add changes

app.py

@@ -65,19 +65,18 @@ def submit_chat(chatbot, text_input):
 def clear_chat():
     return [], None, "", "", 0.2, 0.7, 512
 
-with open(f"{cur_dir}/logo.svg", "r", encoding="utf-8") as svg_file:
-    svg_content = svg_file.read()
-font_size = "2.5em"
-svg_content = re.sub(r'(<svg[^>]*)(>)', rf'\1 height="{font_size}" style="vertical-align: middle; display: inline-block;"\2', svg_content)
-html = f"""
-<p align="center" style="font-size: {font_size}; line-height: 1;">
-    <span style="display: inline-block; vertical-align: middle;">{svg_content}</span>
-    <span style="display: inline-block; vertical-align: middle;">{model_name}</span>
-</p>
-<center><font size=3><b>{model_name}</b> Demo: Upload an image, provide a prompt, and get insights using advanced AI models. <a href='https://huggingface.co/jadechoghari/Ferret-UI-Gemma2b'>😊 Huggingface</a></font></center>
-"""
-with open(f"{cur_dir}/ferretui_icon.png", "rb") as image_file:
-    image_data = image_file.read()
+# with open(f"{cur_dir}/logo.svg", "r", encoding="utf-8") as svg_file:
+#     svg_content = svg_file.read()
+# font_size = "2.5em"
+# svg_content = re.sub(r'(<svg[^>]*)(>)', rf'\1 height="{font_size}" style="vertical-align: middle; display: inline-block;"\2', svg_content)
+# html = f"""
+# <p align="center" style="font-size: {font_size}; line-height: 1;">
+#     <span style="display: inline-block; vertical-align: middle;">{svg_content}</span>
+#     <span style="display: inline-block; vertical-align: middle;">{model_name}</span>
+# </p>
+# <center><font size=3><b>{model_name}</b> Demo: Upload an image, provide a prompt, and get insights using advanced AI models. <a href='https://huggingface.co/jadechoghari/Ferret-UI-Gemma2b'>😊 Huggingface</a></font></center>
+# """
+
 # html = f"""
 # <p align="center">
 #     <img src='data:image/png;base64,{image_data.encode("base64").decode("utf-8")}' alt='Ferret-UI' style='width: 100px; vertical-align: middle; border-radius: 15px; box-shadow: 0px 4px 10px rgba(0, 0, 0, 0.1);'/>
@@ -142,7 +141,7 @@ top_p_input = gr.Slider(minimum=0.0, maximum=1.0, step=0.05, value=0.7, label="T
 max_new_tokens_input = gr.Slider(minimum=1, maximum=1024, step=1, value=512, label="Max New Tokens")
 
 
-chatbot = gr.Chatbot(label="Chat with Ferret-UI", height=400, show_copy_button=True, latex_delimiters=latex_delimiters_set)
+chatbot = gr.Chatbot(label="Chat with Ferret-UI", height=400, show_copy_button=True, latex_delimiters=latex_delimiters_set, type="tuples")
 
 with gr.Blocks(title=model_name, theme=gr.themes.Ocean()) as demo:
     gr.HTML(html)

eval.json

@@ -1 +1 @@
-[{"id": 0, "image": "Screenshot 2024-10-13 at 12.01.05\u202fAM.png", "image_h": 76, "image_w": 90, "conversations": [{"from": "human", "value": "<image>\ndescribe what you see in details"}]}]
+[{"id": 0, "image": "appstore_reminders.png", "image_h": 2532, "image_w": 1170, "conversations": [{"from": "human", "value": "<image>\nWhat's inside the selected region?"}], "box_x1y1x2y2": [["189, 906, 404, 970"]]}]

eval_output.jsonl/0_of_1.jsonl

@@ -1 +1 @@
-{"id": 0, "image_path": "Screenshot 2024-10-13 at 12.01.05\u202fAM.png", "prompt": "describe what you see in details", "text": "The screen contains a large picture that occupies most of the screen, extending from nearly the top to the bottom. In the lower portion of the screen, there is a button labeled \"menu\". The button is relatively large and positioned at the lower part of the screen.", "label": null}
+{"id": 0, "image_path": "appstore_reminders.png", "prompt": "What's inside the selected region?", "text": "Inside the selected region [[28, 284, 217, 338]] there are several icons and text. The text '210k Ratings' is visible and there are two icons, one labeled 'unknown' and the other one is unknown.", "label": null}