Update app.py

app.py

@@ -41,8 +41,8 @@ model_x = Qwen2_5_VLForConditionalGeneration.from_pretrained(
     MODEL_ID_X, trust_remote_code=True,
     torch_dtype=torch.float16).to(device).eval()
 
-# Load R1-Onevision-7B
-MODEL_ID_T = "FriendliAI/R1-Onevision-7B"
+# Load prithivMLmods/WR30a-Deep-7B-0711
+MODEL_ID_T = "prithivMLmods/WR30a-Deep-7B-0711"
 processor_t = AutoProcessor.from_pretrained(MODEL_ID_T, trust_remote_code=True)
 model_t = Qwen2_5_VLForConditionalGeneration.from_pretrained(
     MODEL_ID_T, trust_remote_code=True,
@@ -110,7 +110,7 @@ def generate_image(model_name: str,
     elif model_name == "Visionary-R1-3B":
         processor = processor_o
         model = model_o
-    elif model_name == "R1-Onevision-7B":
+    elif model_name == "WR30a-Deep-7B-0711":
         processor = processor_t
         model = model_t
     elif model_name == "MonkeyOCR-pro-1.2B":
@@ -177,7 +177,7 @@ def generate_video(model_name: str,
     elif model_name == "Visionary-R1-3B":
         processor = processor_o
         model = model_o
-    elif model_name == "R1-Onevision-7B":
+    elif model_name == "WR30a-Deep-7B-0711":
         processor = processor_t
         model = model_t
     elif model_name == "MonkeyOCR-pro-1.2B":
@@ -331,18 +331,18 @@ with gr.Blocks(css=css, theme="bethecloud/storj_theme") as demo:
                 #download_btn = gr.Button("Download Result.md")
 
             model_choice = gr.Radio(choices=[
-                 "Vision-Matters-7B", "R1-Onevision-7B",
+                 "Vision-Matters-7B", "WR30a-Deep-7B-0711",
                  "ViGaL-7B", "MonkeyOCR-pro-1.2B", "Visionary-R1-3B"
             ],
                                     label="Select Model",
                                     value="Vision-Matters-7B")
 
-            gr.Markdown("**Model Info 💻** | [Report Bug](https://huggingface.co/spaces/prithivMLmods/Multimodal-VLMs-5x/discussions)")          
+            gr.Markdown("**Model Info 💻** | [Report Bug](https://huggingface.co/spaces/prithivMLmods/Multimodal-VLMs-5x/discussions)")         
+            gr.Markdown("> [WR30a-Deep-7B-0711](https://huggingface.co/prithivMLmods/WR30a-Deep-7B-0711): wr30a-deep-7b-0711 model is a fine-tuned version of qwen2.5-vl-7b-instruct, optimized for image captioning, visual analysis, and image reasoning. Built on top of the qwen2.5-vl architecture, this experimental model enhances visual comprehension capabilities with focused training on 1,500k image pairs for superior image understanding.")
             gr.Markdown("> [MonkeyOCR-pro-1.2B](https://huggingface.co/echo840/MonkeyOCR-pro-1.2B): 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("> [Vision Matters 7B](https://huggingface.co/Yuting6/Vision-Matters-7B): vision-matters is a simple visual perturbation framework that can be easily integrated into existing post-training pipelines including sft, dpo, and grpo. our findings highlight the critical role of visual perturbation: better reasoning begins with better seeing.")
             gr.Markdown("> [ViGaL 7B](https://huggingface.co/yunfeixie/ViGaL-7B): vigal-7b shows that training a 7b mllm on simple games like snake using reinforcement learning boosts performance on benchmarks like mathvista and mmmu without needing worked solutions or diagrams indicating transferable reasoning skills.")
             gr.Markdown("> [Visionary-R1](https://huggingface.co/maifoundations/Visionary-R1): visionary-r1 is a novel framework for training visual language models (vlms) to perform robust visual reasoning using reinforcement learning (rl). unlike traditional approaches that rely heavily on (sft) or (cot) annotations, visionary-r1 leverages only visual question-answer pairs and rl, making the process more scalable and accessible.")
-            gr.Markdown("> [R1-Onevision-7B](https://huggingface.co/Fancy-MLLM/R1-Onevision-7B): r1-onevision model enhances vision-language understanding and reasoning capabilities, making it suitable for various tasks such as visual reasoning and image understanding. with its robust ability to perform multimodal reasoning, r1-onevision emerges as a powerful ai assistant capable of addressing different domains.")
             gr.Markdown(">⚠️note: all the models in space are not guaranteed to perform well in video inference use cases.")  
 
     # Define the submit button actions