-forecasting-womart-store). Along the way I will discuss some of its strengths, pain-points, and give you my conclusion on whether it's worth using.\n_Note - Although initially developed by Uber, Ludwig is an_[ _open source_](https://github.com/ludwig-ai/ludwig/blob/master/LICENSE) _library under the Apache 2.0 License. The project is hosted by the_[ _Linux Foundation AI & Data_](https://lfaidata.foundation/) _. I have no commercial affiliation with Uber or the developers of Ludwig._\n* * *\n## The project - Demand Forecasting\n**The project brief** : Forecast the final 30 days of orders across each store for the retailer WOMart.\n> Your Client WOMart is a leading nutrition and supplement retail chain that offers a comprehensive range of products for all your wellness and fitness needs.\n> WOMart follows a multi-channel distribution strategy with 350+ retail stores spread across 100+ cities.\n### The Data\nThe dataset had a total of 22,265 observations, each pertaining to one day of sales at a given store. For brevity I won't go into the full details of the data set, but you can see some descriptive statistics [here](https://www.kaggle.com/datasets/shelvigarg/sales-forecasting-womart-store).\n**Note** : The data is free to use for any purposes under the Open Data Commons license.\n  * Link to the [data](https://www.kaggle.com/datasets/shelvigarg/sales-forecasting-womart-store)\n  * [License](https://opendatacommons.org/licenses/dbcl/1-0/) for the data\n\n\n**The data dictionary:**\n* * *\n## A Brief Overview of the Methodology\nI won't go into too much detail on the methodology here as this isn't the main purpose of this article. I'll cover how I framed the problem at a high level to give you some context.\nI framed the forecasting issue as a 'pseudo' sequence-to-sequence deep learning problem. The method entailed utilising 360 days of time series data points to predict the subsequent 30 days of customer orders. I incorporated some categorical variables and had to generate separate predictions for each day of orders, which led to a somewhat unconventional setup - thus the term 'pseudo' sequence-to-sequence for describing the problem. I will expand on the specifics of the feature engineering later in this article.\nBesides this, the methodology I followed was standard for model development. I split my data into a training data set and a holdout dataset and did some re-scaling of features and labels. Model training was done on the training data, and testing on the holdout.\n_Note: Ludwig does provide the capability to natively split data within the API. However, to maintain rigour, I established a separate holdout dataset. The training dataset was subsequently divided into additional training, validation, and testing subsets. The holdout dataset was excluded entirely, used only for analysing the predictions generated by the model._\n* * *\n## Feature Engineering\nAt the time of writing this article, I would say that sequence-to-sequence modelling as it pertains to timeseries forecasting is tricky to do in Ludwig. The reason for this is the feature engineering. The Ludwig API has an excellent ability to handle sequences as inputs, but they haven't (yet) developed a coherent approach for time series sequences as outputs. You can develop a \"pseudo\" sequence-to-sequence model by declaring multiple outputs, but the overall feature engineering experience feels quite \"hacky\".\n**Sequence features** : Except for those features that don't change with time, I engineered all the predictive features to be \"Ludwig-formatted\" sequences. Each input sequence is a horizontal stacking of each \"time series\" feature across a pre-defined timeframe. Each feature sequence is determined at the store level and encapsulated in one cell of the dataframe (it looks as messy as it sounds).\n**Sequence labels** : For sequence labels, you must declare each point in the sequence as an individual label for the model. This ended up with me declaring 30 labels for each store, one label per day of orders to be predicted.\n_Example of the feature engineering process below:_\n* * *\n## Architecting Your Model\nThe Ludwig API allows you to architect fairly complex and customisable models declaratively. Ludwig does this through a .yaml file. Now, I appreciate many data scientists reading this might not have used _.yaml_ files, but generally in software development these are used for configuration. The files might appear scary at first glance, but they are quite friendly. Let's step through the main parts of the file I create to build the model.\n![Image by Author: Model architecture](../images/001.png)\nImage by Author: Model architecture\nBefore we delve into the configurations, it's worth briefly introducing the [architecture](https://ludwig.ai/latest/user_guide/how_ludwig_works/) at the heart of Ludwig's deep learning framework: the Encoder, Combiner, and Decoder. Most of the models you configure in Ludwig will predominantly adhere to this architecture. Understanding this can simplify the process of stacking components to quickly build your deep learning models.\n### Declaring your Model\nRight at the top of the file you declare the model type used. Ludwig provides two options, tree-based models, and deep neural networks for which I chose the latter.\nyaml\n\n```\nmodel_type: ecd\n```\n\n### Declaring your data splits\nYou can split data sets natively by declaring your split percentages, type of split, and column or variable you're splitting on. For my purposes I wanted to ensure that a store could only appear in one of the data sets, hash splitting was perfect for that.\nFor best practice, I would probably advise constructing a holdout set outside of the Ludwig API especially where you are doing some initial feature engineering like one-hot-encoding or normalisation. This should help prevent data leakage.\nyaml\n\n```\nmodel_type: ecd\nsplit:\n    type: hash\n    column: Store_id\n    probabilities:\n    - 0.7\n    - 0.15\n    - 0.15\n#...omitted sections...\n```\n\n### Declaring the Model Inputs\nYou declare inputs by name, type, and encoder. Depending on the type of input to the model you have a variety of options for encoders. Essentially encoders are a way of transforming your inputs such that it can be interpreted by the model. The choice of encoder really depends on the data and the modelling task.\nyaml\n\n```\nmodel_type: ecd\nsplit:\n    type: hash\n    column: Store_id\n    probabilities:\n    - 0.7\n    - 0.15\n    - 0.15\ninput_features:\n  - name: Sales\n    type: sequence\n    encoder: stacked_cnn\n    reduce_output: null\n  - name: Order\n    type: sequence\n    encoder: stacked_cnn\n    reduce_output: null\n  - name: Discount\n    type: sequence\n    encoder: stacked_cnn\n    reduce_output: null\n  - name: DayOfWeek\n    type: sequence\n    encoder: stacked_cnn\n    reduce_output: null\n  - name: MonthOfYear\n    type: sequence\n    encoder: stacked_cnn\n    reduce_output: null\n  - name: Holiday\n    type: sequence\n    encoder: stacked_cnn\n    reduce_output: null\n  - name: Store_Type\n    type: category\n    encoder: dense\n  - name: Location_Type\n    type: category\n    encoder: dense\n  - name: Region_Code\n    type: category\n    encoder: dense\n#...omitted sections...\n```\n\n### Declaring the Combiner\n[Combiners](https://ludwig.ai/latest/configuration/combiner/), as the name suggests, amalgamate the outputs of your encoders. The Ludwig API offers an array of different combiners, each with its own specific use case. The choice of combiner can depend on the structure of your model and the relationships between your features. For instance, you might use a 'concat' combiner if you want to simply concatenate the outputs of your encoders, or a 'sequence' combiner if your features have a sequential relationship.\nyaml\n\n```\nmodel_type: ecd\nsplit:\n    type: hash\n    column: Store_id\n    probabilities:\n    - 0.7\n    - 0.15\n    - 0.15\ninput_features:\n  - name: Sales\n    type: sequence\n    encoder: stacked_cnn\n    reduce_output: null\n  - name: Order\n    type: sequence\n    encoder: stacked_cnn\n    reduce_output: null \n  # ... omitted sections ...\n\n  - name: Location_Type\n    type: category\n    encoder: dense\n  - name: Region_Code\n    type: category\n    encoder: dense\ncombiner:\n    type: sequence\n    main_sequence_feature: Order\n    reduce_output: null\n    encoder:\n    # ... omitted sections ...\n```\n\nAs with many aspects of deep learning, the optimal choice of combiner often depends on the specifics of your dataset and problem, and may require some experimentation.\n### Declaring the Model Outputs\nFinalising your network is as simple as declaring your outputs, which are just your labels. My pet peeve with Ludwig for timeseries is that you can't (yet) declare timeseries outputs. As I mentioned previously, you have to \"hack\" it by declaring each point in your time series separately. This left me with thirty separate declarations, which looks very messy in all honesty. For each output you can specify the loss function too adding additional configurability. Ludwig has a myriad of options pre-built for different output types, however I don't know if you are able to implement custom loss functions as you can with Pytorch.\nyaml\n\n```\nmodel_type: ecd\nsplit:\n    type: hash\n    column: Store_id\n    probabilities:\n    - 0.7\n    - 0.15\n    - 0.15\ninput_features:\n  - name: Sales\n    type: sequence\n    encoder: stacked_cnn\n    reduce_output: null\n  - name: Order\n    type: sequence\n    encoder: stacked_cnn\n    reduce_output: null\n# ...omitted sections...\n\n  - name: Location_Type\n    type: category\n    encoder: dense\n  - name: Region_Code\n    type: category\n    encoder: dense\ncombiner:\n    type: sequence\n    main_sequence_feature: Order\n    reduce_output: null\n    encoder:\n        type: parallel_cnn\noutput_features:\n  - name: Order_sequence_label_2019-05-02\n    type: number\n    loss:\n      type: mean_absolute_error\n  - name: Order_sequence_label_2019-05-03\n    type: number\n    loss:\n      type: mean_absolute_error\n#...omitted sections...\n\n      type: mean_absolute_error\n  - name: Order_sequence_label_2019-05-30\n    type: number\n    loss:\n      type: mean_absolute_error\n  - name: Order_sequence_label_2019-05-31\n    type: number\n    loss:\n      type: mean_absolute_error\n#...omitted sections...