update app

app.py

@@ -1,22 +1,100 @@
 import os
 import sys
+import random
+import uuid
+import json
 import time
 from threading import Thread
+from typing import Iterable
+from huggingface_hub import snapshot_download
 
 import gradio as gr
 import spaces
 import torch
+import numpy as np
 from PIL import Image
+import cv2
 
 from transformers import (
     Qwen2_5_VLForConditionalGeneration,
     Qwen3VLForConditionalGeneration,
+    AutoModelForImageTextToText,
     AutoModelForCausalLM,
     AutoProcessor,
     TextIteratorStreamer,
 )
 
-from theme import steel_blue_theme
+from transformers.image_utils import load_image
+from gradio.themes import Soft
+from gradio.themes.utils import colors, fonts, sizes
+
+colors.steel_blue = colors.Color(
+    name="steel_blue",
+    c50="#EBF3F8",
+    c100="#D3E5F0",
+    c200="#A8CCE1",
+    c300="#7DB3D2",
+    c400="#529AC3",
+    c500="#4682B4",
+    c600="#3E72A0",
+    c700="#36638C",
+    c800="#2E5378",
+    c900="#264364",
+    c950="#1E3450",
+)
+
+class SteelBlueTheme(Soft):
+    def __init__(
+        self,
+        *,
+        primary_hue: colors.Color | str = colors.gray,
+        secondary_hue: colors.Color | str = colors.steel_blue,
+        neutral_hue: colors.Color | str = colors.slate,
+        text_size: sizes.Size | str = sizes.text_lg,
+        font: fonts.Font | str | Iterable[fonts.Font | str] = (
+            fonts.GoogleFont("Outfit"), "Arial", "sans-serif",
+        ),
+        font_mono: fonts.Font | str | Iterable[fonts.Font | str] = (
+            fonts.GoogleFont("IBM Plex Mono"), "ui-monospace", "monospace",
+        ),
+    ):
+        super().__init__(
+            primary_hue=primary_hue,
+            secondary_hue=secondary_hue,
+            neutral_hue=neutral_hue,
+            text_size=text_size,
+            font=font,
+            font_mono=font_mono,
+        )
+        super().set(
+            background_fill_primary="*primary_50",
+            background_fill_primary_dark="*primary_900",
+            body_background_fill="linear-gradient(135deg, *primary_200, *primary_100)",
+            body_background_fill_dark="linear-gradient(135deg, *primary_900, *primary_800)",
+            button_primary_text_color="white",
+            button_primary_text_color_hover="white",
+            button_primary_background_fill="linear-gradient(90deg, *secondary_500, *secondary_600)",
+            button_primary_background_fill_hover="linear-gradient(90deg, *secondary_600, *secondary_700)",
+            button_primary_background_fill_dark="linear-gradient(90deg, *secondary_600, *secondary_800)",
+            button_primary_background_fill_hover_dark="linear-gradient(90deg, *secondary_500, *secondary_500)",
+            button_secondary_text_color="black",
+            button_secondary_text_color_hover="white",
+            button_secondary_background_fill="linear-gradient(90deg, *primary_300, *primary_300)",
+            button_secondary_background_fill_hover="linear-gradient(90deg, *primary_400, *primary_400)",
+            button_secondary_background_fill_dark="linear-gradient(90deg, *primary_500, *primary_600)",
+            button_secondary_background_fill_hover_dark="linear-gradient(90deg, *primary_500, *primary_500)",
+            slider_color="*secondary_500",
+            slider_color_dark="*secondary_600",
+            block_title_text_weight="600",
+            block_border_width="3px",
+            block_shadow="*shadow_drop_lg",
+            button_primary_shadow="*shadow_drop_lg",
+            button_large_padding="11px",
+            color_accent_soft="*primary_100",
+            block_label_background_fill="*primary_200",
+        )
+
+steel_blue_theme = SteelBlueTheme()
 
 css = """
 #main-title h1 {
@@ -44,6 +122,43 @@ if torch.cuda.is_available():
 
 print("Using device:", device)
 
