Update app.py

app.py

@@ -1,63 +1,118 @@
+import os
 import gradio as gr
-from huggingface_hub import InferenceClient
+from http import HTTPStatus
+import openai
+from typing import Generator, List, Optional, Tuple, Dict
+from urllib.error import HTTPError
 
-"""
-For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
-"""
-client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
+API_URL = os.getenv('API_URL')
+API_KEY = os.getenv('API_KEY')
+CUSTOM_JS = os.getenv('CUSTOM_JS', None)
+oai_client = openai.OpenAI(api_key=API_KEY, base_url=API_URL)
 
+History = List[Tuple[str, str]]
+Messages = List[Dict[str, str]]
 
-def respond(
-    message,
-    history: list[tuple[str, str]],
-    system_message,
-    max_tokens,
-    temperature,
-    top_p,
-):
-    messages = [{"role": "system", "content": system_message}]
+def clear_session() -> History:
+    return '', []
 
-    for val in history:
-        if val[0]:
-            messages.append({"role": "user", "content": val[0]})
-        if val[1]:
-            messages.append({"role": "assistant", "content": val[1]})
+def history_to_messages(history: History) -> Messages:
+    messages = []
+    for h in history:
+        messages.append({'role': 'user', 'content': h[0].strip()})
+        messages.append({'role': 'assistant', 'content': h[1].strip()})
+    return messages
 
-    messages.append({"role": "user", "content": message})
+def messages_to_history(messages: Messages) -> Tuple[str, History]:
+    history = []
+    for q, r in zip(messages[0::2], messages[1::2]):
+        history.append([q['content'], r['content']])
+    return history
 
-    response = ""
+def model_chat(query: Optional[str], history: Optional[History]) -> Generator[Tuple[str, History], None, None]:
+    if query is None:
+        query = ''
+    if history is None:
+        history = []
+    if not query.strip():
+        return
+    messages = history_to_messages(history)
+    messages.append({'role': 'user', 'content': query.strip()})
+    gen = oai_client.chat.completions.create(
+        model='dicta-il/dictalm2.0-instruct',
+        messages=messages,
+        temperature=0.7,
+        max_tokens=1024,
+        top_p=0.9,
+        stream=True
+    )
+    full_response = ''
+    for completion in gen:
+        text = completion.choices[0].delta.content
+        full_response += text or ''
+        yield full_response
+        
+with gr.Blocks(css='''
+    .gr-group {direction: rtl;}
+    .chatbot{text-align:right;}
+  .dicta-header {
+    background-color: var(--input-background-fill);  /* Replace with desired background color */
+    border-radius: 10px;
+    padding: 20px;
+    text-align: center;
+    display: flex;
+    flex-direction: row;
+    align-items: center;
+    box-shadow: var(--block-shadow);
+    border-color: var(--block-border-color);
+    border-width: 1px;
+  }
+               
+               
+  @media (max-width: 768px) {
+    .dicta-header {
+      flex-direction: column; /* Change to vertical for mobile devices */
+    }
+  }
+  .chatbot.prose {
+    font-size: 1.2em;
+  }
+  .dicta-logo {
+    width: 150px; /* Replace with actual logo width as desired */
+    height: auto;
+    margin-bottom: 20px;
+  }
+  .dicta-intro-text {
+    margin-bottom: 20px;
+    text-align: center;
+    display: flex;
+    flex-direction: column;
+    align-items: center;
+    width: 100%;
+    font-size: 1.1em;
+  }
+               
+  textarea {
+    font-size: 1.2em;
+  }
+''', js=CUSTOM_JS) as demo:
+    gr.Markdown("""
+<div class="dicta-header">
+ 
+  <div class="dicta-intro-text">
+    <h1>DictaLM 2.0 - Instruct Chat Demo</h1>
+    <p>Welcome to the interactive demo of DictaLM-2.0. Explore the capabilities of our model and see how it can assist with your tasks.<br/>
+    <span dir='rtl'> ברוכים הבאים לדמו האינטראקטיבי של DictaLM-2.0. חקרו את יכולות המודל שלנו וראו כיצד הוא יכול לסייע לכם במשימותיכם.</span><br/>
+    <span dir='rtl'> המודל משוחרר לנחלת הכלל ואפשר להורידו בקישור: <a href="https://huggingface.co/dicta-il/dictalm2.0-instruct">כאן</a></span></p>
+  </div>
+</div>
+""")
+    
+    interface = gr.ChatInterface(model_chat, fill_height=False)
+    interface.chatbot.rtl = True
+    interface.textbox.placeholder = "הכנס שאלה בעברית (או באנגלית!)"
+    interface.textbox.rtl = True
+    interface.textbox.text_align = 'right'
+    interface.theme_css += '.gr-group {direction: rtl !important;}'
 
-    for message in client.chat_completion(
-        messages,
-        max_tokens=max_tokens,
-        stream=True,
-        temperature=temperature,
-        top_p=top_p,
-    ):
-        token = message.choices[0].delta.content
-
-        response += token
-        yield response
-
-"""
-For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
-"""
-demo = gr.ChatInterface(
-    respond,
-    additional_inputs=[
-        gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
-        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
-        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
-        gr.Slider(
-            minimum=0.1,
-            maximum=1.0,
-            value=0.95,
-            step=0.05,
-            label="Top-p (nucleus sampling)",
-        ),
-    ],
-)
-
-
-if __name__ == "__main__":
-    demo.launch()
+demo.queue(api_open=False).launch(max_threads=20, share=False)