Model Registry docs

content/docs/sidebar.json

@@ -544,6 +544,20 @@
               "share-a-project"
             ]
           },
+          {
+            "slug": "model-registry",
+            "source": false,
+            "children": [
+              {
+                "label": "What is a Model Registry",
+                "slug": "what-is-a-model-registry"
+              },
+              "add-a-model",
+              "view-models",
+              "register-version",
+              "assign-stage"
+            ]
+          },
           "teams",
           "change-team-plan-and-size",
           "account-management",

content/docs/studio/get-started.md

@@ -8,6 +8,8 @@ Studio.
 
 https://www.youtube.com/watch?v=hKf4twg832g
 
+## Sign up / sign in
+
 1. Sign in to your [Iterative Studio](https://studio.iterative.ai/) dashboard
    using your GitHub.com, GitLab.com or Bitbucket.org account, or your email
    address.
@@ -29,20 +31,24 @@ https://www.youtube.com/watch?v=hKf4twg832g
 
    </admon>
 
-2. If this is the first time you are signing in to Iterative Studio, you will
+## Explore demo project
+
+1. If this is the first time you are signing in to Iterative Studio, you will
    see that there already exists a `Demo` project that connects to an example
-   DVC repository. Use this project to explore the features that Iterative
-   Studio has to offer.
+   repository. Use this project to explore the features that Iterative Studio
+   has to offer.
 
 ![](https://static.iterative.ai/img/studio/login_home_v3.png) _When you first
 login, an example project is already created for you to explore, and you can add
 more projects._
 
-3. Each project on this dashboard displays the metrics that Iterative Studio
+2. Each project on this dashboard displays the metrics that Iterative Studio
    identified in your Git repository. In the figure above, you can see that
    `avg_prec` and `roc_auc` metrics are displayed.
 
-4. To connect to a new repository, click on `Add a Project`. All the
+## Add your projects
+
+1. To connect to a new repository, click on `Add a Project`. All the
    organizations that you have access to will be listed.
 
     <admon type="info">
@@ -63,12 +69,12 @@ more projects._
 
     </admon>
 
-5. Open the organization whose repository you want to connect to. You can also
+2. Open the organization whose repository you want to connect to. You can also
    use the search bar to directly look for a repository.
 
    ![](https://static.iterative.ai/img/studio/select_repo_v3.png)
 
-6. Specify additional connection settings if required.
+3. Specify additional connection settings if required.
 
    ![](https://static.iterative.ai/img/studio/project_settings.png)
 
@@ -81,24 +87,52 @@ more projects._
 
 You should now see that a project has been added in your dashboard.
 
-7. If your project requires any of the additional settings, then remember to
+4. If your project requires any of the additional settings, then remember to
    configure them by opening the [project settings]. Otherwise, your project may
    not work as expected. To go to project settings, click on the
    ![](https://static.iterative.ai/img/studio/view_open_settings_icon.png) icon
    in the project. In the menu that opens up, click on `Settings`.
 
 ![](https://static.iterative.ai/img/studio/project_open_settings.png)
 
-8. You can also [create multiple projects from a single Git repository].
+5. You can also [create multiple projects from a single Git repository].
 
-9. After creating a project, you can use it to [explore the ML experiments] as
+## Explore and share your projects
+
+1. After creating a project, you can use it to [explore the ML experiments] as
    well as [submit new experiments] from Iterative Studio.
 
-10. You can also make the project public by [sharing it on the web].
+2. You can also make the project public by [sharing it on the web].
+
+## Use model registry
+
+1. Click on the `Models` tab to open the central [Models dashboard]. If any of
+   your projects contain [GTO-annotated models], they will be listed here
+   automatically.
+
+   [models dashboard]:
+     /doc/studio/user-guide/model-registry/view-models#models-dashboard
+   [gto-annotated models]: /doc/studio/user-guide/model-registry/add-a-model
+
+2. You can [add new models](/doc/studio/user-guide/model-registry/add-a-model)
+   to the model registry by clicking on the `Add a model` button and following
+   on-screen instructions.
+
+3. To
+   [view the details](/doc/studio/user-guide/model-registry/view-models#model-details-page)
+   of a model, click on the name of the model.
+
+4. You can
+   [register new versions](/doc/studio/user-guide/model-registry/register-version)
+   of the models and
+   [assign stages to them](/doc/studio/user-guide/model-registry/assign-stage)
+   (e.g., development, staging, production).
+
+## Create and manage teams
 
-11. You can [create a team] and invite collaborators. Each team will have its
-    own projects dashboard. To create teams with more than 2 team members, [sign
-    up for the **Team** plan or **Enterprise** plan].
+1. You can [create a team] and invite collaborators. Each team will have its own
+   projects dashboard. To create teams with more than 2 team members, [sign up
+   for the **Team** plan or **Enterprise** plan].
 
