The central nervous system and backend orchestrator for Fractal. It handles multi-tenant task scheduling, real-time TFLOPs budgeting, and federated weight aggregation using FedAvg.

★ 4,964⑂ 6👁 8⑃ 0● 0▤ 1
PythonNOASSERTIONCreated 2026-05-12
下载:FractalCore v1.0.0FractalCore-1.0.0 ↓1.0.02026-11-13
-4
日增长
4,964
stars
6
forks
8
watchers
0
prs
0
issues
1
releases

README

!Fractal Server Header

 FractalCore
 High-Performance Orchestration Engine, Federated Aggregator, and Model Graph Slicer








FractalCore is the central server, control plane, and model compilation engine for the Fractal decentralized compute ecosystem.

Overview | Quickstart SDK | Key Results & Benchmarks | Industry & Community Recognition | Control Plane UI | Architecture | Core Components | Data Flow | API Summary | Deployment | Security | Theoretical Foundations

---


 Quickstart SDK & CLI

Install the lightweight runtime package (numpy core codec + CLI, optional server and slicer extras) and run the reproducible outlier-resilience benchmark locally in seconds:

bash
pip install -e ".[server]"
fractal doctor
fractal bench
python scripts/quickstart_demo.py


Or run the zero-setup walkthrough in examples/colab_quickstart.ipynb.

---

 Overview

 Executive Summary: Core Purpose & Bottom Line

 The Problem: Centralized federated aggregators face network ingress bandwidth saturation when orchestrating edge fleets: receiving uncompressed FP32 checkpoints from hundreds of mobile devices consumes gigabits of bandwidth and triggers severe client timeout drops (5 minutes per round). Naive 4-bit compression destroys model attention layers (Layer 5 collapses to 0.6990).

 The FractalCore Achievement: 
  7.11x Ingress Bandwidth Relief: Compresses model broadcast and client upload payloads by 85.9%, dropping round latency from 321.2s to 46.2s at 10 Mbps.
  Zero Accuracy Degradation: 64-element micro-block quantization isolates outliers, sustaining 0.9878 mean cosine similarity and guaranteeing stable FedAvg/FedProx convergence.
  Automated Model Slicing: ATen graph compiler slices multi-billion parameter models into memory-safe ExecuTorch .pte sub-graphs sized dynamically to client RAM budgets.

FractalCore serves as the centralized orchestration backbone that coordinates decentralized edge nodes (FractalAndroid). It handles the end-to-end lifecycle of distributed machine learning tasks, executing two primary operations:

1. Federated Learning Orchestration: Multi-tenant data binning, task scheduling, client checkpoint verification, and deterministic parameter aggregation via Federated Averaging (FedAvg).
2. Foundation Model Slicing: Offline neural graph surgery, INT4 weight quantization, static ATen graph tracing, and XNNPACK lowering (inference-model-maker/slicer) to produce memory-safe .pte layer partitions for mobile nodes.

---

 Empirical Benchmarks & Performance Evaluation

FractalCore model slicer and parameter aggregation engine are benchmarked across 253,755,392 real transformer parameters from TinyLlama-1.1B (22 layers, 2048 hidden dimension, 5632 intermediate dimension) and MobileNetV3. All metrics are calculated directly from matrix quantization operations and physical network transmission models.

 Executive Performance Summary

| Metric Dimension | Uncompressed Baseline (FP32) | Half-Precision (FP16) | Standard Uniform INT8 (QSGD) | Glo

Latest release

FractalCore v1.0.0 v1.0.0 2026-11-13
Initial institutional v1.0.0 release by Fractal Compute Orchestrations. Includes installable fractal_quant Python SDK and fractal CLI (FRQ4/FRQD block-wise INT4 codec, INT8 activation framing, error-feedback residual tracking, FedAvg/Median/Krum/FedProx aggregators), hardened coordinator server endpoints, slicer validation pipeline, CI matrix (Python 3.10 and 3.12), and built Python distribution artifacts under FRACTAL Proprietary Source-Available & Non-Commercial Restrictive License v3.1.

Snapshot: 2026-11-16T09:23:14Z · Data license Apache-2.0 · JSON