From f16d911c04e65673bced7a0813c7e835229e83f8 Mon Sep 17 00:00:00 2001 From: tapadipti Date: Wed, 26 May 2021 22:27:06 +0530 Subject: [PATCH 01/17] #2468 Studio docs enhancements - initiated task --- content/docs/sidebar.json | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/content/docs/sidebar.json b/content/docs/sidebar.json index 254cc8c4b3..cb2ae49def 100644 --- a/content/docs/sidebar.json +++ b/content/docs/sidebar.json @@ -525,7 +525,10 @@ "label": "Create a View", "slug": "create-view" }, - "visualize-experiments", + { + "label": "Visualize and Compare", + "slug": "visualize-experiments" + }, "run-experiments", "teams" ] From f112e10cb693d3d88b8c0177785d6ffb7ed949cc Mon Sep 17 00:00:00 2001 From: tapadipti Date: Wed, 26 May 2021 23:57:32 +0530 Subject: [PATCH 02/17] Changed the structure of the "teams" page + Added more details in the "experiments" page --- content/docs/studio/run-experiments.md | 59 +++++++++++++++++++------- content/docs/studio/teams.md | 33 +++++++------- 2 files changed, 61 insertions(+), 31 deletions(-) diff --git a/content/docs/studio/run-experiments.md b/content/docs/studio/run-experiments.md index ee60c25a2e..431843242c 100644 --- a/content/docs/studio/run-experiments.md +++ b/content/docs/studio/run-experiments.md @@ -5,24 +5,53 @@ CI/CD set-up is used to run the experiments. For instance, [CML](https://dvc.org/doc/cml) Github Actions can be used to run ML experiments on each new commit. -> Note that you cannot run experiments on the demo view (`example-get-started`) -> that is provided to you by default. Once you create views for your ML project -> repositories, you can follow the instructions given below and run experiments -> directly from DVC Studio. +> Note that due to access restrictions, you cannot run experiments on the demo +> view (`example-get-started`) that is provided to you by default. Once you +> create views for your ML project repositories, you can follow the instructions +> given below and run experiments directly from DVC Studio. To run experiments from DVC Studio, select the commit that you want to use and -click the `Run` button. A form will let you to specify all the changes that you -want to make to your experiment input files/dirs and parameters. +click the `Run` button. A form will let you specify all the changes that you +want to make to your experiment. On this form, there are 2 types of inputs that +you can change: + +1. **Input data files**: You can change datasets as well as metrics and model + files. The list of files that you can change will depend on your ML project. + For instance, in the `example-get-started` project, you can change the + `data.xml` file. If you select the + `Show all input parameters (including hidden)` option, then you can also + change the hidden files such as the `model.pkl` model file and the + `scores.json` metrics file. You can also choose not to change any input data + files if you only wish to change the values of one or more hyperparameters. +2. **Hyperparameters**: You can change the values of the hyperparameters that + are defined in your ML project. For instance, in the `example-get-started` + project, you can change `max_features` (the maximum number of features that + the model uses), `ngrams`, etc. You can also choose not to change any + hyperparameters if you only wish to change one or more input data files. + +The default values of the input data files and hyperparameters in this form are +extracted from your selected commit. drawing -The default values in this form are extracted from your selected commit. For -instance, in the given example, `max_features` can be increased to 4000, -`ngrams` can be changed to 2, and so on. Enter your commit message and -description, select the branch to commit to (either the base branch or a new -branch), and click on `Commit changes`. +Once you have made all the required changes, enter your Git commit message and +description. Then, select the branch to commit to. You can commit to either the +base branch or a new branch. If you commit to a new branch, a Git pull request +will automatically be created from the new branch to the base branch. Now, click +on `Commit changes`. + +At this point, the new experiment appears in the view table. If you check in +your Git repository, you will find the Git commit that you just made. If you had +committed to a new branch, a new pull request will also have been created from +the new branch to the base branch. + +If your project is integrated with a CI/CD setup (eg, GitHub Actions), the CI/CD +setup will get invoked. If this setup includes a model training process, it will +be triggered, which means that your ML experiment will run automatically. The +model training can happen on any cloud or Kubernetes. For more details on how to +set up CI/CD pipelines for your ML project, refer to +[CML](https://dvc.org/doc/cml) . -Now, the CI/CD pipeline that you have set up (eg, Github Actions) will be -invoked to run the experiment. If you refresh the view (reload the page), the -experiment (i.e. the commit that you just pushed) along with its results -(metrics) will be included in the list of all commits. +Once the experiment completes, its metrics will be available in the view table. +You can then generate plots and trend charts for it, or compare it with the +other experiments. diff --git a/content/docs/studio/teams.md b/content/docs/studio/teams.md index 4b18b28af1..8fb5ecc5d0 100644 --- a/content/docs/studio/teams.md +++ b/content/docs/studio/teams.md @@ -6,21 +6,6 @@ define teams with one or more team members. The team members are also called collaborators, and you can assign different roles to them. The views that you create in your team's page will be accessible to all members of the team. -## Roles - -Team members can have different roles. - -- **View.** Users with the View access cannot edit team settings or create new - views within the team's page. They have read-only access to the views created - by other team members. -- **Edit.** In addition to accessing all the team's views, users with the Edit - role can also create new views for the team, edit the views' settings, and run - experiments. -- **Admin.** Admin users have full access to the team's views and settings. They - can do everything that viewers and editors can do. Additionally, they can add - (invite) and remove collaborators as well as change team settings such as - cloud credentials (data remotes). - ## Create a team To create a team, click on the drop down next to `Personal`. All the teams that @@ -46,11 +31,27 @@ invite will be sent to each invitee, and they will have to join using their Github, Gitlab or Bitbucket account. You can add multiple collaborators. Each collaborator can be assigned the Admin, -Edit, or View role. ![](https://static.iterative.ai/img/studio/team_roles.png) +Edit, or View role. Refer to the [Roles](#roles) section below for more details +about the roles. ![](https://static.iterative.ai/img/studio/team_roles.png) Once you have added the people that you wish to add to your team, click on `Send Invites and Close` on the top right corner. +## Roles + +Team members can have different roles. + +- **View.** Users with the View access cannot edit team settings or create new + views within the team's page. They have read-only access to the views created + by other team members. +- **Edit.** In addition to accessing all the team's views, users with the Edit + role can also create new views for the team, edit the views' settings, and run + experiments. +- **Admin.** Admin users have full access to the team's views and settings. They + can do everything that viewers and editors can do. Additionally, they can add + (invite) and remove collaborators as well as change team settings such as + cloud credentials (data remotes). + ## Manage your team and its views Once you have created the team, the team's page opens up. From 6bae5f270c3e1970977296e49cf818ce3c39b2d1 Mon Sep 17 00:00:00 2001 From: tapadipti Date: Fri, 28 May 2021 11:14:31 +0530 Subject: [PATCH 03/17] Updated images + Some minor style changes --- content/docs/studio/create-view.md | 18 ++++++++---------- content/docs/studio/get-started.md | 4 ++-- content/docs/studio/index.md | 14 +++++++------- content/docs/studio/run-experiments.md | 4 +++- content/docs/studio/teams.md | 10 +++++----- content/docs/studio/visualize-experiments.md | 10 ++++------ 6 files changed, 29 insertions(+), 31 deletions(-) diff --git a/content/docs/studio/create-view.md b/content/docs/studio/create-view.md index 93de45d2cc..919ce0347c 100644 --- a/content/docs/studio/create-view.md +++ b/content/docs/studio/create-view.md @@ -3,37 +3,35 @@ To create a new view for your repository, follow these steps. 