This article is reproduced from the public AI Technology Base Camp.Original address
Top 1: Pytext (4176 Stars)
a build based on PyTorch NLP Modeling (based on deep learning) framework, contributed by Facebook Research open source. The core functions can support neural network models such as text classification and sequence labeling.
Github addressTop 2: Jax (2734 Stars)
Jax collects tools for high-performance machine learning research from Autograd and XLA, contributed by Google Open Source. Many people say that Jax is a replacement for TF and is simpler and easier to use.
Github addressTop 3: Reaver-pysc2 (323 Stars)
Reaver is a proxy for deep reinforcement learning. Yes, StarCraft II is based on this agent training.
Github addressTop 4: PySyft (2552 Stars)
PySyft is a Python library for deep learning security and privacy issues. Open source by OpenMind.
Github addressTop 5: Image-super-resolution (267 Stars)
The Residual Dense Network in the field of image super-resolution implemented by Keras.
Github addressTop 6: BERT-as-service (1703 Stars)
Using Google's BERT In the model application and sentence coding tasks, a variable length sentence is mapped into a fixed length vector.
Github addressTop 7: Transparent_latent_gan (1302 Stars)
Supervised learning and TL-GAN Control image editing and generation.
Github addressTop 8: Self_driving_pi_car (667 Stars)
Self-Driving Pi Ca r is a deep neural network-based autonomous vehicle that combines the computing power of Lego Mindstorms NXT and Raspberry Pi 3.
Github addressTop 9: Gandissect (972 Stars)
Gandissect is a PyTorch-based tool for visualizing and understanding GAN neurons.
Github addressTop 10: Aleph_star (138 Stars)
The realization of the paper "Reinforcement Learning with A* and a Deep Heuristic".
Github address
Comments