diffusers/examples/text_to_image/train_text_to_image.py training SD1.4 question

[OliverLeeXZ]

I using the following script to train SD1.4 model:
export MODEL_NAME="/home/bingxing2/home/scx7kzs/CODE/MODEL/stable-diffusion-v1-4"

export TRAIN_DIR="/home/bingxing2/home/scx7kzs/CODE/DATA/allava_vflan/images"

export TRAIN_DIR="/home/bingxing2/home/scx7kzs/CODE/diffusers/ft_data/cc_sbu/train"
export OUTPUT_DIR="/home/bingxing2/home/scx7kzs/CODE/diffusers/ckpts"

accelerate launch /home/bingxing2/home/scx7kzs/CODE/diffusers/examples/text_to_image/train_text_to_image.py
--pretrained_model_name_or_path=$MODEL_NAME
--train_data_dir=$TRAIN_DIR
--use_ema
--resolution=512 --center_crop --random_flip
--train_batch_size=1
--gradient_accumulation_steps=4
--gradient_checkpointing
--mixed_precision="fp16"
--max_train_steps=15000
--learning_rate=1e-05
--max_grad_norm=1
--lr_scheduler="constant" --lr_warmup_steps=0
--output_dir=${OUTPUT_DIR}

However, I've been stuck here for more than 20 minutes:
12/06/2024 11:25:07 - INFO - main - ***** Running training *****
12/06/2024 11:25:07 - INFO - main - Num examples = 3439
12/06/2024 11:25:07 - INFO - main - Num Epochs = 18
12/06/2024 11:25:07 - INFO - main - Instantaneous batch size per device = 1
12/06/2024 11:25:07 - INFO - main - Total train batch size (w. parallel, distributed & accumulation) = 4
12/06/2024 11:25:07 - INFO - main - Gradient Accumulation steps = 4
12/06/2024 11:25:07 - INFO - main - Total optimization steps = 15000

Steps: 0%| | 0/15000 [00:00<?, ?it/s]

How can I slove it?

[wzhiyuan2016]

I using the following script to train SD1.4 model: export MODEL_NAME="/home/bingxing2/home/scx7kzs/CODE/MODEL/stable-diffusion-v1-4"

export TRAIN_DIR="/home/bingxing2/home/scx7kzs/CODE/DATA/allava_vflan/images"

export TRAIN_DIR="/home/bingxing2/home/scx7kzs/CODE/diffusers/ft_data/cc_sbu/train" export OUTPUT_DIR="/home/bingxing2/home/scx7kzs/CODE/diffusers/ckpts"

accelerate launch /home/bingxing2/home/scx7kzs/CODE/diffusers/examples/text_to_image/train_text_to_image.py --pretrained_model_name_or_path=$MODEL_NAME --train_data_dir=$TRAIN_DIR --use_ema --resolution=512 --center_crop --random_flip --train_batch_size=1 --gradient_accumulation_steps=4 --gradient_checkpointing --mixed_precision="fp16" --max_train_steps=15000 --learning_rate=1e-05 --max_grad_norm=1 --lr_scheduler="constant" --lr_warmup_steps=0 --output_dir=${OUTPUT_DIR}

However, I've been stuck here for more than 20 minutes: 12/06/2024 11:25:07 - INFO - main - ***** Running training ***** 12/06/2024 11:25:07 - INFO - main - Num examples = 3439 12/06/2024 11:25:07 - INFO - main - Num Epochs = 18 12/06/2024 11:25:07 - INFO - main - Instantaneous batch size per device = 1 12/06/2024 11:25:07 - INFO - main - Total train batch size (w. parallel, distributed & accumulation) = 4 12/06/2024 11:25:07 - INFO - main - Gradient Accumulation steps = 4 12/06/2024 11:25:07 - INFO - main - Total optimization steps = 15000

Steps: 0%| | 0/15000 [00:00<?, ?it/s]

How can I slove it?

How did your final training generate results?
I trained my own data and the final results are in this discussion====>
#10195
My training results are very poor!!!!