File size: 7,915 Bytes
d4cac13
 
5427fa5
 
d4cac13
21fd9ba
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
b90bc53
21fd9ba
b90bc53
 
5427fa5
 
 
 
 
 
 
 
 
d4cac13
b90bc53
 
 
 
 
 
 
 
 
 
5427fa5
 
b90bc53
 
f140759
 
 
5427fa5
b90bc53
5427fa5
 
b90bc53
f140759
 
 
5427fa5
 
 
 
 
 
 
d4cac13
5427fa5
 
 
 
 
 
d4cac13
 
 
f140759
5427fa5
b90bc53
f140759
 
 
 
 
5427fa5
 
 
 
 
 
 
 
 
 
a84257c
5427fa5
 
a84257c
 
5427fa5
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
b90bc53
5427fa5
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
b90bc53
5427fa5
f140759
5427fa5
 
 
 
 
 
 
 
 
 
 
 
 
b90bc53
5427fa5
 
f140759
 
 
 
 
 
5427fa5
 
 
f140759
 
 
 
 
 
 
5427fa5
b90bc53
5427fa5
 
 
 
 
 
 
 
 
 
 
 
 
 
f140759
5427fa5
 
 
 
 
 
b90bc53
f140759
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
b90bc53
5427fa5
d4cac13
 
 
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
import gradio as gr
import os
import base64
import threading

login(token=os.environ["HF_TOKEN"])
repo_id = os.environ["REPO_ID"]

# Torch optimizations
import torch
torch.set_num_threads(1)
if torch.cuda.is_available():
    torch.backends.cudnn.benchmark = True

# Download generation logic from private repo
try:
    generate_file = hf_hub_download(
        repo_id=repo_id,
        filename="gen1.py",
        token=os.environ["HF_TOKEN"]
    )
    os.system(f"cp {generate_file} ./gen1.py")
except Exception as e:
    print(f"Error downloading files: {e}")

# Import the generation wrapper
from gen1 import setup_translation, translate_text as gen_translate

# Logo and favicon setup
LOGO_PATH = "static/logo.png"
if os.path.isfile(LOGO_PATH):
    with open(LOGO_PATH, "rb") as f:
        LOGO_B64 = base64.b64encode(f.read()).decode()
    LOGO_HTML = f'<img src="data:image/png;base64,{LOGO_B64}" alt="Maruth Labs Logo" style="height:40px;">'
    FAVICON_HTML = f'<link rel="icon" type="image/png" href="data:image/png;base64,{LOGO_B64}">'
else:
    LOGO_HTML = '<div style="width:40px;height:40px;background:#ccc;border-radius:4px;"></div>'
    FAVICON_HTML = ''

def init_translation_model():
    """Initialize the translation model from private repo"""
    success = setup_translation(
        repo_id=os.environ["REPO_ID"],
        token=os.environ["HF_TOKEN"]
    )
    if success:
        print("Model loaded successfully!")
    else:
        print("Failed to load model")

def translate_text(source_text, source_lang, target_lang, temperature, top_k, repetition_penalty, max_tokens):
    """Handle translation requests"""
    return gen_translate(
        source_text, source_lang, target_lang, 
        temperature, top_k, repetition_penalty, max_tokens
    )

# Language options
languages = ["English", "Hindi", "Bengali", "Tamil", "Telugu", "Kannada", "Panjabi"]

# Custom CSS
css_path = "static/style.css"
custom_css = open(css_path, encoding="utf-8").read() if os.path.isfile(css_path) else ""

theme_lock_css = """
.gradio-container .theme-toggle,
.gradio-container button[aria-label*="theme"],
.gradio-container button[title*="theme"],
.gradio-container .settings button,
.gradio-container [data-testid="theme-toggle"] {
    display: none !important;
}

:root { color-scheme: dark !important; }

body, .gradio-container {
    background-color: #0a1628 !important;
    color: #e6eef8 !important;
}
"""

combined_css = custom_css + theme_lock_css

# Theme configuration
locked_theme = gr.themes.Monochrome(
    primary_hue="blue",
    secondary_hue="slate",
    neutral_hue="slate"
).set(
    background_fill_primary="#0a1628",
    background_fill_secondary="#1f2937", 
    block_background_fill="#374151",
    border_color_primary="#374151",
    color_accent_soft="#2563eb",
    block_title_text_color="#e6eef8",
    block_label_text_color="#e6eef8",
    body_text_color="#e6eef8"
)

