TowstyandCursor 1f10082087 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>

2026-09-20 16:22:55 -05:00

..

cleanup.ts

Retry confirmed Studio 2 host cleanup after failures

2026-09-11 18:43:25 -05:00

export-lora-catalog.py

Read production LoRA names in Studio 2

2026-09-12 21:05:11 -05:00

lora-catalog.ts

Read production LoRA names in Studio 2

2026-09-12 21:05:11 -05:00

media.ts

Tighten needsFit to exact canvas size, guard against empty renders

2026-09-11 23:11:02 -05:00

progress.ts

Fit Studio 2 inputs and report live generation progress

2026-09-10 23:48:25 -05:00

runner.ts

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

2026-09-20 16:22:55 -05:00

size.ts

Wire Qwen 2.1 as Generate-only engine with 25/1 sampler defaults.

2026-09-20 15:53:44 -05:00

store.ts

Queue durable Studio 2 jobs with sticky identity and saved metadata

2026-09-10 20:56:37 -05:00