graffiti-style / train.bat
TheAwakenOne's picture
Upload folder using huggingface_hub
cd60162 verified
raw
history blame
1.66 kB
accelerate launch ^
--mixed_precision bf16 ^
--num_cpu_threads_per_process 1 ^
sd-scripts/flux_train_network.py ^
--pretrained_model_name_or_path "E:\AIOne\pinokio\api\fluxgym.git\models\unet\flux1-dev.sft" ^
--clip_l "E:\AIOne\pinokio\api\fluxgym.git\models\clip\clip_l.safetensors" ^
--t5xxl "E:\AIOne\pinokio\api\fluxgym.git\models\clip\t5xxl_fp16.safetensors" ^
--ae "E:\AIOne\pinokio\api\fluxgym.git\models\vae\ae.sft" ^
--cache_latents_to_disk ^
--save_model_as safetensors ^
--sdpa --persistent_data_loader_workers ^
--max_data_loader_n_workers 2 ^
--seed 42 ^
--gradient_checkpointing ^
--mixed_precision bf16 ^
--save_precision bf16 ^
--network_module networks.lora_flux ^
--network_dim 4 ^
--optimizer_type adafactor ^
--optimizer_args "relative_step=False" "scale_parameter=False" "warmup_init=False" ^
--split_mode ^
--network_args "train_blocks=single" ^
--lr_scheduler constant_with_warmup ^
--max_grad_norm 0.0 ^--sample_prompts="E:\AIOne\pinokio\api\fluxgym.git\outputs\graffiti-style\sample_prompts.txt" --sample_every_n_steps="100" ^
--learning_rate 8e-4 ^
--cache_text_encoder_outputs ^
--cache_text_encoder_outputs_to_disk ^
--fp8_base ^
--highvram ^
--max_train_epochs 16 ^
--save_every_n_epochs 4 ^
--dataset_config "E:\AIOne\pinokio\api\fluxgym.git\outputs\graffiti-style\dataset.toml" ^
--output_dir "E:\AIOne\pinokio\api\fluxgym.git\outputs\graffiti-style" ^
--output_name graffiti-style ^
--timestep_sampling shift ^
--discrete_flow_shift 3.1582 ^
--model_prediction_type raw ^
--guidance_scale 1 ^
--loss_type l2 ^