Pinned Loading

1. cramming cramming Public

   Cramming the training of a (BERT-type) language model into limited compute.

   Python 1.3k 100

2. breaching breaching Public

   Breaching privacy in federated learning scenarios for vision and text

   Python 274 60

3. poisoning-gradient-matching poisoning-gradient-matching Public

   Witches' Brew: Industrial Scale Data Poisoning via Gradient Matching

   Python 95 13

4. invertinggradients invertinggradients Public

   Algorithms to recover input data from their gradient signal through a neural network

   Jupyter Notebook 276 72

5. data-poisoning data-poisoning Public

   Implementations of data poisoning attacks against neural networks and related defenses.

   Python 69 10

6. fullbatchtraining fullbatchtraining Public

   Training vision models with full-batch gradient descent and regularization

   Python 38 5