{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "e2e8f8e9-16cd-4d86-aeb0-99007b9335bb",
   "metadata": {},
   "source": [
    "Since you’ve requested an explanation of diffusion models tailored for a beginner preparing for a master’s in AI, I’ll follow your specified response style, keeping the explanation clear, concise, and beginner-friendly. I’ll assume you have high school math and basic Python knowledge, and I’ll provide a complete breakdown as per your guidelines.\n",
    "\n",
    "---\n",
    "\n",
    "### 1. Simple Explanation of Diffusion Models (100–150 words)\n",
    "\n",
    "Diffusion models are a class of generative AI models that create data, like images or audio, by mimicking a process of adding and then removing noise. Imagine starting with a clear image, gradually adding random noise until it becomes unrecognizable, like static on a TV. The model learns to reverse this process, starting from pure noise and reconstructing the original data step-by-step. This is done by training a neural network to predict how to \"denoise\" the data at each step. Diffusion models, like those powering DALL·E 2 or Stable Diffusion, excel at generating high-quality, realistic outputs. They rely on probability and iterative refinement, making them computationally intensive but powerful for tasks like image generation. With basic math (probability and linear algebra) and Python, you can understand their core mechanics.\n",
    "\n",
    "*Word count: 135*\n",
    "\n",
    "---\n",
    "\n",
    "### 2. Detailed Flow of Diffusion Models (Roadmap of Key Concepts)\n",
    "\n",
    "To fully understand diffusion models, follow this logical progression of subtopics:\n",
    "\n",
    "1. **Generative Modeling Basics**  \n",
    "   - Understand generative models: They create new data samples (e.g., images, audio) similar to a training dataset.  \n",
    "   - Diffusion models are a type of generative model, like GANs or VAEs, but use a unique noise-based approach.\n",
    "\n",
    "2. **Forward Process (Adding Noise)**  \n",
    "   - The forward process gradually adds Gaussian noise to data over multiple steps.  \n",
    "   - Each step makes the data noisier, eventually turning it into pure noise (isotropic Gaussian distribution).  \n",
    "   - Key concept: Markov chain, where each step depends only on the previous one.\n",
    "\n",
    "3. **Reverse Process (Denoising)**  \n",
    "   - The reverse process learns to remove noise, reconstructing the original data from noise.  \n",
    "   - A neural network predicts the noise added at each step, guiding the denoising process.  \n",
    "   - This is also a Markov chain, trained to approximate the true data distribution.\n",
    "\n",
    "4. **Mathematical Foundation**  \n",
    "   - Probability distributions: The model works with Gaussian distributions for noise.  \n",
    "   - Loss function: The model minimizes the difference between predicted and actual noise.  \n",
    "   - Key operations involve matrix manipulations (e.g., mean and variance of distributions).\n",
    "\n",
    "5. **Training the Model**  \n",
    "   - Use a dataset (e.g., images) to train the model.  \n",
    "   - Add noise to data in multiple steps and train the neural network to predict the noise.  \n",
    "   - Optimize using a loss function (e.g., mean squared error).\n",
