2x DGX Spark (GB10) running DeepSeek-V4-Flash: dual-node FP8 TP=2 (~41 tok/s single / ~350 agg), single-node ds4, a 10-model GGUF benchmark suite, and a cross-machine comparison (Mac M2 Ultra, RTX PRO 6000). Settings, recipes, gotchas & benchmarks.
☆30Jun 16, 2026Updated 4 months ago
Alternatives and similar repositories for dgx-spark-deepseek-v4-flash
Users that are interested in dgx-spark-deepseek-v4-flash are comparing it to the libraries listed below. We may earn a commission when you buy through links labeled 'Ad' on this page.
Sorting:
- vLLM + Qwen3.6-27B (BF16) OpenAI-compatible inference server on AMD Strix Halo (Ryzen AI Max+ 395, gfx1151). Vision input, 256K context, …☆19Apr 26, 2026Updated 6 months ago
- What I run on my inference server: Qwen3.8-Flash-Next on two RTX 3090s, and how I got here☆18Oct 12, 2026Updated 3 weeks ago
- Unofficial implementation of HyperFlow for MiniMax H3 in ComfyUI. 8-step speedup with minimal quality loss☆18Oct 1, 2026Updated last month
- Home-enthusiast's guide to fine-tuning 27B+ LLMs on AMD Strix Halo (gfx1151, Ryzen AI MAX+ 395) — the patches and tuning to make Linux ma…☆32Jun 9, 2026Updated 5 months ago
- Live per-node GPU, host, and vLLM stats for DGX Spark clusters, in your terminal☆17Nov 2, 2026Updated last week
- 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.
- Qwen3.6 27B × DFlash — 30-35 tok/s on NVIDIA DGX Spark (GB10) - LLama.Cpp☆17May 2, 2026Updated 6 months ago
- ☆16Feb 3, 2026Updated 9 months ago
- Comparative forensic analysis of LLM abliteration techniques☆63Jul 25, 2026Updated 3 months ago
- ☆12Jul 24, 2019Updated 7 years ago
- Orange Pi 6 Plus GPU bring-up on mainline/Armbian (SCMI ACPI/SMC, resets, Panthor)☆18Oct 30, 2026Updated last week
- Fully local, persona-driven AI companion bot for Telegram (Telethon) with optional Discord bridge. Modular text pipeline, per-user memory…☆23Jun 21, 2026Updated 4 months ago
- MiniMax M2 inference server for NVIDIA DGX Spark☆17Jan 24, 2026Updated 9 months ago
- Model-agnostic LLM benchmarking + tuning in one command. A model is a 15-key profile file. Built in production on a 4x NVIDIA DGX Spark c…☆10Sep 16, 2026Updated last month
- LLM Combinator exports real-time results of multi-modal Ollama executions with input from Frigate cameras and HA sensors into HomeAssista…☆14Nov 10, 2024Updated 2 years ago
- Deploy on Railway without the complexity - Free Credits Offer • AdConnect your repo and Railway handles the rest with instant previews. Quickly provision container image services, databases, and storage volumes.
- Constructive Solid Geometry (CSG) and Quick Hull Library for golang☆10Oct 30, 2019Updated 7 years ago
- Pong in Haskell / Clash, running as software using SDL and as hardware targeting FPGAs☆10Apr 20, 2024Updated 2 years ago
- Easy audit capabilities for EF projects.☆13Oct 24, 2017Updated 9 years ago
- keys- 1M context: Inkling-Small-NVFP4 + DSpark + NVFP4 KV on SGLang, sm_121a-optimized on Two DGX Sparks — one-shot agent-executable rec…☆10Aug 17, 2026Updated 2 months ago
- ☆12Apr 17, 2025Updated last year
- Script to normalize and validate STL files so that they play better with git version control.☆15Mar 15, 2021Updated 5 years ago
- Experimental runtime reclamation of GB10 display-reserved memory into ordinary Linux RAM, with guarded driver, local signing, and reboot …☆13Oct 23, 2026Updated 2 weeks ago
- ☆16Sep 1, 2025Updated last year
- Qwen3.8 Flash Next AWQ INT4 4x3090 v1: 262K, Vision, FP8 QSA KV and CUDA Graphs on 4x RTX 3090☆16Aug 30, 2026Updated 2 months ago
- Deploy on Railway without the complexity - Free Credits Offer • AdConnect your repo and Railway handles the rest with instant previews. Quickly provision container image services, databases, and storage volumes.
- A tiny static site generator, greatly inspired by Zola.☆14Aug 8, 2022Updated 4 years ago
- ☆50Oct 21, 2026Updated 3 weeks ago
- Qwen3.8-27B via vLLM on a single AMD Radeon AI PRO R9700 (gfx1201, 32 GiB) — bare, script-free setup guide, written to be executed by an …☆13Sep 3, 2026Updated 2 months ago
- The core18 base snap☆14Updated this week
- Trackpoint emulator: 3 choc swithes and analog stick☆15Mar 23, 2025Updated last year
- Truthful local registry and event stream for running coding agents☆12Updated this week
- SGLang 0.5.21 for Qwen3.8-Flash-Next on one RTX PRO 6000: 264.5 tok/s decode, W4A16 routed experts (KLD 0.0111 vs BF16), 262K context, ex…☆18Oct 26, 2026Updated 2 weeks ago
- 24 keys ought to be enough for anybody☆15May 28, 2024Updated 2 years ago
- GLM-5.3-Flash with 152 of 288 experts per layer (RCO), for coding and agentic use on a home PC: recipe, mask, patches, evaluation harness☆14Oct 31, 2026Updated last week
- Proton VPN Special Offer - Get 70% off • AdSpecial partner offer. Trusted by over 100 million users worldwide. Tested, Approved and Recommended by Experts.
- Claude Code skill: let your agent drive fleets of local omp (oh-my-pi) model instances toward one goal — provider config, bounded turns, …☆15Aug 28, 2026Updated 2 months ago
- Qwen 3.8 125B-A6B CUDA inference speedup benchmark: ds4 engine, overlay, and submission harness☆15Oct 10, 2026Updated last month
- Universal embedding-space translation library. Plug-and-play adapters that map one embedding model’s vector space into another — locally …☆34Apr 14, 2026Updated 6 months ago
- ffcache(Flutter File Cache) is a file based key value store for caching.☆14May 5, 2026Updated 6 months ago
- Claude Code skill: Build vs Buy advisor that searches for existing solutions before vibe coding☆14Dec 31, 2025Updated 10 months ago
- A Model Context Protocol (MCP) server for the NextDNS API, built with FastMCP and generated from OpenAPI specifications.☆15Updated this week
- Generate architecture diagrams by docker compose files.☆14Feb 22, 2025Updated last year
Are these results useful?