1. Sign into your [DVC Studio](https://studio.iterative.ai/) dashboard. - ![](https://static.iterative.ai/img/studio/login_home.png) + ![](https://static.iterative.ai/img/studio/login_home_v2.png) 2. Click on `Add a View`. All the organizations that you have access to will be - listed. ![](https://static.iterative.ai/img/studio/create_view.png) + listed. 3. Open the organization whose repository you want to connect to. You can also use the search bar to search directly for the repo which you want to connect - to. ![](https://static.iterative.ai/img/studio/select_repo.png) + to. ![](https://static.iterative.ai/img/studio/select_repo_v2.png) 4. Select the Git repository for which you want to create a view. For this, hover over the required repository and click `Connect`. 5. Once you click on `Connect`, you will be able to specify advanced settings for the connection. - ![](https://static.iterative.ai/img/studio/view_settings.png) + ![](https://static.iterative.ai/img/studio/view_settings_v2.png) > In most scenarios, you can skip the advanced settings. Refer to the > [Advanced Settings](#advanced-settings) section below for more details. You should now see that a view has been added in your dashboard. -![](https://static.iterative.ai/img/studio/view_added.png) - If your project requires any of the advanced settings, then remember to configure them by opening the View settings. Otherwise, your view may not work as expected. To go to View settings, click on the -![](https://static.iterative.ai/img/studio/view_open_settings_icon.png) icon in -the view. In the menu that opens up (highlighted in yellow in the screenshot +![](https://static.iterative.ai/img/studio/view_open_settings_icon_v2.png) icon +in the view. In the menu that opens up (highlighted in yellow in the screenshot below), click on `Settings`. -![](https://static.iterative.ai/img/studio/view_open_settings.png) +![](https://static.iterative.ai/img/studio/view_open_settings_v2.png) ## Advanced Settings @@ -61,7 +59,7 @@ You will need to specify advanced settings in the following scenarios. You can share your views on the web. Simply click on the button labelled `Private` next to the name of the view. In the menu that pops up, turn on -`Share to Web`. ![](https://static.iterative.ai/img/studio/view_share.png) +`Share to Web`. ![](https://static.iterative.ai/img/studio/view_share_v2.png) You can change a shared (public) view and make it private by turning off `Share to web`. diff --git a/content/docs/studio/get-started.md b/content/docs/studio/get-started.md index 3057f3ee6f..5bf10c1f7e 100644 --- a/content/docs/studio/get-started.md +++ b/content/docs/studio/get-started.md @@ -19,7 +19,7 @@ experiments. You can also generate plots and compare experiments here. In your browser, open . Sign in with your Github, Gitlab, or Bitbucket account. -![](https://static.iterative.ai/img/studio/login_home.png) _When you first +![](https://static.iterative.ai/img/studio/login_home_v2.png) _When you first login, an example view is already created for you to explore, and you can add more views._ @@ -40,7 +40,7 @@ A table will be generated as shown below. This includes metrics, hyperparameters and information about the datasets. All the data is flattened and neatly presented for you to evaluate and compare the experiments. -![](https://static.iterative.ai/img/studio/view_components.png) +![](https://static.iterative.ai/img/studio/view_components_v2.png) This tabular display has the following components: diff --git a/content/docs/studio/index.md b/content/docs/studio/index.md index b3e63b51c4..555b88276a 100644 --- a/content/docs/studio/index.md +++ b/content/docs/studio/index.md @@ -9,17 +9,17 @@ share results. Project website: https://studio.iterative.ai -![](https://static.iterative.ai/img/studio/main.png) _DVC Studio experiments +![](https://static.iterative.ai/img/studio/main_v2.png) _DVC Studio experiments dashboard_ Use DVC, CML and Studio to: -- **Visualize, collaborate, and do everything that a regular ML tracking tool - does.** +- Visualize, collaborate, and do everything that a regular ML tracking tool + does. -- **Keep your code, data and model connected at all times.** +- Keep your code, data and model connected at all times. -- **Use the power of Git to track and preserve all your experiments.** +- Use the power of Git to track and preserve all your experiments. -- **Run [CI/CD](https://en.wikipedia.org/wiki/CI/CD) for your ML projects on - cloud resources of your choice without any new tools.** +- Run [CI/CD](https://en.wikipedia.org/wiki/CI/CD) for your ML projects on cloud + resources of your choice without any new tools. diff --git a/content/docs/studio/run-experiments.md b/content/docs/studio/run-experiments.md index 431843242c..a1b76ec61d 100644 --- a/content/docs/studio/run-experiments.md +++ b/content/docs/studio/run-experiments.md @@ -32,7 +32,7 @@ you can change: The default values of the input data files and hyperparameters in this form are extracted from your selected commit. -drawing +![](https://static.iterative.ai/img/studio/cml_changes_v2.png) Once you have made all the required changes, enter your Git commit message and description. Then, select the branch to commit to. You can commit to either the @@ -40,6 +40,8 @@ base branch or a new branch. If you commit to a new branch, a Git pull request will automatically be created from the new branch to the base branch. Now, click on `Commit changes`. +![](https://static.iterative.ai/img/studio/cml_commit_v2.png) + At this point, the new experiment appears in the view table. If you check in your Git repository, you will find the Git commit that you just made. If you had committed to a new branch, a new pull request will also have been created from diff --git a/content/docs/studio/teams.md b/content/docs/studio/teams.md index 8fb5ecc5d0..a36af41bdd 100644 --- a/content/docs/studio/teams.md +++ b/content/docs/studio/teams.md @@ -11,11 +11,11 @@ create in your team's page will be accessible to all members of the team. To create a team, click on the drop down next to `Personal`. All the teams that you have created so far will be listed within "Teams" in the drop down menu. If you have not created any team so far, this list will be empty. Now, click on -`Create a team`. ![](https://static.iterative.ai/img/studio/team_create.png) +`Create a team`. ![](https://static.iterative.ai/img/studio/team_create_v2.png) You will be asked to enter the URL namespace for your team. Enter a unique name. The URL for your team will be formed using this name. -![](https://static.iterative.ai/img/studio/team_enter_name.png) +![](https://static.iterative.ai/img/studio/team_enter_name_v2.png) Then, click the `Create team` button on the top right corner. @@ -32,7 +32,7 @@ Github, Gitlab or Bitbucket account. You can add multiple collaborators. Each collaborator can be assigned the Admin, Edit, or View role. Refer to the [Roles](#roles) section below for more details -about the roles. ![](https://static.iterative.ai/img/studio/team_roles.png) +about the roles. ![](https://static.iterative.ai/img/studio/team_roles_v2.png) Once you have added the people that you wish to add to your team, click on `Send Invites and Close` on the top right corner. @@ -56,7 +56,7 @@ Team members can have different roles. Once you have created the team, the team's page opens up. -![](https://static.iterative.ai/img/studio/team_page.png) +![](https://static.iterative.ai/img/studio/team_page_v2.png) On this page, you can perform three types of tasks: @@ -70,4 +70,4 @@ On this page, you can perform three types of tasks: - **Change settings.