mirror of
https://github.com/modelscope/DiffSynth-Studio.git
synced 2026-03-18 13:58:15 +00:00
17 lines
973 B
Bash
17 lines
973 B
Bash
modelscope download --dataset DiffSynth-Studio/diffsynth_example_dataset --include "wanvideo/LongCat-Video/*" --local_dir ./data/diffsynth_example_dataset
|
|
|
|
accelerate launch examples/wanvideo/model_training/train.py \
|
|
--dataset_base_path data/diffsynth_example_dataset/wanvideo/LongCat-Video \
|
|
--dataset_metadata_path data/diffsynth_example_dataset/wanvideo/LongCat-Video/metadata.csv \
|
|
--height 480 \
|
|
--width 832 \
|
|
--dataset_repeat 100 \
|
|
--model_id_with_origin_paths "meituan-longcat/LongCat-Video:dit/diffusion_pytorch_model*.safetensors,Wan-AI/Wan2.1-T2V-14B:models_t5_umt5-xxl-enc-bf16.pth,Wan-AI/Wan2.1-T2V-14B:Wan2.1_VAE.pth" \
|
|
--learning_rate 1e-4 \
|
|
--num_epochs 5 \
|
|
--remove_prefix_in_ckpt "pipe.dit." \
|
|
--output_path "./models/train/LongCat-Video_lora" \
|
|
--lora_base_model "dit" \
|
|
--lora_target_modules "adaLN_modulation.1,attn.qkv,attn.proj,cross_attn.q_linear,cross_attn.kv_linear,cross_attn.proj,ffn.w1,ffn.w2,ffn.w3" \
|
|
--lora_rank 32
|