update

README.md

@@ -41,10 +41,10 @@ To try sliding tile attention (optional), please follow the instruction in [csrc
 ## 🚀 Inference
 
 ### Inference HunyuanVideo with Sliding Tile Attention
-Coming soon... Merging commits now.
 ```bash
 python scripts/huggingface/download_hf.py --repo_id=FastVideo/hunyuan --local_dir=data/hunyuan --repo_type=model 
 # Inference with FA2 full attn
+sh scripts/inference/inference_hunyuan_STA.sh
 ```
 
 ### Inference FastHunyuan on single RTX4090

fastvideo/models/hunyuan/modules/attenion.py

@@ -1,13 +1,19 @@
 import torch
 import torch.nn.functional as F
+
 from einops import rearrange
+
 from fastvideo.models.flash_attn_no_pad import flash_attn_no_pad
 from fastvideo.utils.communications import all_gather, all_to_all_4D
-from fastvideo.utils.parallel_states import (get_sequence_parallel_state,
-                                             nccl_info)
+from fastvideo.utils.parallel_states import (
+    get_sequence_parallel_state,
+    nccl_info,
+)
+
 from st_attn import sliding_tile_attention
 
 
+
 def attention(
     q,
     k,

fastvideo/sample/sample_t2v_hunyuan_STA.py

@@ -1,6 +1,8 @@
 import argparse
+import json
 import os
 from pathlib import Path
+from typing import Any, Dict, Optional, Union
 
 import imageio
 import numpy as np
@@ -8,12 +10,13 @@ import torch
 import torch.distributed as dist
 import torchvision
 from einops import rearrange
-import json
+
 from fastvideo.models.hunyuan.inference import HunyuanVideoSampler
 from fastvideo.models.hunyuan.modules.modulate_layers import modulate
-from typing import Any, Optional, Union, Dict
 from fastvideo.utils.parallel_states import (
-    initialize_sequence_parallel_state, nccl_info)
+    initialize_sequence_parallel_state,
+    nccl_info,
+)
 
 
 def teacache_forward(