Update app.py

app.py

@@ -1,5 +1,6 @@
 import gradio as gr
-from transformers import BitsAndBytesConfig, LlavaNextVideoForConditionalGeneration, LlavaNextVideoProcessor
+from transformers import  LlavaNextVideoForConditionalGeneration, LlavaNextVideoProcessor
+# BitsAndBytesConfig,
 import torch
 import av
 import numpy as np
@@ -9,11 +10,11 @@ import tempfile
 
 
 # Configuration du modèle
-quantization_config = BitsAndBytesConfig(
-    load_in_4bit=True,
-    bnb_4bit_compute_dtype=torch.float16,
-    llm_int8_enable_fp32_cpu_offload=True # Enable CPU offloading for unsupported layers
-)
+# quantization_config = BitsAndBytesConfig(
+#     load_in_4bit=True,
+#     bnb_4bit_compute_dtype=torch.float16,
+#     llm_int8_enable_fp32_cpu_offload=True # Enable CPU offloading for unsupported layers
+# )
 
 
 # Configuration du modèle
@@ -25,7 +26,7 @@ quantization_config = BitsAndBytesConfig(
 processor = LlavaNextVideoProcessor.from_pretrained("llava-hf/LLaVA-NeXT-Video-7B-hf")
 model = LlavaNextVideoForConditionalGeneration.from_pretrained(
     "llava-hf/LLaVA-NeXT-Video-7B-hf",
-    quantization_config=quantization_config,
+    # quantization_config=quantization_config,
     device_map='auto'
 )