updated new version deepseek-r1

Dockerfile

@@ -21,4 +21,4 @@ COPY . .
 EXPOSE 11434 8501
 
 # Start both Ollama and Streamlit
-# CMD ["bash", "-c", "ollama serve & sleep 5 && streamlit run app.py --server.port 8501 --server.enableCORS false --server.enableXsrfProtection false"]
+# CMD ["bash", "-c", "streamlit run app.py --server.port 8501 --server.enableCORS false --server.enableXsrfProtection false"]

app.py

@@ -5,6 +5,15 @@ from services.llm import initialize_llm, initialize_embeddings
 from services.vector_store import create_vector_store, retrive_vector_store, generate_prompt
 from services.pdf_processing import load_and_split_pdf
 from utils.helpers import extract_thoughts, response_generator
+import subprocess
+
+try:
+    print("🚀 Checking and starting Ollama...")
+    subprocess.run(["bash", "install_ollama.sh"], check=True)
+    print("✅ Ollama is running!")
+except subprocess.CalledProcessError as e:
+    print(f"❌ Error: {e}")
+
 
 # Custom CSS for chat styling
 CHAT_CSS = """

doc.txt

@@ -1,14 +0,0 @@
-rag_chatbot/
-│── app.py                    # Main Streamlit app
-│── components/
-│   ├── sidebar.py            # Sidebar UI (model selection, upload, user info)
-│   ├── chat_ui.py            # Chat UI styling and history
-│── services/
-│   ├── pdf_processing.py      # PDF loading and text splitting
-│   ├── vector_store.py        # Vector database (FAISS) setup
-│   ├── llm.py                 # Model initialization and QA chain
-│── utils/
-│   ├── helpers.py             # Helper functions (e.g., get_file_size)
-│── docs/                      # Folder for storing uploaded PDFs (created dynamically)
-│── requirements.txt           # Dependencies
-│── README.md                  # Project documentation

install_ollama.sh

@@ -0,0 +1,32 @@
+#!/bin/bash
+
+# Function to check if Ollama is installed
+is_ollama_installed() {
+    command -v ollama >/dev/null 2>&1
+}
+
+# Install Ollama only if not already installed
+if is_ollama_installed; then
+    echo "✅ Ollama is already installed. Skipping installation..."
+else
+    echo "🚀 Installing Ollama..."
+    sudo apt-get update && sudo apt-get install -y curl
+    curl -fsSL https://ollama.com/install.sh | sh
+    echo "✅ Ollama installation completed!"
+fi
+
+# Start Ollama
+echo "🚀 Starting Ollama..."
+ollama serve &
+sleep 5
+
+# Pull the model if not already present
+if ollama list | grep -q "deepseek-r1:1.5b"; then
+    echo "✅ Model 'deepseek-r1:1.5b' is already available."
+else
+    echo "🚀 Pulling model 'deepseek-r1:1.5b'..."
+    ollama pull deepseek-r1:1.5b
+    echo "✅ Model pulled successfully!"
+fi
+
+echo "🚀 Ollama is running!"

services/llm.py

@@ -8,7 +8,7 @@ def initialize_llm(model_name, temperature, top_p, max_tokens):
     # Configure the LLM with additional parameters
     llm = OllamaLLM(
         model=model_name,
-        base_url="https://deepak7376-ollama-server.hf.space",
+        # base_url="https://deepak7376-ollama-server.hf.space",
         temperature=temperature,   # Controls randomness (0 = deterministic, 1 = max randomness)
         max_tokens=max_tokens,   # Limit the number of tokens in the output
         top_p=top_p          # Nucleus sampling for controlling diversity