Download text encoder

app.py

@@ -9,6 +9,7 @@ import spaces
 import torch
 
 from hyvideo.utils.file_utils import save_videos_grid
+from hyvideo.utils.preprocess_text_encoder_tokenizer_utils import preprocess_text_encoder_tokenizer
 from hyvideo.config import parse_args
 from hyvideo.inference import HunyuanVideoSampler
 from hyvideo.constants import NEGATIVE_PROMPT
@@ -17,6 +18,9 @@ from huggingface_hub import snapshot_download
 
 if torch.cuda.device_count() > 0:
     snapshot_download(repo_id="tencent/HunyuanVideo", repo_type="model", local_dir="ckpts", force_download=True)
+    snapshot_download(repo_id="xtuner/llava-llama-3-8b-v1_1-transformers", repo_type="model", local_dir="ckpts/llava-llama-3-8b-v1_1-transformers", force_download=True)
+    preprocess_text_encoder_tokenizer(input_dir = "ckpts/llava-llama-3-8b-v1_1-transformers", output_dir = "ckpts/text_encoder")
+    snapshot_download(repo_id="openai/clip-vit-large-patch14", repo_type="model", local_dir="ckpts/text_encoder_2", force_download=True)
 
 def initialize_model(model_path):
     print('initialize_model: ' + model_path)