** Finally, you can click on the `Settings` menu item to change the team name, add credentials for the data remotes, and delete the - team. ![](https://static.iterative.ai/img/studio/team_settings.png) + team. ![](https://static.iterative.ai/img/studio/team_settings_v2.png) diff --git a/content/docs/studio/visualize-experiments.md b/content/docs/studio/visualize-experiments.md index b3f2dc87bd..cb978af573 100644 --- a/content/docs/studio/visualize-experiments.md +++ b/content/docs/studio/visualize-experiments.md @@ -15,21 +15,19 @@ are generated. To generate metric plots, select one or more experiments (represented by the commits), and click on the 'Show plots' button. -drawing - The plots will appear below the tabular display. If you have selected more than one experiment, you can use the plots to compare them. -drawing +![](https://static.iterative.ai/img/studio/plots_v2.png) ## Generate trend charts -Click on the Trends button to generate a plot of how the metrics changed over +Click on the `Trends` button to generate a plot of how the metrics changed over the course of the different experiments. For each metric, the trend charts show how the metric changed from one commit to another. You can include one or more branches in the trend chart. -drawing +![](https://static.iterative.ai/img/studio/trends_v2.png) ## Compare experiments @@ -37,4 +35,4 @@ To compare different experiments, select two experiments (represented by the commits), and click on the `Compare` button. The metrics, parameters and files in the selected experiments will be displayed side by side for easy comparison. -drawing +![](https://static.iterative.ai/img/studio/compare_v2.png) From b7915fe0aae272701bf7d60fb4dba4e6494dc1d3 Mon Sep 17 00:00:00 2001 From: tapadipti Date: Fri, 28 May 2021 12:40:32 +0530 Subject: [PATCH 04/17] Text changes as per PR #2455 review comments --- content/docs/studio/create-view.md | 10 ++++------ content/docs/studio/index.md | 3 ++- 2 files changed, 6 insertions(+), 7 deletions(-) diff --git a/content/docs/studio/create-view.md b/content/docs/studio/create-view.md index 919ce0347c..ad5fd5a19b 100644 --- a/content/docs/studio/create-view.md +++ b/content/docs/studio/create-view.md @@ -9,13 +9,11 @@ To create a new view for your repository, follow these steps. listed. 3. Open the organization whose repository you want to connect to. You can also - use the search bar to search directly for the repo which you want to connect - to. ![](https://static.iterative.ai/img/studio/select_repo_v2.png) + use the search bar to search directly for the repository. Click on the + required repository to connect to it. + ![](https://static.iterative.ai/img/studio/select_repo_v2.png) -4. Select the Git repository for which you want to create a view. For this, - hover over the required repository and click `Connect`. - -5. Once you click on `Connect`, you will be able to specify advanced settings +4. Once you click on `Connect`, you will be able to specify advanced settings for the connection. ![](https://static.iterative.ai/img/studio/view_settings_v2.png) diff --git a/content/docs/studio/index.md b/content/docs/studio/index.md index 555b88276a..05d28d5c13 100644 --- a/content/docs/studio/index.md +++ b/content/docs/studio/index.md @@ -1,7 +1,8 @@ # DVC Studio [`DVC Studio`](https://studio.iterative.ai/) is a comprehensive and interactive -collaboration tool for your Machine Learning projects. +web application to manage and collaborate on your Machine Learning projects +visually. Using the power of leading open-source tools DVC, CML and Git, it enables you to seamlessly manage data and models, run and track experiments, and visualize and From 3dace5f3a0628288b67257eca00da184410f8b01 Mon Sep 17 00:00:00 2001 From: tapadipti Date: Fri, 28 May 2021 15:46:07 +0530 Subject: [PATCH 05/17] Added section on GH app installation + other small changes --- content/docs/studio/create-view.md | 49 +++++++++++++++++++++++------- content/docs/studio/get-started.md | 10 +++--- content/docs/studio/teams.md | 2 +- 3 files changed, 44 insertions(+), 17 deletions(-) diff --git a/content/docs/studio/create-view.md b/content/docs/studio/create-view.md index ad5fd5a19b..63a05f3c3f 100644 --- a/content/docs/studio/create-view.md +++ b/content/docs/studio/create-view.md @@ -1,20 +1,27 @@ # Create a View -To create a new view for your repository, follow these steps. +You can create views from your Git repositories, just like the `Demo` view you +saw in the last section. To create a new view, follow these steps. -1. Sign into your [DVC Studio](https://studio.iterative.ai/) dashboard. - ![](https://static.iterative.ai/img/studio/login_home_v2.png) +1. Sign into your [DVC Studio](https://studio.iterative.ai/) dashboard using + GitHub, GitLab or Bitbucket. 2. Click on `Add a View`. All the organizations that you have access to will be listed. +> To create views from your GitHub repositories, you must install the DVC Studio +> GitHub app. Refer to the section on +> [GitHub app installation](#dvc-studio-github-app-installation) for more +> details. + 3. Open the organization whose repository you want to connect to. You can also use the search bar to search directly for the repository. Click on the required repository to connect to it. + ![](https://static.iterative.ai/img/studio/select_repo_v2.png) -4. Once you click on `Connect`, you will be able to specify advanced settings - for the connection. +4. You will be able to specify advanced settings for the connection. + ![](https://static.iterative.ai/img/studio/view_settings_v2.png) > In most scenarios, you can skip the advanced settings. Refer to the @@ -31,6 +38,17 @@ below), click on `Settings`. ![](https://static.iterative.ai/img/studio/view_open_settings_v2.png) +## Share a view + +You can share your views on the web. Simply click on the button labelled +`Private` next to the name of the view. In the menu that pops up, turn on +`Share to Web`. + +![](https://static.iterative.ai/img/studio/view_share_v2.png) + +You can change a shared (public) view and make it private by turning off +`Share to web`. + ## Advanced Settings You will need to specify advanced settings in the following scenarios. @@ -53,11 +71,20 @@ You will need to specify advanced settings in the following scenarios. Enter the full file path, and specify whether the file is for `Metrics` or `Parameters`. -## Share a view +## DVC Studio GitHub app installation -You can share your views on the web. Simply click on the button labelled -`Private` next to the name of the view. In the menu that pops up, turn on -`Share to Web`. ![](https://static.iterative.ai/img/studio/view_share_v2.png) +If you are using a GitHub account, you will have to install the DVC Studio +GitHub app on your GitHub repositories / organizations that you want to use with +DVC Studio. When you try to create a view, if the app is not already installed, +DVC Studio will not be able to fetch the list of your GitHub repositories. In +this case, you will be prompted to configure Git integrations as shown below. -You can change a shared (public) view and make it private by turning off -`Share to web`. +![](https://static.iterative.ai/img/studio/configure_git_integrations.png) + +Click on the link to `Configure Git integrations settings`. This will open the +`Git integrations` section of your profile page. + +![](https://static.iterative.ai/img/studio/configure_github.png) + +Click on the `Configure` button, and you will be redirected to the GitHub page +to install and authorize the DVC Studio GitHub app. diff --git a/content/docs/studio/get-started.md b/content/docs/studio/get-started.md index 5bf10c1f7e..6dd379b776 100644 --- a/content/docs/studio/get-started.md +++ b/content/docs/studio/get-started.md @@ -1,7 +1,7 @@ # Get Started with DVC Studio Here, we will walk you through a tutorial to use DVC Studio for collaboration on -your ML projects. You will need access to a Github, Gitlab or Bitbucket account +your ML projects. You will need access to a GitHub, GitLab or Bitbucket account which has access to the Git repositories you want to connect. DVC Studio creates views from repositories when you connect to them. @@ -17,15 +17,15 @@ experiments. You can also generate plots and compare experiments here. ## DVC Studio Views page In your browser, open . Sign in with your Github, -Gitlab, or Bitbucket account. +GitLab, or Bitbucket account. ![](https://static.iterative.ai/img/studio/login_home_v2.png) _When