\n```\n\n### Declaring the Trainer\nThe trainer configuration in Ludwig, while optional due to Ludwig's provision of sensible defaults, allows for a high degree of customisation. It gives you control over the specifics of how your model is trained. This includes the ability to specify the type of optimiser used, the number of training epochs, the learning rate, and criteria for early stopping, among other parameters.\nyaml\n\n```\nmodel_type: ecd\nsplit:\n    type: hash\n    column: Store_id\n    probabilities:\n    - 0.7\n    - 0.15\n    - 0.15\ninput_features:\n  - name: Sales\n    type: sequence\n    encoder: stacked_cnn\n    reduce_output: null\n  - name: Order\n    type: sequence\n    encoder: stacked_cnn\n    reduce_output: null\n# ...omitted sections...\n\n  - name: Location_Type\n    type: category\n    encoder: dense\n  - name: Region_Code\n    type: category\n    encoder: dense\ncombiner:\n    type: sequence\n    main_sequence_feature: Order\n    reduce_output: null\n    encoder:\n        type: parallel_cnn\noutput_features:\n  - name: Order_sequence_label_2019-05-02\n    type: number\n    loss:\n      type: mean_absolute_error\n  - name: Order_sequence_label_2019-05-03\n    type: number\n    loss:\n      type: mean_absolute_error\n#...omitted sections...\n\n      type: mean_absolute_error\n  - name: Order_sequence_label_2019-05-30\n    type: number\n    loss:\n      type: mean_absolute_error\n  - name: Order_sequence_label_2019-05-31\n    type: number\n    loss:\n      type: mean_absolute_error\ntrainer:\n    epochs: 200\n    learning_rate: 0.0001\n    early_stop: 20\n    evaluate_training_set: true\n    validation_metric: mean_absolute_error\n    validation_field: Order_sequence_label_2019-05-31\n```\n\nFor your particular use case, you might find it beneficial to define these parameters yourself. For instance, you might want to adjust the learning rate, or the number of epochs based on the complexity of your model and the size of your dataset. Similarly, early stopping can be a useful tool to prevent overfitting by halting the training process if the model's performance on a validation set stops improving.\n### Train your Model\nTraining your model can be easily done with Ludwig's python expermiment API. See the script example below:\n### Other Configurations\nOutside of those I mentioned, Ludwig has a myriad of possible configurations. They are all very well documented and well structured. I would advise having a read of their [documentation](https://ludwig.ai/latest/configuration/) to familiarise yourself.\n* * *\n## Model Performance Analysis - A brief view\nThe aim of this article is to explore some of the functionalities of the Ludwig framework through a practical example project. Although demonstrating model performance is part of this, there's no need to delve into the specifics of the metrics. I will limit the discussion to presenting some of the charts generated from the model's analysis. Note that the comprehensive end-to-end script is available on my GitHub, linked at the conclusion of the article.\n![](../images/body-b0bb348c40b7aa6e7138.png)\n![](../images/body-56b4baaa62e3e2ff03dd.png)\n![](../images/body-b0bb348c40b7aa6e7138.png)\n![Image by Author: Model predictions \\(red\\) against actual orders \\(blue\\) for the last 30 days of data. These examples are drawn from the holdout set.](../images/002.png)\nImage by Author: Model predictions (red) against actual orders (blue) for the last 30 days of data. These examples are drawn from the holdout set.\n![Image by Author: Distribution of error, where error is Actual Orders minus Predicted Orders.](../images/003.png)\nImage by Author: Distribution of error, where error is Actual Orders minus Predicted Orders.\n![Image by Author: Loss curve for the model training. Loss metric is Mean Absolute Error](../images/004.png)\nImage by Author: Loss curve for the model training. Loss metric is Mean Absolute Error\n* * *\n## My Verdict\nI'll begin by confessing that I was initially sceptical about Ludwig. However, after experimenting with it myself, I'm convinced of its capabilities and believe it delivers as promised. I think there are several truly impressive features worth highlighting.\n**Coding experience** : The coding experience feels more like constructing an elaborate Lego set. You really can have a lot of fun messing around with components and different architectures to find your perfect model.\n**Documentation** : The documentation is clear and well structured. It was very easy to figure out how to implement different architectures and change your model configurations. Most of the documentation appears to be up to date too, which is a bonus.\n**Backend** : The backend experience is excellent. The library developers have done a great job abstracting away a lot of the usual configurations you have to do to train deep neural networks on GPUs. I trained my model on a Google Collab, and Ludwig automatically moved the workloads onto the GPU.\nAnother great thing about Ludwig is that the backend is highly configurable. For example, if you're running massive workloads and need a cluster of GPUs, you can configure this too!\n**Experiment Tracking** : Ludwig provides you with an Experiment API that can be used to track your model artefacts between experiment runs. I believe there are integrations with MLflow too which is fantastic for commercial scale MLOps.\n### Pet Peeves\nThere are a few areas in which the framework could be enhanced, let's get into those too.\n**Visualisation** : Ludwig does offer a visualisation API to track training losses across your data sets. However, at the time of writing, it does not function particularly well and its usage is unintuitive. I tried running it in a Google Collab, but to no avail. Eventually, I created my own loss curve visualisations by scripting a Python function to parse the training_statistics.json files that Ludwig saves for you after each experiment run.\n**Support** : Whilst there is some support available, Ludwig's community does not seem to be as extensive as those of TensorFlow or Pytorch yet. A few issues have been raised on GitHub, and there are some threads that might offer assistance, but for the most part, it feels like you are left to your own devices. As for ChatGPT, it provides a certain level of support up until 2021.\n**Transparency** : Ludwig excels in eliminating the more challenging aspects of building deep learning models. However, this comes at the expense of transparency, which occasionally renders the logs somewhat cryptic and challenging to debug.\n* * *\n## Conclusion\nIn my opinion, Ludwig is an excellent tool for those looking to start utilising deep learning whether in a commercial setting or just for learning. While it may be too high-level for state-of-the-art research purposes, it is well-suited to solving clearly defined problems swiftly. A certain degree of understanding of deep learning is still necessary, but once you've grasped the concept, the barrier to entry for Ludwig is considerably lower than it is for TensorFlow or Pytorch.\nThe end-to-end notebook is available in my [GitHub](https://github.com/john-adeojo/womartdata/blob/main/ludwig_deep_learning_notebook%20\\(2\\).ipynb) repo, please feel free to experiment.\n* * *\n_Follow me on_[ _LinkedIn_](https://www.linkedin.com/in/john-adeojo/)\n_Subscribe to medium to get more insights from me:_\n> [**Join Medium with my referral link - John Adeojo**](https://johnadeojo.medium.com/membership)\n_Should you be interested in integrating AI or data science into your business operations, we invite you to schedule a complimentary initial consultation with us:_\n> [**Book Online | Data-Centric Solutions**](https://www.data-centric-solutions.com/book-online)\n## Register Now to Continue Reading\n### Unlock Towards Data Science with a Free Account\nAll articles on Towards Data Science are available to registered users. Create a free account today to access this article and thousands more.\nRegister for freeSign in to your account\n* * *\n![](../images/body-ead068aa953e7b688a96.png)\nWritten By\n## [John Adeojo](https://towardsdatascience.com/author/johnadeojo/)\n[See all from John Adeojo](https://towardsdatascience.com/author/johnadeojo/)\n[Data Science](https://towardsdatascience.com/tag/data-science/)[Deep Learning](https://towardsdatascience.com/tag/deep-learning/)[Hands On Tutorials](https://towardsdatascience.com/tag/hands-on-tutorials/)[Machine Learning](https://towardsdatascience.com/tag/machine-learning/)[Programming](https://towardsdatascience.com/tag/programming/)\nShare This Article\n  * [](https://www.facebook.com/sharer/sharer.php?u=https%3A%2F%2Ftowardsdatascience.com%2Fludwig-a-friendlier-deep-learning-framework-946ee3d3b24%2F&title=Ludwig%20-%20A%20%22Friendlier%22%20Deep%20Learning%20Framework)\n  * [](https://www.linkedin.com/shareArticle?mini=true&url=https%3A%2F%2Ftowardsdatascience.com%2Fludwig-a-friendlier-deep-learning-framework-946ee3d3b24%2F&title=Ludwig%20-%20A%20%22Friendlier%22%20Deep%20Learning%20Framework)\n  * [](https://x.com/share?url=https%3A%2F%2Ftowardsdatascience.com%2Fludwig-a-friendlier-deep-learning-framework-946ee3d3b24%2F&text=Ludwig%20-%20A%20%22Friendlier%22%20Deep%20Learning%20Framework)\n\n\nTowards Data Science is a community publication. Submit your insights to reach our global audience and earn through the TDS Author Payment Program.\n[Write for TDS](https://towardsdatascience.com/submissions/)\n## Related Articles\n  * [![](../images/005.jpg)](https://towardsdatascience.com/step-by-step-yolo-model-deployment-in-localhost-using-python-8537e93a1784/)\n## [Deploy Custom Object Detection using Flask & Python, Step by Step](https://towardsdatascience.com/step-by-step-yolo-model-deployment-in-localhost-using-python-8537e93a1784/)\n[Artificial Intelligence](https://towardsdatascience.com/category/artificial-intelligence/)\nCustom Object Detection Localhost Deployment | YOLO\n[Pranjal Saxena](https://towardsdatascience.com/author/pranjalai/)\nSeptember 3, 20205 min read\n  * [![](../images/006.png)](https://towardsdatascience.com/data-science-in-algorithmic-trading-d21a46d1565d/)\n## [Data Science in Algorithmic Trading](https://towardsdatascience.com/data-science-in-algorithmic-trading-d21a46d1565d/)\n[Artificial Intelligence](https://towardsdatascience.com/category/artificial-intelligence/)\nIn this article I plan to give you a glimpse into an asset model for algorithmic trading. This model of the world should allow us to make...