    "\n",
    "6. **Sampling (Generating New Data)**  \n",
    "   - Start with random noise.  \n",
    "   - Iteratively apply the trained model to denoise and generate new data samples.  \n",
    "   - Adjust hyperparameters like the number of steps for quality vs. speed.\n",
    "\n",
    "7. **Advanced Variants**  \n",
    "   - Denoising Diffusion Probabilistic Models (DDPM): The standard framework.  \n",
    "   - Score-based generative models: Alternative formulation using score functions.  \n",
    "   - Latent diffusion: Operates in a compressed (latent) space for efficiency, used in Stable Diffusion.\n",
    "\n",
    "8. **Applications and Limitations**  \n",
    "   - Applications: Image generation, audio synthesis, text-to-image models.  \n",
    "   - Limitations: Computationally expensive, requires many steps for high-quality output.\n",
    "\n",
    "---\n",
    "\n",
    "### 3. Relevant Formulas with Explanations\n",
    "\n",
    "Diffusion models rely on probabilistic and linear algebra concepts. Below are the key formulas, with terms explained for a beginner.\n",
    "\n",
    "#### 1. Forward Process (Adding Noise)\n",
    "The forward process adds Gaussian noise over \\( T \\) steps, defined as a Markov chain:\n",
    "\\[\n",
    "q(\\mathbf{x}_t | \\mathbf{x}_{t-1}) = \\mathcal{N}(\\mathbf{x}_t; \\sqrt{1-\\beta_t} \\mathbf{x}_{t-1}, \\beta_t \\mathbf{I})\n",
    "\\]\n",
    "- **\\(\\mathbf{x}_t\\)**: Data at step \\( t \\) (e.g., a noisy image).\n",
    "- **\\(\\mathbf{x}_{t-1}\\)**: Data at step \\( t-1 \\) (less noisy).\n",
    "- **\\(\\beta_t\\)**: Variance schedule, a small value controlling noise added at step \\( t \\). Typically pre-defined (e.g., linear or cosine schedule).\n",
    "- **\\(\\sqrt{1-\\beta_t}\\)**: Scales the mean to gradually reduce the signal.\n",
    "- **\\(\\mathcal{N}(\\mu, \\sigma^2)\\)**: Gaussian distribution with mean \\(\\mu\\) and variance \\(\\sigma^2\\).\n",
    "- **\\(\\mathbf{I}\\)**: Identity matrix, ensuring noise is added independently to each dimension (e.g., pixel).\n",
    "\n",
    "This formula describes how data becomes noisier each step, with the mean shrinking and variance increasing.\n",
    "\n",
    "#### 2. Simplified Forward Process\n",
    "For efficiency, the forward process can be computed in one step:\n",
    "\\[\n",
    "\\mathbf{x}_t = \\sqrt{\\bar{\\alpha}_t} \\mathbf{x}_0 + \\sqrt{1 - \\bar{\\alpha}_t} \\mathbf{\\epsilon}\n",
    "\\]\n",
    "- **\\(\\mathbf{x}_0\\)**: Original data (e.g., clean image).\n",
    "- **\\(\\bar{\\alpha}_t = \\prod_{s=1}^t (1 - \\beta_s)\\)**: Cumulative product of \\(1 - \\beta_s\\), representing how much of the original signal remains.\n",
    "- **\\(\\mathbf{\\epsilon} \\sim \\mathcal{N}(\\mathbf{0}, \\mathbf{I})\\)**: Random noise sampled from a standard Gaussian.\n",
    "\n",
    "This allows direct sampling of \\(\\mathbf{x}_t\\) at any step \\( t \\) without iterating.\n",
    "\n",
    "#### 3. Reverse Process (Denoising)\n",
    "The reverse process approximates the transition from \\(\\mathbf{x}_t\\) to \\(\\mathbf{x}_{t-1}\\):\n",
    "\\[\n",
    "p_\\theta(\\mathbf{x}_{t-1} | \\mathbf{x}_t) = \\mathcal{N}(\\mathbf{x}_{t-1}; \\mu_\\theta(\\mathbf{x}_t, t), \\Sigma_\\theta(\\mathbf{x}_t, t))\n",
    "\\]\n",
    "- **\\(\\mu_\\theta(\\mathbf{x}_t, t)\\)**: Mean predicted by a neural network parameterized by \\(\\theta\\).\n",
