diff --git a/community-contributions/sach91-bootcamp/week3-exercise.ipynb b/community-contributions/sach91-bootcamp/week3-exercise.ipynb
new file mode 100644
index 0000000..4f04b7e
--- /dev/null
+++ b/community-contributions/sach91-bootcamp/week3-exercise.ipynb
@@ -0,0 +1 @@
+{"nbformat":4,"nbformat_minor":0,"metadata":{"colab":{"provenance":[],"gpuType":"T4","authorship_tag":"ABX9TyOvR3p3rMyPRwqLuduIBmR0"},"kernelspec":{"name":"python3","display_name":"Python 3"},"language_info":{"name":"python"},"accelerator":"GPU"},"cells":[{"cell_type":"code","source":["# A HuggingFace LLAMA travel agent biased to one particular destination, using translation support."],"metadata":{"id":"xsWGWo7YrSPA"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["import gradio as gr\n","import torch\n","from transformers import AutoTokenizer, AutoModelForCausalLM, TextIteratorStreamer\n","from threading import Thread\n","from huggingface_hub import login\n","from google.colab import userdata"],"metadata":{"id":"ZzWgGqk2qPNP"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["# Load model and tokenizer\n","model_name = \"meta-llama/Llama-3.2-1B-Instruct\"\n","print(f\"Loading {model_name}...\")\n","\n","# load_dotenv(override=True)\n","# OPENWEATHER_API_KEY = os.getenv(\"OPENWEATHER_API_KEY\")\n","hf_token = userdata.get('HF_TOKEN')\n","login(hf_token, add_to_git_credential=True)\n","\n","tokenizer = AutoTokenizer.from_pretrained(model_name)\n","tokenizer.pad_token = tokenizer.eos_token\n","model = AutoModelForCausalLM.from_pretrained(\n","    model_name,\n","    torch_dtype=torch.bfloat16,\n","    device_map=\"auto\",\n",")\n","\n","print(\"Model loaded successfully!\")"],"metadata":{"id":"5sYVSW-eqdYj"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["def do_translate(msg):\n","    \"\"\"\n","    Translator function to format the response.\n","    \"\"\"\n","    sys_msg = \"\"\"\n","    You are an expert translator who can translate the given English text to Hindi.\n","    \"\"\"\n","\n","    usr_msg = f\"\"\"\n","    Translate the given English text to Hindi.\\n\n","    {msg}\n","    \"\"\"\n","\n","    # Format the conversation history for the model\n","    messages = [{\"role\": \"system\", \"content\": sys_msg}, {\"role\": \"user\", \"content\": usr_msg}]\n","\n","    # Apply chat template\n","    input_text = tokenizer.apply_chat_template(\n","        messages,\n","        tokenize=False,\n","        add_generation_prompt=True\n","    )\n","\n","    # Tokenize inputchat_with_llama\n","    inputs = tokenizer(input_text, return_tensors=\"pt\").to(model.device)\n","\n","    # Set up the streamer\n","    streamer = TextIteratorStreamer(\n","        tokenizer,\n","        skip_prompt=True,\n","        skip_special_tokens=True\n","    )\n","\n","    # Generation parameters\n","    generation_kwargs = dict(\n","        inputs,\n","        streamer=streamer,\n","        max_new_tokens=512,\n","        temperature=0.7,\n","        top_p=0.9,\n","        do_sample=True,\n","    )\n","\n","    # Start generation in a separate thread\n","    thread = Thread(target=model.generate, kwargs=generation_kwargs)\n","    thread.start()\n","\n","    # Stream the response\n","    partial_response = msg + '\\n**Hindi**: '\n","    yield partial_response\n","    for new_text in streamer:\n","        partial_response += new_text\n","        yield partial_response\n"],"metadata":{"id":"PmgoWUVNqgL7"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["def chat_with_llama(message, history):\n","    \"\"\"\n","    Chat function that streams responses from the Llama model.\n","\n","    Args:\n","        message: The user's current message\n","        history: List of [user_message, assistant_message] pairs\n","\n","    Yields:\n","        Partial responses as they are generated\n","    \"\"\"\n","\n","    sys_msg = \"\"\"\n","      You are a manipulative assistant for a travel company.\n","      Give short answers.\n","      You have the motive to make the user to travel to Spain than any other country.\n","      \"\"\"\n","    # Format the conversation history for the model\n","    messages = [{\"role\": \"system\", \"content\": sys_msg}]\n","\n","    # Add conversation history\n","    for user_msg, assistant_msg in history:\n","        messages.append({\"role\": \"user\", \"content\": user_msg})\n","        messages.append({\"role\": \"assistant\", \"content\": assistant_msg.split('\\n**Hindi**: ')[0]})\n","\n","    # Add the current message\n","    messages.append({\"role\": \"user\", \"content\": message})\n","\n","    # Apply chat template\n","    input_text = tokenizer.apply_chat_template(\n","        messages,\n","        tokenize=False,\n","        add_generation_prompt=True\n","    )\n","\n","    # Tokenize inputchat_with_llama\n","    inputs = tokenizer(input_text, return_tensors=\"pt\").to(model.device)\n","\n","    # Generate response\n","    with torch.no_grad():\n","        outputs = model.generate(\n","            **inputs,\n","            max_new_tokens=512,\n","            temperature=0.7,\n","            top_p=0.9,\n","            do_sample=True,\n","        )\n","\n","    # Decode and return the response\n","    response = tokenizer.decode(outputs[0][inputs['input_ids'].shape[1]:], skip_special_tokens=True)\n","\n","    yield from do_translate(response)\n"],"metadata":{"id":"pONxPaeYqkzg"},"execution_count":null,"outputs":[]},{"cell_type":"code","execution_count":null,"metadata":{"id":"Z6qAHIq1Wbqc"},"outputs":[],"source":["# Create the Gradio interface\n","w_model = model_name.split('/')[-1]\n","demo = gr.ChatInterface(\n","    fn=chat_with_llama,\n","    title = f\"🦙 {w_model} Chat\",\n","    description = f\"Chat with Meta's {w_model} model with streaming responses\",\n","    examples=[\n","        \"What is the capital of France?\",\n","        \"I want to travel to America\",\n","        \"What are some tips for learning a new language?\"\n","    ],\n","    theme=gr.themes.Soft()\n",")\n","\n","demo.launch(share=True, debug=True)\n","\n"]}]}
\ No newline at end of file