Merge branch 'main' of https://huggingface.co/spaces/lmms-lab/LLaVA-NeXT-Interleave-Demo

app.py

@@ -9,7 +9,7 @@ import cv2
 # import copy
 import torch
 
-# import spaces
+import spaces
 import numpy as np
 
 from llava import conversation as conversation_lib
@@ -155,7 +155,7 @@ def add_message(history, message):
     return history, gr.MultimodalTextbox(value=None, interactive=False)
 
 
-# @spaces.GPU
+@spaces.GPU
 def bot(history):
     text = history[-1][0]
     images_this_term = []
@@ -437,10 +437,8 @@ if __name__ == "__main__":
     args = argparser.parse_args()
     model_path = args.model_path
     filt_invalid = "cut"
-    # model_name = get_model_name_from_path(args.model_path)
-    # tokenizer, model, image_processor, context_len = load_pretrained_model(args.model_path, args.model_base, model_name, args.load_8bit, args.load_4bit)
-    # model=model.to(torch.device('cuda'))
+    model_name = get_model_name_from_path(args.model_path)
+    tokenizer, model, image_processor, context_len = load_pretrained_model(args.model_path, args.model_base, model_name, args.load_8bit, args.load_4bit)
+    model=model.to(torch.device('cuda'))
     our_chatbot = None
-    # import pdb;pdb.set_trace()
-    # try:
     demo.launch()

requirements.txt

@@ -4,6 +4,7 @@ opencv-python
 open_clip_torch
 fastapi
 gradio==4.35.0
+gradio_client==1.0.1
 markdown2[all]
 numpy
 requests
@@ -22,8 +23,7 @@ scikit-learn==1.2.2
 sentencepiece~=0.1.99
 einops==0.6.1
 einops-exts==0.0.4
-gradio_client==0.2.9
-pydantic==1.10.8
+pydantic>=2.0
 timm
 hf_transfer
 decord
@@ -31,5 +31,5 @@ datasets==2.16.1
 tyro
 scipy
 rouge
-urllib3<=2.0.0
+urllib3~=2.0