    "- **\\(\\Sigma_\\theta(\\mathbf{x}_t, t)\\)**: Variance, often fixed or learned.\n",
    "- The neural network typically predicts the noise \\(\\mathbf{\\epsilon}_\\theta(\\mathbf{x}_t, t)\\), and the mean is computed as:\n",
    "\\[\n",
    "\\mu_\\theta(\\mathbf{x}_t, t) = \\frac{1}{\\sqrt{1-\\beta_t}} \\left( \\mathbf{x}_t - \\frac{\\beta_t}{\\sqrt{1 - \\bar{\\alpha}_t}} \\mathbf{\\epsilon}_\\theta(\\mathbf{x}_t, t) \\right)\n",
    "\\]\n",
    "\n",
    "#### 4. Loss Function\n",
    "The model is trained by minimizing the mean squared error between predicted and actual noise:\n",
    "\\[\n",
    "L = \\mathbb{E}_{\\mathbf{x}_0, \\mathbf{\\epsilon}, t} \\left[ \\|\\mathbf{\\epsilon} - \\mathbf{\\epsilon}_\\theta(\\mathbf{x}_t, t)\\|^2 \\right]\n",
    "\\]\n",
    "- **\\(\\mathbb{E}\\)**: Expectation over data, noise, and timesteps.\n",
    "- **\\(\\mathbf{\\epsilon}\\)**: True noise added in the forward process.\n",
    "- **\\(\\mathbf{\\epsilon}_\\theta(\\mathbf{x}_t, t)\\)**: Noise predicted by the neural network.\n",
    "- **\\(\\|\\cdot\\|^2\\)**: Squared Euclidean norm (measures difference between vectors).\n",
    "\n",
    "This loss ensures the model learns to predict the noise accurately, enabling effective denoising.\n",
    "\n",
    "---\n",
    "\n",
    "### 4. Step-by-Step Example Calculation\n",
    "\n",
    "Let’s compute one step of the forward and reverse process for a 2D vector (simulating a tiny image with 2 pixels). Assume \\( T = 1000 \\) steps, and at step \\( t = 500 \\), \\(\\beta_t = 0.02\\), \\(\\bar{\\alpha}_t = 0.1\\) (pre-computed).\n",
    "\n",
    "#### Forward Process\n",
    "1. **Start with data**: \\(\\mathbf{x}_0 = [1, 2]\\) (original pixel values).\n",
    "2. **Apply the simplified forward process**:\n",
    "\\[\n",
    "\\mathbf{x}_t = \\sqrt{\\bar{\\alpha}_t} \\mathbf{x}_0 + \\sqrt{1 - \\bar{\\alpha}_t} \\mathbf{\\epsilon}\n",
    "\\]\n",
    "   - \\(\\sqrt{\\bar{\\alpha}_t} = \\sqrt{0.1} \\approx 0.3162\\).\n",
    "   - \\(\\sqrt{1 - \\bar{\\alpha}_t} = \\sqrt{1 - 0.1} = \\sqrt{0.9} \\approx 0.9487\\).\n",
    "   - Sample noise: \\(\\mathbf{\\epsilon} = [0.5, -0.3]\\) (random Gaussian).\n",
    "   - Compute:\n",
    "     \\[\n",
    "     \\mathbf{x}_t = 0.3162 \\cdot [1, 2] + 0.9487 \\cdot [0.5, -0.3]\n",
    "     \\]\n",
    "     \\[\n",
    "     = [0.3162 \\cdot 1, 0.3162 \\cdot 2] + [0.9487 \\cdot 0.5, 0.9487 \\cdot (-0.3)]\n",
    "     \\]\n",
    "     \\[\n",
    "     = [0.3162, 0.6324] + [0.4744, -0.2846]\n",
    "     \\]\n",
    "     \\[\n",
    "     \\mathbf{x}_t \\approx [0.7906, 0.3478]\n",
    "     \\]\n",
    "\n",
    "#### Reverse Process\n",
    "1. **Given**: \\(\\mathbf{x}_t = [0.7906, 0.3478]\\), assume the neural network predicts noise \\(\\mathbf{\\epsilon}_\\theta(\\mathbf{x}_t, t) = [0.48, -0.31]\\).\n",
    "2. **Compute mean**:\n",
    "\\[\n",
    "\\mu_\\theta = \\frac{1}{\\sqrt{1-\\beta_t}} \\left( \\mathbf{x}_t - \\frac{\\beta_t}{\\sqrt{1 - \\bar{\\alpha}_t}} \\mathbf{\\epsilon}_\\theta \\right)\n",
    "\\]\n",
    "   - \\(\\sqrt{1-\\beta_t} = \\sqrt{1-0.02} = \\sqrt{0.98} \\approx 0.9899\\).\n",
    "   - \\(\\frac{1}{\\sqrt{1-\\beta_t}} \\approx \\frac{1}{0.9899} \\approx 1.0102\\).\n",
    "   - \\(\\sqrt{1 - \\bar{\\alpha}_t} = \\sqrt{0.9} \\approx 0.9487\\).\n",