\n[Daniel Shapiro, PhD](https://towardsdatascience.com/author/lemaysolutions/)\nDecember 31, 201816 min read\n  * [![](../images/007.jpg)](https://towardsdatascience.com/stop-blaming-the-data-a-better-way-to-handle-covariance-shift/)\n## [Stop Blaming the Data: A Better Way to Handle Covariance Shift](https://towardsdatascience.com/stop-blaming-the-data-a-better-way-to-handle-covariance-shift/)\n[Data Science](https://towardsdatascience.com/category/data-science/)\nInstead of using data shift as an excuse for poor performance, use Inverse Probability Weighting to estimate how your model should perform in the new environment\n[Alon Lanyado](https://towardsdatascience.com/author/alon.lanyado/)\nJanuary 5, 20267 min read\n  * [![](../images/008.jpg)](https://towardsdatascience.com/8-visualizations-with-python-to-handle-multiple-time-series-data-19b5b2e66dd0/)\n## [8 Visualizations with Python to Handle Multiple Time-Series Data](https://towardsdatascience.com/8-visualizations-with-python-to-handle-multiple-time-series-data-19b5b2e66dd0/)\n[Programming](https://towardsdatascience.com/category/programming/)\nVisualization ideas for coping with overlapping lines in multiple time-series plots.\n[Boriharn K](https://towardsdatascience.com/author/borih.k/)\nApril 28, 20228 min read\n  * [![](../images/009.jpg)](https://towardsdatascience.com/time-series-clustering-and-dimensionality-reduction-5b3b4e84f6a3/)\n## [Time Series Clustering and Dimensionality Reduction](https://towardsdatascience.com/time-series-clustering-and-dimensionality-reduction-5b3b4e84f6a3/)\n[Machine Learning](https://towardsdatascience.com/category/artificial-intelligence/machine-learning/)\nCluster sensor data with Kolmogorv Smirnov Statistic and Machine Learning\n[Marco Cerliani](https://towardsdatascience.com/author/cerlymarco/)\nJuly 25, 20195 min read\n  * [![](../images/010.jpg)](https://towardsdatascience.com/paper-readthrough-gauge-equivariant-convolutional-networks-and-the-icosahedral-cnn-13e9d3d95b85/)\n## [[Paper Readthrough] - Gauge Equivariant Convolutional Networks and the Icosahedral CNN](https://towardsdatascience.com/paper-readthrough-gauge-equivariant-convolutional-networks-and-the-icosahedral-cnn-13e9d3d95b85/)\n[Deep Learning](https://towardsdatascience.com/category/artificial-intelligence/deep-learning/)\nOverview\n[Nicola Bernini](https://towardsdatascience.com/author/nicolabernini_63880/)\nMay 25, 20194 min read\n  * [![](../images/011.jpg)](https://towardsdatascience.com/cliff-walking-problem-with-the-discrete-policy-gradient-algorithm-59d1900d80d8/)\n## [Cliff-Walking Problem With The Discrete Policy Gradient Algorithm](https://towardsdatascience.com/cliff-walking-problem-with-the-discrete-policy-gradient-algorithm-59d1900d80d8/)\n[Artificial Intelligence](https://towardsdatascience.com/category/artificial-intelligence/)\nA full implementation of the REINFORCE algorithm in Python. The steps are performed manually to illustrate the inner workings of the...\n[Wouter van Heeswijk, PhD](https://towardsdatascience.com/author/wvheeswijk/)\nOctober 21, 20218 min read\n  * [![](../images/012.jpg)](https://towardsdatascience.com/stop-using-python-to-aggregate-data-use-sql-instead-7781d6ea4f0c/)\n## [Stop Using Python to Aggregate Data - Use SQL Instead](https://towardsdatascience.com/stop-using-python-to-aggregate-data-use-sql-instead-7781d6ea4f0c/)\n[Programming](https://towardsdatascience.com/category/programming/)\nAre you using Python to extract raw data from the database? It could be a huge bottleneck in your apps. Use SQL instead.\n[Dario Radečić](https://towardsdatascience.com/author/radecicdario/)\nFebruary 17, 20226 min read\n  * [![](../images/013.jpeg)](https://towardsdatascience.com/ai-for-industrial-process-control-ee774267094b/)\n## [AI for Industrial Process Control](https://towardsdatascience.com/ai-for-industrial-process-control-ee774267094b/)\n[Machine Learning](https://towardsdatascience.com/category/artificial-intelligence/machine-learning/)\nTuning a process oven using Reinforcement Learning\n[John Morrow](https://towardsdatascience.com/author/john_morrow/)\nAugust 16, 20193 min read\n\n\n  * [YouTube](https://www.youtube.com/c/TowardsDataScience)\n  * [X](https://x.com/TDataScience)\n  * [LinkedIn](https://www.linkedin.com/company/towards-data-science/)\n  * [Threads](https://www.threads.net/@towardsdatascience)\n  * [Bluesky](https://bsky.app/profile/towardsdatascience.com)\n\n\n[](https://towardsdatascience.com/)\nYour home for data science and AI. The world's leading publication for data science, data analytics, data engineering, machine learning, and artificial intelligence professionals.\nSubscribe to Our Newsletter[Write For TDS](https://towardsdatascience.com/submissions/)•[About](https://towardsdatascience.com/about-towards-data-science-d691af11cc2f/)•[Advertise](https://contact.towardsdatascience.com/advertise-with-towards-data-science)•[Privacy Policy](https://towardsdatascience.com/privacy-policy/)•[Terms of Use](https://towardsdatascience.com/website-terms-of-use/)\n© Insight Media Group, LLC 2026•[The New Stack](https://thenewstack.io)•[roadmap.sh](https://roadmap.sh)•[Creator Network](https://insightmediagroup.io/creator-network)\n####  Recommended Articles \nClose\n  * [ ![](https://h030.towardsdatascience.com/rest/v2/contentStores/d7d3a4ab-7caf-4383-bbd2-fbf7fb4f1787/items/607686/image?etag=1782161464448&width=500&height=500) How Relevance Models Foreshadowed Transformers for NLP ](https://towardsdatascience.com/how-relevance-models-foreshadowed-transformers-for-nlp/ \"How Relevance Models Foreshadowed Transformers for NLP\")\n  * [ ![](https://h030.towardsdatascience.com/rest/v2/contentStores/d7d3a4ab-7caf-4383-bbd2-fbf7fb4f1787/items/607639/image?etag=1781250125848&width=500&height=500) Javascript Fatigue: HTMX Is All You Need to Build ChatGPT — Part 1 ](https://towardsdatascience.com/javascript-fatigue-you-dont-need-js-to-build-chatgpt/ \"Javascript Fatigue: HTMX Is All You Need to Build ChatGPT — Part 1\")\n  * [ ![](https://h030.towardsdatascience.com/rest/v2/contentStores/d7d3a4ab-7caf-4383-bbd2-fbf7fb4f1787/items/607930/image?etag=1782124566054&width=500&height=500) The Subset Sum Problem Solved in Linear Time for Dense Enough Inputs ](https://towardsdatascience.com/subset-sum-problem-solved-in-linear-time-for-dense-enough-inputs/ \"The Subset Sum Problem Solved in Linear Time for Dense Enough Inputs\")\n  * [ ![](https://h030.towardsdatascience.com/rest/v2/contentStores/d7d3a4ab-7caf-4383-bbd2-fbf7fb4f1787/items/3660/image?etag=1780176246070&width=500&height=500) Does Your Company Have a Data Strategy? ](https://towardsdatascience.com/does-your-company-have-a-data-strategy-e8944dbec4ae/ \"Does Your Company Have a Data Strategy?\")\n\n\n",
  "metadata": {
    "title": "Ludwig - A \"Friendlier\" Deep Learning Framework | Towards Data Science",
    "description": "Deep Learning Made Easier with this low-code Declarative Framework",
    "keywords": null,
    "author": "John Adeojo",
    "og:locale": "en_US",
    "og:type": "article",
    "og:title": "Ludwig - A \"Friendlier\" Deep Learning Framework",
    "og:description": "Deep Learning Made Easier with this low-code Declarative Framework",
    "og:url": "https://towardsdatascience.com/ludwig-a-friendlier-deep-learning-framework-946ee3d3b24/",
    "og:site_name": "Towards Data Science",
    "og:image": "https://assets.insightmediagroup.io/media/wp-content/uploads/2023/06/1c7aAb_-ShzIYXrkB1NcYAA.png",
    "og:image:width": "1024",
    "og:image:height": "1024",
    "og:image:type": "image/png",
    "twitter:card": "summary_large_image",
    "twitter:site": "@TDataScience",
    "twitter:creator": "@TDataScience",
    "twitter:label1": "Written by",
    "twitter:data1": "John Adeojo",
    "twitter:label2": "Est. reading time",
    "twitter:data2": "12 min read",
    "article:published_time": "2023-06-26T19:50:58.000Z",
    "article:modified_time": "2023-06-26T19:50:58.000Z",
    "article:author": "https://towardsdatascience.com/author/johnadeojo/"
  },
  "paywall": {
    "listed_paid": false,
    "detected": false,
    "marker": "",
    "public_only": false,
    "policy": "付费/会员内容只保留页面公开可见部分，不绕过付费墙"
  },
  "crawl": {
    "success": true,
    "status_code": 202,
    "error": ""
  },
  "code_blocks": [
    {
      "language": "txt",
      "file": "attachments/e6b19855f20f/code/block_000.txt",
      "chars": 16
    },
    {
      "language": "txt",
      "file": "attachments/e6b19855f20f/code/block_001.txt",
      "chars": 134
    },
    {
      "language": "txt",
      "file": "attachments/e6b19855f20f/code/block_002.txt",
      "chars": 850
    },
    {
      "language": "txt",
      "file": "attachments/e6b19855f20f/code/block_003.txt",
      "chars": 577
    },
    {
      "language": "txt",
      "file": "attachments/e6b19855f20f/code/block_004.txt",
      "chars": 1070
    },
    {
      "language": "txt",
      "file": "attachments/e6b19855f20f/code/block_005.txt",
      "chars": 1245
    }
  ],
  "images": [
    {
      "status": "http_404",
      "url": "https://h030.towardsdatascience.com/rest/v2/contentStores/d7d3a4ab-7caf-4383-bbd2-fbf7fb4f1787/items/607686/image?etag=1782161464448&width=500&height=500",
      "src": "https://h030.towardsdatascience.com/rest/v2/contentStores/d7d3a4ab-7caf-4383-bbd2-fbf7fb4f1787/items/607686/image?etag=1782161464448&width=500&height=500",
      "alt": "",
      "downloaded": false,
      "discovery": "saved_markdown",
      "retried_at": "2026-10-20T18:09:24Z"
    },
    {
      "status": "http_404",
      "url": "https://h030.towardsdatascience.com/rest/v2/contentStores/d7d3a4ab-7caf-4383-bbd2-fbf7fb4f1787/items/607639/image?etag=1781250125848&width=500&height=500",
      "src": "https://h030.towardsdatascience.com/rest/v2/contentStores/d7d3a4ab-7caf-4383-bbd2-fbf7fb4f1787/items/607639/image?etag=1781250125848&width=500&height=500",
      "alt": "",
      "downloaded": false,
      "discovery": "saved_markdown",
      "retried_at": "2026-10-20T18:09:24Z"
    },
    {
      "status": "http_404",
      "url": "https://h030.towardsdatascience.com/rest/v2/contentStores/d7d3a4ab-7caf-4383-bbd2-fbf7fb4f1787/items/607930/image?etag=1782124566054&width=500&height=500",
      "src": "https://h030.towardsdatascience.com/rest/v2/contentStores/d7d3a4ab-7caf-4383-bbd2-fbf7fb4f1787/items/607930/image?etag=1782124566054&width=500&height=500",
