Code for the paper: Why Transformers Need Adam: A Hessian Perspective
☆66Mar 11, 2025Updated last year
Alternatives and similar repositories for hessian-spectrum
Users that are interested in hessian-spectrum are comparing it to the libraries listed below. We may earn a commission when you buy through links labeled 'Ad' on this page.
Sorting:
- Code for Adam-mini: Use Fewer Learning Rates To Gain More https://arxiv.org/abs/2406.16793☆460May 13, 2025Updated last year
- Code for "What really matters in matrix-whitening optimizers?"☆25Oct 31, 2025Updated 11 months ago
- ☆70Aug 28, 2026Updated last month
- Code for ICLR 2022 Paper (HyperDQN: A Randomized Exploration Method for Deep Reinforcement Learning)☆12Nov 28, 2023Updated 2 years ago
- [ICLR 2025] This repository contains the code to reproduce the results from our paper From Sparse Dependence to Sparse Attention: Unveili…☆12Mar 7, 2025Updated last year
- Deploy on Railway without the complexity - Free Credits Offer • AdConnect your repo and Railway handles the rest with instant previews. Quickly provision container image services, databases, and storage volumes.
- Minimal pretraining script for language modeling in PyTorch. Supporting torch compilation and DDP. It includes a model implementation and…☆51Apr 27, 2026Updated 5 months ago
- The official code of "Mano: Restriking Manifold Optimization for LLM Training".☆25Jun 1, 2026Updated 4 months ago
- Using PyTorch autograd to compute Hessian of Perplexity for Large Language Models☆29Apr 17, 2025Updated last year
- ☆284Dec 2, 2024Updated last year
- Official implementation for the paper "Controlled Sparsity via Constrained Optimization"☆12Aug 10, 2022Updated 4 years ago
- ☆15May 2, 2026Updated 5 months ago
- ☆79Dec 7, 2024Updated last year
- Code for "Echo Chamber: RL Post-training Amplifies Behaviors Learned in Pretraining"☆30Oct 14, 2025Updated last year
- My attempt to improve the speed of the newton schulz algorithm, starting from the dion implementation.☆42Apr 30, 2026Updated 5 months ago
- Deploy on Railway without the complexity - Free Credits Offer • AdConnect your repo and Railway handles the rest with instant previews. Quickly provision container image services, databases, and storage volumes.
- Spectral Sphere Optimizer☆136Mar 23, 2026Updated 6 months ago
- An extention to the GaLore paper, to perform Natural Gradient Descent in low rank subspace☆19Oct 21, 2024Updated last year
- GRASS: Compute Efficient Low-Memory LLM Training with Structured Sparse Gradients☆10Jun 19, 2024Updated 2 years ago
- Code for Dayal Kalra's research internship on scalable curvature measures for neural networks.☆32Feb 3, 2026Updated 8 months ago
- APOLLO: SGD-like Memory, AdamW-level Performance; MLSys'25 Oustanding Paper Honorable Mention☆365Nov 29, 2025Updated 10 months ago
- Official implementation for "Mixture of In-Context Experts Enhance LLMs’ Awareness of Long Contexts" (Accepted by Neurips2024)☆14Jan 7, 2025Updated last year
- ☆15May 4, 2024Updated 2 years ago
- An implementation of the penalty-based bilevel gradient descent (PBGD) algorithm and the iterative differentiation (ITD/RHG) methods.☆20Feb 13, 2023Updated 3 years ago
- ☆11Sep 20, 2024Updated 2 years ago
- Deploy to Railway using AI coding agents - Free Credits Offer • AdUse Claude Code, Codex, OpenCode, and more. Autonomous software development now has the infrastructure to match with Railway.
- Kinetics: Rethinking Test-Time Scaling Laws☆87Jul 11, 2025Updated last year
- ☆16Apr 15, 2026Updated 6 months ago
- Code for Paper (Policy Optimization in RLHF: The Impact of Out-of-preference Data)☆29Dec 19, 2023Updated 2 years ago
- LISA: Layerwise Importance Sampling for Memory-Efficient Large Language Model Fine-Tuning☆39Apr 4, 2024Updated 2 years ago
- Implementation of various equivariant models in JAX☆18Apr 12, 2024Updated 2 years ago
- DoG is SGD's Best Friend: A Parameter-Free Dynamic Step Size Schedule☆61Aug 23, 2023Updated 3 years ago
- NeurIPS'24 - LLM Safety Landscape☆41Oct 21, 2025Updated 11 months ago
- Official Pytorch Implementation of "Outlier-weighed Layerwise Sampling for LLM Fine-tuning" by Pengxiang Li, Lu Yin, Xiaowei Gao, Shiwei …☆35Jun 3, 2025Updated last year
- Code for reproducing figures and results in the paper ``Early stopping in deep networks: Double descent and how to eliminate it''☆15Jun 27, 2022Updated 4 years ago
- Proton VPN Special Offer - Get 70% off • AdSpecial partner offer. Trusted by over 100 million users worldwide. Tested, Approved and Recommended by Experts.
- A clean and crisp CV template☆59Oct 6, 2020Updated 6 years ago
- Benchmarking Optimizers for LLM Pretraining☆61May 3, 2026Updated 5 months ago
- Code for the paper "The Journey, Not the Destination: How Data Guides Diffusion Models"☆27Dec 12, 2023Updated 2 years ago
- Official code for "Algorithmic Capabilities of Random Transformers" (NeurIPS 2024)☆16Sep 28, 2024Updated 2 years ago
- ☆13Mar 22, 2023Updated 3 years ago
- NeurIPS 2026-Pion: A Spectrum-Preserving Optimizer via Orthogonal Equivalence Transformation☆40Sep 25, 2026Updated 3 weeks ago
- ☆284Updated this week