metadata
license: apache-2.0
It's a merged model by me
Backed with these loras :
SVI_v2_PRO_Wan2.2-I2V-A14B_LOW_lora_rank_128_fp16
SVI_v2_PRO_Wan2.2-I2V-A14B_hIGH_lora_rank_128_fp16
lightx2v_T2V_14B_cfg_step_distill_v2_lora_rank256_bf16
Set 8 steps, cfg 2 on high noise and you'll get a nice result
Also I put a special workflow for those models and HunyuanVideoFoley to add sound towards videos