KhoomeiK / complexity-scalingLinks
gzip Predicts Data-dependent Scaling Laws
โ34Updated last year
Alternatives and similar repositories for complexity-scaling
Users that are interested in complexity-scaling are comparing it to the libraries listed below
Sorting:
- โ28Updated last year
- A MAD laboratory to improve AI architecture designs ๐งชโ131Updated 10 months ago
- โ142Updated last month
- Scaling is a distributed training library and installable dependency designed to scale up neural networks, with a dedicated module for trโฆโ64Updated 3 weeks ago
- โ61Updated last year
- some common Huggingface transformers in maximal update parametrization (ยตP)โ86Updated 3 years ago
- โ53Updated last year
- The simplest, fastest repository for training/finetuning medium-sized GPTs.โ168Updated 4 months ago
- Experiments for efforts to train a new and improved t5โ75Updated last year
- โ81Updated last year
- Learning Universal Predictorsโ79Updated last year
- Extract full next-token probabilities via language model APIsโ247Updated last year
- โ91Updated last year
- nanoGPT-like codebase for LLM trainingโ109Updated 5 months ago
- Simple Transformer in Jaxโ139Updated last year
- An introduction to LLM Samplingโ79Updated 10 months ago
- Code to reproduce "Transformers Can Do Arithmetic with the Right Embeddings", McLeish et al (NeurIPS 2024)โ193Updated last year
- Automatically take good care of your preemptible TPUsโ37Updated 2 years ago
- A set of Python scripts that makes your experience on TPU betterโ54Updated last month
- Understanding how features learned by neural networks evolve throughout trainingโ39Updated last year
- โ53Updated last year
- train with kittens!โ63Updated last year
- โ69Updated last year
- Minimal (400 LOC) implementation Maximum (multi-node, FSDP) GPT trainingโ132Updated last year
- Code implementing "Efficient Parallelization of a Ubiquitious Sequential Computation" (Heinsen, 2023)โ95Updated 10 months ago
- Understand and test language model architectures on synthetic tasks.โ233Updated last month
- โ22Updated 2 years ago
- โ58Updated last year
- Notebooks accompanying Anthropic's "Toy Models of Superposition" paperโ129Updated 3 years ago
- Yet another random morning idea to be quickly tried and architecture shared if it works; to allow the transformer to pause for any amountโฆโ52Updated 2 years ago