Havuz

kimi-k3-in-c, Kimi K3 modelinde inference çalıştırmayı hedefleyen bağımlılıksız bir C99 motorudur.

Neden yükseliyor · 57 günde 8.683 yıldız alması ve Kimi K3'ü 8 GB RAM'li CPU'da çalıştırma iddiası dikkat çekiyor; ancak son 28 gündeki 12 commit'in tek yazardan gelmesi, ilginin bağımsız teknik doğrulama veya geniş katkıcı tabanıyla desteklendiğini göstermiyor.

:sunrise: A comprehensive, cutting-edge open-source framework for next-generation Transformers, with compute-matched benchmarking and a plat

Spark-X2.5, yerel çalıştırma ve agent iş akışları için tasarlanmış 1.7B ve 4B boyutlarında açık dil modelleri serisidir.

Neden yükseliyor · Yeni model serisi; uzun context, küçük model boyutu ve agent/coding yeteneklerini aynı anda öne çıkarıyor. README'de 34 günlük yaş, 386 yıldız ve son 28 günde 9 commit ile 2 farklı yazar bilgisi var; bunlar erken ilgi ve güncel çalışma gösteriyor, ancak model performansını bağımsız olarak kanıtlamıyor.

A language model trained from scratch and running offline on an ESP32-S3 with a touchscreen chat interface.

A vLLM-inspired inference engine for looped transformers, with loop-level scheduling and execution across composable stages

Bilingual (中文+EN) ML / LLM / diffusion / agent interview cheat sheets for AI 秋招 — generated by ARIS /interview-cheatsheet, rendered by /rend

LLM inference with 7x longer context. Pure C, zero dependencies. Lossless KV cache compression + single-header library.

A typescript-go toolchain, with evidence graph

This repo is a alternative solution for standard self-attention. It uses tree routing with learnable functions for cheaper self-attention wh

从 MiniMind 源码读起,再延伸到现代大模型技术体系的中文学习笔记。主线逐行精读预训练 / SFT / DPO / PPO / GRPO 与训练机制;附录 17 篇进阶卷覆盖量化、投机解码、RLHF 全景、模型代际史等 MiniMind 没涉及、但进阶绕不开的主题。

Distributed 56M-parameter LLM inference across 3 ESP32-S3 boards via ESP-NOW , Split-PLE + KV cache, fully offline.

Full-stack open-source AI engine for building language models — tokenizer training, transformer architecture, cognitive reasoning and chat p

600 KB WASM runtime for Cactus Compute's Needle AI tool-calling models, Needle 3, 2 and 1 from one build. Browser, Cloudflare Workers, Node.

FSQ K-line foundation model for China A-share stock selection, with a containerised multi-factor research platform: data governance, factor

Stanford CS336《从头构建大语言模型》2026 中配版：18 篇中文技术博客 + 19 篇「图-字幕」原始文稿，基于 video2blog-skill 整理。

Build world models from scratch — from perception and latent dynamics to planning, JEPA, VLA, and 3D/4D intelligence.

Cooking AGI / AGI 大锅烩 · Bilingual (English / 中文) ML & LLM notes with worked examples, step-by-step figures and coding practice. Free to read

First open-source implementation of Google TurboQuant (ICLR 2026) -- near-optimal KV cache compression for LLM inference. 5x compression wit

Pyramis-L3D: hierarchical sparse attention research prototype (code only; weights on Hugging Face)

Run the native 284B-A13B DeepSeek-V4-Flash-0731 LLM locally on one laptop CPU: pure C, 8 GB RAM minimum, no GPU, best TPOT 0.892 s/token, re

Cross-architecture LLM internal observation database (23 models, 13 architecture families). Exposed as MCP tools for any AI coding agent.

TurboQuant KV cache compression plugin for vLLM — asymmetric K/V, 8 models validated, consumer GPUs

Implementation of LatentMoE,Toward Optimal Accuracy per FLOP and Parameter in Mixture of Experts (Elango et al., NVIDIA 2026) — in Pytorch.

a LLM inference engine to run on consumer hardware

A tiny language model that talks like a house cat named Miso.

Watch a real AI model think - live 3D visualization of a transformer's actual forward pass: tokens, embeddings, attention, neurons, sampling

End-to-end pipeline for seeing how LLMs actually process your prompts. Capture attention across every layer, render heatmaps and cooking cur

Building an Image-to-Video (I2V) Model from Scratch

Dynamic per-token early exit for LLM inference. Skip layers tokens don't need

Unified KV-cache compression for LLM inference: 12 Python-native methods, guarded add-on composition and routing, analytical capacity simula

A minimal RTL implementation of llama2.c stories260K forward inference on FPGA

3D explorable atlas of every neuron in a transformer LLM. Click neurons, trace residual highways, compare cross-architecture conservation. L

Transformer inference via unconventional compute: a single algebraic primitive, a tag-system scheduler, and Apple's Neural Engine through co

A compact, readable PyTorch reference implementation of the DeepSeek-V4 architecture (Flash and Pro), in the spirit of nanoGPT. ~4,300 LOC.