File size: 8,064 Bytes
81c9f4d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
e22a631
c9f8b16
81c9f4d
e22a631
 
 
c9f8b16
e22a631
 
 
 
 
 
 
 
 
 
 
81c9f4d
 
 
 
e22a631
81c9f4d
 
 
 
 
7c5d7b1
e22a631
81c9f4d
dfadffa
81c9f4d
 
 
 
dfadffa
81c9f4d
 
 
 
e22a631
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
81c9f4d
e22a631
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
81c9f4d
 
e22a631
81c9f4d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
c9f8b16
81c9f4d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
c9f8b16
81c9f4d
 
 
c9f8b16
81c9f4d
c9f8b16
81c9f4d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
e22a631
81c9f4d
e22a631
81c9f4d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
#!/usr/bin/env python3
"""
Gradio app for Sanskrit text transcription using Qwen2.5-VL model
Based on quick_test_improved.py
"""

import gradio as gr
import torch
import base64
import io
from PIL import Image
from transformers import Qwen2_5_VLForConditionalGeneration, AutoProcessor
from qwen_vl_utils import process_vision_info
import os
import logging
import spaces

# Set up logging
logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)


# Load model at module level (global scope)
model_path = 'diabolic6045/Sanskrit-Qwen2.5-VL-7B-Instruct-OCR'

logger.info("Loading processor...")
processor = AutoProcessor.from_pretrained(model_path)

logger.info("Loading Sanskrit OCR model...")
# Check if CUDA is available, otherwise use CPU
device_map = "auto" if torch.cuda.is_available() else "cpu"
model = Qwen2_5_VLForConditionalGeneration.from_pretrained(
    model_path,
    torch_dtype=torch.bfloat16 if torch.cuda.is_available() else torch.float32,
    device_map=device_map
)

model.eval()
device = next(model.parameters()).device
logger.info(f"Model loaded on device: {device}")

def check_model_status():
    """Check if model is loaded and ready"""
    try:
        if model is not None and processor is not None:
            return "βœ… Model loaded and ready"
        else:
            return "⏳ Model not loaded yet"
    except Exception as e:
        return f"❌ Model error: {str(e)}"

@spaces.GPU
def transcribe_sanskrit(image, custom_prompt, progress=gr.Progress()):
    """Gradio interface function for transcription using pre-loaded model"""
    if image is None:
        return "Please upload an image first."
    
    try:
        progress(0.1, desc="Processing image...")
        
        # Use custom prompt if provided, otherwise use default
        prompt = custom_prompt if custom_prompt.strip() else "Please transcribe the Sanskrit text shown in this image:"
        
        # Format the conversation using chat template
        messages = [
            {
                "role": "user",
                "content": [
                    {"type": "image", "image": image},
                    {"type": "text", "text": prompt}
                ]
            }
        ]
        
        # Preparation for inference
        text = processor.apply_chat_template(
            messages, tokenize=False, add_generation_prompt=True
        )
        image_inputs, video_inputs = process_vision_info(messages)
        inputs = processor(
            text=[text],
            images=image_inputs,
            videos=video_inputs,
            padding=True,
            return_tensors="pt",
        )
        
        # Get model device and move inputs there
        model_device = next(model.parameters()).device
        inputs = {k: v.to(model_device) for k, v in inputs.items()}
        
        progress(0.5, desc="Generating transcription...")
        with torch.no_grad():
            generated_ids = model.generate(
                **inputs,
                max_new_tokens=512,
                do_sample=False,
                pad_token_id=processor.tokenizer.eos_token_id,
                use_cache=True,
                repetition_penalty=1.1
            )
        
        # Extract only the generated part
        generated_ids_trimmed = [
            out_ids[len(in_ids) :] for in_ids, out_ids in zip(inputs['input_ids'], generated_ids)
        ]
        output_text = processor.batch_decode(
            generated_ids_trimmed, skip_special_tokens=True, clean_up_tokenization_spaces=False
        )
        
        progress(1.0, desc="Complete!")
        return output_text[0] if output_text else ""
        
    except Exception as e:
        logger.error(f"Error in transcribe_sanskrit: {e}")
        return f"❌ Error occurred: {str(e)}\n\nPlease try again or check if the model files are properly loaded."

def create_gradio_interface():
    """Create and configure the Gradio interface"""
    
    with gr.Blocks(
        title="Sanskrit Text Transcription",
        theme=gr.themes.Soft()
    ) as app:
        
        gr.HTML("""
        <div class="main-header">
            <h1>πŸ•‰οΈ Sanskrit Text Transcription</h1>
            <p>Upload an image containing Sanskrit text and get an accurate transcription using the specialized Sanskrit OCR model</p>
            <p><strong>πŸš€ Powered by ZeroGPU:</strong> Dynamic GPU allocation for efficient processing</p>
        </div>
        """)
        
        with gr.Row():
            with gr.Column(scale=1):
                gr.Markdown("### Upload Image")
                image_input = gr.Image(
                    type="pil",
                    label="Sanskrit Text Image",
                    height=400
                )
                
                gr.Markdown("### Custom Prompt (Optional)")
                custom_prompt = gr.Textbox(
                    label="Custom transcription prompt",
                    placeholder="Please transcribe the Sanskrit text shown in this image:",
                    lines=2,
                    value="Please transcribe the Sanskrit text shown in this image:"
                )
                
                transcribe_btn = gr.Button(
                    "πŸ•‰οΈ Transcribe Sanskrit Text",
                    variant="primary",
                    size="lg"
                )
                
                gr.Markdown("""
                ### Instructions:
                1. Upload an image containing Sanskrit text
                2. Optionally modify the prompt for better results
                3. Click the transcribe button
                4. View the transcribed text below
                """)
            
            with gr.Column(scale=1):
                gr.Markdown("### Transcription Result")
                output_text = gr.Textbox(
                    label="Transcribed Sanskrit Text",
                    lines=10,
                    max_lines=20,
                    show_copy_button=True
                )
                
                gr.Markdown("### Model Information")
                model_status = gr.Textbox(
                    label="Model Status",
                    value="Checking...",
                    interactive=False
                )
                
                check_status_btn = gr.Button("πŸ”„ Check Model Status", size="sm")
                
                gr.Markdown("""
                **Model:** diabolic6045/Sanskrit-Qwen2.5-VL-7B-Instruct-OCR
                
                **Features:**
                - Multimodal vision-language model
                - Pre-trained specifically for Sanskrit OCR
                - Supports various Sanskrit scripts
                - High accuracy Sanskrit text transcription
                """)
        
        
        # Event handlers
        transcribe_btn.click(
            fn=transcribe_sanskrit,
            inputs=[image_input, custom_prompt],
            outputs=output_text,
            show_progress=True
        )
        
        # Auto-transcribe when image is uploaded
        image_input.change(
            fn=transcribe_sanskrit,
            inputs=[image_input, custom_prompt],
            outputs=output_text,
            show_progress=True
        )
        
        # Model status check
        check_status_btn.click(
            fn=check_model_status,
            outputs=model_status
        )
        
        # Check model status on app load
        app.load(
            fn=check_model_status,
            outputs=model_status
        )
    
    return app

def main():
    """Main function to launch the Gradio app"""
    logger.info("Starting Sanskrit Transcription Gradio App...")
    
    # Create the interface
    app = create_gradio_interface()
    
    # Launch the app
    app.launch(
        server_name="0.0.0.0",  # Allow external access
        server_port=7860,       # Default Gradio port
        share=False,      # Enable request queuing
        max_threads=4           # Limit concurrent requests
    )

if __name__ == "__main__":
    main()