    "   - \\(\\frac{\\beta_t}{\\sqrt{1 - \\bar{\\alpha}_t}} = \\frac{0.02}{0.9487} \\approx 0.0211\\).\n",
    "   - Compute noise term:\n",
    "     \\[\n",
    "     \\frac{\\beta_t}{\\sqrt{1 - \\bar{\\alpha}_t}} \\mathbf{\\epsilon}_\\theta = 0.0211 \\cdot [0.48, -0.31] \\approx [0.0101, -0.0065]\n",
    "     \\]\n",
    "   - Compute:\n",
    "     \\[\n",
    "     \\mathbf{x}_t - \\frac{\\beta_t}{\\sqrt{1 - \\bar{\\alpha}_t}} \\mathbf{\\epsilon}_\\theta = [0.7906, 0.3478] - [0.0101, -0.0065] \\approx [0.7805, 0.3543]\n",
    "     \\]\n",
    "     \\[\n",
    "     \\mu_\\theta = 1.0102 \\cdot [0.7805, 0.3543] \\approx [0.7885, 0.3579]\n",
    "     \\]\n",
    "\n",
    "3. **Sample \\(\\mathbf{x}_{t-1}\\)**:\n",
    "   - Assume \\(\\Sigma_\\theta = \\beta_t \\mathbf{I} = 0.02 \\mathbf{I}\\).\n",
    "   - Sample from \\(\\mathcal{N}(\\mu_\\theta, 0.02 \\mathbf{I})\\), adding small noise (for simplicity, we’ll use \\(\\mu_\\theta\\) as the result): \\(\\mathbf{x}_{t-1} \\approx [0.7885, 0.3579]\\).\n",
    "\n",
    "This shows one denoising step, moving closer to the original \\(\\mathbf{x}_0 = [1, 2]\\).\n",
    "\n",
    "---\n",
    "\n",
    "### 5. Python Implementation\n",
    "\n",
    "Below is a beginner-friendly Python implementation of a simplified Denoising Diffusion Probabilistic Model (DDPM) for a toy dataset. We’ll use a small 2D dataset and a basic neural network to predict noise.\n",
    "\n",
    "#### Library Installations\n",
    "```bash\n",
    "pip install torch numpy\n",
    "```\n",
    "\n",
    "#### Full Code\n",
    "```python\n",
    "import torch\n",
    "import torch.nn as nn\n",
    "import numpy as np\n",
    "\n",
    "# Set random seed for reproducibility\n",
    "torch.manual_seed(42)\n",
    "np.random.seed(42)\n",
    "\n",
    "# Toy dataset: 2D points forming a small cluster\n",
    "data = np.array([[1.0, 2.0], [1.1, 2.1], [0.9, 1.9], [1.2, 2.2]], dtype=np.float32)\n",
    "data = torch.tensor(data)  # Convert to tensor\n",
    "\n",
    "# Hyperparameters\n",
    "T = 1000  # Number of diffusion steps\n",
    "beta = torch.linspace(0.0001, 0.02, T)  # Linear beta schedule\n",
    "alpha = 1. - beta  # Alpha = 1 - beta\n",
    "alpha_bar = torch.cumprod(alpha, dim=0)  # Cumulative product for alpha_bar\n",
    "\n",
    "# Simple neural network to predict noise\n",
    "class NoisePredictor(nn.Module):\n",
    "    def __init__(self):\n",
    "        super(NoisePredictor, self).__init__()\n",
    "        self.model = nn.Sequential(\n",
    "            nn.Linear(4, 64),  # Input: 2D data + timestep (encoded as 2D)\n",
    "            nn.ReLU(),\n",
    "            nn.Linear(64, 64),\n",
    "            nn.ReLU(),\n",
    "            nn.Linear(64, 2)  # Output: 2D noise\n",
    "        )\n",
    "    \n",
    "    def forward(self, x, t):\n",
    "        # Encode timestep as a simple scaling factor\n",
    "        t_embed = torch.ones_like(x) * t.view(-1, 1) / T\n",
    "        input = torch.cat([x, t_embed], dim=1)  # Concatenate data and timestep\n",
    "        return self.model(input)\n",
    "\n",
    "# Forward process: Add noise to data\n",
    "def forward_process(x0, t):\n",
    "    epsilon = torch.randn_like(x0)  # Sample random noise\n",
    "    sqrt_alpha_bar = torch.sqrt(alpha_bar[t]).view(-1, 1)\n",
