Hey HN! Builder of Model Zoo here.<p>I've worked as a software engineer in machine learning for a few years, bringing models to production in a variety of areas. Although there exists an open-source ecosystem for deploying ML models, most of these tools are targeted towards infrastructure engineers -- Kubernetes, Docker, and web server frameworks. As a result, there exists a gap today between some of the data scientists and machine learning engineers that develop these models and the skills required to deploy them.<p>I built Model Zoo to address that gap. Deploy your model to an HTTP endpoint with a single line of code, from any Python environment. Plus, you get all the features you'll need from a production ML system for free (monitoring features / predictions, autoscaling, web interface for documentation).<p>Test it out with one of our quickstarts here for free. You can experiment with it in-browser via Google Colaboratory or in your own Python environment:<p><a href="https://docs.modelzoo.dev/quickstart/tensorflow.html" rel="nofollow">https://docs.modelzoo.dev/quickstart/tensorflow.html</a> or <a href="https://docs.modelzoo.dev/quickstart/transformers.html" rel="nofollow">https://docs.modelzoo.dev/quickstart/transformers.html</a>
This is pretty cool! Do you have support for GPU/TPU acceleration? I saw you can deploy in your own AWS VPC, can you do the same on GCP? Also, what model frameworks do you support right now? TF 1.x/2.x and PyTorch?
Neat idea.<p>I'm wondering what ML model formats you use and what is specification for inputs\results? Do you try to use some common format or it is fine for you to use a proprietary one?<p>I'm building <a href="https://DVC.org" rel="nofollow">https://DVC.org</a> - Git for data. We think about how to save and version ML models properly (through GitFlow). One of the biggest challenges - there are not common formats for ML models, inputs and scorings formats.<p>I'd appreciate if you could share your opinion about models format generalization.
Ahem -- Name conflicts with the collection of deep learning models by Intel: <a href="https://www.intel.com/content/www/us/en/artificial-intelligence/posts/model-zoo-ia.html" rel="nofollow">https://www.intel.com/content/www/us/en/artificial-intellige...</a>
This is very interesting. Do you support Azure or even custom hosting?
Once the model is deployed do you monitor things besides performance? Such as I/O to the model, data drift, etc?