Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/tomconte/learnml
Azure ML learning experiment
https://github.com/tomconte/learnml
Last synced: 12 days ago
JSON representation
Azure ML learning experiment
- Host: GitHub
- URL: https://github.com/tomconte/learnml
- Owner: tomconte
- Created: 2021-01-28T13:36:35.000Z (almost 4 years ago)
- Default Branch: master
- Last Pushed: 2021-03-04T08:39:39.000Z (over 3 years ago)
- Last Synced: 2024-10-26T09:47:31.777Z (25 days ago)
- Language: Python
- Size: 457 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Topic Modeling with LDA on Azure Machine Learning
Experimenting with Topic Modeling using [Gensim's LDA implementation](https://radimrehurek.com/gensim//auto_examples/tutorials/run_lda.html) and the [20 Newsgroups](http://qwone.com/~jason/20Newsgroups/) data set; running on [Azure Machine Learning](https://docs.microsoft.com/en-us/azure/machine-learning/).
![LDA-Viz](https://github.com/tomconte/learnml/raw/media/lda-viz.png)
## `azureml`
Contains the Python scripts used to train and deploy the model, using the Azure ML SDK.
- `create_environment.py`: create the Environment, mostly specifying the required Conda packages (`nltk` and `gensim`).
- `run_training_once.py`: train the model within an Experiment, with configurable parameters. Registers the resulting Model.
- `run_training_hyperdrive.py`: train the model using HyperDrive to compare different paramter values (e.g. number of topics).
- `deploy_model_aci.py`: deploy the model to ACI (suitable for testing).
- `deploy_model_function.py`: deploy the model to Functions (suitable for production).
- `download_model.py`: download the model locally; used to create the Docker image for custom Functions deployment.
- `model_pipeline.py`: create and run a pipeline consisting of data preparation and model training. Registers the resulting model.## `topicmodel`
The data preparation, training and scoring scripts for the LDA model.
## `fn-topic-model`
A Function definition and build script that creates a custom Docker container image, ready to be deployed to a Function host. Alternative to deploying "natively" using Azure ML deployment capabilities.
## `terraform`
Terraform configuration for the Function host.