samsja Claim this profile

Agentic RL Training at Scale

(Core maintainer) 🧬 Represent, send, and store multimodal data · Neural Search · Vector Search · Document Store

Pytorch like framework in rust: Autograd with dynamic graphs for deep learning

Manage ML configuration with pydantic

Muon fsdp 2

Belote libre: Implementation of a open source version of the french game Belote. The repo contains the code for the backend server, the frontend website and the reinforcement learning models for the IA in game

cmd-k converts written instruction into executable shell commands, cursor command k that works from the terminal.

Config files for k8s cluster of raspberry pi

No description.

No description.

LLaMa implementation fork from lit-llama fully typed with jaxtyping

repo for github page files

Fast inference of LLM on cpu written in rust

GPU implementation of the dimension reduction algorithm Empirical Interpolation Method (EIM) in pytorch

Code for the paper : Improving traffic sign recognition by active search

some flash attention pre-build wheel

prime is a framework for efficient, globally distributed training of AI models over the internet.

bigger context window, less weight = good local model

Langchain-based Agent utilizing OpenAI Functions to help you with a couple of Git commands

No description.

some utils script

Pre-built wheels for flash-attn, deep-ep, deep-gemm, nixl and other CUDA packages

An open-source database of AI models.

No description.

Verifiers for LLM Reinforcement Learning

Fault tolerance for PyTorch (HSDP, LocalSGD, DiLoCo, Streaming DiLoCo)

Serve Open Flamingo via gRPC

No description.

My dotfiles for my manjaro i3 linux configuration,

My implementation of transformers related papers for computer vision in pytorch

No description.

No description.

No description.

agent multiplexer that lives in your terminal.

AI agent toolkit: coding agent CLI, unified LLM API, TUI & web UI libraries, Slack bot, vLLM pods

A high-performance and light-weight router for vLLM large scale deployment

A Datacenter Scale Distributed Inference Serving Framework

Dion optimizer algorithm

A native PyTorch Library for large model training

No description.

No description.

Real-time terminal monitor for InfiniBand networks - htop for high-speed interconnects

🤗 Transformers: State-of-the-art Machine Learning for Pytorch, TensorFlow, and JAX.

No description.

torch file for trainer llm, simple but not simple enough

No description.

An extremely fast Python package and project manager, written in Rust.

An Easy-to-use, Scalable and High-performance RLHF Framework (70B+ PPO Full Tuning & Iterative DPO & LoRA & RingAttention & RFT)

Minimal hackable GRPO implementation

No description.

Tensors and Dynamic neural networks in Python with strong GPU acceleration

No description.

8-bit CUDA functions for PyTorch

A JAX research toolkit for building, editing, and visualizing neural networks.

No description.

my sandbox

Self-hosted AI coding assistant

Model and training recipe for image retrieval. Both selfsupervised and supervised learning

No description.

🤗 PEFT: State-of-the-art Parameter-Efficient Fine-Tuning.

A minimal PyTorch re-implementation of the OpenAI GPT (Generative Pretrained Transformer) training

ndarray: an N-dimensional array with array views, multidimensional slicing, and efficient operations

No description.

No description.

No description.

FORK of PyTorch implementation of SimSiam https//arxiv.org/abs/2011.10566

Runtime inspection utilities for Python typing module

No description.

No description.

A latent text-to-image diffusion model

My blog_post

Mono repo for Belote libre v2: Implementation of a open source version of the french game Belote.

Masked Siamese Networks for Label-Efficient Learning (https://arxiv.org/abs/2204.07141)

Small example of how to serve python code with a rust server

contains all the docker-compose stacks I require for my testing and development work

No description.

Yet An Other Rust Lox implementation

scikit-learn: machine learning in Python

Mono repo for findmybike.com

Run `black` on python code blocks in documentation files

Cloud-native neural search framework for 𝙖𝙣𝙮 kind of data

easily manage wallpaper for i3wm

This is the repository of the book "SmartPoop".

No description.

A Julia clone of https://pypi.python.org/pypi/tqdm

My personnal boilerplate for pyorch project, include jupyterlab dev inside Docker, precommit hook, jupytext sync

No description.

No description.

No description.