- To complete this action, to create, edit and manage this new resource in your project, you must be a Workspace Manager. Only this role on the platform can allow to configure resources inside a workspace.
How to create a new resource into your workspace step by step
Step 1: Access to the "Resources" section
This step allows you to review, edit, and manage resources inside your workspace
Step 2: Create a new resource
A MT is considerated as a resource in Wezen, you need to create a new one, so click on "Create Resource"
Step 3: Inside the configuration page, create a new MT (Machine Translation) resource
You must select the "Resource type" with the field "MT"
You can edit the name of your new MT in the "resource name”, please make sure that you are naming this engine identical to its Datalake name (can be seen in Datalake engines page provided to you by MT team)
You must select one or several groups where this MT can be available.
A group is defined by:
- a list of resources
- a list of user and their role(s)
- All users in the group will then have the defined roles on all the resources of the group.
Step 4: Complete the required fields
This part of the set up is mandatory to activate a functional MT Datalake in your project(s)
First of all, you must select the choice "Datalake" from the list
Then, you have to complete several fields (like the template below), this information can be found in Datalake engine pages provided to you by MT team. If you don’t have this information page below, please contact MT team:
- URL: Add the following link, https://datalake.datawords.io/api/v1/translation/bulk_translate/
- Key: For this field, you must always add first the word Token and a space, and then paste the code like this example
- Token abcdefghijkl1234567890
- Project uuid: Add the code only.
- Engine uuid: Add the code only.
When all fields are completed, click on "Create resource". When this is done, the first part of the configuration is done, to create the Datalake Machine Translation Resource in your workspace.
The second step of the configuration is to activate it in the translation project.
Here is a video of the first part of the configuration:
Step 5: Activate the MT in your project(s)
The second part of the configuration is now to activate the MT in your translation project(s). For this, click on the project where you want to apply your MT
Click on the section "Settings" to access to the set up of your project to link your MT to the project
Step 6: Activation of the MT Datalake for one or several languages in your project
Inside the "Settings", within the section "Local Configurations",find the section "VIII Machine Translation" where you can activate your MT Datalake for one or several pair of languages with the local configurations.
In the below example, the MT Datalake will be applied for all languages translated in the project.
You need to set it up this way:
- "Select another machine": Select the Machine Translation you just created before.
- "Scope": Select "Project" to make sure the MT will work for all the project
- "Daily chat Limit": Add a large number, this will allow the MT to not have limit to cover the need of important project
When it is done, validate your update with the button.
Now, all the set up is done, your Machine Translation Datalake is activated. You can now upload content on Wezen to allow the platform to use your set up and MT.
Here is a video of the second part of the configuration:
When you add a MT to a local configuration, please look if this local configuration concerns the default local configuration or a specific local configuration (by pair language).
If you create a new local configuration with another pair language, this one is created blank with basic settings. Don't forget to copy the other settings from the default local configuration like:
- The workflow
- The segmentation
- The repetition scope
- Enabled WYSIWYG
All must be set up as well and identical from the initial local configuration