    "    sqrt_one_minus_alpha_bar = torch.sqrt(1 - alpha_bar[t]).view(-1, 1)\n",
    "    xt = sqrt_alpha_bar * x0 + sqrt_one_minus_alpha_bar * epsilon  # Noisy data\n",
    "    return xt, epsilon\n",
    "\n",
    "# Training loop\n",
    "model = NoisePredictor()\n",
    "optimizer = torch.optim.Adam(model.parameters(), lr=0.001)\n",
    "num_epochs = 1000\n",
    "\n",
    "for epoch in range(num_epochs):\n",
    "    total_loss = 0\n",
    "    for i in range(len(data)):\n",
    "        x0 = data[i:i+1]  # Single data point\n",
    "        t = torch.randint(0, T, (1,))  # Random timestep\n",
    "        xt, epsilon = forward_process(x0, t)  # Add noise\n",
    "        epsilon_pred = model(xt, t)  # Predict noise\n",
    "        loss = ((epsilon - epsilon_pred) ** 2).mean()  # MSE loss\n",
    "        optimizer.zero_grad()\n",
    "        loss.backward()\n",
    "        optimizer.step()\n",
    "        total_loss += loss.item()\n",
    "    if epoch % 100 == 0:\n",
    "        print(f\"Epoch {epoch}, Loss: {total_loss / len(data):.4f}\")\n",
    "\n",
    "# Sampling: Generate new data from noise\n",
    "def sample(model, n_samples=4):\n",
    "    x = torch.randn(n_samples, 2)  # Start with random noise\n",
    "    for t in reversed(range(T)):\n",
    "        t_tensor = torch.tensor([t]).float()\n",
    "        epsilon_pred = model(x, t_tensor)  # Predict noise\n",
    "        alpha_t = alpha[t].view(-1, 1)\n",
    "        beta_t = beta[t].view(-1, 1)\n",
    "        sqrt_one_minus_alpha_bar = torch.sqrt(1 - alpha_bar[t]).view(-1, 1)\n",
    "        mu = (x - (beta_t / sqrt_one_minus_alpha_bar) * epsilon_pred) / torch.sqrt(alpha_t)\n",
    "        if t > 0:\n",
    "            x = mu + torch.sqrt(beta_t) * torch.randn_like(x)  # Add noise\n",
    "        else:\n",
    "            x = mu  # Final step: no noise\n",
    "    return x\n",
    "\n",
    "# Generate samples\n",
    "generated_samples = sample(model)\n",
    "print(\"Generated samples:\\n\", generated_samples.detach().numpy())\n",
    "\n",
    "# Comments for each line:\n",
    "# torch.manual_seed(42) - Ensures reproducible results\n",
    "# data = np.array(...) - Creates a small 2D dataset\n",
    "# data = torch.tensor(data) - Converts data to PyTorch tensor\n",
    "# beta = torch.linspace(...) - Defines noise schedule\n",
    "# alpha = 1. - beta - Computes alpha for each step\n",
    "# alpha_bar = torch.cumprod(...) - Cumulative product of alpha\n",
    "# class NoisePredictor(nn.Module) - Defines a simple neural network\n",
    "# nn.Linear(4, 64) - Input layer: 2D data + 2D timestep embedding\n",
    "# forward(self, x, t) - Forward pass, concatenates data and timestep\n",
    "# forward_process(x0, t) - Adds noise to data at step t\n",
    "# xt, epsilon = forward_process(x0, t) - Gets noisy data and true noise\n",
    "# epsilon_pred = model(xt, t) - Predicts noise using the model\n",
    "# loss = ((epsilon - epsilon_pred) ** 2).mean() - Computes MSE loss\n",
    "# sample(model, n_samples=4) - Generates new samples from noise\n",
    "# x = torch.randn(n_samples, 2) - Starts with random noise\n",
    "# mu = ... - Computes mean for reverse process\n",
    "# x = mu + ... - Updates sample with predicted mean and noise\n",
    "# print(\"Generated samples:\\n\", ...) - Prints generated 2D points\n",
    "```\n",
    "\n",