      "alt": "",
      "downloaded": false,
      "discovery": "saved_markdown",
      "retried_at": "2026-10-20T18:09:24Z"
    },
    {
      "status": "http_404",
      "url": "https://h030.towardsdatascience.com/rest/v2/contentStores/d7d3a4ab-7caf-4383-bbd2-fbf7fb4f1787/items/3660/image?etag=1780176246070&width=500&height=500",
      "src": "https://h030.towardsdatascience.com/rest/v2/contentStores/d7d3a4ab-7caf-4383-bbd2-fbf7fb4f1787/items/3660/image?etag=1780176246070&width=500&height=500",
      "alt": "",
      "downloaded": false,
      "discovery": "saved_markdown",
      "retried_at": "2026-10-20T18:09:24Z"
    }
  ],
  "audio": [],
  "videos": [
    {
      "url": "https://www.youtube.com/c/TowardsDataScience",
      "kind": "channel_reference",
      "status": "reference_only"
    }
  ],
  "links": [
    {
      "type": "internal",
      "href": "https://towardsdatascience.com/"
    },
    {
      "type": "internal",
      "href": "https://towardsdatascience.com/latest/"
    },
    {
      "type": "internal",
      "href": "https://towardsdatascience.com/tag/editors-pick/"
    },
    {
      "type": "internal",
      "href": "https://towardsdatascience.com/tag/the-variable/"
    },
    {
      "type": "internal",
      "href": "https://towardsdatascience.com/shipai/"
    },
    {
      "type": "internal",
      "href": "https://towardsdatascience.com/submissions/"
    },
    {
      "type": "internal",
      "href": "https://towardsdatascience.com/category/data-science/"
    },
    {
      "type": "internal",
      "href": "https://towardsdatascience.com/author/johnadeojo/"
    },
    {
      "type": "internal",
      "href": "https://towardsdatascience.com/tag/data-science/"
    },
    {
      "type": "internal",
      "href": "https://towardsdatascience.com/tag/deep-learning/"
    },
    {
      "type": "internal",
      "href": "https://towardsdatascience.com/tag/hands-on-tutorials/"
    },
    {
      "type": "internal",
      "href": "https://towardsdatascience.com/tag/machine-learning/"
    },
    {
      "type": "internal",
      "href": "https://towardsdatascience.com/tag/programming/"
    },
    {
      "type": "internal",
      "href": "https://towardsdatascience.com/step-by-step-yolo-model-deployment-in-localhost-using-python-8537e93a1784/"
    },
    {
      "type": "internal",
      "href": "https://towardsdatascience.com/category/artificial-intelligence/"
    },
    {
      "type": "internal",
      "href": "https://towardsdatascience.com/author/pranjalai/"
    },
    {
      "type": "internal",
      "href": "https://towardsdatascience.com/data-science-in-algorithmic-trading-d21a46d1565d/"
    },
    {
      "type": "internal",
      "href": "https://towardsdatascience.com/author/lemaysolutions/"
    },
    {
      "type": "internal",
      "href": "https://towardsdatascience.com/stop-blaming-the-data-a-better-way-to-handle-covariance-shift/"
    },
    {
      "type": "internal",
      "href": "https://towardsdatascience.com/author/alon.lanyado/"
    },
    {
      "type": "external",
      "href": "https://write.insightmediagroup.io"
    },
    {
      "type": "external",
      "href": "https://www.linkedin.com/company/towards-data-science/"
    },
    {
      "type": "external",
      "href": "https://x.com/TDataScience"
    },
    {
      "type": "external",
      "href": "https://roadmap.sh/ai-data-scientist"
    },
    {
      "type": "external",
      "href": "https://roadmap.sh/computer-science"
    },
    {
      "type": "external",
      "href": "https://roadmap.sh/machine-learning"
    },
    {
      "type": "external",
      "href": "https://pytorch.org/"
    },
    {
      "type": "external",
      "href": "https://www.tensorflow.org/"
    },
    {
      "type": "external",
      "href": "https://ludwig.ai/latest/"
    },
    {
      "type": "external",
      "href": "https://www.kaggle.com/datasets/shelvigarg/sales-forecasting-womart-store"
    },
    {
      "type": "external",
      "href": "https://github.com/ludwig-ai/ludwig/blob/master/LICENSE"
    },
    {
      "type": "external",
      "href": "https://lfaidata.foundation/"
    },
    {
      "type": "external",
      "href": "https://opendatacommons.org/licenses/dbcl/1-0/"
    },
    {
      "type": "external",
      "href": "https://ludwig.ai/latest/user_guide/how_ludwig_works/"
    },
    {
      "type": "external",
      "href": "https://ludwig.ai/latest/configuration/combiner/"
    },
    {
      "type": "external",
      "href": "https://ludwig.ai/latest/configuration/"
    },
    {
      "type": "external",
      "href": "https://github.com/john-adeojo/womartdata/blob/main/ludwig_deep_learning_notebook%20(2).ipynb"
    },
    {
      "type": "external",
      "href": "https://www.linkedin.com/in/john-adeojo/"
    },
    {
      "type": "external",
      "href": "https://johnadeojo.medium.com/membership"
    },
    {
      "type": "external",
      "href": "https://www.data-centric-solutions.com/book-online"
    }
  ],
  "repair_version": 2,
  "repaired_at": "2026-10-20T17:34:01Z",
  "source_url": "https://towardsdatascience.com/ludwig-a-friendlier-deep-learning-framework-946ee3d3b24/",
  "delivery_status": "unverified",
  "discovery_status": "http_403",
  "missing_reason": "源页面未重新核验:http_403；已按保留正文补采图片",
  "repair_error": "",
  "actual_video_files": [],
  "expected_saved_body_images": 4,
  "image_discovery": "saved_markdown",
  "source_body_markdown": "![Revisit consent button](https://cdn-cookieyes.com/assets/images/revisit.svg)\nWe value your privacy\nWe use cookies to enhance your browsing experience, serve personalised ads or content, and analyse our traffic. By clicking \"Accept All\", you consent to our use of cookies.\nCustomise Reject All Accept All\nCustomise Consent Preferences ![](https://cdn-cookieyes.com/assets/images/close.svg)\nWe use cookies to help you navigate efficiently and perform certain functions. You will find detailed information about all cookies under each consent category below.\nThe cookies that are categorised as \"Necessary\" are stored on your browser as they are essential for enabling the basic functionalities of the site. ... Show more\nNecessaryAlways Active\nNecessary cookies are required to enable the basic features of this site, such as providing secure log-in or adjusting your consent preferences. These cookies do not store any personally identifiable data.\n  * Cookie\nBCTempID\n  * Duration\n10 minutes\n  * Description\nNo description available.\n\n\n  * Cookie\n__cf_bm\n  * Duration\n1 hour\n  * Description\nThis cookie, set by Cloudflare, is used to support Cloudflare Bot Management. \n\n\n  * Cookie\nAWSALBCORS\n  * Duration\n7 days\n  * Description\nAmazon Web Services set this cookie for load balancing.\n\n\n  * Cookie\n_cfuvid\n  * Duration\nsession\n  * Description\nCloudflare sets this cookie to track users across sessions to optimize user experience by maintaining session consistency and providing personalized services\n\n\n  * Cookie\nli_gc\n  * Duration\n6 months\n  * Description\nLinkedin set this cookie for storing visitor's consent regarding using cookies for non-essential purposes.\n\n\n  * Cookie\n__hssrc\n  * Duration\nsession\n  * Description\nThis cookie is set by Hubspot whenever it changes the session cookie. The __hssrc cookie set to 1 indicates that the user has restarted the browser, and if the cookie does not exist, it is assumed to be a new session.\n\n\n  * Cookie\n__hssc\n  * Duration\n1 hour\n  * Description\nHubSpot sets this cookie to keep track of sessions and to determine if HubSpot should increment the session number and timestamps in the __hstc cookie.\n\n\n  * Cookie\nwpEmojiSettingsSupports\n  * Duration\nsession\n  * Description\nWordPress sets this cookie when a user interacts with emojis on a WordPress site. It helps determine if the user's browser can display emojis properly.\n\n\n  * Cookie\nBCSessionID\n  * Duration\n1 year 1 month 4 days\n  * Description\nBlueconic sets this cookie as a unique identifier for the BlueConic profile.\n\n\n  * Cookie\n_octo\n  * Duration\n1 year\n  * Description\nNo description available.\n\n\n  * Cookie\nlogged_in\n  * Duration\n1 year\n  * Description\nNo description available.\n\n\n  * Cookie\n__Secure-YEC\n  * Duration\npast\n  * Description\nYouTube sets this cookie to stores the user's video player preferences using embedded YouTube video\n\n\n  * Cookie\n__eoi\n  * Duration\n6 months\n  * Description\nDescription is currently not available.\n\n\n  * Cookie\nAWSALBTGCORS\n  * Duration\n7 days\n  * Description\nNo description available.\n\n\n  * Cookie\nlogin-status-p\n  * Duration\npast\n  * Description\nDescription is currently not available.\n\n\n  * Cookie\nAWSALBTG\n  * Duration\n7 days\n  * Description\nNo description available.\n\n\n  * Cookie\ncsrf_token\n  * Duration\nsession\n  * Description\nNo description available.\n\n\n  * Cookie\ntoken_v2\n  * Duration\n1 day\n  * Description\nDescription is currently not available.\n\n\n  * Cookie\nD\n  * Duration\n1 year\n  * Description\nDescription is currently not available.\n\n\n  * Cookie\nPHPSESSID\n  * Duration\nsession\n  * Description\nThis cookie is native to PHP applications. The cookie stores and identifies a user's unique session ID to manage user sessions on the website. The cookie is a session cookie and will be deleted when all the browser windows are closed.\n\n\n  * Cookie\nVISITOR_PRIVACY_METADATA\n  * Duration\n6 months\n  * Description\nYouTube sets this cookie to store the user's cookie consent state for the current domain. \n\n\n  * Cookie\ncookietest\n  * Duration\nsession\n  * Description\nThe cookietest cookie is typically used to determine whether the user's browser accepts cookies, essential for website functionality and user experience.\n\n\n  * Cookie\n__Host-airtable-session\n  * Duration\n1 year\n  * Description\nThis cookie is used to enable us to integrate the services of Airtable.\n\n\n  * Cookie\n__Host-airtable-session.sig\n  * Duration\n1 year\n  * Description\nThis cookie is used to enable us to integrate the services of Airtable.\n\n\n  * Cookie\nm\n  * Duration\n1 year 1 month 4 days\n  * Description\nStripe sets this cookie for fraud prevention purposes. It identifies the device used to access the website, allowing the website to be formatted accordingly.