 [project settings]:
   /doc/studio/user-guide/projects-and-experiments/configure-a-project

content/docs/studio/how-it-works.md

@@ -1,19 +1,8 @@
 # How Iterative Studio Works
 
-[Iterative Studio](https://studio.iterative.ai/) works with the data, metrics
-and hyperparameters that you add to your ML project Git repositories. It works
-very closely with your Git ecosystem. Using DVC and Git, you will push all your
-ML experiments to your GitHub, GitLab or Bitbucket repositories as Git commits.
-
-When you connect to these repositories from Iterative Studio, the project's
-`dvc.yaml` is used to identify all the data, metrics and hyperparameters in your
-experiments. If you are not using DVC, you can also add the metrics and
-hyperparameters to your Git repositories manually.
-
-Iterative Studio then creates an interactive, tabular representation of all the
-identified values. You can also
-[run new experiments](/doc/studio/user-guide/run-experiments) from Iterative
-Studio using your regular CI/CD setup (e.g. GitHub Actions).
+[Iterative Studio](https://studio.iterative.ai/) works with the data, metrics,
+hyperparameters and model metadata that you add to your ML project Git
+repositories. It works very closely with your Git ecosystem.
 
 This video illustrates how Iterative Studio works closely with your Git
 ecosystem.
@@ -22,3 +11,56 @@ ecosystem.
 > Projects.
 
 https://www.youtube.com/watch?v=5xM5az78Lrg
+
+## How your project data is saved
+
+- Using [DVC] and [Git], you will push all your ML experiments to your GitHub,
+  GitLab or Bitbucket repositories as Git commits.
+- Using Iterative Studio, or using the command line interface (CLI) of [GTO],
+  and possibly [MLEM], you will push all your ML model details to the Git
+  repositories as Git commits and Git tags.
+
+## How Iterative Studio extracts your project data
+
+- When you connect to these Git repositories from Iterative Studio, the
+  project's `dvc.yaml` is used to identify all the data, metrics and
+  hyperparameters in your experiments.
+- If you are not using DVC, you can
+  [add the metrics and hyperparameters to your Git repositories manually](/doc/studio/user-guide/projects-and-experiments/configure-a-project#custom-metrics-and-parameters).
+- Details of your ML models, including versions and stage assignments, are
+  extracted from the Git commits and tags.
+
+## How Iterative Studio presents your project data
+
+- Iterative Studio creates an
+  [interactive, tabular representation](/doc/studio/user-guide/projects-and-experiments/explore-ml-experiments#components-of-a-project)
+  of all the identified values.
+- All the projects that you have created are presented in a central projects
+  dashboard.
+- All identified models are included in an interactive
+  [models dashboard](/doc/studio/user-guide/model-registry/view-models#models-dashboard)
+  and individual
+  [model details pages](/doc/studio/user-guide/model-registry/view-models#model-details-page).
+
+## How Iterative Studio saves updates to your ML projects
+
+- When you
+  [run new experiments](/doc/studio/user-guide/projects-and-experiments/run-experiments)
+  or
+  [add models to your model registry](/doc/studio/user-guide/model-registry/add-a-model)
+  in Iterative Studio, it creates Git commits and pull requests with the
+  changes.
+- You can set up your CI/CD actions (e.g. GitHub Actions) to run model training
+  upon the creation of Git commits, tags or pull requests. You can use [CML] in
+  your CI/CD actions for continuous machine learning.
+- When you
+  [register new versions](/doc/studio/user-guide/model-registry/register-version)
+  of your ML models or
+  [assign stages](/doc/studio/user-guide/model-registry/assign-stage) to them,
+  Iterative Studio creates annotated Git tags representing the actions.
+
+[dvc]: https://dvc.org/
+[cml]: https://cml.dev
+[mlem]: https://mlem.ai/
+[gto]: https://github.com/iterative/gto
+[git]: https://git-scm.com/

content/docs/studio/index.md

@@ -1,11 +1,11 @@
 # Iterative Studio
 
 [Iterative Studio](https://studio.iterative.ai/) is a web application that you
-can access online or even host on-prem. It works with the data, metrics, and
-hyperparameters that you add to your ML project repositories. Using the power of
-leading open-source tools DVC, [CML](https://cml.dev), and Git, it enables you
-to seamlessly manage data and models, run and track experiments, and visualize
-and share results.
+can access online or even host on-prem. Using the power of leading open-source
+tools [DVC](https://dvc.org/), [CML](https://cml.dev), [MLEM](https://mlem.ai/),
+[GTO](https://github.com/iterative/gto) and [Git](https://git-scm.com/), it
+enables you to seamlessly manage data and machine learning models, run and track
+experiments, and visualize and share results.
 
 <cards>
 
@@ -35,3 +35,7 @@ and share results.
 - Apply your existing software engineering stack for ML teams.
 - Automate your ML process by transitioning to a no-code interface for running
   experiments on cloud resources of your choice.
+- Build an ML model registry for transparency and discovery across all your
+  projects.
+- Because your model registry is built on top of Git, you can use GitOps for
+  model deployment.

content/docs/studio/troubleshooting.md

@@ -12,6 +12,8 @@ settings_ are now called _Project settings_; and so on.
 Here we provide help for some of the problems that you may encounter when using
 Iterative Studio.
 
