I tweaked my own version of @ostrisai ai-toolkit this morning to include it for training testing, but im getting a lot of burned in sensitivity filters even with the json prompts for even the most basic of images - the standard sample set.
Will be monitoring to see how the issue progresses.
Mind you I didn't do mflux... Maybe I'll try that.
Today we're shipping our biggest MLX-VLM release yet: v0.6.0
...and we are raising ๐ธ
This one's about turning your Apple devices into real local agent machines. From your desk to your pocket.
What's new:
โก Speculative decoding everywhere โ Gemma 4 EAGLE3 + DFlash, Qwen MTP, DeepSeek V4 MTP. Faster tokens, less waiting.
๐ค Agent-ready server โ native Anthropic /v1/messages API, stateful /v1/responses, tool calls, Codex context budgets. Plug Claude Code & Codex straight into local models.
๐๏ธ New models galore โ DeepSeek V4, ZAYA1-VL, MiniCPM-V 4.6, LFM2 MoE, Step-3.7 Flash, Laguna + more.
๐จ Image gen & editing โ FLUX.2 (base + klein), PrismML Bonsai.
๐ Audio in โ Qwen3 Omni, Gemma 4 audio, base64 chat audio.
๐งฎ TurboQuant KV cache โ RHT-correct fast paths for leaner memory.
๐ฆ Modular server, better metrics, cleaner streaming.
Run real agents on the hardware already in your hands.
Github: https://t.co/1T06ur6LU5
huge thanks to @jtdavies for ongoing encouragement, @Prince_Canuma for everything MLX and @pollenrobotics for creating this pocket sized bot for the community to experiment with
@ivanfioravanti@pollenrobotics@huggingface As promised.
Early days but hopefully something you can steal or at least give you something to think about
https://t.co/8NlPztaxFV
I trained an ACEStep 1.5 XL LoRA on "some obscure 60s English rock band". Then I wrote a song about LoRA training and had them play it. Absolutely wonderful experience. I still have some UI work before I can make training public in AI Toolkit, but working on it as fast as I can.