update README

README.md

@@ -45,9 +45,9 @@ First, download the model:
 ```
 python scripts/huggingface/download_hf.py --repo_id=FastVideo/hunyuan --local_dir=data/hunyuan --repo_type=model 
 ```
-You can use the following script to run inference with STA + [TeaCache](https://github.com/ali-vilab/TeaCache) to enjoy training-free 3x speed up. Turn on ```--enable_torch_compile``` when only use STA, and use ```--enable_teacache``` when using both. 
+We provide two examples in the following script to run inference with STA + [TeaCache](https://github.com/ali-vilab/TeaCache) and STA only.
 ```
-sh scripts/inference/inference_hunyuan_STA.sh
+scripts/inference/inference_hunyuan_STA.sh
 ```
 ### Video Demos using STA + Teacache
 Visit our [demo website](https://fast-video.github.io/) to explore our complete collection of examples. We shorten a single video generation process from 945s to 317s on a single H100.

scripts/inference/inference_hunyuan_STA.sh

@@ -1,9 +1,10 @@
 #!/bin/bash
 
+# Inference with STA + Teacache
 num_gpus=1
 mask_strategy_file_path=assets/mask_strategy.json
 export MODEL_BASE=data/hunyuan
-rel_l1_thresh=0.2
+rel_l1_thresh=0.15
 CUDA_VISIBLE_DEVICES=1 torchrun --nnodes=1 --nproc_per_node=$num_gpus --master_port 29603 \
     fastvideo/sample/sample_t2v_hunyuan_STA.py \
     --height 768 \
@@ -22,4 +23,27 @@ CUDA_VISIBLE_DEVICES=1 torchrun --nnodes=1 --nproc_per_node=$num_gpus --master_p
     --rel_l1_thresh $rel_l1_thresh \
     --dit-weight ${MODEL_BASE}/hunyuan-video-t2v-720p/transformers/mp_rank_00_model_states.pt \
     --vae-sp \
-    --enable_teacache 
+    --enable_teacache 
+
+# Inference with STA only
+num_gpus=1
+mask_strategy_file_path=assets/mask_strategy.json
+export MODEL_BASE=data/hunyuan
+CUDA_VISIBLE_DEVICES=1 torchrun --nnodes=1 --nproc_per_node=$num_gpus --master_port 29603 \
+    fastvideo/sample/sample_t2v_hunyuan_STA.py \
+    --height 768 \
+    --width 1280 \
+    --num_frames 117 \
+    --num_inference_steps 50 \
+    --guidance_scale 1 \
+    --embedded_cfg_scale 6 \
+    --flow_shift 7 \
+    --flow-reverse \
+    --prompt ./assets/prompt.txt \
+    --seed 12345 \
+    --output_path outputs_video/hunyuan_STA/ \
+    --model_path $MODEL_BASE \
+    --mask_strategy_file_path $mask_strategy_file_path \
+    --dit-weight ${MODEL_BASE}/hunyuan-video-t2v-720p/transformers/mp_rank_00_model_states.pt \
+    --vae-sp \
+    --enable_torch_compile