Update app.py

app.py

@@ -2,16 +2,17 @@ from huggingface_hub import InferenceClient
 import gradio as gr
 from deep_translator import GoogleTranslator
 
+# Initialize the Hugging Face client with your model
 client = InferenceClient("mistralai/Mixtral-8x7B-Instruct-v0.1")
 
-# Function to translate Arabic text to English
-def translate_to_english(text):
-    return GoogleTranslator(source='arabic', target='english').translate(text)
-
-# Function to translate English text to Arabic
+# Define the translation functions
 def translate_to_arabic(text):
-    return GoogleTranslator(source='english', target='arabic').translate(text)
+    return GoogleTranslator(source='auto', target='ar').translate(text)
+
+def translate_to_english(text):
+    return GoogleTranslator(source='auto', target='en').translate(text)
 
+# Format the prompt for the model
 def format_prompt(message, history):
     prompt = "<s>"
     for user_prompt, bot_response in history:
@@ -20,88 +21,50 @@ def format_prompt(message, history):
     prompt += f"[INST] {message} [/INST]"
     return prompt
 
-def generate(
-    prompt, history, temperature=0.1, max_new_tokens=256, top_p=0.95, repetition_penalty=1.0,
-):
-    # Initialize history as an empty list if it's None
-    if history is None:
-        history = []
-
-    temperature = float(temperature)
-    if temperature < 1e-2:
-        temperature = 1e-2
-    top_p = float(top_p)
-
-    generate_kwargs = dict(
-        temperature=temperature,
-        max_new_tokens=max_new_tokens,
-        top_p=top_p,
-        repetition_penalty=repetition_penalty,
-        do_sample=True,
-        seed=42,
-    )
-
-    # Translate the Arabic prompt to English
-    english_prompt = translate_to_english(prompt)
-
-    formatted_prompt = format_prompt(english_prompt, history)
-
+# Generate a response from the model
+def generate(prompt, history=[], temperature=0.1, max_new_tokens=256, top_p=0.95, repetition_penalty=1.0):
+    # Translate the Arabic prompt to English before sending to the model
+    prompt_in_english = translate_to_english(prompt)
+    formatted_prompt = format_prompt(prompt_in_english, history)
+
+    generate_kwargs = {
+        "temperature": temperature,
+        "max_new_tokens": max_new_tokens,
+        "top_p": top_p,
+        "repetition_penalty": repetition_penalty,
+        "do_sample": True,
+        "seed": 42,  # Seed for reproducibility, remove or change if randomness is preferred
+    }
+
+    # Generate the response
     stream = client.text_generation(formatted_prompt, **generate_kwargs, stream=True, details=True, return_full_text=False)
+    
     output = ""
-
     for response in stream:
-        output += response.token.text
+        output += response["token"]["text"]
 
     # Translate the English response back to Arabic
-    arabic_output = translate_to_arabic(output)
-
-    # Update the history state with the latest exchange
-    history.append((prompt, arabic_output))
-
-    return arabic_output, history
-
-additional_inputs=[
-    gr.Slider(
-        label="Temperature",
-        value=0.9,
-        minimum=0.0,
-        maximum=1.0,
-        step=0.05,
-        interactive=True,
-        info="Higher values produce more diverse outputs",
-    ),
-    gr.Slider(
-        label="Max new tokens",
-        value=256,
-        minimum=0,
-        maximum=1048,
-        step=64,
-        interactive=True,
-        info="The maximum numbers of new tokens",
-    ),
-    gr.Slider(
-        label="Top-p (nucleus sampling)",
-        value=0.90,
-        minimum=0.0,
-        maximum=1,
-        step=0.05,
-        interactive=True,
-        info="Higher values sample more low-probability tokens",
-    ),
-    gr.Slider(
-        label="Repetition penalty",
-        value=1.2,
-        minimum=1.0,
-        maximum=2.0,
-        step=0.05,
-        interactive=True,
-        info="Penalize repeated tokens",
-    )
+    response_in_arabic = translate_to_arabic(output)
+    return response_in_arabic
+
+# Define additional inputs for Gradio interface
+additional_inputs = [
+    gr.Slider(label="Temperature", value=0.9, minimum=0.0, maximum=1.0, step=0.05),
+    gr.Slider(label="Max new tokens", value=256, minimum=0, maximum=1048, step=64),
+    gr.Slider(label="Top-p (nucleus sampling)", value=0.90, minimum=0.0, maximum=1.0, step=0.05),
+    gr.Slider(label="Repetition penalty", value=1.2, minimum=1.0, maximum=2.0, step=0.05)
 ]
 
-gr.ChatInterface(
+# Set up the Gradio interface
+iface = gr.Interface(
     fn=generate,
-    chatbot=gr.Chatbot(show_label=False, show_share_button=False, show_copy_button=True, likeable=True, layout="panel"),
-    additional_inputs=additional_inputs,
-    title="DorjGPT interface"
-).launch(show_api=True)
+    inputs=[
+        gr.Textbox(lines=5, placeholder='Type your Arabic query here...', label='Arabic Query'),
+        *additional_inputs
+    ],
+    outputs='text',
+    title="DorjGPT Arabic-English Translation Chatbot",
+)
+
+# Launch the Gradio interface
+iface.launch()