Autonomous self-improving 4x DGX Spark (GB10) MoA stack + LoRA loop (DSV4F router, Qwen3.6/Omni/TwoTower/Gemma). Hermes MoA routing, ~90% local, self-LoRA. Includes a SINGLE-SPARK recipe (all-Gemma-4: 31B brain + DiffusionGemma + 12B) to build the same on 1 GB10.
☆15Jul 5, 2026Updated last week
Alternatives and similar repositories for Keys-Setup-Autonomous-Self-Improving-Local-Inference-Stack
Users that are interested in Keys-Setup-Autonomous-Self-Improving-Local-Inference-Stack are comparing it to the libraries listed below. We may earn a commission when you buy through links labeled 'Ad' on this page.
Sorting:
- MiniMax-M3 (428B, no pruning) at 36 tok/s on 2× NVIDIA DGX Spark — W4A16 GPTQ + NVFP4 KV + EAGLE-3 speculative decoding on vLLM. Three se…☆26Updated this week
- MiMo-V2.5 NVFP4 4-bit weights + NVFP4 4-bit KV cache + DFlash speculative decoding on 2× NVIDIA DGX Spark — 1M context, 3.4M-token KV poo…☆17Updated this week
- Recipe: GLM-5.2 (unpruned QuantTrio Int4-Int8Mix) at 200K ctx with MTP spec decode on a 4x NVIDIA DGX Spark (GB10) cluster☆51Updated this week
- DeepSeek V4 Flash @ 1M token context on 2x NVIDIA DGX Spark — production-tested recipe (45 tok/s decode, real 800K prompts served)☆34Updated this week
- AEON vLLM Ultimate — vLLM 0.25.0 built from source for DGX Spark / Blackwell (sm_121a/GB10). One image serves the whole AEON fleet (Gemma…☆71Updated this week
- GPU virtual machines on DigitalOcean Gradient AI • AdGet to production fast with high-performance AMD and NVIDIA GPUs you can spin up in seconds. The definition of operational simplicity.
- Working recipe to serve DeepSeek-V4-Flash across two NVIDIA DGX Spark (GB10) nodes with vLLM (TP=2, FP8 KV, MTP) over a RoCE/RDMA link — …☆51Updated this week
- Run on TWO-DGX-Spark - vLLm-0.24.0 dual cache optimized DSV4F+DSpark+NVFP4 KV (Concurrency 12 with 1.5M context/3M KV token Pool) >0.58-0…☆19Jul 7, 2026Updated last week
- Recipe: GLM-5.2 (unpruned QuantTrio Int4-Int8Mix) at 655,360-token context + MTP spec decode (DCP4, fp8_ds_mla) on a 4x NVIDIA DGX Spark …☆22Updated this week
- Give any blind LLM eyes — on any machine — in one shot. Tiny local VLM (Qwen3.5-0.8B) as an OpenAI-compatible vision service. Mac (MLX) /…☆24Jul 7, 2026Updated last week
- Deploy DeepSeek V4 Flash (MoE reasoning model) on dual DGX Spark nodes with 1M token context, InfiniBand, and FP8 KV-cache☆77Updated this week
- Run the AEON Bench suite on your own hardware: verified HuggingFace pull → serve → benchmark (text · agentic ×3 harnesses · vision · audi…☆20Updated this week
- Local diagnostic CLI for NVIDIA DGX Spark (GB10). Detects power caps, UMA pressure, thermal risk, CUDA 13/SM_121 wheel mismatches, Docker…☆90Updated this week
- DFlash vLLM for DGX Spark — Plug & Play Block-Diffusion Speculative Decoding☆52Jun 28, 2026Updated 2 weeks ago
- How much experts do we need to serve a model?☆153Mar 18, 2026Updated 3 months ago
- Deploy to Railway using AI coding agents - Free Credits Offer • AdUse Claude Code, Codex, OpenCode, and more. Autonomous software development now has the infrastructure to match with Railway.
- Mixed-capability LLM benchmark for DGX Spark — 57 scenarios, 10 domains, partial-credit grading, trial statistics☆128Updated this week
- Open Source Desktop App to help with Previs for AI-native filmmaking — stage grey-box scenes, choreograph camera & cast with marks, expor…☆79Updated this week
- ☆65Updated this week
- Git-backed long-term memory for AI coding agents☆73May 25, 2026Updated last month
- my version of the models dev site which is way faster and has the filters I want☆18Apr 22, 2026Updated 2 months ago
- ☆12Feb 19, 2017Updated 9 years ago
- Perplexity AI search, reasoning, research, and compute - MCP server, dashboard, and multi-IDE auto-config for VS Code.☆59Updated this week
- Interface boards for the Neato LDS - the Lidar sensor of the XV-11.☆14Jun 21, 2014Updated 12 years ago
- Swift package for reading and writing Safetensors files.☆13Feb 6, 2026Updated 5 months ago
- Managed hosting for WordPress and PHP on Cloudways • AdManaged hosting for WordPress, Magento, Laravel, or PHP apps, on multiple cloud providers. Deploy in minutes on Cloudways by DigitalOcean.
- Biologically-inspired persistent memory engine for Claude Code. 26 cognitive subsystems, Hopfield networks, predictive coding, causal dis…☆61Apr 1, 2026Updated 3 months ago
- Qwen3.6-35B-A3B-heretic NVFP4 + DFlash speculative decoding on DGX Spark (GB10/sm_121a). Source-built vLLM image + 7 patches + comprehens…☆118Jun 28, 2026Updated 2 weeks ago
- A self-evolving agent in rust with one tool that only uses skills☆22Jun 11, 2026Updated last month
- DeepSeek V4 Flash DSpark Abliterated (Uncensored) — 2× DGX Spark serving☆19Updated this week
- Interactive chat application that enables users to have conversations with any website's content using Groq's fast inference capabilities…☆25Sep 25, 2025Updated 9 months ago
- Local generative video panel for Apple Silicon. Wraps LTX-2 MLX, joint audio+video, one-click Pinokio install.☆94Updated this week
- DGX Spark / GB10 vLLM image for Gemma 4 31B Deckard Heretic Uncensored NVFP4 with z-lab DFlash speculative decoding.☆32May 15, 2026Updated last month
- ☆68Jun 4, 2026Updated last month
- ☆16Sep 5, 2021Updated 4 years ago
- Deploy to Railway using AI coding agents - Free Credits Offer • AdUse Claude Code, Codex, OpenCode, and more. Autonomous software development now has the infrastructure to match with Railway.
- Serial buffer for Particle Photon and Electron☆13Nov 2, 2024Updated last year
- LLM-managed personal knowledge base. Auto-extracts from git repos, Claude Code sessions (via ccrider FTS5), and iMessage bookmarks. Cross…☆26Updated this week
- A curated collection of production-ready Hermes Agent skills — brainstorming, PRD workflows, debugging, Apple integrations, MLOps, docume…☆71May 2, 2026Updated 2 months ago
- If I had my druthers....a programming language made during Khan Academy's 2014 Hack Week.☆15Jan 26, 2015Updated 11 years ago
- sparkrun - launch, manage, and stop LLM inference workloads on NVIDIA DGX Spark systems☆383Updated this week
- Project Doxa is an advanced, autonomous agent-based simulation engine where AI agents act as citizens of evolving civilizations.☆27Jun 14, 2026Updated last month
- AseeVR fix for Droolon Pi1☆14Mar 16, 2024Updated 2 years ago