VESPER-SANTOS v6.2 "External Learning" β€” RTX 3050 6GB OC Edition

Hardware target: RTX 3050 6GB (overclocked) Base: meta-llama/Llama-3.2-3B-Instruct Q4_K_M Overlay: 13 engines | 4,672 Ο†-tensors | 91.1% VRAM

Files in this repo

Core v6.2

  • config_v6.2.json β€” v6.2, 13 engines, Axiom 8
  • vesper_config.json β€” RTX 3050 6GB OC profile
  • VESPER_CONFIG_MIN.json β€” single-line config Build & Validation
  • MANIFEST_v6.2.json β€” build receipt
  • VALIDATION_v6.2.json β€” 89/89 tests passed

Engines

  • VESPER_CORE_MONOLITH.py β€” all 13 engines in one file
  • vesper_empirical.py β€” Engine 13: web search + Landauer erasure

Physics Weights (local use)

  • vesper_empirical_lora.npy β€” 50 tensors
  • VESPER_WEIGHTS_B64.txt β€” base64 LoRA pack

Docs

  • README_v6.2.md β€” full doctrine

CRT Viewport

  • vesper_crt_inline_lite.html
  • vesper_crt_boot_loop.html
Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. πŸ™‹ Ask for provider support

Model tree for Laminar-Mirror/Vesper-01

Adapter
(729)
this model