erfanzar / jax-flash-attn2
A flexible and efficient implementation of Flash Attention 2.0 for JAX, supporting multiple backends (GPU/TPU/CPU) and platforms (Triton/Pallas/JAX).
☆23Updated 3 weeks ago
Alternatives and similar repositories for jax-flash-attn2:
Users that are interested in jax-flash-attn2 are comparing it to the libraries listed below
- (EasyDel Former) is a utility library designed to simplify and enhance the development in JAX☆26Updated 2 weeks ago
- Minimal but scalable implementation of large language models in JAX☆34Updated 4 months ago
- Train a SmolLM-style llm on fineweb-edu in JAX/Flax with an assortment of optimizers.☆17Updated last week
- Parallel Associative Scan for Language Models☆18Updated last year
- ☆51Updated 10 months ago
- ☆20Updated last year
- ☆76Updated 8 months ago
- If it quacks like a tensor...☆57Updated 4 months ago
- supporting pytorch FSDP for optimizers☆79Updated 3 months ago
- Machine Learning eXperiment Utilities☆46Updated 9 months ago
- Triton Implementation of HyperAttention Algorithm☆47Updated last year
- ☆47Updated last year
- Custom triton kernels for training Karpathy's nanoGPT.☆18Updated 5 months ago
- ☆58Updated 3 years ago
- ☆87Updated last week
- A set of Python scripts that makes your experience on TPU better☆50Updated 8 months ago
- Pytorch/XLA SPMD Test code in Google TPU☆23Updated 11 months ago
- Engineering the state of RNN language models (Mamba, RWKV, etc.)☆32Updated 10 months ago
- Yet another random morning idea to be quickly tried and architecture shared if it works; to allow the transformer to pause for any amount…☆53Updated last year
- Large scale 4D parallelism pre-training for 🤗 transformers in Mixture of Experts *(still work in progress)*☆81Updated last year
- Inference code for LLaMA models in JAX☆116Updated 10 months ago
- Experiment of using Tangent to autodiff triton☆78Updated last year
- Accelerate, Optimize performance with streamlined training and serving options with JAX.☆257Updated last week
- Transformer with Mu-Parameterization, implemented in Jax/Flax. Supports FSDP on TPU pods.☆30Updated 3 months ago
- ☆30Updated 4 months ago
- some common Huggingface transformers in maximal update parametrization (µP)☆80Updated 3 years ago
- Simple and efficient pytorch-native transformer training and inference (batched)☆71Updated 11 months ago
- An implementation of the Llama architecture, to instruct and delight☆21Updated 2 months ago
- ☆19Updated 5 months ago
- Efficient PScan implementation in PyTorch☆16Updated last year