push ui

.gitignore

@@ -0,0 +1 @@
+.env

README.md

@@ -1,14 +1,71 @@
----
-title: SLM Chatbot
-emoji: 💬
-colorFrom: yellow
-colorTo: purple
-sdk: gradio
-sdk_version: 5.0.1
-app_file: app.py
-pinned: false
-license: mit
-short_description: A web UI to chat with SLMs via the HF Inference API.
----
-
-An example chatbot using [Gradio](https://gradio.app), [`huggingface_hub`](https://huggingface.co/docs/huggingface_hub/v0.22.2/en/index), and the [Hugging Face Inference API](https://huggingface.co/docs/api-inference/index).
+# Gradio Chatbot : HuggingFace SLMs
+
+A modular Gradio-based application for interacting with various small language models through the Hugging Face API.
+
+## Project Structure
+
+```
+slm-poc/
+├── main.py                 # Main application entry point
+├── modules/
+│   ├── __init__.py         # Package initialization
+│   ├── config.py           # Configuration settings and constants
+│   ├── document_processor.py  # Document handling and processing
+│   └── model_handler.py    # Model interaction and response generation
+├── Dockerfile              # Docker configuration
+├── requirements.txt        # Python dependencies
+└── README.md               # Project documentation
+```
+
+## Features
+
+- Interactive chat interface with multiple language model options
+- Document processing (PDF, DOCX, TXT) for question answering
+- Adjustable model parameters (temperature, top_p, max_length)
+- Streaming responses for better user experience
+- Docker support for easy deployment
+
+## Setup and Running
+
+### Local Development
+
+1. Clone the repository
+2. Install dependencies:
+   ```
+   pip install -r requirements.txt
+   ```
+3. Create a `.env` file with your HuggingFace API token:
+   ```
+   HF_TOKEN=hf_your_token_here
+   ```
+4. Run the application:
+   ```
+   python main.py
+   ```
+
+### Docker Deployment
+
+1. Build the Docker image:
+   ```
+   docker build -t slm-poc .
+   ```
+2. Run the container:
+   ```
+   docker run -p 7860:7860 -e HF_TOKEN=hf_your_token_here slm-poc
+   ```
+
+## Usage
+
+1. Access the web interface at http://localhost:7860
+2. Enter your HuggingFace API token if not provided via environment variables
+3. Select your preferred model and adjust parameters
+4. Start chatting with the model
+5. Optionally upload documents for document-based Q&A
+
+## Supported Models
+
+T2T Inference models provided by Hugging Face via the Inference API
+
+## License
+
+This project is licensed under the MIT License - see the LICENSE file for details.

dockerfile

@@ -0,0 +1,28 @@
+FROM python:3.10-slim
+
+WORKDIR /app
+
+# Install system dependencies
+RUN apt-get update && apt-get install -y \
+    build-essential \
+    python3-dev \
+    && rm -rf /var/lib/apt/lists/*
+
+# Copy requirements first for better caching
+COPY requirements.txt .
+RUN pip install --no-cache-dir -r requirements.txt
+
+# Copy application code
+COPY main.py .
+COPY modules/ ./modules/
+
+# Environment variables
+ENV HF_TOKEN=""
+ENV GRADIO_SERVER_NAME="0.0.0.0"
+ENV GRADIO_SERVER_PORT=7860
+
+# Run the application
+CMD ["python", "main.py"]
+
+# Expose port for the application
+EXPOSE 7860