\n\n\n  * Cookie\nBIGipServer*\n  * Duration\nsession\n  * Description\nMarketo sets this cookie to collect information about the user's online activity and build a profile about their interests to provide advertisements relevant to the user. \n\n\n  * Cookie\n__cfruid\n  * Duration\nsession\n  * Description\nCloudflare sets this cookie to identify trusted web traffic.\n\n\n  * Cookie\n_GRECAPTCHA\n  * Duration\n6 months\n  * Description\nGoogle Recaptcha service sets this cookie to identify bots to protect the website against malicious spam attacks.\n\n\n  * Cookie\n__Secure-YNID\n  * Duration\n6 months\n  * Description\nGoogle cookie used to protect user security and prevent fraud, especially during the login process.\n\n\n  * Cookie\ncookieyes-consent\n  * Duration\n1 year\n  * Description\nCookieYes sets this cookie to remember users' consent preferences so that their preferences are respected on subsequent visits to this site. It does not collect or store any personal information about the site visitors.\n\n\nFunctional\nFunctional cookies help perform certain functionalities like sharing the content of the website on social media platforms, collecting feedback, and other third-party features.\n  * Cookie\nlidc\n  * Duration\n1 day\n  * Description\nLinkedIn sets the lidc cookie to facilitate data center selection.\n\n\n  * Cookie\nbrw\n  * Duration\n1 year\n  * Description\nNo description available.\n\n\n  * Cookie\nbrwConsent\n  * Duration\n5 minutes\n  * Description\nDescription is currently not available.\n\n\n  * Cookie\nWMF-Uniq\n  * Duration\n1 year\n  * Description\nDescription is currently not available.\n\n\n  * Cookie\nloom_anon_comment\n  * Duration\n1 year\n  * Description\nNo description available.\n\n\n  * Cookie\nloom_referral_video\n  * Duration\nsession\n  * Description\nDescription is currently not available.\n\n\n  * Cookie\nVISITOR_INFO1_LIVE\n  * Duration\n6 months\n  * Description\nA cookie set by YouTube to measure bandwidth that determines whether the user gets the new or old player interface.\n\n\n  * Cookie\nyt-remote-connected-devices\n  * Duration\nNever Expires\n  * Description\nYouTube sets this cookie to store the user's video preferences using embedded YouTube videos.\n\n\n  * Cookie\nytidb::LAST_RESULT_ENTRY_KEY\n  * Duration\nNever Expires\n  * Description\nThe cookie ytidb::LAST_RESULT_ENTRY_KEY is used by YouTube to store the last search result entry that was clicked by the user. This information is used to improve the user experience by providing more relevant search results in the future.\n\n\n  * Cookie\nyt-remote-device-id\n  * Duration\nNever Expires\n  * Description\nYouTube sets this cookie to store the user's video preferences using embedded YouTube videos.\n\n\n  * Cookie\nyt-remote-session-name\n  * Duration\nsession\n  * Description\nThe yt-remote-session-name cookie is used by YouTube to store the user's video player preferences using embedded YouTube video.\n\n\n  * Cookie\nyt-remote-fast-check-period\n  * Duration\nsession\n  * Description\nThe yt-remote-fast-check-period cookie is used by YouTube to store the user's video player preferences for embedded YouTube videos.\n\n\n  * Cookie\nyt-remote-session-app\n  * Duration\nsession\n  * Description\nThe yt-remote-session-app cookie is used by YouTube to store user preferences and information about the interface of the embedded YouTube video player.\n\n\n  * Cookie\nyt-remote-cast-available\n  * Duration\nsession\n  * Description\nThe yt-remote-cast-available cookie is used to store the user's preferences regarding whether casting is available on their YouTube video player.\n\n\n  * Cookie\nyt-remote-cast-installed\n  * Duration\nsession\n  * Description\nThe yt-remote-cast-installed cookie is used to store the user's video player preferences using embedded YouTube video.\n\n\n  * Cookie\ncp_session\n  * Duration\n3 months\n  * Description\nCodepen sets this cookie for Help systems found in the website.\n\n\n  * Cookie\nloid\n  * Duration\n1 year 1 month 4 days\n  * Description\nThis cookie is set by the Reddit. The cookie enables the sharing of content from the website onto the social media platform.\n\n\nAnalytics\nAnalytical cookies are used to understand how visitors interact with the website. These cookies help provide information on metrics such as the number of visitors, bounce rate, traffic source, etc.\n  * Cookie\n__hstc\n  * Duration\n6 months\n  * Description\nHubspot set this main cookie for tracking visitors. It contains the domain, initial timestamp (first visit), last timestamp (last visit), current timestamp (this visit), and session number (increments for each subsequent session).\n\n\n  * Cookie\nhubspotutk\n  * Duration\n6 months\n  * Description\nHubSpot sets this cookie to keep track of the visitors to the website. This cookie is passed to HubSpot on form submission and used when deduplicating contacts.\n\n\n  * Cookie\n_ga\n  * Duration\n1 year 1 month 4 days\n  * Description\nGoogle Analytics sets this cookie to calculate visitor, session and campaign data and track site usage for the site's analytics report. The cookie stores information anonymously and assigns a randomly generated number to recognise unique visitors.\n\n\n  * Cookie\n_ga_*\n  * Duration\n1 year 1 month 4 days\n  * Description\nGoogle Analytics sets this cookie to store and count page views.\n\n\n  * Cookie\n__Host-psifi.analyticsTrace\n  * Duration\n6 hours\n  * Description\nDescription is currently not available.\n\n\n  * Cookie\n__Host-psifi.analyticsTraceV2\n  * Duration\n6 hours\n  * Description\nDescription is currently not available.\n\n\n  * Cookie\n_gh_sess\n  * Duration\nsession\n  * Description\nGitHub sets this cookie for temporary application and framework state between pages like what step the user is on in a multiple step form.\n\n\n  * Cookie\nYSC\n  * Duration\nsession\n  * Description\nYSC cookie is set by Youtube and is used to track the views of embedded videos on Youtube pages.\n\n\n  * Cookie\najs_anonymous_id\n  * Duration\n1 year\n  * Description\nThis cookie is set by Segment to count the number of people who visit a certain site by tracking if they have visited before.\n\n\n  * Cookie\nvuid\n  * Duration\n1 year 1 month 4 days\n  * Description\nVimeo installs this cookie to collect tracking information by setting a unique ID to embed videos on the website. \n\n\nPerformance\nPerformance cookies are used to understand and analyse the key performance indexes of the website which helps in delivering a better user experience for the visitors.\n  * Cookie\nAWSALB\n  * Duration\n7 days\n  * Description\nAWSALB is an application load balancer cookie set by Amazon Web Services to map the session to the target.\n\n\n  * Cookie\nacq\n  * Duration\npast\n  * Description\nDescription is currently not available.\n\n\n  * Cookie\nacq.sig\n  * Duration\npast\n  * Description\nDescription is currently not available.\n\n\n  * Cookie\nptc\n  * Duration\n2 years\n  * Description\nNo description available.\n\n\nAdvertisement\nAdvertisement cookies are used to provide visitors with customised advertisements based on the pages you visited previously and to analyse the effectiveness of the ad campaigns.\n  * Cookie\nmuc_ads\n  * Duration\n1 year 1 month 4 days\n  * Description\nTwitter sets this cookie to collect user behaviour and interaction data to optimize the website.\n\n\n  * Cookie\nguest_id_marketing\n  * Duration\n1 year 1 month 4 days\n  * Description\nTwitter sets this cookie to identify and track the website visitor.\n\n\n  * Cookie\nguest_id_ads\n  * Duration\n1 year 1 month 4 days\n  * Description\nTwitter sets this cookie to identify and track the website visitor.\n\n\n  * Cookie\npersonalization_id\n  * Duration\n1 year 1 month 4 days\n  * Description\nTwitter sets this cookie to integrate and share features for social media and also store information about how the user uses the website, for tracking and targeting.\n\n\n  * Cookie\nguest_id\n  * Duration\n1 year 1 month 4 days\n  * Description\nTwitter sets this cookie to identify and track the website visitor. It registers if a user is signed in to the Twitter platform and collects information about ad preferences.\n\n\n  * Cookie\nbcookie\n  * Duration\n1 year\n  * Description\nLinkedIn sets this cookie from LinkedIn share buttons and ad tags to recognize browser IDs.\n\n\n  * Cookie\n__Secure-ROLLOUT_TOKEN\n  * Duration\n6 months\n  * Description\nYouTube sets this cookie to manage feature rollout and experimentation. It helps Google control which new features or interface changes are shown to users as part of testing and staged rollouts, ensuring consistent experience for a given user during an experiment.\n\n\n  * Cookie\nyt.innertube::nextId\n  * Duration\nNever Expires\n  * Description\nYouTube sets this cookie to register a unique ID to store data on what videos from YouTube the user has seen.\n\n\n  * Cookie\nyt.innertube::requests\n  * Duration\nNever Expires\n  * Description\nYouTube sets this cookie to register a unique ID to store data on what videos from YouTube the user has seen.\n\n\n  * Cookie\nsession_tracker\n  * Duration\nsession\n  * Description\nThis cookie is set by the Reddit. This cookie is used to identify trusted web traffic. It also helps in adverstising on the website.\n\n\n  * Cookie\nedgebucket\n  * Duration\nsession\n  * Description\nReddit sets this cookie to save the information about a log-on Reddit user, for the purpose of advertisement recommendations and updating the content.\n\n\n  * Cookie\ndid\n  * Duration\n1 year\n  * Description\nArbor sets this cookie to show targeted ads to site visitors.This cookie expires after 2 months or 1 year.\n\n\nUncategorised\nOther uncategorised cookies are those that are being analysed and have not been classified into a category as yet.\nNo cookies to display.\nReject All Save My Preferences Accept All\nToggle Search[](https://towardsdatascience.com/)Toggle Mobile Navigation\n[](https://towardsdatascience.com/)\nPublish AI, ML & data-science insights to a global community of data professionals.