Hello, im currently learning about OCR and document processing and im currently stuck at the extracting stage. At this stage, im trying to use machine learning to extract data from my invoices and receipts and i dont know if what im doing wrong is the endpoint or something else. Ive tried the https://du.uipath.com/ie/invoices/info/model as the endpoint and https://invoices.uipath.com but neither worked.
The error i received is: “Data Extraction Scope: An error occurred while sending the request.”
The error message that popped up while im trying to configure the extractors is:
“An error occurred while retrieving capabilities from the server. Please check the logs for more details”
For more info, im currently using studio 2020. 10.0-beta.149 with a community license. I installed the UiPath.IntelligentOCR.Activities package version 4.9.0 preview.
Also, i have another slightly unrelated question but its still regarding OCR. I cant seem to digitalise some areas in a scanned document like it doesnt show as a selection option, so (as i said, im still learning haha) what can i do to fix? this. thank you in advance for answering my queries!
There are a couple things to check
First try entering in this as the endpoint if you are a community user
or for enterprise user enter this:
Second thing to try is make sure that your ApiKey for document understanding is correct. Go to Admin->Licenses->Other Services->Document Understanding and copy that key or create a new one if it still isn’t working. Let me know if either of these don’t solve your issue here
hello, ive tried all that you mentioned but none are working.
Ok there is one more option you can try but it is a temporary fix (due to the trial time ending).
The way that I access and use the ML extractor is through AiFabric which I believe requires a enterprise trail. You can sign up for a 60 day enterprise trial through your cloud portal.
From there you can begin to access Ai Fabric. You will need to create a new project on Ai Fabric and deploy a package (most likely you will want the “out of box”->“Document understanding”->“invoice”. after you have deployed the backage you need to create a skill. When creating make sure to select the package you just created. After the skill is finished deploying then you should be able to access this in your uipath programs. Instead of providing an endpoint you will select your ML skill inside of your ML extractor. Once you do this it should connect up and the ML extractor should function properly. If you need any help setting this up please reach out again!
I would like to ask that except invoices, ML extractor would be able to work other categories such as ESG reports? I am a newbie, please kindly help me.
The activity you are using may be obsolete. Try using activity from “Document understanding ML activities”.
Thank you JosephNehl, This tip worked for me
ok, but what if I just select Document Understanding (not invoice). I have labeled all my pdf’s and now going back to UIPath Studio-> Machine Learning Extractor; I select my skill but when clicking get capabilities it says “an error ocurred while retrieving capabilities from the server”. If I test with another skill (previously done by someone else) I get his fields. So the API key is not an issue
I have the same problem does someone has a solution
I just experienced the same problem as you and found a potential solution.
When configuring the ML Skill in AI Center you need to select the trained version of the ML package, the trained one. So if you chose minor version 0, please try redeploying with the trained version. (Example: minor version 1).
I hope this helps!
Thanku it it worked for me! i used enterprise endpoint as i had enterprise trail!