Guys
As Google Cloud ML request (AutoML Entity Extraction Project)
Each label must have at least 50 annotations. Though ideally, each label should have at least 100 annotations. Fewer annotations often result in inaccurate precision and recall. You must also have at least 10 annotations each assigned to your Train, Validation and Test sets.
More than 50 annotations, and at least 10 annotations
But we still can't click the Start Training Button.
For Entity Extraction, AutoML requires at least 100 training items to be able to train your model as is mentioned in the documentation 1.
Here is a more detailed reference regarding Entity Extraction 2
You supply between 50 and 100,000 documents to use for training your custom model. You use between one and 100 unique labels to annotate the entities you want the model to learn to extract. Each annotation is a span of text and an associated label. Label names can be between 2 and 30 characters, and can be used to annotate between one and 10 words. We recommend using each label at least 200 times in your training data set.