you first login, an example view is already created for you to explore, and you can add more views._ -When you first login, you will find that there already exists a view connecting -to an example DVC project. Use this view to explore the features that DVC Studio -has to offer. +When you first login, you will find that there already exists a `Demo` view +connecting to an example DVC project. Use this view to explore the features that +DVC Studio has to offer. DVC Studio automatically identifies datasets, metrics and hyperparameters in your ML experiments. Each view on the dashboard displays the metrics. In the diff --git a/content/docs/studio/teams.md b/content/docs/studio/teams.md index a36af41bdd..9d711c505b 100644 --- a/content/docs/studio/teams.md +++ b/content/docs/studio/teams.md @@ -28,7 +28,7 @@ collaborators by accessing team settings later. If you wish to add collaborators now, enter their email addresses. An email invite will be sent to each invitee, and they will have to join using their -Github, Gitlab or Bitbucket account. +GitHub, GitLab or Bitbucket account. You can add multiple collaborators. Each collaborator can be assigned the Admin, Edit, or View role. Refer to the [Roles](#roles) section below for more details From 5718e3b91debbc5aaee72fecc4f96a0eabdb5926 Mon Sep 17 00:00:00 2001 From: tapadipti Date: Fri, 28 May 2021 17:21:47 +0530 Subject: [PATCH 06/17] Added clarification in the GitHub app section --- content/docs/studio/create-view.md | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/content/docs/studio/create-view.md b/content/docs/studio/create-view.md index 63a05f3c3f..f57fb28c0f 100644 --- a/content/docs/studio/create-view.md +++ b/content/docs/studio/create-view.md @@ -88,3 +88,7 @@ Click on the link to `Configure Git integrations settings`. This will open the Click on the `Configure` button, and you will be redirected to the GitHub page to install and authorize the DVC Studio GitHub app. + +> Note that you do not have to authorize DVC Studio on all the repositories in +> your GitHub organization. You can limit access to only those repositories +> which you want to use with DVC Studio. From 7c7c4a5c0b1f23fe67b2f5401ed27c41da739276 Mon Sep 17 00:00:00 2001 From: tapadipti Date: Fri, 28 May 2021 17:32:04 +0530 Subject: [PATCH 07/17] Minor correction --- content/docs/studio/create-view.md | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/content/docs/studio/create-view.md b/content/docs/studio/create-view.md index f57fb28c0f..f08979ee83 100644 --- a/content/docs/studio/create-view.md +++ b/content/docs/studio/create-view.md @@ -33,8 +33,7 @@ If your project requires any of the advanced settings, then remember to configure them by opening the View settings. Otherwise, your view may not work as expected. To go to View settings, click on the ![](https://static.iterative.ai/img/studio/view_open_settings_icon_v2.png) icon -in the view. In the menu that opens up (highlighted in yellow in the screenshot -below), click on `Settings`. +in the view. In the menu that opens up, click on `Settings`. ![](https://static.iterative.ai/img/studio/view_open_settings_v2.png) From ccd23274106387646fbe494ac518cf449965f0bb Mon Sep 17 00:00:00 2001 From: tapadipti Date: Tue, 1 Jun 2021 16:56:43 +0530 Subject: [PATCH 08/17] Added a separate page for view settings --- content/docs/sidebar.json | 1 + content/docs/studio/create-view.md | 28 +-------- content/docs/studio/get-started.md | 7 ++- content/docs/studio/teams.md | 4 +- content/docs/studio/view-settings.md | 92 ++++++++++++++++++++++++++++ 5 files changed, 103 insertions(+), 29 deletions(-) create mode 100644 content/docs/studio/view-settings.md diff --git a/content/docs/sidebar.json b/content/docs/sidebar.json index cb2ae49def..c41c8831ac 100644 --- a/content/docs/sidebar.json +++ b/content/docs/sidebar.json @@ -525,6 +525,7 @@ "label": "Create a View", "slug": "create-view" }, + "view-settings", { "label": "Visualize and Compare", "slug": "visualize-experiments" diff --git a/content/docs/studio/create-view.md b/content/docs/studio/create-view.md index f08979ee83..62a1beb159 100644 --- a/content/docs/studio/create-view.md +++ b/content/docs/studio/create-view.md @@ -20,12 +20,12 @@ saw in the last section. To create a new view, follow these steps. ![](https://static.iterative.ai/img/studio/select_repo_v2.png) -4. You will be able to specify advanced settings for the connection. +4. You will be able to specify additional settings for the connection. ![](https://static.iterative.ai/img/studio/view_settings_v2.png) -> In most scenarios, you can skip the advanced settings. Refer to the -> [Advanced Settings](#advanced-settings) section below for more details. +> In most scenarios, you can skip these additional View settings. Refer to the +> [View Settings](/doc/studio/view-settings) section for more details. You should now see that a view has been added in your dashboard. @@ -48,28 +48,6 @@ You can share your views on the web. Simply click on the button labelled You can change a shared (public) view and make it private by turning off `Share to web`. -## Advanced Settings - -You will need to specify advanced settings in the following scenarios. - -> These settings can be skipped when creating the view -- they can be edited -> later. - -- **Monorepo:** If you have connected to a - [monorepo](https://en.wikipedia.org/wiki/Monorepo), then specify the full path - to the root directory of the project that you want to connect. - -- **Data remotes:** The metrics and parameters that you want to include in the - view may be present in a data remote (cloud storage or another - location outside of the Git repo). If you want to include such data in your - views, then you will have to grant DVC Studio access to the data remote. - -- **Custom metrics and parameters:** DVC Studio automatically detects metrics, - plots, and parameters files specified in the project's `dvc.yaml`. If you want - to connect custom files, you can add them by clicking the `Add file` button. - Enter the full file path, and specify whether the file is for `Metrics` or - `Parameters`. - ## DVC Studio GitHub app installation If you are using a GitHub account, you will have to install the DVC Studio diff --git a/content/docs/studio/get-started.md b/content/docs/studio/get-started.md index 6dd379b776..b0f7ac28dc 100644 --- a/content/docs/studio/get-started.md +++ b/content/docs/studio/get-started.md @@ -10,9 +10,10 @@ views from repositories when you connect to them. A _view_ is an interactive representation of the experiments run in your ML project. DVC Studio automatically identifies datasets, metrics and hyperparameters in your ML experiments. You can also add custom files with the -required data (more on this later). Using this data, DVC Studio creates a -summary view. This includes tables of all the metrics across all your -experiments. You can also generate plots and compare experiments here. +required data (more on this [here](/doc/studio/view-settings)). Using this data, +DVC Studio creates a summary view. This includes tables of all the metrics +across all your experiments. You can also generate plots and compare experiments +here. ## DVC Studio Views page diff --git a/content/docs/studio/teams.md b/content/docs/studio/teams.md index 9d711c505b..c54caa2a6c 100644 --- a/content/docs/studio/teams.md +++ b/content/docs/studio/teams.md @@ -70,4 +70,6 @@ On this page, you can perform three types of tasks: - **Change settings.