File size: 8,354 Bytes
d43c6a1
 
 
 
 
 
 
3dfc0a6
d43c6a1
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
494d93f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
bd0b898
d43c6a1
494d93f
 
 
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
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
import gradio as gr
from PIL import Image
from inference.main import MultiModalPhi2

messages = []

multimodal_phi2 = MultiModalPhi2(
    modelname_or_path="Navyabhat/Llava-Phi2",
    temperature=0.2,
    max_new_tokens=1024,
    device="cpu",
)


def add_content(chatbot, text, image, audio_upload, audio_mic) -> gr.Chatbot:
    textflag, imageflag, audioflag = False, False, False
    if text not in ["", None]:
        chatbot.append((text, None))
        textflag = True
    if image is not None:
        chatbot.append(((image,), None))
        imageflag = True
    if audio_mic is not None:
        chatbot.append(((audio_mic,), None))
        audioflag = True
    else:
        if audio_upload is not None:
            chatbot.append(((audio_upload,), None))
            audioflag = True
    if not any([textflag, imageflag, audioflag]):
        # Raise an error if neither text nor file is provided
        raise gr.Error("Enter a valid text, image or audio")
    return chatbot


def clear_data():
    return {prompt: None, image: None, audio_upload: None, audio_mic: None, chatbot: []}


def run(history, text, image, audio_upload, audio_mic):
    if text in [None, ""]:
        text = None

    if audio_upload is not None:
        audio = audio_upload
    elif audio_mic is not None:
        audio = audio_mic
    else:
        audio = None

    print("text", text)
    print("image", image)
    print("audio", audio)

    if image is not None:
        image = Image.open(image)
    outputs = multimodal_phi2(text, audio, image)
    # outputs = ""

    history.append((None, outputs.title()))
    return history, None, None, None, None


# # Custom styling
# interface_style = {
#     "box": {
#         "backgroundColor": "#f9f9f9",
#         "padding": "20px",
#         "borderRadius": "10px",
#         "boxShadow": "0 0 10px rgba(0, 0, 0, 0.1)",
#     },
#     "button": {
#         "backgroundColor": "#4caf50",
#         "color": "#fff",
#         "padding": "10px",
#         "border": "none",
#         "borderRadius": "5px",
#         "cursor": "pointer",
#     },
#     "textbox": {
#         "width": "100%",
#         "padding": "10px",
#         "marginBottom": "10px",
#         "boxSizing": "border-box",
#     },
#     "image": {
#         "width": "100%",
#         "marginBottom": "10px",
#     },
#     "audio": {
#         "width": "100%",
#         "marginBottom": "10px",
#     },
#     "chatbox": {
#         "height": "550px",
#         "backgroundColor": "#f0f0f0",
#         "borderRadius": "5px",
#         "padding": "10px",
#         "overflowY": "auto",
#     },
# }

# with gr.Blocks() as demo:
#     gr.Markdown("## MultiModal Phi2 Model Pretraining and Finetuning from Scratch")

#     with gr.Row():
#         with gr.Column(scale=4):
#             with gr.Box(style=interface_style["box"]):
#                 with gr.Row():
#                     prompt = gr.Textbox(
#                         placeholder="Enter Prompt",
#                         lines=2,
#                         label="Query",
#                         value=None,
#                         style=interface_style["textbox"],
#                     )
#                 with gr.Row():
#                     image = gr.Image(
#                         type="filepath", value=None, style=interface_style["image"]
#                     )
#                 with gr.Row():
#                     audio_upload = gr.Audio(
#                         source="upload", type="filepath", style=interface_style["audio"]
#                     )
#                     audio_mic = gr.Audio(
#                         source="microphone",
#                         type="filepath",
#                         format="mp3",
#                         style=interface_style["audio"],
#                     )