\nSign in[Submit an Article](https://write.insightmediagroup.io)\n[Latest](https://towardsdatascience.com/latest/)[Editor's Picks](https://towardsdatascience.com/tag/editors-pick/)[Newsletter](https://towardsdatascience.com/tag/the-variable/)[ShipAI](https://towardsdatascience.com/shipai/)\n* * *\n[Write For TDS](https://towardsdatascience.com/submissions/)\n[LinkedIn](https://www.linkedin.com/company/towards-data-science/)[X](https://x.com/TDataScience)Toggle Search\n[Data Science](https://towardsdatascience.com/category/data-science/)\n# Ludwig - A \"Friendlier\" Deep Learning Framework\nDeep Learning Made Easier with this low-code Declarative Framework\n[John Adeojo](https://towardsdatascience.com/author/johnadeojo/)\nJune 26, 202312 min read\n### Deep Learning Made Easy with this Low-Code, Declarative Framework\n![Image by Author: Generated with Midjourney](https://assets.insightmediagroup.io/media/wp-content/uploads/2023/06/1c7aAb_-ShzIYXrkB1NcYAA.png)\nImage by Author: Generated with Midjourney\nLearn this step by step with the interactive [AI and Data Scientist](https://roadmap.sh/ai-data-scientist), [Computer Science](https://roadmap.sh/computer-science) and [Machine Learning](https://roadmap.sh/machine-learning) roadmaps.\n## Background - Deep Learning, too Complex?\nI have tended to shy away from deep learning for industry use cases. Not due to a lack of interest, rather I find the popular deep learning frameworks cumbersome. I appreciate [Pytorch](https://pytorch.org/) and [TensorFlow](https://www.tensorflow.org/) are fantastic tools for research purposes, but the APIs are not the most user friendly. In situations where I need to deliver a quick proof of concept for a client, the last thing I want to be doing is tinkering with Pytorch tensors.\nWhile attending the AI summit in London, I stumbled across a team claiming to have a solution to my deep learning problem. They have been leveraging a different approach, which they described as \"the mid-point between TensorFlow and AutoML\", a framework called Ludwig.\n### What is Ludwig?\nDeveloped by Uber, [Ludwig](https://ludwig.ai/latest/) is an open-source framework for building deep learning models. It is declarative, which means that instead of constructing complex models layer-by-layer like you would in TensorFlow, you simply declare the structure of the model with a configuration file. This sounded too good to be true, so I had to see for myself. In the remainder of this article, I detail my experience with Ludwig through an example project I have taken from [Kaggle](https://www.kaggle.com/datasets/shelvigarg/sales-forecasting-womart-store). Along the way I will discuss some of its strengths, pain-points, and give you my conclusion on whether it's worth using.\n_Note - Although initially developed by Uber, Ludwig is an_[ _open source_](https://github.com/ludwig-ai/ludwig/blob/master/LICENSE) _library under the Apache 2.0 License. The project is hosted by the_[ _Linux Foundation AI & Data_](https://lfaidata.foundation/) _. I have no commercial affiliation with Uber or the developers of Ludwig._\n* * *\n## The project - Demand Forecasting\n**The project brief** : Forecast the final 30 days of orders across each store for the retailer WOMart.\n> Your Client WOMart is a leading nutrition and supplement retail chain that offers a comprehensive range of products for all your wellness and fitness needs.\n> WOMart follows a multi-channel distribution strategy with 350+ retail stores spread across 100+ cities.\n### The Data\nThe dataset had a total of 22,265 observations, each pertaining to one day of sales at a given store. For brevity I won't go into the full details of the data set, but you can see some descriptive statistics [here](https://www.kaggle.com/datasets/shelvigarg/sales-forecasting-womart-store).\n**Note** : The data is free to use for any purposes under the Open Data Commons license.\n  * Link to the [data](https://www.kaggle.com/datasets/shelvigarg/sales-forecasting-womart-store)\n  * [License](https://opendatacommons.org/licenses/dbcl/1-0/) for the data\n\n\n**The data dictionary:**\n* * *\n## A Brief Overview of the Methodology\nI won't go into too much detail on the methodology here as this isn't the main purpose of this article. I'll cover how I framed the problem at a high level to give you some context.\nI framed the forecasting issue as a 'pseudo' sequence-to-sequence deep learning problem. The method entailed utilising 360 days of time series data points to predict the subsequent 30 days of customer orders. I incorporated some categorical variables and had to generate separate predictions for each day of orders, which led to a somewhat unconventional setup - thus the term 'pseudo' sequence-to-sequence for describing the problem. I will expand on the specifics of the feature engineering later in this article.\nBesides this, the methodology I followed was standard for model development. I split my data into a training data set and a holdout dataset and did some re-scaling of features and labels. Model training was done on the training data, and testing on the holdout.\n_Note: Ludwig does provide the capability to natively split data within the API. However, to maintain rigour, I established a separate holdout dataset. The training dataset was subsequently divided into additional training, validation, and testing subsets. The holdout dataset was excluded entirely, used only for analysing the predictions generated by the model._\n* * *\n## Feature Engineering\nAt the time of writing this article, I would say that sequence-to-sequence modelling as it pertains to timeseries forecasting is tricky to do in Ludwig. The reason for this is the feature engineering. The Ludwig API has an excellent ability to handle sequences as inputs, but they haven't (yet) developed a coherent approach for time series sequences as outputs. You can develop a \"pseudo\" sequence-to-sequence model by declaring multiple outputs, but the overall feature engineering experience feels quite \"hacky\".\n**Sequence features** : Except for those features that don't change with time, I engineered all the predictive features to be \"Ludwig-formatted\" sequences. Each input sequence is a horizontal stacking of each \"time series\" feature across a pre-defined timeframe. Each feature sequence is determined at the store level and encapsulated in one cell of the dataframe (it looks as messy as it sounds).\n**Sequence labels** : For sequence labels, you must declare each point in the sequence as an individual label for the model. This ended up with me declaring 30 labels for each store, one label per day of orders to be predicted.\n_Example of the feature engineering process below:_\n* * *\n## Architecting Your Model\nThe Ludwig API allows you to architect fairly complex and customisable models declaratively. Ludwig does this through a .yaml file. Now, I appreciate many data scientists reading this might not have used _.yaml_ files, but generally in software development these are used for configuration. The files might appear scary at first glance, but they are quite friendly. Let's step through the main parts of the file I create to build the model.\n![Image by Author: Model architecture](https://assets.insightmediagroup.io/media/wp-content/uploads/2023/06/1eSRESUee1wx0RFgf_rOCZQ.png)\nImage by Author: Model architecture\nBefore we delve into the configurations, it's worth briefly introducing the [architecture](https://ludwig.ai/latest/user_guide/how_ludwig_works/) at the heart of Ludwig's deep learning framework: the Encoder, Combiner, and Decoder. Most of the models you configure in Ludwig will predominantly adhere to this architecture. Understanding this can simplify the process of stacking components to quickly build your deep learning models.\n### Declaring your Model\nRight at the top of the file you declare the model type used. Ludwig provides two options, tree-based models, and deep neural networks for which I chose the latter.\nyaml\n\n```\nmodel_type: ecd\n```\n\n### Declaring your data splits\nYou can split data sets natively by declaring your split percentages, type of split, and column or variable you're splitting on. For my purposes I wanted to ensure that a store could only appear in one of the data sets, hash splitting was perfect for that.\nFor best practice, I would probably advise constructing a holdout set outside of the Ludwig API especially where you are doing some initial feature engineering like one-hot-encoding or normalisation. This should help prevent data leakage.\nyaml\n\n```\nmodel_type: ecd\nsplit:\n    type: hash\n    column: Store_id\n    probabilities:\n    - 0.7\n    - 0.15\n    - 0.15\n#...omitted sections...\n```\n\n### Declaring the Model Inputs\nYou declare inputs by name, type, and encoder. Depending on the type of input to the model you have a variety of options for encoders. Essentially encoders are a way of transforming your inputs such that it can be interpreted by the model. The choice of encoder really depends on the data and the modelling task.\nyaml\n\n```\nmodel_type: ecd\nsplit:\n    type: hash\n    column: Store_id\n    probabilities:\n    - 0.7\n    - 0.15\n    - 0.15\ninput_features:\n  - name: Sales\n    type: sequence\n    encoder: stacked_cnn\n    reduce_output: null\n  - name: Order\n    type: sequence\n    encoder: stacked_cnn\n    reduce_output: null\n  - name: Discount\n    type: sequence\n    encoder: stacked_cnn\n    reduce_output: null\n  - name: DayOfWeek\n    type: sequence\n    encoder: stacked_cnn\n    reduce_output: null\n  - name: MonthOfYear\n    type: sequence\n    encoder: stacked_cnn\n    reduce_output: null\n  - name: Holiday\n    type: sequence\n    encoder: stacked_cnn\n    reduce_output: null\n  - name: Store_Type\n    type: category\n    encoder: dense\n  - name: Location_Type\n    type: category\n    encoder: dense\n  - name: Region_Code\n    type: category\n    encoder: dense\n#...omitted sections...\n```\n\n### Declaring the Combiner\n[Combiners](https://ludwig.ai/latest/configuration/combiner/), as the name suggests, amalgamate the outputs of your encoders. The Ludwig API offers an array of different combiners, each with its own specific use case. The choice of combiner can depend on the structure of your model and the relationships between your features. For instance, you might use a 'concat' combiner if you want to simply concatenate the outputs of your encoders, or a 'sequence' combiner if your features have a sequential relationship.\nyaml\n\n```\nmodel_type: ecd\nsplit:\n    type: hash\n    column: Store_id\n    probabilities:\n    - 0.7\n    - 0.15\n    - 0.15\ninput_features:\n  - name: Sales\n    type: sequence\n    encoder: stacked_cnn\n    reduce_output: null\n  - name: Order\n    type: sequence\n    encoder: stacked_cnn\n    reduce_output: null \n  # ... omitted sections ...