    "#### Notes\n",
    "- **Dataset**: A toy 2D dataset simulating a small cluster of points.\n",
    "- **Model**: A simple feedforward neural network predicts noise, taking noisy data and timestep as input.\n",
    "- **Environment**: This code runs in Google Colab or a local Python environment with PyTorch installed.\n",
    "- **Output**: Generates 2D points close to the original dataset (e.g., around \\([1, 2]\\)).\n",
    "- **Simplifications**: Uses a basic network and small dataset for clarity. Real diffusion models (e.g., for images) use convolutional networks like U-Net and larger datasets.\n",
    "\n",
    "---\n",
    "\n",
    "### 6. Practical AI Use Case\n",
    "\n",
    "Diffusion models are widely used in **text-to-image generation**. For example, Stable Diffusion, a popular diffusion model, powers tools like DALL·E 2 and Midjourney. Given a text prompt (e.g., “a cat in a spacesuit”), the model generates a high-quality image by starting with random noise and iteratively denoising it, guided by the text embedding. This is achieved using latent diffusion, where the model operates in a compressed latent space to reduce computational cost. Applications include:\n",
    "- **Art and Design**: Generating artwork or concept designs.\n",
    "- **Entertainment**: Creating visuals for games or movies.\n",
    "- **Research**: Simulating scientific data (e.g., molecular structures).\n",
    "The ability to generate realistic, diverse images from text makes diffusion models invaluable in creative and commercial AI systems.\n",
    "\n",
    "---\n",
    "\n",
    "### 7. Tips for Mastering Diffusion Models\n",
    "\n",
    "1. **Practice Problems**:\n",
    "   - Implement a diffusion model for a 1D dataset (e.g., Gaussian points) to understand the forward and reverse processes.\n",
    "   - Experiment with different \\(\\beta_t\\) schedules (e.g., linear vs. cosine) and observe their impact on generation quality.\n",
    "   - Try generating images using a pre-trained Stable Diffusion model (available via Hugging Face).\n",
    "\n",
    "2. **Additional Resources**:\n",
    "   - **Paper**: “Denoising Diffusion Probabilistic Models” by Ho et al. (2020) – foundational paper, beginner-accessible with math explained.\n",
    "   - **Tutorial**: Lilian Weng’s blog post on diffusion models (lilianweng.github.io) – clear explanations with visuals.\n",
    "   - **Course**: Hugging Face’s “Diffusion Models from Scratch” tutorial – includes Python code.\n",
    "   - **Book**: “Deep Learning” by Goodfellow et al. – Chapter on generative models for background.\n",
    "\n",
    "3. **Hands-On**:\n",
    "   - Use Google Colab to run pre-trained diffusion models (e.g., via Hugging Face’s `diffusers` library).\n",
    "   - Experiment with hyperparameters like \\( T \\) or network architecture.\n",
    "   - Visualize the denoising process by saving intermediate samples during generation.\n",
    "\n",
    "4. **Math Practice**:\n",
    "   - Work through Gaussian distribution properties (mean, variance) using high school probability.\n",
    "   - Practice matrix operations (e.g., scaling vectors) in Python with NumPy.\n",
    "\n",
    "By combining theoretical study, coding practice, and experimentation with pre-trained models, you’ll build a strong foundation in diffusion models for your AI master’s program.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5fe2bd3c-285e-424b-b247-56d091e9ee09",