#         with gr.Column(scale=8):
#             with gr.Box(style=interface_style["box"]):
#                 with gr.Row():
#                     chatbot = gr.Chatbot(
#                         avatar_images=("🧑", "🤖"),
#                         height=550,
#                         style=interface_style["chatbox"],
#                     )
#                 with gr.Row():
#                     submit = gr.Button(style=interface_style["button"])
#                     clear = gr.Button(value="Clear", style=interface_style["button"])

#     submit.click(
#         add_content,
#         inputs=[chatbot, prompt, image, audio_upload, audio_mic],
#         outputs=[chatbot],
#     ).success(
#         run,
#         inputs=[chatbot, prompt, image, audio_upload, audio_mic],
#         outputs=[chatbot, prompt, image, audio_upload, audio_mic],
#     )

#     clear.click(
#         clear_data,
#         outputs=[prompt, image, audio_upload, audio_mic, chatbot],
#     )

# demo.launch()

custom_html = """
<!DOCTYPE html>
<html lang="en">
<head>
    <meta charset="UTF-8">
    <meta name="viewport" content="width=device-width, initial-scale=1.0">
    <style>
        body {
            font-family: 'Arial', sans-serif;
            background-color: #f4f4f4;
            margin: 0;
            padding: 0;
        }
        #container {
            max-width: 800px;
            margin: 20px auto;
            padding: 20px;
            background-color: #fff;
            box-shadow: 0 0 10px rgba(0, 0, 0, 0.1);
            border-radius: 10px;
            display: flex;
            flex-direction: column;
            align-items: center;
        }
        label {
            font-weight: bold;
            margin-bottom: 5px;
            display: block;
        }
        textarea, input[type="file"] {
            width: 100%;
            padding: 10px;
            margin-bottom: 10px;
            box-sizing: border-box;
        }
        button {
            background-color: #4caf50;
            color: #fff;
            padding: 10px;
            border: none;
            border-radius: 5px;
            cursor: pointer;
        }
        button:hover {
            background-color: #45a049;
        }
        #chat-container {
            max-height: 550px;
            overflow-y: auto;
            border: 1px solid #ddd;
            padding: 10px;
            border-radius: 5px;
            background-color: #f0f0f0;
            margin-top: 10px;
        }
    </style>
</head>
<body>
    <div id="container">
        <h2 style="text-align: center;">MultiModal Phi2 Model Pretraining and Finetuning from Scratch</h2>
        <div style="display: flex; width: 100%;">
            <div style="flex: 1; margin-right: 10px;">
                <label for="inputText">Text Input:</label>
                <textarea id="inputText" name="inputText" placeholder="Enter Prompt" rows="4"></textarea>

                <label for="inputImage">Image Input:</label>
                <input type="file" id="inputImage" name="inputImage" accept="image/*">

                <label for="inputAudio">Audio Input:</label>
                <input type="file" id="inputAudio" name="inputAudio" accept="audio/*">
            </div>
            <div style="flex: 2;">
                <div id="chat-container"></div>
            </div>
        </div>
        <button id="submitBtn">Submit</button>
        <button id="clearBtn" style="background-color: #f44336;">Clear</button>
    </div>

    <script>
        document.getElementById('submitBtn').addEventListener('click', function() {
            // Add your logic to process and update the chat
            // Example: Update the chat with a placeholder response
            var inputText = document.getElementById('inputText').value;
            var chatContainer = document.getElementById('chat-container');
            var response = "<strong>User:</strong> " + inputText + "<br><strong>Model:</strong> This is a placeholder response.<br>";
            chatContainer.innerHTML += response;
        });

        document.getElementById('clearBtn').addEventListener('click', function() {
            // Clear input and chat
            document.getElementById('inputText').value = '';
            document.getElementById('inputImage').value = '';
            document.getElementById('inputAudio').value = '';
            document.getElementById('chat-container').innerHTML = '';
        });
    </script>
</body>
</html>
"""

with gr.Blocks() as demo:
    gr.HTML(custom_html)
    
demo.launch()