Files
hao-ai-lab-FastVideo/scripts/inference/v1_inference_wan_STA.sh

22 lines
989 B
Bash
Executable File

#!/bin/bash
num_gpus=2
export FASTVIDEO_ATTENTION_CONFIG=assets/mask_strategy_wan.json
export FASTVIDEO_ATTENTION_BACKEND=SLIDING_TILE_ATTN
export MODEL_BASE=Wan-AI/Wan2.1-T2V-14B-Diffusers
fastvideo generate \
--model-path $MODEL_BASE \
--sp-size $num_gpus \
--tp-size 1 \
--num-gpus $num_gpus \
--height 768 \
--width 1280 \
--num-frames 69 \
--num-inference-steps 50 \
--fps 16 \
--guidance-scale 5.0 \
--prompt "A beautiful woman in a red dress walking down a street" \
--negative-prompt "Bright tones, overexposed, static, blurred details, subtitles, style, works, paintings, images, static, overall gray, worst quality, low quality, JPEG compression residue, ugly, incomplete, extra fingers, poorly drawn hands, poorly drawn faces, deformed, disfigured, misshapen limbs, fused fingers, still picture, messy background, three legs, many people in the background, walking backwards" \
--seed 1024 \
--output-path outputs_video/