Emil Bogomolov
Emil Bogomolov
 In this formula we have an expectation of $w_i$. That means for each pair of $(w_I, w_O)$ we should calculate this expectation. But as I can...
Initial version of the repo has possibility to quantize neural network. After vast refactoring done by @rakhimovv to address the problem that results from the paper cannot be reproduced, we've...
to get_scheduler func, * add it to the train_dreambooth.py
Hello, Thank you for your great work. I've noticed some traces of DLIB in your repo, can you give an advice how one can replace facelib with DLIB in that...
The dependency on invisible-watermark adds dependency on opencv. This line breaks a lot of dockerfiles. Can we switch to headless?
Trying to make a prediction with LoRA weights trained and stored locally within `/training_out/*`. Command: `cog predict -i prompt="a photo of TOK" -i lora_url=@/training_out/` or `cog predict -i prompt="a photo...
Command ```cog train -i input_images=@example_datasets/ellen_sdxl.zip -i use_face_detection_instead=True``` Results in: ``` ... File "/root/.pyenv/versions/3.9.18/lib/python3.9/site-packages/cog/predictor.py", line 158, in get_predictor_ref raise PredictorNotSet( cog.errors.PredictorNotSet: Can't run predictions: 'train' option not found in cog.yaml ⅹ...