⚡ All JANG models are meant to be run in vMLX
Ornith-1.0-9B · JANG_6M
Vision-language · Qwen3.5 hybrid backbone · 6-bit near-lossless mixed precision · ~8.0 GB
⚠️ Requires MLX Studio (or the vMLX runtime) to run. Standard
mlx_lmcannot load JANG bundles correctly — they store the Qwen3.5 RMSNorm un-shifted and rely on the runtime's +1 scale_shift + per-layer bit detection. MLX Studio includes the JANG loader.
JANG_6M = 8-bit attention + 6-bit MLP (affine mixed precision, group-size 64); vision tower kept fp16. A near-lossless JANG profile from JANGQ-AI.
Architecture
| Family | qwen3_5 (dense, hybrid) |
| Text layers | 32 — 24 Gated-DeltaNet + 8 full-attention |
| MoE / dims | hidden 4096 · untied lm_head |
| Vision | ViT tower (model.visual) preserved fp16 |
| Cache | hybrid (GDN state + KV for attention layers) |
| Parsers | reasoning qwen3 · tools qwen |
Provenance
- Base: deepreinforce-ai/Ornith-1.0-9B © DeepReinforce — MIT (Qwen3.5-based)
- Quantization: JANG · JANG_6M (8-bit attention + 6-bit MLP, group-size 64; vision tower fp16) · eric@jangq.ai