search

huggingface / diffusers

🤗 Diffusers: State-of-the-art diffusion models for image and audio generation in PyTorch and FLAX.
https://huggingface.co/docs/diffusers
Apache License 2.0
25.06k stars 5.18k forks source link

train_dreambooth_sd3.py still broken #8918

Open neuron-party opened 1 month ago

neuron-party commented 1 month ago

Describe the bug

incorrect code/logic, referencing variables issue

Reproduction

run train_dreambooth_sd3.py with valid args

Logs

text_input_ids_list referenced before variable assignment

System Info

diffusers 30 dev (source)

Who can help?

No response

DN6 commented 1 month ago

@neuron-party Can you please provide the specific issue you're running into and provide the exact command you're running

neuron-party commented 1 month ago

@DN6 

accelerate launch --config_file config_file_path train_dreambooth_lora_sd3.py (or train_dreambooth_sd3.py)
 --mixed_precision=fp16
 --resolution=1024
 --pretrained_model_name_or_path=stabilityai/stable-diffusion-3-medium-diffusers
 --num_validation_images=8
 --validation_epochs=100
 --train_batch_size=2
 --lr_scheduler=cosine
 --checkpointing_steps=200
 --report_to=tensorboard
 --train_text_encoder
 --instance_prompt=xyz
 --instance_data_dir=xyz
 --validation_prompt=xyz
 --max_train_steps=1600
 --rank=64
 --learning_rate=5e-6
 --text_encoder_lr=1e-6
 --output_dir=xyz
 --logging_dir=xyz
DN6 commented 1 month ago

At @neuron-party what is error/issue you are encountering? Can you provide a traceback or example of it?