PolymathicAI / xVal
Repository for code used in the xVal paper
โ110Updated 5 months ago
Related projects: โ
- A MAD laboratory to improve AI architecture designs ๐งชโ84Updated 4 months ago
- Code to reproduce "Transformers Can Do Arithmetic with the Right Embeddings", McLeish et al (2024)โ169Updated 3 months ago
- โ73Updated 5 months ago
- โ53Updated 8 months ago
- Explorations into the proposal from the paper "Grokfast, Accelerated Grokking by Amplifying Slow Gradients"โ82Updated 3 weeks ago
- ฯ-GPT: A New Approach to Autoregressive Modelsโ53Updated last month
- Understand and test language model architectures on synthetic tasks.โ156Updated 4 months ago
- Tree Attention: Topology-aware Decoding for Long-Context Attention on GPU clustersโ94Updated 2 weeks ago
- Scalable neural net training via automatic normalization in the modular norm.โ108Updated 3 weeks ago
- Explorations into the recently proposed Taylor Series Linear Attentionโ85Updated last month
- Latent Diffusion Language Modelsโ66Updated last year
- Experiment of using Tangent to autodiff tritonโ66Updated 7 months ago
- Minimal (400 LOC) implementation Maximum (multi-node, FSDP) GPT trainingโ110Updated 5 months ago
- Code for exploring Based models from "Simple linear attention language models balance the recall-throughput tradeoff"โ206Updated last month
- Diffusion models in PyTorchโ81Updated last week
- โ42Updated 3 months ago
- Unofficial but Efficient Implementation of "Mamba: Linear-Time Sequence Modeling with Selective State Spaces" in JAXโ74Updated 7 months ago
- some common Huggingface transformers in maximal update parametrization (ยตP)โ76Updated 2 years ago
- 94% on CIFAR-10 in 3.09 seconds ๐จ 96% in 27 secondsโ127Updated last month
- Yet another random morning idea to be quickly tried and architecture shared if it works; to allow the transformer to pause for any amountโฆโ50Updated 10 months ago
- โ35Updated 5 months ago
- โ55Updated 9 months ago
- Collection of autoregressive model implementationโ62Updated 2 weeks ago
- โ42Updated this week
- PyTorch implementation of models from the Zamba2 series.โ63Updated last month
- โ38Updated 8 months ago
- Transformer with Mu-Parameterization, implemented in Jax/Flax. Supports FSDP on TPU pods.โ29Updated 3 weeks ago
- โ160Updated last year
- Implementation of GateLoop Transformer in Pytorch and Jaxโ86Updated 3 months ago
- A fast implementation of T5/UL2 in PyTorch using Flash Attentionโ60Updated this week