haraschax / nogradLinks
Gradient descent is cool and all, but what if we could delete it?
☆104Updated this week
Alternatives and similar repositories for nograd
Users that are interested in nograd are comparing it to the libraries listed below
Sorting:
- Simple Transformer in Jax☆138Updated last year
- look how they massacred my boy☆63Updated 9 months ago
- Plotting (entropy, varentropy) for small LMs☆97Updated last month
- The history files when recording human interaction while solving ARC tasks☆113Updated this week
- ComplexTensor: Machine Learning By Bridging Classical and Quantum Computation☆76Updated 7 months ago
- A graph visualization of attention☆56Updated last month
- MLX port for xjdr's entropix sampler (mimics jax implementation)☆64Updated 8 months ago
- Following master Karpathy with GPT-2 implementation and training, writing lots of comments cause I have memory of a goldfish☆173Updated 11 months ago
- smol models are fun too☆93Updated 8 months ago
- i will automate factorio☆106Updated 11 months ago
- ☆115Updated 6 months ago
- A tree-based prefix cache library that allows rapid creation of looms: hierarchal branching pathways of LLM generations.☆70Updated 5 months ago
- ☆97Updated last year
- SIMD quantization kernels☆73Updated last week
- could we make an ml stack in 100,000 lines of code?☆46Updated 11 months ago
- smolLM with Entropix sampler on pytorch☆150Updated 8 months ago
- Reasoning Computers. Lambda Calculus, Fully Differentiable. Also Neural Stacks, Queues, Arrays, Lists, Trees, and Latches.☆264Updated 8 months ago
- Modify Entropy Based Sampling to work with Mac Silicon via MLX☆49Updated 8 months ago
- in this repository, i'm going to implement increasingly complex llm inference optimizations☆63Updated last month
- papers.day☆91Updated last year
- a highly efficient compression algorithm for the n1 implant (neuralink's compression challenge)☆47Updated last year
- A reinforcement learning framework based on MLX.☆233Updated 4 months ago
- The open-source implementation of Q*, achieved in context as a zero-shot reprogramming of the attention mechanism. (synthetic data)☆1Updated 7 months ago
- Just large language models. Hackable, with as little abstraction as possible. Done for my own purposes, feel free to rip.☆44Updated last year
- an open source reproduction of NVIDIA's nGPT (Normalized Transformer with Representation Learning on the Hypersphere)☆101Updated 4 months ago
- noise_step: Training in 1.58b With No Gradient Memory☆220Updated 6 months ago
- ☆323Updated this week
- A really tiny autograd engine☆94Updated last month
- Solve puzzles to improve your tinygrad skills!☆135Updated 4 months ago
- ☆96Updated 7 months ago