In this repository it was rebuilt as an operable ML system: the data was audited and repaired, the baseline was re-measured, data and model were iterated on with tracked experiments, and versioned ...

Eye, a 303M-parameter image moderation model using free compute. On my 3,000-image benchmark, it outscored AWS and Google.

Identifies 37 cat and dog breeds by fine-tuning a pretrained ResNet-50 on Oxford-IIIT Pet: 93.6% test accuracy, with tests and a Gradio demo - Gauhet/pet-breed-classifier ...

A new text-driven multiple instance learning model, TedMIL, improved fine-grained melanoma classification and FISH status ...

A deep reinforcement learning model achieved 88.67 percent accuracy in identifying disclosure quality and 86.50 percent in ...