** Finally, you can click on the `Settings` menu item to change the team name, add credentials for the data remotes, and delete the - team. ![](https://static.iterative.ai/img/studio/team_settings_v2.png) + team. + + ![](https://static.iterative.ai/img/studio/team_settings_v2.png) diff --git a/content/docs/studio/view-settings.md b/content/docs/studio/view-settings.md new file mode 100644 index 0000000000..642d9643aa --- /dev/null +++ b/content/docs/studio/view-settings.md @@ -0,0 +1,92 @@ +# Additional Settings for a View + +If you are creating a view for a DVC repo, and if the DVC repo is at the root of +the Git repository and does not reference remote/cloud storage, then you can +successfully visualize it without specifying additional settings. + +In other cases, you could be trying to create views from non-DVC repositories. +Or, your project may be in a sub-directory inside a monorepo. Or, the data to be +visualized could be in custom files in your repository or even in remote / cloud +storages. In all these scenarios, you will need to specify advanced settings so +that DVC Studio is able to access the data required for visualization. These +scenarios have been explained in details below. + +## Non-DVC repositories + +DVC Studio creates views by identifying datasets, metrics and hyperparameters +defined in your Git repositories. These details (datasets, metrics and +hyperparameters) are stored in your Git repositories as CSV, JSON or YAML files. +You can add these details to your Git repositories in two ways: + +1. **Set up DVC repositories**: You can use [DVC](https://dvc.org/) and Git to + version your code, data and models all within your Git repositories. By using + DVC, you can be sure not to bloat your repositories with large volumes of + data or huge models. These large assets reside in cloud or other remote + storage locations. You will simply track their version info in Git. + + Refer to the [DVC documentation](https://dvc.org/doc) to initialize a DVC + repository. You can then connect to this DVC repository and create a view as + described [earlier](/doc/studio/create-view). DVC Studio automatically + detects metrics, plots, and hyperparameters files specified in the project's + `dvc.yaml`. And every time you push a new commit to this DVC repository, your + view will reflect the new changes. + +2. **Specify custom files with your metrics and parameters**: If you are working + with a non-DVC repository, you can still create views for it if your + repository contains CSV, JSON or YAML files with the metrics and + hyperparameters that you want to visualize. To enable DVC Studio to visualize + such custom data, simply + [specify the custom files](#specifying-view-settings) with your metrics and + parameters, and DVC Studio will efficiently generate tables and plots for + your custom input. For instance, if you have an ML project for which you + generate and save metrics either manually or using some ML tracking tools, + then you can create a view for this project by specifying the file (within + your Git repo) which contains your saved metrics. + +## Monorepo + +Depending on how you have set up your Git repositories, your DVC repo (for which +you are trying to create the view) may not be in the root of your Git repo. +Instead, it could be in a sub-directory of a +[monorepo](https://en.wikipedia.org/wiki/Monorepo). If this is the case, you +will need to specify the full path to the sub-directory that you want to use +with your view. + +## Data remotes (cloud / remote storage) + +The metrics and parameters that you want to include in the view may also be +present in a data remote(cloud storage or another location outside of the Git +repo). If you want to include such data in your views, then you will have to +grant DVC Studio access to the data remote. + +# Specifying View settings + +For any of the scenarios defined above, specify the additional settings as +described below. You can access these settings at any time after creating the +view. For this, click on the +![](https://static.iterative.ai/img/studio/view_open_settings_icon_v2.png) icon +in the view. In the menu that opens up, click on `Settings`. + +- **Custom metrics and parameters:** If you want to connect custom files, you + can add them by clicking the `Add file` button. Enter the full file path, and + specify whether the file is for `Metrics` or `Parameters`. + +- **Monorepo:** If you have connected to a + [monorepo](https://en.wikipedia.org/wiki/Monorepo), then specify the full path + to the sub-directory that contains the DVC repo for which you want to create + the view. + +- **Data remotes:** If you need to set up DVC data remotes for your view, you + will need to do it after your view has been created. So, first create your + view without specifying the data remotes. Once your view is created, open the + View settings. Open the `Data remotes / cloud storage credentials` section. + The data remotes that are used in your DVC repo will be listed. Now, click on + `Add new credentials`. In the form that opens up, select the provider (Amazon + S3, GCP, etc.). depending on the provider, you will be asked for more details + such as the credentials name, username, password etc. + + ![](https://static.iterative.ai/img/studio/s3_remote_settings.png) + + For details on what permissions are required, refer to the DVC documentation + on + [supported storage types](/doc/command-reference/remote/add#supported-storage-types). From 40c621985167915c0b82690e6103c67ceb40747a Mon Sep 17 00:00:00 2001 From: tapadipti Date: Tue, 1 Jun 2021 17:11:19 +0530 Subject: [PATCH 09/17] Clarified that DVC or non-DVC repos can be used with Studio --- content/docs/studio/get-started.md | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/content/docs/studio/get-started.md b/content/docs/studio/get-started.md index b0f7ac28dc..2ae6a9387d 100644 --- a/content/docs/studio/get-started.md +++ b/content/docs/studio/get-started.md @@ -9,9 +9,11 @@ views from repositories when you connect to them. A _view_ is an interactive representation of the experiments run in your ML project. DVC Studio automatically identifies datasets, metrics and -hyperparameters in your ML experiments. You can also add custom files with the -required data (more on this [here](/doc/studio/view-settings)). Using this data, -DVC Studio creates a summary view. This includes tables of all the metrics +hyperparameters in your ML experiments. The data can either be in DVC +repositories or you can add custom files with the required data. Refer to +[View settings](/doc/studio/view-settings) to understand the different ways in +which you can prepare your Git repository for use with DVC Studio. Using this +data, DVC Studio creates a summary view. This includes tables of all the metrics across all your experiments. You can also generate plots and compare experiments here. From 079b7f44db3c27e2a132f63f2b808c802c6e688a Mon Sep 17 00:00:00 2001 From: tapadipti Date: Tue, 1 Jun 2021 18:02:54 +0530 Subject: [PATCH 10/17] Added links to Studio in the DVC docs and banner --- content/docs/start/index.md | 5 +++++ src/components/LayoutHeader/alert.tsx | 4 ++-- 2 files changed, 7 insertions(+), 2 deletions(-) diff --git a/content/docs/start/index.md b/content/docs/start/index.md index ecb46dceb8..9f1951c318 100644 --- a/content/docs/start/index.md +++ b/content/docs/start/index.md @@ -76,3 +76,8 @@ by one in the next few pages: comparison across many ML experiments. Track your experiments with automatic versioning and checkpoint logging. Compare differences in parameters, metrics, code, and data. Apply, drop, roll back, resume, or share any experiment. + +**New!** Once you set up your DVC repository, you can also interact with it +using DVC Studio, the online UI for DVC. +[Here's a demo](https://studio.iterative.ai/team/Iterative/views/example-get-started-zde16i6c4g) +of how that looks like! diff --git a/src/components/LayoutHeader/alert.tsx b/src/components/LayoutHeader/alert.tsx index 69ac432ee6..8649405d0b 100644 --- a/src/components/LayoutHeader/alert.tsx +++ b/src/components/LayoutHeader/alert.tsx @@ -10,8 +10,8 @@ const LayoutAlert: React.FC<{ collapsed: boolean }> = ({ collapsed }) => ( 🛡 {' '} - DVC for Enterprise - data access control & - compliance!{' '} + DVC Studio, the online UI + for DVC, is live!{' '} ) From 3f4aa1e641df728d36e50032f42c7661714a0d56 Mon Sep 17 00:00:00 2001 From: tapadipti Date: Tue, 1 Jun 2021 18:46:34 +0530 Subject: [PATCH 11/17] Added Studio icon in the banner --- src/components/LayoutHeader/alert.tsx | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/src/components/LayoutHeader/alert.tsx b/src/components/LayoutHeader/alert.tsx index 8649405d0b..57e3e2bdab 100644 --- a/src/components/LayoutHeader/alert.tsx +++ b/src/components/LayoutHeader/alert.tsx @@ -7,8 +7,8 @@ import styles from './styles.module.css' const LayoutAlert: React.FC<{ collapsed: boolean }> = ({ collapsed }) => (
