-
Notifications
You must be signed in to change notification settings - Fork 440
Description
Git commit
Operating System & Version
Ubuntu 24.04
GGML backends
CUDA
Command-line arguments used
sd -m ~/SD_models/sdxl/sd_xl_base_1.0_0.9vae.safetensors --vae ~/SD_models/sdxl/sdxl_vae.safetensors --clip-on-cpu --sampling-method euler --diffusion-fa -H 1024 -W 1024 --seed 5 --steps 40 --cfg-scale 5 --photo-maker ~/SD_models/loras_sdxl/photomaker-v2.safetensors --pm-id-images-dir ./photomaker_examples/newton_man --pm-style-strength 15 -p "half portrait of a man img, playing guitar, wearing blue shirt, dark background, RAW photo, 8k uhd, high quality" -n "Cropped, out of frame, out of focus, worst quality, greyscale, bad anatomy, bad hands" -o PM2.png
Steps to reproduce
Try above command line :
- "as is"
- with additional arg : --pm-id-embed-path ./photomaker_examples/newton_man/id_embeds.bin
What you expected to happen
sd should warn user and continue w/o using PM2 ....
(...or maybe clean exit)
What actually happened
crash / core dump
Logs / error messages / stack trace
....
[INFO ] stable-diffusion.cpp:2615 - TXT2IMG
[INFO ] stable-diffusion.cpp:947 - attempting to apply 0 LoRAs
[INFO ] stable-diffusion.cpp:967 - apply_loras completed, taking 0.00s
[INFO ] stable-diffusion.cpp:2177 - pmid_lora apply completed, taking 0.15s
[ERROR] ggml_extend.hpp:293 - failed to open ''
....... Speicherzugriffsfehler (Speicherabzug geschrieben) ../build/bin/sd -m ~/SD_models/sdxl/sd_xl_base_1.0_0.9vae.safetensors --vae ~/SD_models/sdxl/sdxl_vae.s
afetensors --clip-on-cpu --sampling-method euler --diffusion-fa -H 1024 -W 1024 --seed 5 --steps 40 --cfg-scale 5 --photo-maker ~/SD_models/loras_sdxl/photomaker-v2.safetensors --pm-id-image
s-dir ./photomaker_examples/newton_man --pm-style-strength 15 -p "half portrait of a man img, playing guitar, wearing blue shirt, dark background, RAW photo, 8k uhd, high quality" -n "Croppe
d, out of frame, out of focus, worst quality, greyscale, bad anatomy, bad hands" -o PM2.png
Hint: "Speicherzugriffsfehler (Speicherabzug geschrieben)" is german
and stands simply for "core dumped"
Additional context / environment details
No response