w32zhong

llm.c

LLM training in simple, raw C/CUDA

w32zhong.github.io

BitDelta

blackmamba-fork

causal-conv1d

Causal depthwise conv1d in CUDA, with a PyTorch interface

cllm_fork

CS-Drafting

Cascade Speculative Drafting

DeepSeek-Math

EAGLE

EAGLE: Lossless Acceleration of LLM Decoding by Feature Extrapolation

filebrowser

📂 Web File Browser

laser

The Truth Is In There: Improving Reasoning in Language Models with Layer-Selective Rank Reduction

mamba

marker

Convert PDF to markdown quickly with high accuracy

MCSD

Multi-Candidate Speculative Decoding

Medusa

Medusa: Simple Framework for Accelerating LLM Generation with Multiple Decoding Heads

minbpe

Minimal, clean code for the Byte Pair Encoding (BPE) algorithm commonly used in LLM tokenization.

Ouroboros

Ouroboros: Speculative Decoding with Large Model Enhanced Drafting

Progressive-Hint

This is the official implementation of "Progressive-Hint Prompting Improves Reasoning in Large Language Models"

pytorch-that-I-successfully-built

Tensors and Dynamic neural networks in Python with strong GPU acceleration

rerope

Rectified Rotary Position Embeddings

search_with_lepton

Building a quick conversation-based search demo with Lepton AI.

SeeAct

SeeAct is a system for generalist web agents that autonomously carry out tasks on any given website, with a focus on large multimodal models (LMMs) such as GPT-4V(ision).

self-speculative-decoding

Code associated with the paper **Draft & Verify: Lossless Large Language Model Acceleration via Self-Speculative Decoding**

Sequoia

scalable and robust tree-based speculative decoding algorithm

tkblog

my blog.

toolformer-pytorch

Implementation of Toolformer, Language Models That Can Use Tools, by MetaAI

trl

Train transformer language models with reinforcement learning.

vivado-risc-v

Xilinx Vivado block designs for FPGA RISC-V SoC running Debian Linux distro

vllm

A high-throughput and memory-efficient inference and serving engine for LLMs

webarena

Code repo for "WebArena: A Realistic Web Environment for Building Autonomous Agents"