Uploads overview
Last updated
Last updated
When navigation to the training uploads, you arrive upon an overview of all your uploads with training data. Each line in this overview shows an upload with its properties. An upload can consist of several documents.
The overview consists of several columns:
Status
Processing Document(s) in the upload are being processed by Metamaze AI
Input required Document(s) in the upload have been processed but some or all documents require human input to be finalised
Processed Document(s) in the upload have been fully processed
Failed Processing failed for document(s) in the upload
Number of documents
Shows the number of document in the upload with a breakdown by status
Upload ID
The unique ID associated with the upload
Uploaded on
The date the files were uploaded
Uploaded by
User who made the upload, in case the upload was done through the UI
Filtering uploads based on criteria is handy to find your upload(s) quickly. Note: When extra fields for your organisation have been defined, they will also be shown in the filter options.
Selecting uploads allows you to perform an action on multiple uploads:
Retry upload(s) This will re-send an upload through your AI pipeline, performing a new OCR, re-creating documents and making predictions when possible
Delete Deletes the selected upload(s) from the training data set
When you click an upload, you can view extra details such as when it was uploaded, which steps of the pipeline it went through and what the current status is of the upload and the documents in the upload.
When you click on an upload row, the panel with a detail view of the pipeline steps appears.
In this example, we see input, OCR and output at the upload level and for each document, we see the step document classification, entity extraction and business rules. When page management is enabled you get an extra section in between the OCR and documents.
For the following example we can see the following:
The upload was done 18/08/2022
OCR was successful
2 documents were created of which 1 was processed and 1 requires input from a human
The first document is recognised as an English bill of lading(s) document for which 6 entities were found and 3 couldn't be found. It also has 1 failing business rule.
You can click on each of the steps in the 'detail view pipeline' to see more details of that step. If you click on the 'input' step:
- If the files were uploaded via the UI - the file names (and link to the actual file) and show the user who uploaded the files via the UI.
- If the files have been uploaded via the API - show the body payload of your HTTP request to Metamaze with the corresponding response payload and response code.
If you click on the page management step in the 'detail view pipeline' it will show the result of the page management model and what happened in the manual intervention. Check out the page on Page management.
When you click on a document, you get extra details on the following:
Document classification
Entity extraction
Business rules
Clicking on "View document" button will give you all the details on the document, check out the page Document overview.
Detail view output
When you click on the 'output' step you will see the body payload and the response code of the result that was sent to your service, just like the 'input' step. More information about the payload can be found in the API documentation.
