update app

app.py

@@ -56,15 +56,6 @@ model_q = Qwen2_5_VLForConditionalGeneration.from_pretrained(
     torch_dtype=torch.float16
 ).to(device).eval()
 
-# Load allenai/olmOCR-7B-0825
-MODEL_ID_F = "allenai/olmOCR-7B-0825"
-processor_f = AutoProcessor.from_pretrained(MODEL_ID_F, trust_remote_code=True)
-model_f = Qwen2_5_VLForConditionalGeneration.from_pretrained(
-    MODEL_ID_F,
-    trust_remote_code=True,
-    torch_dtype=torch.float16
-).to(device).eval()
-
 # Load R-4B
 MODEL_ID_Y = "YannQi/R-4B"
 processor_y = AutoProcessor.from_pretrained(MODEL_ID_Y, trust_remote_code=True)
@@ -115,9 +106,6 @@ def generate_image(model_name: str, text: str, image: Image.Image,
     elif model_name == "Qwen2.5-VL-7B-Abliterated-Caption-it":
         processor = processor_q
         model = model_q
-    elif model_name == "olmOCR-7B-0825":
-        processor = processor_f
-        model = model_f
     elif model_name == "R-4B":
         processor = processor_y
         model = model_y
@@ -175,9 +163,6 @@ def generate_video(model_name: str, text: str, video_path: str,
     elif model_name == "Qwen2.5-VL-7B-Abliterated-Caption-it":
         processor = processor_q
         model = model_q
-    elif model_name == "olmOCR-7B-0825":
-        processor = processor_f
-        model = model_f
     elif model_name == "R-4B":
         processor = processor_y
         model = model_y
@@ -295,7 +280,7 @@ with gr.Blocks(css=css, theme="bethecloud/storj_theme") as demo:
                     markdown_output = gr.Markdown()
                 
             model_choice = gr.Radio(
-                choices=["Qwen2.5-VL-7B-Instruct", "Qwen2.5-VL-3B-Instruct", "R-4B", "Qwen2.5-VL-7B-Abliterated-Caption-it", "olmOCR-7B-0825"],
+                choices=["Qwen2.5-VL-7B-Instruct", "Qwen2.5-VL-3B-Instruct", "R-4B", "Qwen2.5-VL-7B-Abliterated-Caption-it"],
                 label="Select Model",
                 value="Qwen2.5-VL-7B-Instruct"
             )
@@ -305,12 +290,6 @@ with gr.Blocks(css=css, theme="bethecloud/storj_theme") as demo:
                 """
             > [Qwen2.5-VL-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-VL-7B-Instruct): The Qwen2.5-VL-7B-Instruct model is a multimodal AI model developed by Alibaba Cloud that excels at understanding both text and images. It's a Vision-Language Model (VLM) designed to handle various visual understanding tasks, including image understanding, video analysis, and even multilingual support.  
             >
-            > [Qwen2.5-VL-3B-Instruct](https://huggingface.co/Qwen/Qwen2.5-VL-3B-Instruct): Qwen2.5-VL-3B-Instruct is an instruction-tuned vision-language model from Alibaba Cloud, built upon the Qwen2-VL series. It excels at understanding and generating text related to both visual and textual inputs, making it capable of tasks like image captioning, visual question answering, and object localization. The model also supports long video understanding and structured data extraction.
-                """
-            )
-            
-            gr.Markdown(
-                """
             > [Qwen2.5-VL-7B-Abliterated-Caption-it](prithivMLmods/Qwen2.5-VL-7B-Abliterated-Caption-it): Qwen2.5-VL-7B-Abliterated-Caption-it is a fine-tuned version of Qwen2.5-VL-7B-Instruct, optimized for Abliterated Captioning / Uncensored Captioning. This model excels at generating detailed, context-rich, and high-fidelity captions across diverse image categories and variational aspect ratios, offering robust visual understanding without filtering or censorship.  
             >
             > [olmOCR-7B-0825](https://huggingface.co/allenai/olmOCR-7B-0825): olmOCR-7B-0825 is a 7B parameter open large model designed for OCR tasks with robust text extraction, especially in complex document layouts. Multimodal model emphasizing strong document reading and extraction capabilities combined with vision-language understanding to support detailed document parsing tasks.