   "metadata": {},
   "source": [
    "---\n",
    "---\n",
    "---\n",
    "\n",
    "I'll explain diffusion models in a simple, step-by-step way that's easy to understand.\n",
    "\n",
    "## What is a Diffusion Model? (Simple Analogy)\n",
    "\n",
    "Imagine you have a **clear photograph** and you want to teach a computer to create similar photos. Here's what diffusion models do:\n",
    "\n",
    "1. **Take the clear photo** → **Gradually add noise** (like TV static) → **Until it becomes pure noise**\n",
    "2. **Train a neural network** to do the reverse: **Take noise** → **Gradually remove noise** → **Get back a clear photo**\n",
    "\n",
    "Think of it like **learning to clean a dirty window step by step**.\n",
    "\n",
    "## The Complete Flow (Visual Step-by-Step)\n",
    "\n",
    "```\n",
    "Original Image → Add Noise → Add More Noise → ... → Pure Noise\n",
    "     ↑                                                    ↓\n",
    "Generated Image ← Remove Noise ← Remove More Noise ← ... ← Start Here\n",
    "```\n",
    "\n",
    "## Step 1: Forward Process (Adding Noise)\n",
    "\n",
    "**What happens:** We gradually destroy the image by adding noise\n",
    "\n",
    "**Simple Formula:**\n",
    "```\n",
    "Noisy_Image = √(keep_signal) × Original_Image + √(add_noise) × Random_Noise\n",
    "```\n",
    "\n",
    "**In Math Notation:**\n",
    "$$x_t = \\sqrt{\\bar{\\alpha}_t} x_0 + \\sqrt{1-\\bar{\\alpha}_t} \\epsilon$$\n",
    "\n",
    "**What each part means:**\n",
    "- $x_0$ = Original clean image\n",
    "- $x_t$ = Image after t steps of adding noise\n",
    "- $\\bar{\\alpha}_t$ = How much of original image to keep (gets smaller over time)\n",
    "- $\\epsilon$ = Random noise we add\n",
    "- $\\sqrt{\\bar{\\alpha}_t}$ = \"Keep signal\" amount\n",
    "- $\\sqrt{1-\\bar{\\alpha}_t}$ = \"Add noise\" amount\n",
    "\n",
    "**Example with numbers:**\n",
    "- Step 0: `keep_signal = 1.0, add_noise = 0.0` → Perfect image\n",
    "- Step 500: `keep_signal = 0.3, add_noise = 0.7` → Mostly noise\n",
    "- Step 1000: `keep_signal = 0.0, add_noise = 1.0` → Pure noise\n",
    "\n",
    "## Step 2: Reverse Process (Removing Noise)\n",
    "\n",
    "**What happens:** We train a neural network to predict and remove the noise\n",
    "\n",
    "**Simple Formula:**\n",
    "```\n",
    "Cleaner_Image = (Noisy_Image - Predicted_Noise) / scaling_factor\n",
    "```\n",
    "\n",
    "**In Math Notation:**\n",
    "$$x_{t-1} = \\frac{1}{\\sqrt{\\alpha_t}} \\left( x_t - \\frac{\\beta_t}{\\sqrt{1-\\bar{\\alpha}_t}} \\epsilon_\\theta(x_t, t) \\right)$$\n",
    "\n",
    "**What each part means:**\n",
    "- $x_t$ = Current noisy image\n",
    "- $x_{t-1}$ = Less noisy image (one step cleaner)\n",
    "- $\\epsilon_\\theta(x_t, t)$ = Neural network's prediction of noise\n",
    "- $\\frac{\\beta_t}{\\sqrt{1-\\bar{\\alpha}_t}}$ = How much noise to remove\n",
    "- $\\frac{1}{\\sqrt{\\alpha_t}}$ = Scaling factor to adjust brightness\n",
    "\n",
    "## Step 3: Training the Neural Network\n",
    "\n",
    "**What we train:** A neural network that can look at a noisy image and predict what noise was added\n",
    "\n",
    "**Loss Function (Super Simple):**\n",
    "```\n",
