I want to create a model to extract info for PDFs containing purchase orders. I thought that I could create an AutoML text entity model for that task. The main doubt is what's the best way to handle with the article lists. How can I label each cell in order I can have a list of rows in the reuslt
Thanks
The labeling is very important, less than 10 labels to start would make it easy. As you will need at least a 100 entities labeled per label to train. remember you have three sets to label, train, test, validate. 100 for train, 30 for test and 30 for validate should suffice.
check the label tab often, it shows the break down of what has been labeled so far.
google's documentation is a good start. https://cloud.google.com/natural-language/automl/docs/prepare
I ended up building a Java client to call predict on the model, sending it a list of files to process. the returned JSON has the entities by label for each file.
Related
We're trying to build an imaged based product search for our webshop using the vertex ai image classification model (single label).
Currently we have around 20k products with xx images per product.
So our dataset containing 20k of labels (one for each product - product number), but on import we receive the following error message:
There are too many AnnotationSpecs in the dataset. Up to 5000 AnnotationSpecs are allowed in one Dataset. Check your csv/jsonl format with our public documentation.
Looks like not more than 5000 labels are allowed per Dataset... This quota is not really visible in the documentation - or we didn't find it.
Anyway, any ideas how we can make it work? Does we have to build 5 Datasets with 5 different Endpoints and than query every Enpoint for matching?
You can find those limits in the AutoML quotas and limits documentation.
It is possible to have multiple models for group of products -- Maybe even something like: one initial model to classify the product category (jewery, watches, shoes, toys, etc) and a second step for a specific model (to identify the specific product belong toys, or belong shoes, etc). But to be honest, it seems a bit hard to support - but certainly worth trying.
A second option would be training a custom model where you could do a fine tuning on some larger model (ie. inception, resnet, etc) do know all your 20k+ classes (products). It could add a little bit more work at first, but after established, it will become a single model for inference and re-training would be simpler using MLOps mechanisms (ie. Vertex Pipelines).
I am trying to make something like this in Filemaker Database views.
The database will contain employees and projects. There need to be two combined views. One displaying the list of employees and which projects they are on in which week and by which percentage. The other one displaying the list of projects and which employees are on it in which week by which percentage. For each employee and each project there need to be a sum of percentages for all assigned projects / employees.
I have made three tables in my database:
employee
projects
employee_on_projects
Now I want to create these views. Can anybody tell me how to do this from scratch?
Thanks in advance.
FileMaker is not a good tool for creating cross-tab reports, as it has no native functionality to build these. It has to be done manually. Do a search on cross tab reporting in FileMaker and you will get some ideas.
Traditionally you need relationships and fields to hold and display your data.
On the top of my head, in this case I would probably use merge variables for displaying progress and other calculated data to avoid having fields for all those weeks. A script would gather the data using ExecuteSQL and calculate the variables. The whole thing could even be generated in HTML and displayed in a web viewer.
This way you should get away without too many relationships and fields.
If you want a simpler, but more cluttered solution, you could go with regular fields for all 53 weeks and use a sub summary report. Each field could be set up with ExecuteSQL if possible, to avoid having a relationship for each period.
I am trying to create a report which groups on a column called "Legal Entity." When the output is directed to Excel, a separate tab will be created for each distinct entity in the query resultset.
For each Excel tab/Legal Entity, there will be two "sections." The first is a repeating section that breaks on a column "Funding Arrangement Type." After all of the Funding Arrangement Types are exhausted, there will be a single "Totals" grid which will summarize the data on the tab for the current Legal Entity. The data will be summarized across all Funding Arrangement Types within the current Legal Entity.
Because the Totals (lower) grid is really just a summarization of the same source query, Query1, I thought that I would also bind the Totals grid to it. However, if I do that, I get a run time error that tells me that I need to establish a Master-Detail relationship (If I decide to use a separate query for the Totals grid, the Totals grid "will not be aware" of the current Legal Entity/tab that must be considered when summarizing.)
Therefore, I continued with my guess at how the Master-Detail relationship should be defined. I made various attempts to link the two grids, including:
On all of the dimension (non-summarized) columns.
On Legal Entity
On Legal Entity and Funding Arrangement Type
Doing so affected previously correct totals reported in the upper cross tab results/
This Master-Detail approach is foreign and as a result I don't understand what it is doing.
I also tried to use a separate query, Query2, for the lower totals grid and adding a filter to filter SQL2 where SQL2. LegalEntity = SQL1.LegalEntity in an effort to get the totals grid to summarize within the current LeglEntity grouping. This resulted in a cross join error.
I’m a real noob with Cognos. Suggestions are welcomed. Thank you!
You can use mouse+scroll wheel to zoom in:
I was able to get it working by binding both grids to a single query and for both grids, establish a Master-Detail Replationship on Legal Entity. Prior to doing that, I added these columns to both grids and hide them, not sure if this was necessary.
In my project we define threats/risks and countermeasures. I want to keep track and refer to both types of entities in Sphinx, as well as generating a list of both threats/risks and the countermeasures. Let's say I have 30 risks and 50 countermeasures (many-to-many relationship).
I'd be happy just to have a lists of both and the ability to refer to each other by numbers (e.g. "risk #23", "countermeasure #12"). It would be even better if the system could display the relationship automatically.
The content of both is let's say a single paragraph or even shorter, so that's why I dislike to use regular headings. And I cannot refer to items in lists or table rows. So, I'm looking for something like a Figure in Sphinx (numbered, with caption), but then for arbitrary types of entities.
My current approach is to create a custom RST role for this. Is this the right approach? If so, where to start?
I'm creating an application in which I have several entities and now I need to filter the content of third combobox dynamically. I explain myself better. I have 3 combobox (building, floor and department), I would like first to show me all the buildings included, but the second should show only selected before the plans for the building, the last I should be select only the departments of the building and the plan you choose. How can I do this? To simplify attaching some photos.
You simply drill down with predicates, if you use single fetch requests to Core Data.
However, your relationships are not set up correctly. For example, there is an edificio attribute in Particelle. If it refers to an building, it should be a relationship to a Edifici object, not some kind of foreign key. There are no foreign keys in Core Data, just relationships.
If you do this, everything becomes much easier by using a NSFetchedResultsController. You can now simply traverse the object graph without any specific fetching.
The scheme could be something like this (maybe need to change the order):
Anno <--->> Particella <---->> Edificio <---->> AreaRischio
Now you can simply tell the fetched results controller to start fetching all Anno entities. Then you drill down with simple dot notation:
NSSet *listForNextTable = selectedAnnoObject.particelle;
and further with
NSSet *listForNextTable = selectedParticellaObject.edifici;
etc. You see, it gets really simple.