Update app.py

app.py

@@ -81,6 +81,18 @@ model_a = AutoModelForImageTextToText.from_pretrained(
     torch_dtype=torch.float16
 ).to(device).eval()
 
+#-----------------------------subfolder-----------------------------#
+# Load MonkeyOCR-1.2B-0709
+MODEL_ID_W = "echo840/MonkeyOCR-1.2B-0709"
+SUBFOLDER = "Recognition"
+processor_w = AutoProcessor.from_pretrained(MODEL_ID_W, trust_remote_code=True, subfolder=SUBFOLDER)
+model_w = Qwen2_5_VLForConditionalGeneration.from_pretrained(
+    MODEL_ID_W, trust_remote_code=True,
+    subfolder=SUBFOLDER,
+    torch_dtype=torch.float16).to(device).eval()
+#-----------------------------subfolder-----------------------------#
+
+
 def downsample_video(video_path):
     """
     Downsamples the video to evenly spaced frames.
@@ -113,7 +125,7 @@ def generate_image(model_name: str, text: str, image: Image.Image,
     Generates responses using the selected model for image input.
     Yields raw text and Markdown-formatted text.
     """
-    if model_name == "RolmOCR":
+    if model_name == "RolmOCR-7B":
         processor = processor_m
         model = model_m
     elif model_name == "Qwen2-VL-OCR":
@@ -122,9 +134,12 @@ def generate_image(model_name: str, text: str, image: Image.Image,
     elif model_name == "Nanonets-OCR-s":
         processor = processor_v
         model = model_v
-    elif model_name == "Aya-Vision":
+    elif model_name == "Aya-Vision-8B":
         processor = processor_a
         model = model_a
+    elif model_name == "MonkeyOCR-1.2B-0709":
+        processor = processor_w
+        model = model_w
     else:
         yield "Invalid model selected.", "Invalid model selected."
         return
@@ -171,7 +186,7 @@ def generate_video(model_name: str, text: str, video_path: str,
     Generates responses using the selected model for video input.
     Yields raw text and Markdown-formatted text.
     """
-    if model_name == "RolmOCR":
+    if model_name == "RolmOCR-7B":
         processor = processor_m
         model = model_m
     elif model_name == "Qwen2-VL-OCR":
@@ -180,9 +195,12 @@ def generate_video(model_name: str, text: str, video_path: str,
     elif model_name == "Nanonets-OCR-s":
         processor = processor_v
         model = model_v
-    elif model_name == "Aya-Vision":
+    elif model_name == "Aya-Vision-8B":
         processor = processor_a
         model = model_a
+    elif model_name == "MonkeyOCR-1.2B-0709":
+        processor = processor_w
+        model = model_w
     else:
         yield "Invalid model selected.", "Invalid model selected."
         return
@@ -284,6 +302,7 @@ with gr.Blocks(css=css, theme="bethecloud/storj_theme") as demo:
                 top_p = gr.Slider(label="Top-p (nucleus sampling)", minimum=0.05, maximum=1.0, step=0.05, value=0.9)
                 top_k = gr.Slider(label="Top-k", minimum=1, maximum=1000, step=1, value=50)
                 repetition_penalty = gr.Slider(label="Repetition penalty", minimum=1.0, maximum=2.0, step=0.05, value=1.2)
+                
         with gr.Column():
             with gr.Column(elem_classes="canvas-output"):
                 gr.Markdown("## Result.Md")
@@ -292,15 +311,17 @@ with gr.Blocks(css=css, theme="bethecloud/storj_theme") as demo:
                 with gr.Accordion("Formatted Result (Result.md)", open=False):
                     markdown_output = gr.Markdown(label="Formatted Result (Result.Md)")
             model_choice = gr.Radio(
-                choices=["Nanonets-OCR-s", "Qwen2-VL-OCR", "RolmOCR", "Aya-Vision"],
+                choices=["Nanonets-OCR-s", "MonkeyOCR-1.2B-0709", "Qwen2-VL-OCR", "RolmOCR-7B", "Aya-Vision-8B"],
                 label="Select Model",
                 value="Nanonets-OCR-s"
             )
             gr.Markdown("**Model Info 💻** | [Report Bug](https://huggingface.co/spaces/prithivMLmods/Multimodal-OCR/discussions)")
-            gr.Markdown("> [Qwen2-VL-OCR-2B-Instruct](https://huggingface.co/prithivMLmods/Qwen2-VL-OCR-2B-Instruct): qwen2-vl-ocr-2b-instruct model is a fine-tuned version of qwen2-vl-2b-instruct, tailored for tasks that involve [messy] optical character recognition (ocr), image-to-text conversion, and math problem solving with latex formatting.")
             gr.Markdown("> [Nanonets-OCR-s](https://huggingface.co/nanonets/Nanonets-OCR-s): nanonets-ocr-s is a powerful, state-of-the-art image-to-markdown ocr model that goes far beyond traditional text extraction. it transforms documents into structured markdown with intelligent content recognition and semantic tagging.")
+            gr.Markdown("> [MonkeyOCR-1.2B-0709](https://huggingface.co/echo840/MonkeyOCR-1.2B-0709): MonkeyOCR adopts a structure-recognition-relation (SRR) triplet paradigm, which simplifies the multi-tool pipeline of modular approaches while avoiding the inefficiency of using large multimodal models for full-page document processing.")
+            gr.Markdown("> [Qwen2-VL-OCR-2B-Instruct](https://huggingface.co/prithivMLmods/Qwen2-VL-OCR-2B-Instruct): qwen2-vl-ocr-2b-instruct model is a fine-tuned version of qwen2-vl-2b-instruct, tailored for tasks that involve [messy] optical character recognition (ocr), image-to-text conversion, and math problem solving with latex formatting.")
             gr.Markdown("> [RolmOCR](https://huggingface.co/reducto/RolmOCR): rolmocr, high-quality, openly available approach to parsing pdfs and other complex documents optical character recognition. it is designed to handle a wide range of document types, including scanned documents, handwritten text, and complex layouts.")
             gr.Markdown("> [Aya-Vision](https://huggingface.co/CohereLabs/aya-vision-8b): cohere labs aya vision 8b is an open weights research release of an 8-billion parameter model with advanced capabilities optimized for a variety of vision-language use cases, including ocr, captioning, visual reasoning, summarization, question answering, code, and more.")
+            
             gr.Markdown(">⚠️note: all the models in space are not guaranteed to perform well in video inference use cases.")
         
     image_submit.click(