Fix Qwen 2.1 GGUF load: promote Q8 norms to F32 and wire TextEncode latent.

abenzerps Q8_0 ships 1D RMSNorms as packed Q8 (136 vs 128), which breaks
Comfy rms_rope; tagger now dequantizes small tensors and the graph uses
TextEncodeQwenImage21's 64-ch latent plus AuraFlow shift.

Co-authored-by: Cursor <cursoragent@cursor.com>

docs/qwen21.md

@@ -21,6 +21,14 @@ Custom node: `custom_nodes\ComfyUI-GGUF` (`UnetLoaderGGUF`). Do not install a se
 powershell -ExecutionPolicy Bypass -File scripts\setup-qwen21.ps1
 ```
 
+DiT download uses the exact host invocation:
+
+```powershell
+hf download hf://abenzerps/Qwen-Image-2.1-Uncensored-GGUF/qwen-image-2.1-Q8_0.gguf
+```
+
+That lands in the Hugging Face hub cache; the setup script copies it into Shared `diffusion_models`. The abenzerps GGUF ships with `kv_count=0` (no `general.architecture`) **and** Q8_0-quantized 1D RMSNorm weights (logical 128 → packed 136), which breaks Comfy’s fused `rms_rope`. `scripts/tag-qwen21-gguf.py` rewrites the file with `general.architecture=qwen_image` and promotes small/1D tensors to F32 so city96 `UnetLoaderGGUF` can load it.
+
 Restart Comfy **only when idle** (`COMFY_CONTROL_URL/status` → `gpu.busy=false`). Confirm `object_info` lists `UnetLoaderGGUF` and the three filenames.
 
 ## App
@@ -28,7 +36,7 @@ Restart Comfy **only when idle** (`COMFY_CONTROL_URL/status` → `gpu.busy=false
 - Engine key: `qwen21` · UI label: **Qwen 2.1**
 - Generate (T2I) only in this build. Edit / Compose / Iterate / Video disable with: “Qwen 2.1 is T2I in this build”.
 - Sampler defaults: euler / simple / cfg **1** / steps **25** · `ModelSamplingAuraFlow` shift **3.1**
-- Default canvas **1024×1024** (aspect 16:9 / 9:16 / 1:1 honored, multiples of 32). Drop to 768 if VRAM errors.
+- Default canvas **1024×1024** (aspect 16:9 / 9:16 / 1:1 → long-edge square via `TextEncodeQwenImage21` resolution; multiples of 32). Drop to 768 if VRAM errors.
 - No hero / locks / Klein LoRA stack on this engine.
 
 Graph: `server/assets/studio2_qwen21_t2i.json`.