Code to reproduce the paper "Deconstructing the Goldilocks Zone of Neural Network Initialization"
-
Updated
May 31, 2024 - Python
Code to reproduce the paper "Deconstructing the Goldilocks Zone of Neural Network Initialization"
Generic derivative objects (gradients, Jacobians, Hessians, and more) by finite differences
Different exercises for finding extremas of a function, approximating roots, finding a minimizer, for univariate and multivariate problems.
Hessian spectral analysis with tensorflow1.x
Implementation of automatic differentiation (AD) in forward and backward modes with mathematical explanations
code for blog post https://gebob19.github.io/natural-gradient/
Particle Metropolis-Hastings using gradient and Hessian information
maptool unauthenticated rce exploit <1.8.0 beta2b
A python3 script computing bond and valence angle force constants using the Seminario (projected hessian) method.
Some Hessian based analysis for practical deep models with tensorflow2
Official code for the paper "Hybrid Quantum-Classical Scheduling for Accelerating Neural Network Training with Newton's Gradient Descent"
Implementation of EPTQ - an Enhanced Post-Training Quantization algorithm for DNN compression
Measuring generalization properties of fine-tuning using Hessian
Implementation of Numerical Analysis algorithms/methods in Python
Scalable Computation of Hessian Diagonals
TensorFlow implementations of losses for sequence to sequence machine learning models
Measuring generalization properties of graph neural networks
Full loss Hessian spectrum approximation tool.
Add a description, image, and links to the hessian topic page so that developers can more easily learn about it.
To associate your repository with the hessian topic, visit your repo's landing page and select "manage topics."