diffusers
Fix: training resume from fp16 for SDXL Consistency Distillation
#6840
Merged
Merged
Fix: training resume from fp16 for SDXL Consistency Distillation #6840
sayakpaul
merged
9 commits into
huggingface:main
from
fix-fp16-train-resume-lcm-sdxl
Fix: training resume from fp16 for lcm distill lora sdxl
b5322915
sriagastya
changed the title 
Merge branch 'main' into fix-fp16-train-resume-lcm-sdxl
53c0a072
Fix coding quality - run linter
cbea2b13
Merge branch 'huggingface:main' into fix-fp16-train-resume-lcm-sdxl
df94b622
Fix 1 - shift mixed precision cast before optimizer
d5ed3352
Fix 2 - State dict errors by removing load_lora_into_unet
e6a1f827
Merge branch 'main' into fix-fp16-train-resume-lcm-sdxl
efa505ee
sayakpaul
commented
on 2024-02-08
Update train_lcm_distill_lora_sdxl.py - Revert default cache dir to None
99e52902
Merge branch 'main' into fix-fp16-train-resume-lcm-sdxl
c539ac7c
Login to write a write a comment.
Login via GitHub