Eternal Reclaimer's Projects
32 times longer context window than vanilla Transformers and up to 4 times longer than memory efficient Transformers.
Blockwise Parallel Transformer for Long Context Large Models
Omni-Modality Processing, Understanding, and Generation
An open source implementation of "Scaling Autoregressive Multi-Modal Models: Pretraining and Instruction Tuning", an all-new multi modal AI that uses just a decoder to generate both text and images
Conditional Flash Attention Fusing Flash Attention and CoLT5 architecture
Plug in and Play Prompt Technique to Boost Model reasoning by 40%
An plug in and play suite of embodied multi-modal robotic transformers for Robotics
A dark flat theme for color scheme for iterm2, designed for Powerlevel9k and oh-my-zsh
Agora's Data Room for investors
A PyTorch implementation of the Decoupled LionW optimizer, an adaptive optimization algorithm with customizable learning rates and weight decay. This optimizer combines gradient updates and momentum while providing metrics such as cosine similarity and l2 norm for analysis
This repository contains implementations and illustrative code to accompany DeepMind publications
PyTorch code and models for the DINOv2 self-supervised learning method.
Data and tools for generating and inspecting OLMo pre-training data.
My implementation of "Dual-Stream Diffusion Net "
The repository for the largest and most comprehensive empirical study of visual foundation models for Embodied AI (EAI).
A powerhouse of multi-modal AI tools designed to understand, analyze, and predict environmental changes. This open-source repository leverages text, images, and structured data, combining insights from each modality to offer a comprehensive understanding of our environment 🌲 🌳 🌏
Unleash the full potential of exascale LLMs on consumer-class GPUs, proven by extensive benchmarks, with no long-term adjustments and minimal learning curve.
An EXA-Scale repository of Multi-Modality AI resources from papers and models, to foundational libraries!
A simple package for leveraging Falcon 180B and the HF ecosystem's tools, including training/inference scripts, safetensors, integrations with bitsandbytes, PEFT, GPTQ, assisted generation, RoPE scaling support, and rich generation parameters.
Finetune any model on HF in less than 30 seconds
Flamingo-with-lion
Implementation of Flash Attention in Jax
Fast and memory-efficient exact attention coupled with the LION Optimizer for ultra fast performance
Get down and dirty with FlashAttention2.0 in pytorch, plug in and play no complex CUDA kernels
Triton implementation of Flash Attention2.0
FlashAttention2.0 with Lora
An simple pytorch implementation of Flash MultiHead Attention
A forest of autonomous agents.
MPT-30B With Multi-Modal Adapters
Plug in and play Implementation of "A Generalist Agent" by Deepmind.