Update app.py

app.py

@@ -1,61 +1,117 @@
+import gradio as gr
+from huggingface_hub import InferenceClient
 from fastai.text.all import *
 from transformers import AutoTokenizer, AutoModelForSequenceClassification
 import torch
-import gradio as gr
+
+# Initialize Hugging Face Client
+client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
 
 # Load the medical model
 medical_learn = load_learner('model.pkl')
 
 # Medical model configuration
-medical_description = "Medical Diagnosis"
 medical_categories = ['Allergy', 'Anemia', 'Bronchitis', 'Diabetes', 'Diarrhea', 'Fatigue', 'Flu', 'Malaria', 'Stress']
 
 def classify_medical_text(txt):
-    pred, idx, probs = medical_learn.predict(txt)
-    return dict(zip(medical_categories, map(float, probs)))
+    try:
+        pred, idx, probs = medical_learn.predict(txt)
+        return dict(zip(medical_categories, map(float, probs)))
+    except Exception as e:
+        return {"error": str(e)}
 
-# Load the psychiatric model from Hugging Face
+# Load the psychiatric model
 psychiatric_model_name = "nlp4good/psych-search"  # Replace with the appropriate model
 psychiatric_tokenizer = AutoTokenizer.from_pretrained(psychiatric_model_name)
 psychiatric_model = AutoModelForSequenceClassification.from_pretrained(psychiatric_model_name)
 
 # Psychiatric model configuration
-psychiatric_description = "Psychiatric Analysis"
-psychiatric_labels = ['Depression', 'Anxiety', 'Bipolar Disorder', 'PTSD', 'OCD', 'Stress', 'Schizophrenia']  # Adjust based on the model
+psychiatric_labels = ['Depression', 'Anxiety', 'Bipolar Disorder', 'PTSD', 'OCD', 'Stress', 'Schizophrenia']
 
 def classify_psychiatric_text(txt):
-    inputs = psychiatric_tokenizer(txt, return_tensors="pt", truncation=True, padding=True)
-    with torch.no_grad():
-        outputs = psychiatric_model(**inputs)
-    logits = outputs.logits
-    probabilities = torch.softmax(logits, dim=1).squeeze().tolist()
-    return dict(zip(psychiatric_labels, probabilities))
+    try:
+        inputs = psychiatric_tokenizer(txt, return_tensors="pt", truncation=True, padding=True)
+        with torch.no_grad():
+            outputs = psychiatric_model(**inputs)
+        logits = outputs.logits
+        probabilities = torch.softmax(logits, dim=1).squeeze().tolist()
+        return dict(zip(psychiatric_labels, probabilities))
+    except Exception as e:
+        return {"error": str(e)}
 
-# Gradio Interfaces
-medical_text = gr.Textbox(lines=2, label='Describe your symptoms in detail')
-medical_label = gr.Label()
-medical_examples = ['I feel short of breath and have a high fever.', 'My throat hurts and I keep sneezing.', 'I am always thirsty.']
+# Chat-based Interface
+def respond(
+    message,
+    history: list[tuple[str, str]],
+    system_message,
+    max_tokens,
+    temperature,
+    top_p,
+):
+    messages = [{"role": "system", "content": system_message}]
+    for val in history:
+        if val[0]:
+            messages.append({"role": "user", "content": val[0]})
+        if val[1]:
+            messages.append({"role": "assistant", "content": val[1]})
+    messages.append({"role": "user", "content": message})
 
-psychiatric_text = gr.Textbox(lines=2, label='Describe your mental health concerns in detail')
-psychiatric_label = gr.Label()
-psychiatric_examples = ['I feel hopeless and have no energy.', 'I am unable to concentrate and feel anxious all the time.', 'I have recurring intrusive thoughts.']
+    response = ""
+    try:
+        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
+    except Exception as e:
+        yield f"Error: {str(e)}"
 
+# Gradio Interfaces
 medical_interface = gr.Interface(
     fn=classify_medical_text,
-    inputs=medical_text,
-    outputs=medical_label,
-    examples=medical_examples,
-    description=medical_description,
+    inputs=gr.Textbox(lines=2, label="Describe your symptoms in detail"),
+    outputs=gr.Label(label="Medical Diagnosis"),
+    examples=["I feel short of breath and have a high fever.", "My throat hurts and I keep sneezing.", "I am always thirsty."],
+    description="Identify potential medical conditions based on symptoms."
 )
 
 psychiatric_interface = gr.Interface(
     fn=classify_psychiatric_text,
-    inputs=psychiatric_text,
-    outputs=psychiatric_label,
-    examples=psychiatric_examples,
-    description=psychiatric_description,
+    inputs=gr.Textbox(lines=2, label="Describe your mental health concerns in detail"),
+    outputs=gr.Label(label="Psychiatric Analysis"),
+    examples=["I feel hopeless and have no energy.", "I am unable to concentrate and feel anxious all the time.", "I have recurring intrusive thoughts."],
+    description="Analyze potential mental health concerns based on input."
 )
 
-# Combine interfaces using Tabs
-app = gr.TabbedInterface([medical_interface, psychiatric_interface], ["Medical Diagnosis", "Psychiatric Analysis"])
-app.launch(inline=False)
+chat_interface = 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)"),
+    ],
+    description="Chat with an AI assistant for general inquiries or extended conversation."
+)
+
+# Unified Gradio App with Tabs
+with gr.Blocks() as app:
+    gr.Markdown("# Unified Medical and Psychiatric Assistant")
+
+    with gr.Tab("Chat Assistant"):
+        chat_interface.render()
+
+    with gr.Tab("Medical Diagnosis"):
+        medical_interface.render()
+
+    with gr.Tab("Psychiatric Analysis"):
+        psychiatric_interface.render()
+
+# Launch the App
+if __name__ == "__main__":
+    app.launch()