- - 🛡 + + {' '} DVC Studio, the online UI for DVC, is live!{' '} From 7d247893a81c8b9b13de17c7f4bc8740a403f83d Mon Sep 17 00:00:00 2001 From: tapadipti <32855442+tapadipti@users.noreply.github.com> Date: Tue, 1 Jun 2021 23:31:48 +0530 Subject: [PATCH 12/17] Apply suggestions from code review Co-authored-by: Casper da Costa-Luis --- content/docs/studio/create-view.md | 15 ++++---- content/docs/studio/run-experiments.md | 14 ++++---- content/docs/studio/view-settings.md | 47 +++++++++++++------------- 3 files changed, 37 insertions(+), 39 deletions(-) diff --git a/content/docs/studio/create-view.md b/content/docs/studio/create-view.md index 62a1beb159..1c9c5349d2 100644 --- a/content/docs/studio/create-view.md +++ b/content/docs/studio/create-view.md @@ -3,7 +3,7 @@ You can create views from your Git repositories, just like the `Demo` view you saw in the last section. To create a new view, follow these steps. -1. Sign into your [DVC Studio](https://studio.iterative.ai/) dashboard using +1. Sign in to your [DVC Studio](https://studio.iterative.ai/) dashboard using GitHub, GitLab or Bitbucket. 2. Click on `Add a View`. All the organizations that you have access to will be @@ -15,19 +15,18 @@ saw in the last section. To create a new view, follow these steps. > details. 3. Open the organization whose repository you want to connect to. You can also - use the search bar to search directly for the repository. Click on the - required repository to connect to it. + use the search bar to directly look for a repository. ![](https://static.iterative.ai/img/studio/select_repo_v2.png) -4. You will be able to specify additional settings for the connection. +4. Specify additional connection settings if required. ![](https://static.iterative.ai/img/studio/view_settings_v2.png) > In most scenarios, you can skip these additional View settings. Refer to the > [View Settings](/doc/studio/view-settings) section for more details. -You should now see that a view has been added in your dashboard. +You should now see that a View has been added in your dashboard. If your project requires any of the advanced settings, then remember to configure them by opening the View settings. Otherwise, your view may not work @@ -51,7 +50,7 @@ You can change a shared (public) view and make it private by turning off ## DVC Studio GitHub app installation If you are using a GitHub account, you will have to install the DVC Studio -GitHub app on your GitHub repositories / organizations that you want to use with +GitHub app on the GitHub repositories/organizations that you want to use with DVC Studio. When you try to create a view, if the app is not already installed, DVC Studio will not be able to fetch the list of your GitHub repositories. In this case, you will be prompted to configure Git integrations as shown below. @@ -67,5 +66,5 @@ Click on the `Configure` button, and you will be redirected to the GitHub page to install and authorize the DVC Studio GitHub app. > Note that you do not have to authorize DVC Studio on all the repositories in -> your GitHub organization. You can limit access to only those repositories -> which you want to use with DVC Studio. +> your GitHub organization. You can limit access to only repositories that +> you want to use with DVC Studio. diff --git a/content/docs/studio/run-experiments.md b/content/docs/studio/run-experiments.md index a1b76ec61d..f0d5261757 100644 --- a/content/docs/studio/run-experiments.md +++ b/content/docs/studio/run-experiments.md @@ -8,7 +8,7 @@ on each new commit. > Note that due to access restrictions, you cannot run experiments on the demo > view (`example-get-started`) that is provided to you by default. Once you > create views for your ML project repositories, you can follow the instructions -> given below and run experiments directly from DVC Studio. +> given below to run experiments directly from DVC Studio. To run experiments from DVC Studio, select the commit that you want to use and click the `Run` button. A form will let you specify all the changes that you @@ -42,17 +42,15 @@ on `Commit changes`. ![](https://static.iterative.ai/img/studio/cml_commit_v2.png) -At this point, the new experiment appears in the view table. If you check in -your Git repository, you will find the Git commit that you just made. If you had -committed to a new branch, a new pull request will also have been created from -the new branch to the base branch. +At this point, the new experiment appears in the view table. If you just +committed to a new branch, then a new pull request will also have been created +from the new branch to the base branch. -If your project is integrated with a CI/CD setup (eg, GitHub Actions), the CI/CD +If your project is integrated with a CI/CD setup (e.g. GitHub Actions), the CI/CD setup will get invoked. If this setup includes a model training process, it will be triggered, which means that your ML experiment will run automatically. The model training can happen on any cloud or Kubernetes. For more details on how to -set up CI/CD pipelines for your ML project, refer to -[CML](https://dvc.org/doc/cml) . +set up CI/CD pipelines for your ML project, refer to [CML](https://cml.dev). Once the experiment completes, its metrics will be available in the view table. You can then generate plots and trend charts for it, or compare it with the diff --git a/content/docs/studio/view-settings.md b/content/docs/studio/view-settings.md index 642d9643aa..735b692504 100644 --- a/content/docs/studio/view-settings.md +++ b/content/docs/studio/view-settings.md @@ -4,12 +4,15 @@ If you are creating a view for a DVC repo, and if the DVC repo is at the root of the Git repository and does not reference remote/cloud storage, then you can successfully visualize it without specifying additional settings. -In other cases, you could be trying to create views from non-DVC repositories. -Or, your project may be in a sub-directory inside a monorepo. Or, the data to be -visualized could be in custom files in your repository or even in remote / cloud -storages. In all these scenarios, you will need to specify advanced settings so -that DVC Studio is able to access the data required for visualization. These -scenarios have been explained in details below. +Alternatively, you could create views from: + +- Non-DVC repositories +- Project sub-directories in a monorepo +- Custom files in your repository or remote/cloud storage + +In each of these scenarios, you will need to specify advanced settings for DVC +Studio to be able to access the data required for visualization. Details are +given below. ## Non-DVC repositories @@ -21,27 +24,25 @@ You can add these details to your Git repositories in two ways: 1. **Set up DVC repositories**: You can use [DVC](https://dvc.org/) and Git to version your code, data and models all within your Git repositories. By using DVC, you can be sure not to bloat your repositories with large volumes of - data or huge models. These large assets reside in cloud or other remote + data or huge models. These large assets reside in the cloud or other remote storage locations. You will simply track their version info in Git. Refer to the [DVC documentation](https://dvc.org/doc) to initialize a DVC repository. You can then connect to this DVC repository and create a view as described [earlier](/doc/studio/create-view). DVC Studio automatically detects metrics, plots, and hyperparameters files specified in the project's - `dvc.yaml`. And every time you push a new commit to this DVC repository, your - view will reflect the new changes. + `dvc.yaml`. Each time you push a commit to this DVC repository, your view + will reflect the new changes. 