# Create Gradio interface
with gr.Blocks(
    title="Madhuram Translation - MaruthLabs",
    css=combined_css,
    theme=locked_theme,
    js=f"""
    function() {{
        {f'document.head.insertAdjacentHTML("beforeend", `{FAVICON_HTML}`);' if FAVICON_HTML else ''}
        
        document.documentElement.setAttribute('data-theme', 'dark');
        document.body.classList.add('dark');
        document.body.classList.remove('light');

        const observer = new MutationObserver(function(mutations) {{
            mutations.forEach(function(mutation) {{
                mutation.addedNodes.forEach(function(node) {{
                    if (node.nodeType === 1) {{
                        const toggles = node.querySelectorAll('.theme-toggle, button[aria-label*="theme"], button[title*="theme"]');
                        toggles.forEach(toggle => toggle.style.display = 'none');
                    }}
                }});
            }});
        }});
        observer.observe(document.body, {{ childList: true, subtree: true }});
    }}
    """
) as demo:

    # Header section
    with gr.Row(elem_classes="main-header"):
        with gr.Column():
            gr.HTML(f"""
            <div style="display:flex;align-items:center;justify-content:space-between;width:100%;">
                <!-- left: logo + text on one line -->
                <div style="display:flex;align-items:center;">
                    {LOGO_HTML}
                    <h3 style="margin-left:8px;margin-top:0;margin-bottom:0;">Maruth Labs</h3>
                </div>

                <!-- center title -->
                <div class="main-title"><h1>Madhuram Translation Model</h1></div>

                <!-- spacer to balance flex -->
                <div style="width:120px;"></div>
            </div>
            """)

    # Main interface
    with gr.Row(equal_height=False):
        # Settings panel
        with gr.Column(scale=1.5, elem_classes="settings-panel"):
            gr.Markdown("## Translation Settings")
            with gr.Row():
                source_lang = gr.Dropdown(choices=languages, label="Source Language", value="English")
                target_lang = gr.Dropdown(choices=languages, label="Target Language", value="Hindi")
            swap_btn = gr.Button("Swap Languages", variant="secondary", size="sm")

            with gr.Accordion("Advanced Settings", open=False):
                temperature = gr.Slider(0.001, 1.001, 0.001, step=0.1, label="Temperature")
                top_k = gr.Slider(1, 100, 10, step=1, label="Top-k")
                repetition_penalty = gr.Slider(1.0, 2.0, 1.2, step=0.1, label="Repetition Penalty")
                max_tokens = gr.Slider(100, 2000, 400, step=50, label="Max Tokens")

        # Translation interface
        with gr.Column(scale=2, elem_classes="translation-card"):
            gr.Markdown("## Translation Interface")
            source_text = gr.Textbox(
                label="Enter text to translate", 
                placeholder="Type or paste your text here", 
                lines=6, 
                max_lines=12
            )
            with gr.Row():
                translate_btn = gr.Button("Translate", variant="primary", size="lg")
                clear_btn = gr.Button("Clear All", variant="secondary", size="lg")
            translated_text = gr.Textbox(
                label="Translation Result", 
                lines=6, 
                max_lines=12, 
                interactive=False, 
                placeholder="Translation will appear here"
            )

    # Examples section
    with gr.Row():
        with gr.Column():
            gr.Markdown("### Quick Examples")
            gr.Examples(
                examples=[
                    ["Hello, how are you today?", "English", "Hindi"],
                    ["তুমি কোথায় যাচ্ছ?", "Bengali", "English"],
                    ["நீங்கள் எப்படி இருக்கிறீர்கள்?", "Tamil", "Telugu"], 
                    ["ನಿನ್ನ ಹೆಸರು ಏನು?", "Kannada", "English"],
                    ["ਸਤ ਸ੍ਰੀ ਅਕਾਲ", "Panjabi", "Hindi"],   
                ],
                inputs=[source_text, source_lang, target_lang],
            )

    # Event handlers
    def swap_languages(src, tgt):
        return tgt, src
    
    def clear_all():
        return "", ""

    # Connect event handlers
    swap_btn.click(
        fn=swap_languages, 
        inputs=[source_lang, target_lang], 
        outputs=[source_lang, target_lang]
    )
    
    clear_btn.click(
        fn=clear_all, 
        outputs=[source_text, translated_text]
    )
    
    translate_btn.click(
        fn=translate_text, 
        inputs=[source_text, source_lang, target_lang, temperature, top_k, repetition_penalty, max_tokens], 
        outputs=[translated_text]
    )
    
    # Load model when demo starts
    demo.load(fn=init_translation_model)

if __name__ == "__main__":
    demo.launch()