\n\n  - name: Location_Type\n    type: category\n    encoder: dense\n  - name: Region_Code\n    type: category\n    encoder: dense\ncombiner:\n    type: sequence\n    main_sequence_feature: Order\n    reduce_output: null\n    encoder:\n    # ... omitted sections ...\n```\n\nAs with many aspects of deep learning, the optimal choice of combiner often depends on the specifics of your dataset and problem, and may require some experimentation.\n### Declaring the Model Outputs\nFinalising your network is as simple as declaring your outputs, which are just your labels. My pet peeve with Ludwig for timeseries is that you can't (yet) declare timeseries outputs. As I mentioned previously, you have to \"hack\" it by declaring each point in your time series separately. This left me with thirty separate declarations, which looks very messy in all honesty. For each output you can specify the loss function too adding additional configurability. Ludwig has a myriad of options pre-built for different output types, however I don't know if you are able to implement custom loss functions as you can with Pytorch.\nyaml\n\n```\nmodel_type: ecd\nsplit:\n    type: hash\n    column: Store_id\n    probabilities:\n    - 0.7\n    - 0.15\n    - 0.15\ninput_features:\n  - name: Sales\n    type: sequence\n    encoder: stacked_cnn\n    reduce_output: null\n  - name: Order\n    type: sequence\n    encoder: stacked_cnn\n    reduce_output: null\n# ...omitted sections...\n\n  - name: Location_Type\n    type: category\n    encoder: dense\n  - name: Region_Code\n    type: category\n    encoder: dense\ncombiner:\n    type: sequence\n    main_sequence_feature: Order\n    reduce_output: null\n    encoder:\n        type: parallel_cnn\noutput_features:\n  - name: Order_sequence_label_2019-05-02\n    type: number\n    loss:\n      type: mean_absolute_error\n  - name: Order_sequence_label_2019-05-03\n    type: number\n    loss:\n      type: mean_absolute_error\n#...omitted sections...\n\n      type: mean_absolute_error\n  - name: Order_sequence_label_2019-05-30\n    type: number\n    loss:\n      type: mean_absolute_error\n  - name: Order_sequence_label_2019-05-31\n    type: number\n    loss:\n      type: mean_absolute_error\n#...omitted sections...\n```\n\n### Declaring the Trainer\nThe trainer configuration in Ludwig, while optional due to Ludwig's provision of sensible defaults, allows for a high degree of customisation. It gives you control over the specifics of how your model is trained. This includes the ability to specify the type of optimiser used, the number of training epochs, the learning rate, and criteria for early stopping, among other parameters.\nyaml\n\n```\nmodel_type: ecd\nsplit:\n    type: hash\n    column: Store_id\n    probabilities:\n    - 0.7\n    - 0.15\n    - 0.15\ninput_features:\n  - name: Sales\n    type: sequence\n    encoder: stacked_cnn\n    reduce_output: null\n  - name: Order\n    type: sequence\n    encoder: stacked_cnn\n    reduce_output: null\n# ...omitted sections...\n\n  - name: Location_Type\n    type: category\n    encoder: dense\n  - name: Region_Code\n    type: category\n    encoder: dense\ncombiner:\n    type: sequence\n    main_sequence_feature: Order\n    reduce_output: null\n    encoder:\n        type: parallel_cnn\noutput_features:\n  - name: Order_sequence_label_2019-05-02\n    type: number\n    loss:\n      type: mean_absolute_error\n  - name: Order_sequence_label_2019-05-03\n    type: number\n    loss:\n      type: mean_absolute_error\n#...omitted sections...\n\n      type: mean_absolute_error\n  - name: Order_sequence_label_2019-05-30\n    type: number\n    loss:\n      type: mean_absolute_error\n  - name: Order_sequence_label_2019-05-31\n    type: number\n    loss:\n      type: mean_absolute_error\ntrainer:\n    epochs: 200\n    learning_rate: 0.0001\n    early_stop: 20\n    evaluate_training_set: true\n    validation_metric: mean_absolute_error\n    validation_field: Order_sequence_label_2019-05-31\n```\n\nFor your particular use case, you might find it beneficial to define these parameters yourself. For instance, you might want to adjust the learning rate, or the number of epochs based on the complexity of your model and the size of your dataset. Similarly, early stopping can be a useful tool to prevent overfitting by halting the training process if the model's performance on a validation set stops improving.\n### Train your Model\nTraining your model can be easily done with Ludwig's python expermiment API. See the script example below:\n### Other Configurations\nOutside of those I mentioned, Ludwig has a myriad of possible configurations. They are all very well documented and well structured. I would advise having a read of their [documentation](https://ludwig.ai/latest/configuration/) to familiarise yourself.\n* * *\n## Model Performance Analysis - A brief view\nThe aim of this article is to explore some of the functionalities of the Ludwig framework through a practical example project. Although demonstrating model performance is part of this, there's no need to delve into the specifics of the metrics. I will limit the discussion to presenting some of the charts generated from the model's analysis. Note that the comprehensive end-to-end script is available on my GitHub, linked at the conclusion of the article.\n![](https://assets.insightmediagroup.io/media/wp-content/uploads/2023/06/1Ctm2Y8eYo_qV4LTfvkY2Tg.png)\n![](https://assets.insightmediagroup.io/media/wp-content/uploads/2023/06/1ZRh9Ov21Z997_JryIV7Z2g.png)\n![](https://assets.insightmediagroup.io/media/wp-content/uploads/2023/06/1Ctm2Y8eYo_qV4LTfvkY2Tg.png)\n![Image by Author: Model predictions \\(red\\) against actual orders \\(blue\\) for the last 30 days of data. These examples are drawn from the holdout set.](https://assets.insightmediagroup.io/media/wp-content/uploads/2023/06/1bC_f58290lrzDngBLoxh8A.png)\nImage by Author: Model predictions (red) against actual orders (blue) for the last 30 days of data. These examples are drawn from the holdout set.\n![Image by Author: Distribution of error, where error is Actual Orders minus Predicted Orders.](https://assets.insightmediagroup.io/media/wp-content/uploads/2023/06/14kcUgoUDImys5IFb4Saxkw.png)\nImage by Author: Distribution of error, where error is Actual Orders minus Predicted Orders.\n![Image by Author: Loss curve for the model training. Loss metric is Mean Absolute Error](https://assets.insightmediagroup.io/media/wp-content/uploads/2023/06/1jrlFOdpwHZEG_RMlnpWGIg.png)\nImage by Author: Loss curve for the model training. Loss metric is Mean Absolute Error\n* * *\n## My Verdict\nI'll begin by confessing that I was initially sceptical about Ludwig. However, after experimenting with it myself, I'm convinced of its capabilities and believe it delivers as promised. I think there are several truly impressive features worth highlighting.\n**Coding experience** : The coding experience feels more like constructing an elaborate Lego set. You really can have a lot of fun messing around with components and different architectures to find your perfect model.\n**Documentation** : The documentation is clear and well structured. It was very easy to figure out how to implement different architectures and change your model configurations. Most of the documentation appears to be up to date too, which is a bonus.\n**Backend** : The backend experience is excellent. The library developers have done a great job abstracting away a lot of the usual configurations you have to do to train deep neural networks on GPUs. I trained my model on a Google Collab, and Ludwig automatically moved the workloads onto the GPU.\nAnother great thing about Ludwig is that the backend is highly configurable. For example, if you're running massive workloads and need a cluster of GPUs, you can configure this too!\n**Experiment Tracking** : Ludwig provides you with an Experiment API that can be used to track your model artefacts between experiment runs. I believe there are integrations with MLflow too which is fantastic for commercial scale MLOps.\n### Pet Peeves\nThere are a few areas in which the framework could be enhanced, let's get into those too.\n**Visualisation** : Ludwig does offer a visualisation API to track training losses across your data sets. However, at the time of writing, it does not function particularly well and its usage is unintuitive. I tried running it in a Google Collab, but to no avail. Eventually, I created my own loss curve visualisations by scripting a Python function to parse the training_statistics.json files that Ludwig saves for you after each experiment run.\n**Support** : Whilst there is some support available, Ludwig's community does not seem to be as extensive as those of TensorFlow or Pytorch yet. A few issues have been raised on GitHub, and there are some threads that might offer assistance, but for the most part, it feels like you are left to your own devices. As for ChatGPT, it provides a certain level of support up until 2021.\n**Transparency** : Ludwig excels in eliminating the more challenging aspects of building deep learning models. However, this comes at the expense of transparency, which occasionally renders the logs somewhat cryptic and challenging to debug.\n* * *\n## Conclusion\nIn my opinion, Ludwig is an excellent tool for those looking to start utilising deep learning whether in a commercial setting or just for learning. While it may be too high-level for state-of-the-art research purposes, it is well-suited to solving clearly defined problems swiftly. A certain degree of understanding of deep learning is still necessary, but once you've grasped the concept, the barrier to entry for Ludwig is considerably lower than it is for TensorFlow or Pytorch.\nThe end-to-end notebook is available in my [GitHub](https://github.com/john-adeojo/womartdata/blob/main/ludwig_deep_learning_notebook%20\\(2\\).ipynb) repo, please feel free to experiment.\n* * *\n_Follow me on_[ _LinkedIn_](https://www.linkedin.com/in/john-adeojo/)\n_Subscribe to medium to get more insights from me:_\n> [**Join Medium with my referral link - John Adeojo**](https://johnadeojo.medium.com/membership)\n_Should you be interested in integrating AI or data science into your business operations, we invite you to schedule a complimentary initial consultation with us:_\n> [**Book Online | Data-Centric Solutions**](https://www.data-centric-solutions.com/book-online)\n## Register Now to Continue Reading\n### Unlock Towards Data Science with a Free Account\nAll articles on Towards Data Science are available to registered users. Create a free account today to access this article and thousands more.