+**Projects and experiments**
+
 - [Error: No data found to visualize](#error-no-data-found-to-visualize)
 - [Error: No DVC repo was found at the root](#error-no-dvc-repo-was-found-at-the-root)
 - [Error: Non-DVC sub-directory of a monorepo](#error-non-dvc-sub-directory-of-a-monorepo)
@@ -21,6 +23,16 @@ Iterative Studio.
 - [Project contains columns that I did not import](#project-contains-columns-that-i-did-not-mark-as-mandatory-to-import)
 - [Project does not contain some of my commits or branches](#project-does-not-contain-some-of-my-commits-or-branches)
 - [Error: Failed to push experiment to repository](#error-failed-to-push-experiment-to-repository)
+
+**Model registry**
+
+- [I cannot find my desired Git repository in the form to add a model](#i-cannot-find-my-desired-git-repository-in-the-form-to-add-a-model)
+- [Model registry does not display the models in my Git repositories](#model-registry-does-not-display-the-models-in-my-git-repositories)
+- [How can I remove models from my model registry](#how-can-i-remove-models-from-my-model-registry)
+- [How can I un-assign stages from model versions](#how-can-i-un-assign-stages-from-model-versions)
+
+**Billing and payment**
+
 - [Questions or problems with billing and payment](#questions-or-problems-with-billing-and-payment)
 
 ## Support
@@ -42,6 +54,10 @@ For more details, refer to the section on how to [prepare your Git repositories]
 for use with Iterative Studio. Instructions on how to specify custom files can
 be found [here][project-settings].
 
+Note that if you're connecting to a repository just to fetch models for the
+model registry, and you are not working with DVC repositories, you can ignore
+this error.
+
 [prepare your git repositories]:
   /doc/studio/user-guide/prepare-your-repositories
 [project-settings]:
@@ -66,6 +82,10 @@ To solve this, you can either:
 Instructions on how to specify the sub-directory or custom files can be found
 [here][project-settings].
 
+Note that if you're connecting to a repository just to fetch models for the
+model registry, and you are not working with DVC repositories, you can ignore
+this error.
+
 ## Error: Non-DVC sub-directory of a monorepo
 
 If you get this message when you try to add a project:
@@ -116,6 +136,11 @@ Refer to the [DVC documentation](https://dvc.org/doc) for help on making commits
 to a DVC repository. Instructions on how to specify custom files can be found
 [here][project-settings].
 
+Note that if you're connecting to a repository just to fetch models for the
+model registry, and your repository is not expected to contain experiment data,
+metrics or hyperparameters, your project will appear empty. This is ok - you
+will still be able to work with your models in the model registry.
+
 ## Project does not contain the columns that I want
 
 There are two possible reasons for this:
@@ -179,10 +204,10 @@ or branches you do not see in your project were manually hidden by you or
 someone else in your team.
 
 You can unhide commits and branches to display them. For details, refer to
-Display preferences -> [Hide commits]. However, if the missing commit/branch is
+[Display preferences -> Hide commits]. However, if the missing commit/branch is
 not in the hidden commits list, then please [raise a support request](#support).
 
-[hide commits]:
+[display preferences -> hide commits]:
   /doc/studio/user-guide/projects-and-experiments/explore-ml-experiments#hide-commits
 
 ## Error: Failed to push experiment to repository
@@ -206,9 +231,42 @@ please check:
 If you get this error and none of the above applies, please
 [get in touch with us](#support).
 
+## I cannot find my desired Git repository in the form to add a model
+
+Only repositories that you have connected to Iterative Studio are available in
+the `Add a model` form. To connect your desired repository to Iterative Studio,
+go to the `Projects` tab and
+[create a project that connects to this Git repository](/doc/studio/user-guide/projects-and-experiments/create-a-project).
+Then you can come back to the model registry and add the model.
+
+## Model registry does not display the models in my Git repositories
+
+For a model to be displayed in the model registry, it has to be registered using
+[GTO]. You can [register the model] from Iterative Studio or with GTO's CLI.
+
+## How can I remove models from my model registry
+
+To remove models from the model registry, you should remove the associated
+project from your projects dashboard.
+
+## How can I un-assign stages from model versions
+
+To un-assign any stage from a model, open the corresponding Git repository (in
+GitHub, GitLab or Bitbucket). Then, remove the Git tag pertaining to the stage
+that you want to un-assign.
+
+After removing the Git tag, reparse the repository. To do this, open the project
+in Iterative Studio and then click on `Force import` as shown below.
+
+![Showing and hiding columns](https://static.iterative.ai/img/studio/force_import.gif)
+
+Currently, you cannot un-assign stages from the Iterative Studio user interface.
+
 ## Questions or problems with billing and payment
 
-If your questions are not answered by
-[Billing and Payment ](/doc/studio/user-guide/billing-and-payment) or by the
-Iterative Studio [FAQ page](https://studio.iterative.ai/faq), please
-[contact us](#support).
+Check out the [Frequently Asked Questions](https://studio.iterative.ai/faq) to
+see if your questions have already been answered. If you still have problems,
+please [contact us](#support).
+
+[gto]: https://github.com/iterative/gto
+[register the model]: /doc/studio/user-guide/model-registry/add-a-model