Update flux.bat

flux.bat

@@ -1,4 +1,4 @@
-accelerate launch  --mixed_precision bf16 --num_cpu_threads_per_process 6 flux_train_network.py --pretrained_model_name_or_path "D:/SDXL/webui_forge_cu121_torch231/webui/models/Stable-diffusion/flux1-dev.safetensors" --clip_l "D:/SDXL/webui_forge_cu121_torch231/webui/models/VAE/clip_l.safetensors" --t5xxl "D:/SDXL/webui_forge_cu121_torch231/webui/models/VAE/t5xxl_fp16.safetensors" --ae "D:/SDXL/webui_forge_cu121_torch231/webui/models/VAE/ae.safetensors" --cache_latents --cache_latents_to_disk --save_model_as safetensors --sdpa --persistent_data_loader_workers --max_data_loader_n_workers 6 --seed 42 --gradient_checkpointing --mixed_precision bf16 --save_precision bf16 --network_module networks.lora_flux --network_dim 16 --learning_rate 1e-3 --network_train_unet_only --cache_text_encoder_outputs --cache_text_encoder_outputs_to_disk --fp8_base --resolution="512,512" --save_every_n_steps="100" --train_data_dir="F:/0_FAV/glory wall 1/" --output_dir "D:/SDXL/webui_forge_cu121_torch231/webui/models/Lora/glory_wall" --logging_dir "D:/SDXL/webui_forge_cu121_torch231/webui/models/Lora/glory_wall" --output_name glory_wall_4 --timestep_sampling shift --discrete_flow_shift 3.1582 --model_prediction_type raw --guidance_scale 1.0 --loss_type l2  --optimizer_type adammini --optimizer_args "weight_decay=0.1" "betas=0.9,0.99" "eps=1e-6"  --max_train_steps="3000" --enable_bucket --caption_extension=".txt" --train_batch_size=4 --apply_t5_attn_mask --noise_offset 0.1 --lr_scheduler "REX" --lr_warmup_steps 300 --split_mode --network_args "train_blocks=single" --network_weights="D:/SDXL/webui_forge_cu121_torch231/webui/models/Lora/glory_wall\glory_wall_3-step00001000.safetensors"
+accelerate launch  --mixed_precision bf16 --num_cpu_threads_per_process 6 flux_train_network.py --pretrained_model_name_or_path "D:/SDXL/webui_forge_cu121_torch231/webui/models/Stable-diffusion/flux1-dev.safetensors" --clip_l "D:/SDXL/webui_forge_cu121_torch231/webui/models/VAE/clip_l.safetensors" --t5xxl "D:/SDXL/webui_forge_cu121_torch231/webui/models/VAE/t5xxl_fp16.safetensors" --ae "D:/SDXL/webui_forge_cu121_torch231/webui/models/VAE/ae.safetensors" --cache_latents --cache_latents_to_disk --save_model_as safetensors --sdpa --persistent_data_loader_workers --max_data_loader_n_workers 6 --seed 42 --gradient_checkpointing --mixed_precision bf16 --save_precision bf16 --network_module networks.lora_flux --network_dim 16 --learning_rate 1e-3 --network_train_unet_only --cache_text_encoder_outputs --cache_text_encoder_outputs_to_disk --fp8_base --resolution="512,512" --save_every_n_steps="100" --train_data_dir="F:/0_FAV/glory wall 1/" --output_dir "D:/SDXL/webui_forge_cu121_torch231/webui/models/Lora/glory_wall" --logging_dir "D:/SDXL/webui_forge_cu121_torch231/webui/models/Lora/glory_wall" --output_name glory_wall_4 --timestep_sampling shift --discrete_flow_shift 3.1582 --model_prediction_type raw --guidance_scale 1.0 --loss_type l2  --optimizer_type adammini --optimizer_args "weight_decay=0.1" "betas=0.9,0.99" "eps=1e-6"  --max_train_steps="3000" --enable_bucket --caption_extension=".txt" --train_batch_size=4 --apply_t5_attn_mask --noise_offset 0.1 --lr_scheduler "REX" --lr_warmup_steps 300 --split_mode --network_args "train_blocks=single" 
 
 :: 先按照GUI的標準方式安裝 然後自己修改FLUX.bat的參數
 :: 用.\venv\Scripts\activate 進入venv再執行FLUX.bat