\nRegister for freeSign in to your account\n* * *\n![](https://assets.insightmediagroup.io/media/1787657091747_phu3q1.png)\nWritten By\n## [John Adeojo](https://towardsdatascience.com/author/johnadeojo/)\n[See all from John Adeojo](https://towardsdatascience.com/author/johnadeojo/)\n[Data Science](https://towardsdatascience.com/tag/data-science/)[Deep Learning](https://towardsdatascience.com/tag/deep-learning/)[Hands On Tutorials](https://towardsdatascience.com/tag/hands-on-tutorials/)[Machine Learning](https://towardsdatascience.com/tag/machine-learning/)[Programming](https://towardsdatascience.com/tag/programming/)\nShare This Article\n  * [](https://www.facebook.com/sharer/sharer.php?u=https%3A%2F%2Ftowardsdatascience.com%2Fludwig-a-friendlier-deep-learning-framework-946ee3d3b24%2F&title=Ludwig%20-%20A%20%22Friendlier%22%20Deep%20Learning%20Framework)\n  * [](https://www.linkedin.com/shareArticle?mini=true&url=https%3A%2F%2Ftowardsdatascience.com%2Fludwig-a-friendlier-deep-learning-framework-946ee3d3b24%2F&title=Ludwig%20-%20A%20%22Friendlier%22%20Deep%20Learning%20Framework)\n  * [](https://x.com/share?url=https%3A%2F%2Ftowardsdatascience.com%2Fludwig-a-friendlier-deep-learning-framework-946ee3d3b24%2F&text=Ludwig%20-%20A%20%22Friendlier%22%20Deep%20Learning%20Framework)\n\n\nTowards Data Science is a community publication. Submit your insights to reach our global audience and earn through the TDS Author Payment Program.\n[Write for TDS](https://towardsdatascience.com/submissions/)\n## Related Articles\n  * [![](https://assets.insightmediagroup.io/media/wp-content/uploads/2020/09/05ESOLmxZT2sYvH24-scaled.jpg)](https://towardsdatascience.com/step-by-step-yolo-model-deployment-in-localhost-using-python-8537e93a1784/)\n## [Deploy Custom Object Detection using Flask & Python, Step by Step](https://towardsdatascience.com/step-by-step-yolo-model-deployment-in-localhost-using-python-8537e93a1784/)\n[Artificial Intelligence](https://towardsdatascience.com/category/artificial-intelligence/)\nCustom Object Detection Localhost Deployment | YOLO\n[Pranjal Saxena](https://towardsdatascience.com/author/pranjalai/)\nSeptember 3, 20205 min read\n  * [![](https://assets.insightmediagroup.io/media/wp-content/uploads/2018/12/1qsvGjgwgBT_iU1ge-uh_xA.png)](https://towardsdatascience.com/data-science-in-algorithmic-trading-d21a46d1565d/)\n## [Data Science in Algorithmic Trading](https://towardsdatascience.com/data-science-in-algorithmic-trading-d21a46d1565d/)\n[Artificial Intelligence](https://towardsdatascience.com/category/artificial-intelligence/)\nIn this article I plan to give you a glimpse into an asset model for algorithmic trading. This model of the world should allow us to make...\n[Daniel Shapiro, PhD](https://towardsdatascience.com/author/lemaysolutions/)\nDecember 31, 201816 min read\n  * [![](https://assets.insightmediagroup.io/media/wp-content/uploads/2026/01/apples_to_apples.jpg)](https://towardsdatascience.com/stop-blaming-the-data-a-better-way-to-handle-covariance-shift/)\n## [Stop Blaming the Data: A Better Way to Handle Covariance Shift](https://towardsdatascience.com/stop-blaming-the-data-a-better-way-to-handle-covariance-shift/)\n[Data Science](https://towardsdatascience.com/category/data-science/)\nInstead of using data shift as an excuse for poor performance, use Inverse Probability Weighting to estimate how your model should perform in the new environment\n[Alon Lanyado](https://towardsdatascience.com/author/alon.lanyado/)\nJanuary 5, 20267 min read\n  * [![](https://assets.insightmediagroup.io/media/wp-content/uploads/2022/04/0bzmBi8HS3AgvM4js-scaled.jpg)](https://towardsdatascience.com/8-visualizations-with-python-to-handle-multiple-time-series-data-19b5b2e66dd0/)\n## [8 Visualizations with Python to Handle Multiple Time-Series Data](https://towardsdatascience.com/8-visualizations-with-python-to-handle-multiple-time-series-data-19b5b2e66dd0/)\n[Programming](https://towardsdatascience.com/category/programming/)\nVisualization ideas for coping with overlapping lines in multiple time-series plots.\n[Boriharn K](https://towardsdatascience.com/author/borih.k/)\nApril 28, 20228 min read\n  * [![](https://assets.insightmediagroup.io/media/wp-content/uploads/2019/07/0wk44fqAFOzM78UMs-scaled.jpg)](https://towardsdatascience.com/time-series-clustering-and-dimensionality-reduction-5b3b4e84f6a3/)\n## [Time Series Clustering and Dimensionality Reduction](https://towardsdatascience.com/time-series-clustering-and-dimensionality-reduction-5b3b4e84f6a3/)\n[Machine Learning](https://towardsdatascience.com/category/artificial-intelligence/machine-learning/)\nCluster sensor data with Kolmogorv Smirnov Statistic and Machine Learning\n[Marco Cerliani](https://towardsdatascience.com/author/cerlymarco/)\nJuly 25, 20195 min read\n  * [![](https://assets.insightmediagroup.io/media/wp-content/uploads/2019/05/0CE7fVdGB4qb8pPqb-scaled.jpg)](https://towardsdatascience.com/paper-readthrough-gauge-equivariant-convolutional-networks-and-the-icosahedral-cnn-13e9d3d95b85/)\n## [[Paper Readthrough] - Gauge Equivariant Convolutional Networks and the Icosahedral CNN](https://towardsdatascience.com/paper-readthrough-gauge-equivariant-convolutional-networks-and-the-icosahedral-cnn-13e9d3d95b85/)\n[Deep Learning](https://towardsdatascience.com/category/artificial-intelligence/deep-learning/)\nOverview\n[Nicola Bernini](https://towardsdatascience.com/author/nicolabernini_63880/)\nMay 25, 20194 min read\n  * [![](https://assets.insightmediagroup.io/media/wp-content/uploads/2021/10/0ll4M5yh1mslZoea3-scaled.jpg)](https://towardsdatascience.com/cliff-walking-problem-with-the-discrete-policy-gradient-algorithm-59d1900d80d8/)\n## [Cliff-Walking Problem With The Discrete Policy Gradient Algorithm](https://towardsdatascience.com/cliff-walking-problem-with-the-discrete-policy-gradient-algorithm-59d1900d80d8/)\n[Artificial Intelligence](https://towardsdatascience.com/category/artificial-intelligence/)\nA full implementation of the REINFORCE algorithm in Python. The steps are performed manually to illustrate the inner workings of the...\n[Wouter van Heeswijk, PhD](https://towardsdatascience.com/author/wvheeswijk/)\nOctober 21, 20218 min read\n  * [![](https://assets.insightmediagroup.io/media/wp-content/uploads/2022/02/0zzQ-GSGIMPUIvkGu-scaled.jpg)](https://towardsdatascience.com/stop-using-python-to-aggregate-data-use-sql-instead-7781d6ea4f0c/)\n## [Stop Using Python to Aggregate Data - Use SQL Instead](https://towardsdatascience.com/stop-using-python-to-aggregate-data-use-sql-instead-7781d6ea4f0c/)\n[Programming](https://towardsdatascience.com/category/programming/)\nAre you using Python to extract raw data from the database? It could be a huge bottleneck in your apps. Use SQL instead.\n[Dario Radečić](https://towardsdatascience.com/author/radecicdario/)\nFebruary 17, 20226 min read\n  * [![](https://assets.insightmediagroup.io/media/wp-content/uploads/2019/08/1oKh4cMArrQfvYFwihH8Egw-scaled.jpeg)](https://towardsdatascience.com/ai-for-industrial-process-control-ee774267094b/)\n## [AI for Industrial Process Control](https://towardsdatascience.com/ai-for-industrial-process-control-ee774267094b/)\n[Machine Learning](https://towardsdatascience.com/category/artificial-intelligence/machine-learning/)\nTuning a process oven using Reinforcement Learning\n[John Morrow](https://towardsdatascience.com/author/john_morrow/)\nAugust 16, 20193 min read\n\n\n  * [YouTube](https://www.youtube.com/c/TowardsDataScience)\n  * [X](https://x.com/TDataScience)\n  * [LinkedIn](https://www.linkedin.com/company/towards-data-science/)\n  * [Threads](https://www.threads.net/@towardsdatascience)\n  * [Bluesky](https://bsky.app/profile/towardsdatascience.com)\n\n\n[](https://towardsdatascience.com/)\nYour home for data science and AI. The world's leading publication for data science, data analytics, data engineering, machine learning, and artificial intelligence professionals.\nSubscribe to Our Newsletter[Write For TDS](https://towardsdatascience.com/submissions/)•[About](https://towardsdatascience.com/about-towards-data-science-d691af11cc2f/)•[Advertise](https://contact.towardsdatascience.com/advertise-with-towards-data-science)•[Privacy Policy](https://towardsdatascience.com/privacy-policy/)•[Terms of Use](https://towardsdatascience.com/website-terms-of-use/)\n© Insight Media Group, LLC 2026•[The New Stack](https://thenewstack.io)•[roadmap.sh](https://roadmap.sh)•[Creator Network](https://insightmediagroup.io/creator-network)\n####  Recommended Articles \nClose\n  * [ ![](https://h030.towardsdatascience.com/rest/v2/contentStores/d7d3a4ab-7caf-4383-bbd2-fbf7fb4f1787/items/607686/image?etag=1782161464448&width=500&height=500) How Relevance Models Foreshadowed Transformers for NLP ](https://towardsdatascience.com/how-relevance-models-foreshadowed-transformers-for-nlp/ \"How Relevance Models Foreshadowed Transformers for NLP\")\n  * [ ![](https://h030.towardsdatascience.com/rest/v2/contentStores/d7d3a4ab-7caf-4383-bbd2-fbf7fb4f1787/items/607639/image?etag=1781250125848&width=500&height=500) Javascript Fatigue: HTMX Is All You Need to Build ChatGPT — Part 1 ](https://towardsdatascience.com/javascript-fatigue-you-dont-need-js-to-build-chatgpt/ \"Javascript Fatigue: HTMX Is All You Need to Build ChatGPT — Part 1\")\n  * [ ![](https://h030.towardsdatascience.com/rest/v2/contentStores/d7d3a4ab-7caf-4383-bbd2-fbf7fb4f1787/items/607930/image?etag=1782124566054&width=500&height=500) The Subset Sum Problem Solved in Linear Time for Dense Enough Inputs ](https://towardsdatascience.com/subset-sum-problem-solved-in-linear-time-for-dense-enough-inputs/ \"The Subset Sum Problem Solved in Linear Time for Dense Enough Inputs\")\n  * [ ![](https://h030.towardsdatascience.com/rest/v2/contentStores/d7d3a4ab-7caf-4383-bbd2-fbf7fb4f1787/items/3660/image?etag=1780176246070&width=500&height=500) Does Your Company Have a Data Strategy? ](https://towardsdatascience.com/does-your-company-have-a-data-strategy-e8944dbec4ae/ \"Does Your Company Have a Data Strategy?\")\n\n\n",
  "body_base_path": "attachments/e6b19855f20f/pages/",
  "body_markdown_path": "attachments/e6b19855f20f/pages/archived-body.md",
  "local_html": "attachments/e6b19855f20f/pages/archived-body.html",
  "body_source": "prior_capture"
}