2. **Specify custom files with your metrics and parameters**: If you are working - with a non-DVC repository, you can still create views for it if your - repository contains CSV, JSON or YAML files with the metrics and - hyperparameters that you want to visualize. To enable DVC Studio to visualize - such custom data, simply - [specify the custom files](#specifying-view-settings) with your metrics and - parameters, and DVC Studio will efficiently generate tables and plots for - your custom input. For instance, if you have an ML project for which you - generate and save metrics either manually or using some ML tracking tools, - then you can create a view for this project by specifying the file (within - your Git repo) which contains your saved metrics. + with a non-DVC repository, you can still create views for it provided that + metrics and hyperparameters are stored in CSV, JSON or YAML files. To + visualize such custom data, simply + [specify the custom files](#specifying-view-settings) to use, and DVC Studio + will efficiently generate tables and plots for your custom input. For instance, + if you have an ML project for which you generate and save metrics either manually + or using some ML tracking tools, then you can create a view for this project by + specifying the file (within your Git repo) which contains your saved metrics. ## Monorepo @@ -52,10 +53,10 @@ Instead, it could be in a sub-directory of a will need to specify the full path to the sub-directory that you want to use with your view. -## Data remotes (cloud / remote storage) +## Data remotes (cloud/remote storage) The metrics and parameters that you want to include in the view may also be -present in a data remote(cloud storage or another location outside of the Git +present in a data remote (cloud storage or another location outside the Git repo). If you want to include such data in your views, then you will have to grant DVC Studio access to the data remote. @@ -77,12 +78,12 @@ in the view. In the menu that opens up, click on `Settings`. the view. - **Data remotes:** If you need to set up DVC data remotes for your view, you - will need to do it after your view has been created. So, first create your + will need to do it after your view has been created. First, create your view without specifying the data remotes. Once your view is created, open the View settings. Open the `Data remotes / cloud storage credentials` section. The data remotes that are used in your DVC repo will be listed. Now, click on `Add new credentials`. In the form that opens up, select the provider (Amazon - S3, GCP, etc.). depending on the provider, you will be asked for more details + S3, GCP, etc.). Depending on the provider, you will be asked for more details such as the credentials name, username, password etc. ![](https://static.iterative.ai/img/studio/s3_remote_settings.png) From d4ce0bbca22b5b377c9896f68e269fce45614084 Mon Sep 17 00:00:00 2001 From: "Restyled.io" Date: Tue, 1 Jun 2021 18:12:03 +0000 Subject: [PATCH 13/17] Restyled by prettier --- content/docs/studio/create-view.md | 4 ++-- content/docs/studio/run-experiments.md | 11 ++++++----- content/docs/studio/view-settings.md | 17 +++++++++-------- 3 files changed, 17 insertions(+), 15 deletions(-) diff --git a/content/docs/studio/create-view.md b/content/docs/studio/create-view.md index 1c9c5349d2..f136115032 100644 --- a/content/docs/studio/create-view.md +++ b/content/docs/studio/create-view.md @@ -66,5 +66,5 @@ Click on the `Configure` button, and you will be redirected to the GitHub page to install and authorize the DVC Studio GitHub app. > Note that you do not have to authorize DVC Studio on all the repositories in -> your GitHub organization. You can limit access to only repositories that -> you want to use with DVC Studio. +> your GitHub organization. You can limit access to only repositories that you +> want to use with DVC Studio. diff --git a/content/docs/studio/run-experiments.md b/content/docs/studio/run-experiments.md index f0d5261757..d1e53e2bf2 100644 --- a/content/docs/studio/run-experiments.md +++ b/content/docs/studio/run-experiments.md @@ -46,11 +46,12 @@ At this point, the new experiment appears in the view table. If you just committed to a new branch, then a new pull request will also have been created from the new branch to the base branch. -If your project is integrated with a CI/CD setup (e.g. GitHub Actions), the CI/CD -setup will get invoked. If this setup includes a model training process, it will -be triggered, which means that your ML experiment will run automatically. The -model training can happen on any cloud or Kubernetes. For more details on how to -set up CI/CD pipelines for your ML project, refer to [CML](https://cml.dev). +If your project is integrated with a CI/CD setup (e.g. GitHub Actions), the +CI/CD setup will get invoked. If this setup includes a model training process, +it will be triggered, which means that your ML experiment will run +automatically. The model training can happen on any cloud or Kubernetes. For +more details on how to set up CI/CD pipelines for your ML project, refer to +[CML](https://cml.dev). Once the experiment completes, its metrics will be available in the view table. You can then generate plots and trend charts for it, or compare it with the diff --git a/content/docs/studio/view-settings.md b/content/docs/studio/view-settings.md index 735b692504..03d8a2dabc 100644 --- a/content/docs/studio/view-settings.md +++ b/content/docs/studio/view-settings.md @@ -39,10 +39,11 @@ You can add these details to your Git repositories in two ways: metrics and hyperparameters are stored in CSV, JSON or YAML files. To visualize such custom data, simply [specify the custom files](#specifying-view-settings) to use, and DVC Studio - will efficiently generate tables and plots for your custom input. For instance, - if you have an ML project for which you generate and save metrics either manually - or using some ML tracking tools, then you can create a view for this project by - specifying the file (within your Git repo) which contains your saved metrics. + will efficiently generate tables and plots for your custom input. For + instance, if you have an ML project for which you generate and save metrics + either manually or using some ML tracking tools, then you can create a view + for this project by specifying the file (within your Git repo) which contains + your saved metrics. ## Monorepo @@ -78,10 +79,10 @@ in the view. In the menu that opens up, click on `Settings`. the view. - **Data remotes:** If you need to set up DVC data remotes for your view, you - will need to do it after your view has been created. First, create your - view without specifying the data remotes. Once your view is created, open the - View settings. Open the `Data remotes / cloud storage credentials` section. - The data remotes that are used in your DVC repo will be listed. Now, click on + will need to do it after your view has been created. First, create your view + without specifying the data remotes. Once your view is created, open the View + settings. Open the `Data remotes / cloud storage credentials` section. The + data remotes that are used in your DVC repo will be listed. Now, click on `Add new credentials`. In the form that opens up, select the provider (Amazon S3, GCP, etc.). Depending on the provider, you will be asked for more details such as the credentials name, username, password etc. From 39cbc0a3ec56d8ae9aa939c6e12d55384cc91a85 Mon Sep 17 00:00:00 2001 From: tapadipti Date: Wed, 2 Jun 2021 00:25:07 +0530 Subject: [PATCH 14/17] Added videos --- content/docs/studio/get-started.md | 4 ++++ content/docs/studio/run-experiments.md | 5 +++++ content/docs/studio/view-settings.md | 6 ++++++ 3 files changed, 15 insertions(+) diff --git a/content/docs/studio/get-started.md b/content/docs/studio/get-started.md index 2ae6a9387d..1af4f97bfe 100644 --- a/content/docs/studio/get-started.md +++ b/content/docs/studio/get-started.md @@ -5,6 +5,10 @@ your ML projects. You will need access to a GitHub, GitLab or Bitbucket account which has access to the Git repositories you want to connect. DVC Studio creates views from repositories when you connect to them. +The following video provides you a quick overview of DVC Studio. + +https://www.youtube.com/watch?v=hKf4twg832g + ## What is a view? A _view_ is an interactive representation of the experiments run in your ML diff --git a/content/docs/studio/run-experiments.md b/content/docs/studio/run-experiments.md index d1e53e2bf2..eb778e436f 100644 --- a/content/docs/studio/run-experiments.md +++ b/content/docs/studio/run-experiments.md @@ -10,6 +10,11 @@ on each new commit. > create views for your ML project repositories, you can follow the instructions > given below to run experiments directly from DVC Studio. +Watch this video for an overview of how you can run experiments from DVC Studio, +or read below for details. + +https://www.youtube.com/watch?v=nXJXR-zBvHQ + To run experiments