In v35, models for flows created in v33 or v34 are forward compatible, meaning that you can use them in v35 without having to retrain them during the upgrade process. As a result, forward-compatible models allow you to:
- upgrade v33 or v34 models after upgrading the application to v35,
- maintain the automation levels achieved in v33 or v34, and
- continue existing training efforts for any use cases that may be in implementation when upgrading to v35.
Which models are forward compatible
In v35, the following types of models are forward compatible:
- Locator models
- Transcription models (i.e., sets of finetuning models)
- Classification models
The following types of models are not forward compatible, so you need to train v35 versions of these models in order to use them in v35:
- Text Classification models
- Text Classification is currently available in its preview version. To learn more, see Text Classification (Preview).
Forward-compatible models and upgrades
When upgrading to v35, keep in mind that some aspects of the v35-upgrade process differ from upgrades to previous versions.
How the upgrade process changes
- When upgrading to v35, you do not need to attach a v35 version of the trainer to your previous application version.
-
Models for flows created in v33 or v34 are not removed from the system during the upgrade process; they remain intact and work "out of the box" after upgrading to v35, with no retraining required.
- Note that you can use only v33 or v34 models, not both. To learn more about this limitation, see the “Upgrading from v33 to v35” section in the Upgrade Considerations and Known Issues article.
-
In v35, Hyperscience supports flows created in one of the previous two versions, but not both.
- Upgrading from v33 to v35 makes all v33 flows compatible with v35.
- Upgrading from v33 to v34 to v35 or from v34 to v35 makes all v34 flows compatible with v35.
- If you want to use flows created in v32 or earlier, you need to upgrade those flows prior to upgrading to v35 — they are not supported in v35, nor are their models forward compatible.
- If you are upgrading from v31 or earlier to v35, you still need to attach trainers from future versions when upgrading to v33 on your way to upgrading to v35. For more detailed instructions, see the "Upgrading from v31" section of the Upgrading to v35 article.
Using v33 or v34 flows and models in v35
As mentioned earlier, you can continue using v33 or v34 flows in v35 without interruption or loss of automation. However, these older flows can only use the models they were using before upgrading to v35, and you cannot import models for these flows. You cannot import a v33 or v34 model for use with a v35 flow.
If you need to change a model for a v33 or v34 flow (e.g., train it for a new Semi-structured layout), you must create a new model, along with a new flow in v35 to go with it. When changing a model, all of the models for the model's flow must be upgraded for the new v35 flow. In other words, you cannot upgrade a flow's models selectively or in a piecemeal fashion. You also cannot attach a v33 or v34 trainer to update the models for a v33 or v34 flow.
When using v33 or v34 flows, you can use cloned versions of the releases you used in those versions. However, if you add Semi-structured layouts to those releases, there is no automation for the submissions matched to these layouts. That is, the system generates Supervision tasks for these submissions.
Memory Management
If you are using v33 or v34 flows and out-of-memory errors occur, you should enable the Memory Management feature. This feature assigns flows that were created in specific application versions to specific application machines in your instance.
To learn more about Memory Management, see Memory Management.
How to know when to train forward-compatible models
When you open a v33 or v34 flow in Flow Studio, the system lets you know which models for the flow need to be trained.
From there, you can click on an issue's Go to Model Management button.
- If the affected model is a Locator Model, you are redirected to the Model Library. There, you can find the model mentioned in the issue's description and click on its name. On the model’s details page, you can take action to correct the issue.
- If the model mentioned in the issue is a Classification or Transcription mode, clicking Go to Model Management takes you directly to the model’s details page, where you can take corrective action.
Note that on the model details page, the model's Last Training date does not show the date v33 or v34 models were last trained. This field applies only to v35 models.
Models from v32 or earlier do not appear in the Model Library.