237 kB

Ctrl+K

1 contributor

load order: DiT (quantize/fuse peak) first, VAEs, text encoder LAST + gc between stages - encoder-first stacked 24GB under the fuse peak and caused an mmap access violation at VAE load on a 96GB box

9fb07f5 verified about 15 hours ago

libs
LoRA fusion: map kohya-FLATTENED keys (lora_unet_*, dots->underscores) onto model paths via exact flat-key lookup - such LoRAs previously matched zero keys 1 day ago
prompts
Upload 11 files 2 days ago
.gitattributes

1.52 kB
initial commit 2 days ago
README.md

10.9 kB
LoRA hardening: kohya lora_down/up support + alpha scaling, fuse telemetry + zero-match warning, refuse comfy_quant checkpoints with clear error; lora_file dropdown 1 day ago
__init__.py

4.51 kB
Upload 11 files 2 days ago
joyecho_prompt_source.py

5.86 kB
Upload 11 files 2 days ago
joyecho_ref_batch.py

2.11 kB
Upload 11 files 2 days ago
joyecho_ref_picker.py

14.5 kB
RefPicker auto mode: emit ONE ref per character named in prose (batch output) - solo scene 1 ref, two-character 2 refs, none 0; dialogue mentions still excluded; single picker now covers every queue case with zero mode switching about 16 hours ago
joyecho_script_picker.py

3.16 kB
Upload 11 files 2 days ago
nodes.py

88.6 kB
load order: DiT (quantize/fuse peak) first, VAEs, text encoder LAST + gc between stages - encoder-first stacked 24GB under the fuse peak and caused an mmap access violation at VAE load on a 96GB box about 15 hours ago
rebels_loaders.py

42.5 kB
Upload 11 files 2 days ago