Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[formrecognizer] Add migration guide #20960

Merged
merged 17 commits into from
Oct 1, 2021
Merged
Prev Previous commit
Next Next commit
add manage models section
catalinaperalta committed Oct 1, 2021

Verified

This commit was created on GitHub.com and signed with GitHub’s verified signature. The key has expired.
commit 5981aca5e04f2182f566402ca5c755f114b230eb
10 changes: 8 additions & 2 deletions sdk/formrecognizer/azure-ai-formrecognizer/MIGRATION_GUIDE.md
Original file line number Diff line number Diff line change
@@ -11,6 +11,7 @@ Familiarity with `azure-ai-formrecognizer (3.1.x and below)` package is assumed.
- [Analyzing document](#analyzing-documents)
- [Analyzing a custom model](#analyzing-a-custom-model)
catalinaperalta marked this conversation as resolved.
Show resolved Hide resolved
- [Training a custom model](#training-a-custom-model)
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Building a custom model?

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I left it as training since the guide is catering more towards people coming from the previous version. I commented on this here: #20960 (comment)

- [Manage models](#manage-models)
- [Additional samples](#additional-samples)

## Migration benefits
@@ -567,8 +568,6 @@ Differences between the versions:
- Files for building a new model for version `3.2.x` can be created using the labeling tool found [here][fr_labeling_tool].
- In version `3.1.x` the `use_training_labels` keyword argument was used to indicate whether to use labeled data when creating the custom model.
- In version `3.2.x` the `use_training_labels` keyword argument is not supported since training must be carried out with labeled training documents. In order to extract key-value pairs from a document, please refer to the prebuilt model "prebuilt-document" which extracts entities, key-value pairs, and layout from a document.
catalinaperalta marked this conversation as resolved.
Show resolved Hide resolved
- When using API version `2021-09-30-preview` and later models no longer include submodels, instead a model can analyze different document types.
- When building or composing new models users can now assign their own model IDs and specify a description.

catalinaperalta marked this conversation as resolved.
Show resolved Hide resolved
Train a custom model with `3.1.x`:
```python
@@ -624,6 +623,13 @@ for name, doc_type in model.doc_types.items():
print("Field: '{}' has confidence score {}".format(field_name, confidence))
```

catalinaperalta marked this conversation as resolved.
Show resolved Hide resolved
### Manage models

Differences between the versions:
- When using API version `2021-09-30-preview` and later models no longer include submodels, instead a model can analyze different document types.
- When building or composing new models users can now assign their own model IDs and specify a description.
catalinaperalta marked this conversation as resolved.
Show resolved Hide resolved
- In version `3.2.x` of the library, only models that succeeded can be retreived from get and list calls. In version `3.1.x` of the library, models that had not succeeded were still created and had to be deleted.
catalinaperalta marked this conversation as resolved.
Show resolved Hide resolved

## Additional samples

For additional samples please take a look at the [Form Recognizer Samples][samples_readme] for more guidance.