+# CACHE_PATH = "./model_cache"
+# if not os.path.exists(CACHE_PATH):
+#     os.makedirs(CACHE_PATH)
+#
+# model_path_d_local = snapshot_download(
+#     repo_id='rednote-hilab/dots.ocr',
+#     local_dir=os.path.join(CACHE_PATH, 'dots.ocr'),
+#     max_workers=20,
+#     local_dir_use_symlinks=False
+# )
+#
+# config_file_path = os.path.join(model_path_d_local, "configuration_dots.py")
+#
+# if os.path.exists(config_file_path):
+#     with open(config_file_path, 'r') as f:
+#         input_code = f.read()
+#
+#     lines = input_code.splitlines()
+#     if "class DotsVLProcessor" in input_code and not any("attributes = " in line for line in lines):
+#         output_lines = []
+#         for line in lines:
+#             output_lines.append(line)
+#             if line.strip().startswith("class DotsVLProcessor"):
+#                 output_lines.append("    attributes = [\"image_processor\", \"tokenizer\"]")
+#
+#         with open(config_file_path, 'w') as f:
+#             f.write('\n'.join(output_lines))
+#         print("Patched configuration_dots.py successfully.")
+#
+#sys.path.append(model_path_d_local)
+
+MAX_MAX_NEW_TOKENS = 4096
+DEFAULT_MAX_NEW_TOKENS = 2048
+MAX_INPUT_TOKEN_LENGTH = int(os.getenv("MAX_INPUT_TOKEN_LENGTH", "4096"))
+
+device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
+
 # Load Chandra-OCR
 MODEL_ID_V = "datalab-to/chandra"
 processor_v = AutoProcessor.from_pretrained(MODEL_ID_V, trust_remote_code=True)
@@ -82,7 +197,6 @@ model_m = Qwen2_5_VLForConditionalGeneration.from_pretrained(
     torch_dtype=torch.float16
 ).to(device).eval()
 
-
 @spaces.GPU
 def generate_image(model_name: str, text: str, image: Image.Image,
                    max_new_tokens: int, temperature: float, top_p: float,
@@ -146,7 +260,6 @@ def generate_image(model_name: str, text: str, image: Image.Image,
         time.sleep(0.01)
         yield buffer, buffer
 
-
 image_examples = [
     ["OCR the content perfectly.", "examples/3.jpg"],
     ["Perform OCR on the image.", "examples/1.jpg"],
@@ -167,30 +280,27 @@ with gr.Blocks(css=css, theme=steel_blue_theme) as demo:
             )
             
             with gr.Accordion("Advanced options", open=False):
-                max_new_tokens = gr.Slider(label="Max new tokens", minimum=1, maximum=MAX_MAX_NEW_TOKENS, step=1,
-                                           value=DEFAULT_MAX_NEW_TOKENS)
+                max_new_tokens = gr.Slider(label="Max new tokens", minimum=1, maximum=MAX_MAX_NEW_TOKENS, step=1, value=DEFAULT_MAX_NEW_TOKENS)
                 temperature = gr.Slider(label="Temperature", minimum=0.1, maximum=4.0, step=0.1, value=0.7)
                 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=1.1)
+                repetition_penalty = gr.Slider(label="Repetition penalty", minimum=1.0, maximum=2.0, step=0.05, value=1.1)
                 
         with gr.Column(scale=3):
-            gr.Markdown("## Output", elem_id="output-title")
-            output = gr.Textbox(label="Raw Output Stream", interactive=False, lines=11, show_copy_button=True)
-            with gr.Accordion("(Result.md)", open=False):
-                markdown_output = gr.Markdown(label="(Result.Md)")
-
-            model_choice = gr.Radio(
-                choices=["Nanonets-OCR2-3B", "Chandra-OCR", "Dots.OCR", "olmOCR-2-7B-1025"],
-                label="Select Model",
-                value="Nanonets-OCR2-3B"
-            )
+                gr.Markdown("## Output", elem_id="output-title")
+                output = gr.Textbox(label="Raw Output Stream", interactive=False, lines=11, show_copy_button=True)
+                with gr.Accordion("(Result.md)", open=False):
+                    markdown_output = gr.Markdown(label="(Result.Md)")
+
+                model_choice = gr.Radio(
+                    choices=["Nanonets-OCR2-3B", "Chandra-OCR", "Dots.OCR", "olmOCR-2-7B-1025"],
+                    label="Select Model",
+                    value="Nanonets-OCR2-3B"
+                )
 
     image_submit.click(
         fn=generate_image,
-        inputs=[model_choice, image_query, image_upload, max_new_tokens, temperature, top_p, top_k,
-                repetition_penalty],
+        inputs=[model_choice, image_query, image_upload, max_new_tokens, temperature, top_p, top_k, repetition_penalty],
         outputs=[output, markdown_output]
     )