main.py

@@ -0,0 +1,168 @@
+# main.py
+import os
+import gradio as gr
+import tempfile
+from dotenv import load_dotenv
+from modules.document_processor import process_document
+from modules.model_handler import get_model_response, get_qa_response
+from modules.config import MODEL_MAPPING, DEFAULT_PARAMETERS
+
+# Load environment variables
+load_dotenv()
+
+def setup_api_key(api_key=None):
+    """Set up the HuggingFace API key from input or environment variables."""
+    if api_key and api_key.startswith('hf_'):
+        os.environ['HF_TOKEN'] = api_key
+        return True, "API key set successfully! ✅"
+    elif os.getenv('HF_TOKEN') and os.getenv('HF_TOKEN').startswith('hf_'):
+        return True, "API key already available! ✅"
+    else:
+        return False, "Please enter a valid HuggingFace API key. ⚠️"
+
+def create_chat_interface():
+    """Create the main chat interface for the application."""
+    with gr.Blocks(title="💬 Small Language Models - POC") as demo:
+        # Application header
+        gr.Markdown("# 💬 Small Language Models - POC")
+        gr.Markdown("This chatbot uses various Language Models such as Llama 3.2, Gemma 2, Gemma 3, Phi 3.5, DeepSeek-V3, and DeepSeek-R1.")
+        
+        with gr.Row():
+            with gr.Column(scale=1):
+                # Sidebar configuration
+                with gr.Group():
+                    api_key_input = gr.Textbox(
+                        label="HuggingFace API Token", 
+                        placeholder="Enter your HF API token (hf_...)", 
+                        type="password"
+                    )
+                    api_key_status = gr.Markdown("Please enter your API key.")
+                    api_key_button = gr.Button("Set API Key")
+                
+                with gr.Group():
+                    gr.Markdown("## Models and Parameters")
+                    model_dropdown = gr.Dropdown(
+                        choices=list(MODEL_MAPPING.keys()),
+                        label="Select Model",
+                        value=list(MODEL_MAPPING.keys())[0]
+                    )
+                    
+                    temperature_slider = gr.Slider(
+                        label="Temperature", 
+                        minimum=0.01, 
+                        maximum=1.0, 
+                        value=DEFAULT_PARAMETERS["temperature"],
+                        step=0.01
+                    )
+                    
+                    top_p_slider = gr.Slider(
+                        label="Top P", 
+                        minimum=0.01, 
+                        maximum=1.0, 
+                        value=DEFAULT_PARAMETERS["top_p"],
+                        step=0.01
+                    )
+                    
+                    max_length_slider = gr.Slider(
+                        label="Max Length", 
+                        minimum=20, 
+                        maximum=2040, 
+                        value=DEFAULT_PARAMETERS["max_length"],
+                        step=5
+                    )
+                    
+                    clear_button = gr.Button("Clear Chat History")
+                
+                with gr.Group():
+                    gr.Markdown("## Document Upload")
+                    file_upload = gr.File(
+                        label="Upload Document (PDF, DOCX, TXT)",
+                        file_types=["pdf", "docx", "txt"]
+                    )
+                    upload_status = gr.Markdown("")
+
+            with gr.Column(scale=2):
+                # Chat interface
+                chatbot = gr.Chatbot(
+                    label="Conversation",
+                    height=500,
+                    bubble_full_width=False
+                )
+                msg = gr.Textbox(
+                    label="Enter your message",
+                    placeholder="Type your message here...",
+                    show_label=False
+                )
+                
+                # State variables to track conversation and document processing
+                conversation_state = gr.State([])
+                document_store = gr.State(None)
+                api_key_state = gr.State(False)
+        
+        # Set up event handlers
+        api_key_button.click(
+            setup_api_key, 
+            inputs=[api_key_input],
+            outputs=[api_key_state, api_key_status]
+        )
+        
+        file_upload.upload(
+            process_document,
+            inputs=[file_upload, api_key_state],
+            outputs=[document_store, upload_status]
+        )
+        
+        # Function to handle chat messages
+        def respond(message, conversation, model_name, temp, top_p, max_len, doc_store, api_ready):
+            if not api_ready:
+                return conversation, conversation, "Please set a valid API key first. ⚠️"
+            
+            if not message.strip():
+                return conversation, conversation, upload_status.value
+            
+            # Update conversation with user message
+            conversation.append([message, None])
+            yield conversation, conversation, upload_status.value
+            
+            # Generate response based on whether document is uploaded
+            if doc_store is not None:
+                response = get_qa_response(
+                    message, 
+                    model_name, 
+                    doc_store, 
+                    {"temperature": temp, "top_p": top_p, "max_length": max_len}
+                )
+            else:
+                response = get_model_response(
+                    message, 
+                    conversation, 
+                    model_name, 
+                    {"temperature": temp, "top_p": top_p, "max_length": max_len}
+                )
+            
+            # Update conversation with assistant response
+            conversation[-1][1] = response
+            yield conversation, conversation, upload_status.value
+        
+        # Function to clear chat history
+        def clear_history():
+            return [], gr.update(value="Chat history cleared.")
+        
+        # Connect events
+        msg.submit(
+            respond,
+            [msg, conversation_state, model_dropdown, temperature_slider, top_p_slider, max_length_slider, document_store, api_key_state],
+            [chatbot, conversation_state, upload_status]
+        )
+        
+        clear_button.click(
+            clear_history,
+            outputs=[conversation_state, upload_status]
+        )
+        
+    return demo
+
+if __name__ == "__main__":
+    # Create and launch the application
+    app = create_chat_interface()
+    app.launch(share=False)

requirements.txt

@@ -1 +1,10 @@
-huggingface_hub==0.25.2
+# requirements.txt
+gradio>=4.0.0
+huggingface_hub>=0.22.0
+langchain>=0.1.0
+langchain_community>=0.0.10
+faiss-cpu>=1.7.4
+python-dotenv>=1.0.0
+pypdf>=4.0.0
+docx2txt>=0.8
+sentence-transformers>=2.2.2