Posts
Everything published
All posts
Project AI
Aeon ULTIMATE Fine Tune of Qwen Next Flash
Raising the compute for the Aeon ULTIMATE fine-tune of Qwen Next Flash: B300 and B200 hardware reserved for the whole run. Larger models need more VRAM and more time, and subscriptions help fund the hardware too.
The Aeon Forge$0 of $25,000Project AI
Aeon ULTIMATE Fine Tune of Qwen 4 27B
Raising the compute for the Aeon ULTIMATE fine-tune of Qwen 4 27B: B300 and B200 hardware reserved for the whole run. Larger models need more VRAM and more time, and subscriptions help fund the hardware too.
The Aeon Forge$0 of $2,000Project AI
Aeon ULTIMATE Fine Tune of MiniMax M3.1 Flash
Raising the compute for the Aeon ULTIMATE fine-tune of MiniMax M3.1 Flash: B300 and B200 hardware reserved for the whole run. Larger models need more VRAM and more time, and subscriptions help fund the hardware too.
The Aeon Forge$0 of $50,000Project AI
Aeon ULTIMATE Fine Tune of DeepSeek V4.1 Flash
Raising the compute for the Aeon ULTIMATE fine-tune of DeepSeek V4.1 Flash: B300 and B200 hardware reserved for the whole run. Larger models need more VRAM and more time, and subscriptions help fund the hardware too.
The Aeon Forge$0 of $300,000Post AI
UniFi network management for AI agents: safe keys, playbooks, backups
A safety-first UniFi skill for OpenClaw and Hermes Agent: keys in a 0600 env file outside the repository, ten playbooks, timestamped backups and a dry-run restore. How I let an agent read my network freely and write to it only on narrow, confirmed instruction.
The Aeon Forge7 min readPost AI
The AEON Forge catalog: every project and what to read first
A guided tour of the AEON Forge catalog: NVFP4 models for the DGX Spark, MLX builds for Apple Silicon, the voice stack, the media tools, the Orb and the Bench, with the article to read first in each group.
The Aeon Forge11 min read
Post AI
Run your own AI benchmark: AEON Bench and the Bench Pod
AEON Bench measures any open model's quality and real speed on your own hardware, signs the result and ranks it with its exact recipe. Why a stranger's score can be trusted, how to read the boards, and the five steps to a first run.
The Aeon Forge14 min readPost Software
Regex Builder: a simple page for building regular expressions
One HTML file that turns a value from your logs into a correctly escaped regular expression, with toggles for match type, case, URL encoding and negation. What the Regex Builder does, how to use it, and one lesson in reading a pattern before you deploy it.
The Aeon Forge5 min readPost AI
Qwen3-TTS server: streaming speech that starts in 0.4 seconds
An OpenAI-compatible Qwen3-TTS server for the DGX Spark that streams audio while the GPU is still generating: first audio in about 0.4 s, voice design and voice cloning loaded side by side, a voices library endpoint, and the version drift to check before you deploy.
The Aeon Forge8 min readPost AI
Qwen3-ASR server: transcription at 16 times real time on a DGX Spark
Qwen3-ASR on vLLM gives a DGX Spark a self-hosted, Whisper-compatible transcription endpoint for 30 languages and 22 Chinese dialects at about 16 times real time. Here is the pre-built image, the 0.6B and 1.7B variants, and the lessons from running it.
The Aeon Forge7 min readPost AI
Qwen 3.8 27B AEON Ultimate: the current flagship and how to serve it
Qwen3.8-27B AEON Ultimate Uncensored is the model I now serve on one Spark, two Sparks and the RTX box: what coherent abliteration removes and keeps, the NVFP4-MIXED weight lattice, the three seat recipes with their commands, and the rules that keep them running.
The Aeon Forge14 min readPost AI
Serving Qwen 3.6 35B-A3B heretic at NVFP4 with DFlash on a DGX Spark
How I serve an uncensored Qwen 3.6 35B-A3B at NVFP4 with DFlash speculative decoding on one DGX Spark: the seven container fixes, the mixture-of-experts angle, the drafter, the measured numbers and the exact command that works today.
The Aeon Forge11 min readPost AI
Qwen 3.6 27B AEON Ultimate on Apple Silicon: native MLX quants
Two native MLX builds of Qwen 3.6 27B AEON Ultimate for Apple Silicon: which Mac fits the 29.5 GB 8-bit or the 16 GB FP4, why the SSM, vision tower and MTP head stay BF16, the 1.78x lossless self-speculation, and a uv quick start.
The Aeon Forge9 min readPost Software
Presto Cockpit turns a Pimoroni Presto into a home-lab dashboard
A Pimoroni Presto touchscreen as a desk-side home-lab dashboard: ten swipeable screens, two optional backends for a GPU box and an agent gateway, and rear LEDs that match the screen. The parts, the wiring, the setup, and what broke.
The Aeon Forge6 min readPost AI
Ornith 1.0 35B AEON Ultimate keeps the coding and drops the refusals
I abliterated DeepReinforce's Ornith 1.0 35B, an agentic coding MoE, and kept the coder: 0 of 80 refusals, pass@1 identical to the base, a three-way BF16, NVFP4 and DFlash benchmark on a DGX Spark, and the quick start to serve it.
The Aeon Forge10 min readPost AI
Nemotron 3 Nano Omni AEON Ultimate, served uncensored on a DGX Spark
An abliterated Nemotron 3 Nano Omni that reasons over text, images and audio on one DGX Spark. How the twelve-dimension refusal removal was found, what NVFP4 costs against BF16, what the source-built vLLM image patches, and how to deploy it.
The Aeon Forge9 min readPost AI
modelopt-fast-moe cuts AWQ calibration time for MoE models at NVFP4
modelopt-fast-moe replaces the one-sample-at-a-time calibration loop that leaves a GPU mostly idle while nvidia-modelopt quantizes a Mixture-of-Experts model to NVFP4. What broke on a 26B MoE, what the batched loop fixes, and what it still leaves to you.
The Aeon Forge6 min readPost AI
Matrix VoIP agent: call your AI, and let it see through your camera
A headless Matrix VoIP agent that answers WebRTC calls from any Matrix client and talks back through local speech recognition, an LLM and streaming TTS, about 1.0 s to first speech on one DGX Spark, plus an optional camera look tool that keeps every frame in RAM.
The Aeon Forge9 min readPost AI
Magick Orb Docker: the headless Orb for any server
One docker run puts the headless Magick Orb on an x86 server, a Mac or a DGX Spark: Model Share over IPFS, the Agent Dashboard, MCP and REST, with no Raspberry Pi. What the container lacks versus the Pi, and why port 4001 matters.
The Aeon Forge6 min readPost AI
Gemma 4 31B DECKARD HERETIC at NVFP4 with DFlash on a DGX Spark
A dense, uncensored Gemma 4 31B with thinking, quantized to NVFP4 and served through a DFlash drafter on one DGX Spark. Single-stream decode went from 11 to 39 tokens per second, 32 concurrent requests reach 427, and reasoning and tool calling stay intact.
The Aeon Forge11 min readPost AI
Gemma 4 26B-A4B at NVFP4: a mixture-of-experts recipe for a DGX Spark
How I serve Gemma 4 26B-A4B, a mixture-of-experts model at NVFP4, on one DGX Spark with a DFlash drafter: the production recipe, the bench recipe behind 816 tok/s at c64, and the drafter defects that made my earlier numbers wrong.
The Aeon Forge10 min read