from DVC Studio, select the commit that you want to use and click the `Run` button. A form will let you specify all the changes that you want to make to your experiment. On this form, there are 2 types of inputs that diff --git a/content/docs/studio/view-settings.md b/content/docs/studio/view-settings.md index 03d8a2dabc..75ba29d684 100644 --- a/content/docs/studio/view-settings.md +++ b/content/docs/studio/view-settings.md @@ -45,6 +45,12 @@ You can add these details to your Git repositories in two ways: for this project by specifying the file (within your Git repo) which contains your saved metrics. + So as you can see, DVC Studio simply requires your metrics and + hyperparameters to be available in data files in your Git repositories. This + video further illustrates this concept. + + https://www.youtube.com/watch?v=5xM5az78Lrg + ## Monorepo Depending on how you have set up your Git repositories, your DVC repo (for which From 731e8a5514faa0d5945e50babb2365bd6dd39910 Mon Sep 17 00:00:00 2001 From: tapadipti Date: Wed, 2 Jun 2021 08:13:45 +0530 Subject: [PATCH 15/17] Changed the intro of Studio docs as per Ivan's comment in the PR --- content/docs/studio/index.md | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/content/docs/studio/index.md b/content/docs/studio/index.md index 05d28d5c13..c68431ff07 100644 --- a/content/docs/studio/index.md +++ b/content/docs/studio/index.md @@ -1,8 +1,9 @@ # DVC Studio -[`DVC Studio`](https://studio.iterative.ai/) is a comprehensive and interactive -web application to manage and collaborate on your Machine Learning projects -visually. +[`DVC Studio`](https://studio.iterative.ai/) is a web application that you can +[access online](https://studio.iterative.ai/) 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 and Git, it enables you to seamlessly manage data and models, run and track experiments, and visualize and From c0d5214bba3d5261fbb9362c0aac1bb7ef73abb2 Mon Sep 17 00:00:00 2001 From: tapadipti Date: Wed, 2 Jun 2021 10:41:09 +0530 Subject: [PATCH 16/17] Added DVC benefits, data remotes access details + some other changes --- content/docs/studio/get-started.md | 16 ++++++++-------- content/docs/studio/view-settings.md | 13 ++++++++++++- 2 files changed, 20 insertions(+), 9 deletions(-) diff --git a/content/docs/studio/get-started.md b/content/docs/studio/get-started.md index 1af4f97bfe..8c524fc251 100644 --- a/content/docs/studio/get-started.md +++ b/content/docs/studio/get-started.md @@ -3,7 +3,7 @@ Here, we will walk you through a tutorial to use DVC Studio for collaboration on your ML projects. You will need access to a GitHub, GitLab or Bitbucket account which has access to the Git repositories you want to connect. DVC Studio creates -views from repositories when you connect to them. +[views](#what-is-a-view) from repositories when you connect to them. The following video provides you a quick overview of DVC Studio. @@ -12,14 +12,14 @@ https://www.youtube.com/watch?v=hKf4twg832g ## What is a view? A _view_ is an interactive representation of the experiments run in your ML -project. DVC Studio automatically identifies datasets, metrics and -hyperparameters in your ML experiments. The data can either be in DVC -repositories or you can add custom files with the required data. Refer to +project. DVC Studio identifies datasets, metrics and hyperparameters in your ML +experiments. These details can either be in DVC repositories or you can add +custom files with the required data. Refer to [View settings](/doc/studio/view-settings) to understand the different ways in -which you can prepare your Git repository for use with DVC Studio. Using this -data, DVC Studio creates a summary view. This includes tables of all the metrics -across all your experiments. You can also generate plots and compare experiments -here. +which you can prepare your Git repository for use with DVC Studio. Using these +details, DVC Studio creates a View, which is a tabular presentation of all your +experiments along with their datasets, metrics ad hyperparameters. You can also +generate plots and compare experiments here. ## DVC Studio Views page diff --git a/content/docs/studio/view-settings.md b/content/docs/studio/view-settings.md index 75ba29d684..e73f0a44e8 100644 --- a/content/docs/studio/view-settings.md +++ b/content/docs/studio/view-settings.md @@ -25,7 +25,13 @@ You can add these details to your Git repositories in two ways: version your code, data and models all within your Git repositories. By using DVC, you can be sure not to bloat your repositories with large volumes of data or huge models. These large assets reside in the cloud or other remote - storage locations. You will simply track their version info in Git. + storage locations. You will simply track their version info in Git. DVC also + enables you to + [share your data and model files](/doc/use-cases/sharing-data-and-model-files), + [create data registries](/doc/use-cases/data-registries), + [create data pipelines](/doc/start/data-pipelines), connect them with + [CML](/doc/cml) for CI/CD in machine learning, and so on. Find more about the + features and benefits of DVC [here](/doc/start). Refer to the [DVC documentation](https://dvc.org/doc) to initialize a DVC repository. You can then connect to this DVC repository and create a view as @@ -98,3 +104,8 @@ in the view. In the menu that opens up, click on `Settings`. For details on what permissions are required, refer to the DVC documentation on [supported storage types](/doc/command-reference/remote/add#supported-storage-types). + + Note that DVC Studio uses the credentials only to read plots/metrics files if + they are not saved into Git. It does not access any other data in your remote + storage. And you do not need to provide the credentials if any DVC data remote + in not used in your Git repository. From 8edab0d5df699cc867e7de728e1b2bfd9a815884 Mon Sep 17 00:00:00 2001 From: tapadipti Date: Wed, 2 Jun 2021 10:52:48 +0530 Subject: [PATCH 17/17] Small changes as per PR comments --- content/docs/studio/get-started.md | 6 +++--- content/docs/studio/index.md | 4 ++-- 2 files changed, 5 insertions(+), 5 deletions(-) diff --git a/content/docs/studio/get-started.md b/content/docs/studio/get-started.md index 8c524fc251..0bb550b335 100644 --- a/content/docs/studio/get-started.md +++ b/content/docs/studio/get-started.md @@ -13,11 +13,11 @@ https://www.youtube.com/watch?v=hKf4twg832g A _view_ is an interactive representation of the experiments run in your ML project. DVC Studio identifies datasets, metrics and hyperparameters in your ML -experiments. These details can either be in DVC repositories or you can add +experiments. These values can either be in DVC repositories or you can add custom files with the required data. Refer to [View settings](/doc/studio/view-settings) to understand the different ways in which you can prepare your Git repository for use with DVC Studio. Using these -details, DVC Studio creates a View, which is a tabular presentation of all your +values, DVC Studio creates a View, which is a tabular presentation of all your experiments along with their datasets, metrics ad hyperparameters. You can also generate plots and compare experiments here. @@ -44,7 +44,7 @@ You can dive deep into all the experiments committed to the repo. For this, open the view by clicking the view name (in this case, `example-get-started`). A table will be generated as shown below. This includes metrics, hyperparameters -and information about the datasets. All the data is flattened and neatly +and information about the datasets. All these values are flattened and neatly presented for you to evaluate and compare the experiments. ![](https://static.iterative.ai/img/studio/view_components_v2.png) diff --git a/content/docs/studio/index.md b/content/docs/studio/index.md index c68431ff07..ef7d96316c 100644 --- a/content/docs/studio/index.md +++ b/content/docs/studio/index.md @@ -23,5 +23,5 @@ Use DVC, CML and Studio to: - Use the power of Git to track and preserve all your experiments. -- Run [CI/CD](https://en.wikipedia.org/wiki/CI/CD) for your ML projects on cloud - resources of your choice without any new tools. +- Run experiments for your ML projects on cloud resources of your choice without + any new tools.