Jose Esteban Merino

GOOGLE COLAB: !apt-get update !apt-get install cuda-toolkit-11-8 import os os.environ["LD_LIBRARY_PATH"] += ":" + "/usr/local/cuda-11/lib64" os.environ["LD_LIBRARY_PATH"] += ":" + "/usr/local/cuda-11.8/lib64"

!apt-get update !apt-get install cuda-toolkit-11-8 import os os.environ["LD_LIBRARY_PATH"] += ":" + "/usr/local/cuda-11/lib64" os.environ["LD_LIBRARY_PATH"] += ":" + "/usr/local/cuda-11.8/lib64"

1. I never said it was a solution 2. For me, I work at Google Colab, it is a solution I hope you find the "solution"

I am a basic user If I were a professional, I would read the comment, check what was proposed, and appreciate the comment, even if it is a partial solution,...

@linoytsaban @DN6 Hello, I tested the changes in the branch --branch dreambooth-advanced I still get the error when saving But only when I use the **--train_text_encoder** parameter When I don't...

> @josemerinom Could you share the exact traceback here? Not a screenshot. Here is test 2 that I did with the changes that were made in the code https://colab.research.google.com/github/josemerinom/test/blob/master/test2.ipynb here...

@DN6 Test 3: **--branch main** ### Reproduction https://colab.research.google.com/github/josemerinom/test/blob/master/test3.ipynb ### Results training start: OK save checkpoint: OK training completed: OK test no lora / step 50 / step 100: OK The...

Thanks, I'll try with fp8 (not scaled) (01-aug) It works! (01-aug)

LoraFlux: was trained using the fp8 model FLUX.1-dev LoraKrea: was trained using the fp8 model FLUX.1-Krea-dev 

> ¿Cuál es la configuración necesaria para entrenar Flux1-Krea Lora? ¿Es la misma que Flux1-Dev? No obtengo buenos resultados al entrenar Krea Lora con la misma configuración que Dev. I...