Erfan Zare Chavoshi

EasyDeL. Accelerate, Optimize performance with streamlined training and serving options with JAX.

Spectrax. SpecTrax is a JAX-native library for neural networks and graph learning, built for performance, composability and modularity.

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).

eformer. (EasyDel Former) is a utility library designed to simplify and enhance the development in JAX

ejkernel. easydel jax kernels writen in triton for gpus and pallas for tpus

Xerxes-Agents. Agents for intelligence and coordination

OST-OpenSourceTransformers. OST Collection: An AI-powered suite of models that predict the next word matches with remarkable accuracy (Text Generative Models). OST Collection is based on a novel approach to work as a full and intelligent NLP Model.

InstinctiveDiffuse. A cutting-edge text-to-image generator model that leverages state-of-the-art Stable Diffusion Model Type to produce high-quality, realistic images based on textual input.

EOpod. EOpod is a streamlined command execution tool designed to run and manage operations on Google Cloud Pods efficiently

Llama-Inference-JAX. Llama-inference-jax: Accelerated inference with Llama Models in JAX for high-speed, pure JAX implementation.

FJDiffusion. implementation of in Jax/Flax !

easymlx. Inference-only port of EasyDeL for Apple Silicon via MLX

eray. Ray-based distributed execution, scaling, and resource management for ML workloads on CPUs, GPUs, and TPUs.

Erutils. A powerful and versatile Python package that includes over 80 pre-built and customized neural networks for natural language processing, object detection, and many other applications. ErUtils also includes a rich set of utilities, such as file downloading and logging functions, that make it easy for developers to build AI models and applications

erfanzar. Config files for my GitHub profile.

dukron. a faster implementation of PSGD Kron second-order optimizer

levanter. Legible, Scalable, Reproducible Foundation Models with Named Tensors and Jax

psgd_jax. Implementation of PSGD optimizer in JAX

jax. Composable transformations of Python+NumPy programs: differentiate, vectorize, JIT to GPU/TPU, and more