[Request for the Community] Add your ML.NET sample to the Community-samples page
CESARDELATORRE opened this issue Β· comments
Do you have any interesting ML.NET sample that you'd like to share with the community?
If so, you can respond to this issue with info about your ML.NET samples.
Eventually, when reviewed, it'll be showcased in the COMMUNITY-SAMPLES page:
Thanks for your contribution! :)
@jwood803 - I'm guessing you might have any sample you'd like us to point to from the new COMMUNITY-SAMPLES page? :)
https://github.com/dotnet/machinelearning-samples/blob/master/docs/COMMUNITY-SAMPLES.md
@CESARDELATORRE I have a recent, small sample that goes over the new API for a regression task. Would that be applicable to put in the doc?
@jwood803 - Sure! You might just want to "dress" it on a little bit? - I mean, like a short description of the problem and a specific name instead "NewApiExample". :)
Ping me when you think it is in a "shareable" state and I'll point to it, ok?
Eventually, if you have more/better samples, feel free to ask me to update it or add additional samples.
Thanks for your contribution! π
@CESARDELATORRE I dressed it up a bit in another repository. If it needs more info, I'll be glad to update it. π
@jwood803 Cool! - I just added it to the Community-Samples page
However, when you have the time, you'll need to migrate it to the dynamic API in v07, as we have our samples, please. π
@CESARDELATORRE Thanks! I didn't realize 0.7 had already been released. π I just updated the sample to it.
@CESARDELATORRE Not sure if this counts when I borrowed some from a similar sample in this repository, but I made a sample that uses a Goodreads data set to make recommendations based off book ratings.
Last night in response to a question I built a fully-worked custom mapping sample: https://github.com/endintiers/Unearth.Demo.MLCustomTransform - is this of any interest?
endintiers, it is just what I needed. I would love to see many more transform examples.
@endintiers - Sure! I'll point to it from the Community page. π
I build a fully working WPF app that can search photos based on classification. The WPF interface with ONNX model that was build using Keras and later exported. I showed both how to build the model using python and how to interface the ONNX model using ML.net.
https://github.com/Tak-Au/Photo-Search
-is this of any interest?
@Tak-Au - Of course! That WPF sample sounds great!
I just added to the COMMUNITY SAMPLES page here:
https://github.com/dotnet/machinelearning-samples/blob/master/docs/COMMUNITY-SAMPLES.md
Btw, it would be great if you add any screenshot of the WPF app in your README.MD in your repo. The more you can explain in the README.MD, the more engaging it'll be! π
Cool, I updated the readme.md file per your recommendations. π
Just mentioning that the UWP sample app was upgraded from v0.6 to v0.11 - thanks to your team.
@XamlBrewer - Updated here: https://github.com/dotnet/machinelearning-samples/blob/master/docs/COMMUNITY-SAMPLES.md
Thanks for the heads-up!
Feel free to submit a PR for changes in the .MDs at any time! π
I have posted few tutorials for ML.NET 1.0.0-preview version, check out on https://github.com/sthakuri/ML.NET-1.0.0-preview
How to build an example of blog recommendation, through user ID, blog ID, interest score.
@freefer There are different approaches mentioned in ReadMe file for different kinds of features in our Movie Recommender sample https://github.com/dotnet/machinelearning-samples/tree/master/samples/csharp/end-to-end-apps/Recommendation-MovieRecommender
Probably, you can try Matrix Factorization algorithm option based on the features user ID, blog ID, interest score.
Please try once and let me know if you need more details
@sthakuri - Great! We'll add that sample (https://github.com/sthakuri/ML.NET-1.0.0-preview) to the COMMUNITY SAMPLES page! π
@sthakuri As per the link https://github.com/sthakuri/ML.NET-1.0.0-preview the samples for sentiment analysis and
Taxifare look same as our Machine learning samples. Is there anything different than our samples to add to community samples? If you have any new sample other than our Machine Learning samples, then we can add it to community sample.
Adding @CESARDELATORRE for reference.
There is a working sample of feeding LoadFromEnumerable from a database with ML.NET 1.2.0 and EF Core 3.0 preview 3 (doesn't work in EF Core 2.x because of threading issues) at: https://github.com/endintiers/Unearth.Demo.ML.FromDB
I updated the code for https://github.com/endintiers/Unearth.Demo.MLCustomTransform to ML.NET 1.2.0 and simplified the code (which is now possible).
I have created a small working sample for a classification problem in ML.Net v 1.3.1. Would it be applicable to add to the community samples?
https://github.com/deepak21188/Mushroom-Classification-using-C-Sharp-and-ML.Net
@deepak21188 - Sure! I just added it to the ML.NET Community Samples page here:
https://github.com/dotnet/machinelearning-samples/blob/master/docs/COMMUNITY-SAMPLES.md
Thanks! π
Hey everyone, does any project implements Image classification with the Bitmap as input value. I knew how to predict with filepath as input, but i dont know how to call predict with System.Drawing.Bitmap :(
I just did data analysis on COVID-19 dataset shared by Johns Hopkinton university using .Net DataFrame API in C#.
Github: https://github.com/praveenraghuvanshi1512/covid-19
@CESARDELATORRE Please have a look at it and see if it could be part of community-samples.
I've created a sample that analyzes the sentiment of movie reviews. Is it suitable for the community samples?
I have a workshop that includes how to build a movie a movie recommendation engine with ML.NET's Matrix factorization.
Would it be applicable to add to the community samples?
Here are the workshop materials, including the code in ML.NET:
https://github.com/Ellyster/STW2020-MovieWars
I've created samples using 2 different YOLO V3 (ONNX) models for object detection, using Bitmap as input value (also answers #86 (comment))
https://github.com/BobLd/YOLOv3MLNet
Related to #575?
EDIT: YOLO V4 available here
I have created a sample that uses the cat/dog image classifier created from this Keras example.
Might be useful for #779
BERT Q&A onnx model https://github.com/GerjanVlot/BERT-ML.NET
Does anyone have a sample that may be used in the Numer.Ai tournament for Market Prediction using their obfuscated datasets?