    "Loss = (Actual_Noise - Predicted_Noise)²\n",
    "```\n",
    "\n",
    "**In Math Notation:**\n",
    "$$L = \\mathbb{E}[\\|\\epsilon - \\epsilon_\\theta(x_t, t)\\|^2]$$\n",
    "\n",
    "**Training Process:**\n",
    "1. Take a clean image\n",
    "2. Add random noise to it\n",
    "3. Ask the network: \"What noise was added?\"\n",
    "4. Compare the network's guess with the actual noise\n",
    "5. Adjust the network to guess better\n",
    "6. Repeat millions of times\n",
    "\n",
    "## Visual Example with Numbers\n",
    "\n",
    "Let's say we have a simple 1D \"image\" with value `5.0`:\n",
    "\n",
    "### Forward Process (Adding Noise):\n",
    "```\n",
    "Step 0:   Image = 5.0                    (clean)\n",
    "Step 1:   Image = 4.5 + 0.5×noise       (little noisy)\n",
    "Step 2:   Image = 4.0 + 1.0×noise       (more noisy)\n",
    "...\n",
    "Step 1000: Image = 0.0 + 1.0×noise      (pure noise)\n",
    "```\n",
    "\n",
    "### Reverse Process (Removing Noise):\n",
    "```\n",
    "Step 1000: Start with pure noise = 2.3\n",
    "Step 999:  Network says \"remove 0.1\" → 2.2\n",
    "Step 998:  Network says \"remove 0.15\" → 2.05\n",
    "...\n",
    "Step 0:    Final result ≈ 5.0 (close to original!)\n",
    "```\n",
    "\n",
    "## Key Formulas Explained Simply\n",
    "\n",
    "### 1. Beta Schedule (β)\n",
    "**What it controls:** How much noise to add at each step\n",
    "```python\n",
    "beta = [0.0001, 0.0002, 0.0003, ..., 0.02]  # Gradually increase\n",
    "```\n",
    "- **Small β:** Add little noise\n",
    "- **Large β:** Add lots of noise\n",
    "\n",
    "### 2. Alpha (α)\n",
    "**What it is:** How much signal to keep\n",
    "```python\n",
    "alpha = 1 - beta  # If we add 0.01 noise, keep 0.99 signal\n",
    "```\n",
    "\n",
    "### 3. Alpha Bar (ᾱ)\n",
    "**What it is:** Total signal remaining after t steps\n",
    "```python\n",
    "alpha_bar = alpha[0] × alpha[1] × alpha[2] × ... × alpha[t]\n",
    "```\n",
    "\n",
    "## The Complete Algorithm\n",
    "\n",
    "### Training:\n",
    "```python\n",
    "for each_image in dataset:\n",
    "    1. Pick random timestep t\n",
    "    2. Add noise: noisy_image = √(alpha_bar[t]) × image + √(1-alpha_bar[t]) × random_noise\n",
    "    3. Ask network: predicted_noise = network(noisy_image, t)\n",
    "    4. Loss = (random_noise - predicted_noise)²\n",
    "    5. Update network to reduce loss\n",
    "```\n",
    "\n",
    "### Sampling (Generation):\n",
    "```python\n",
    "start_with_pure_noise = random_noise\n",
    "for t from 1000 down to 0:\n",
    "    1. Ask network: predicted_noise = network(current_image, t)\n",
    "    2. Remove noise: cleaner_image = (current_image - predicted_noise) / scale\n",
    "    3. Add tiny bit of randomness (except last step)\n",
    "    4. current_image = cleaner_image\n",
    "return final_clean_image\n",
    "```\n",
    "\n",
    "## Why This Works\n",
    "\n",
    "1. **Forward process is easy:** Adding noise is simple math\n",
    "2. **Reverse process is learned:** Neural network learns to remove noise\n",
    "3. **Step by step:** Small changes are easier to learn than big jumps\n",
    "4. **Probabilistic:** Uses randomness to create variety\n",
    "\n",
    "## Real-World Example: Text-to-Image\n",
    "\n",
    "When you type \"a cat in a hat\":\n",
    "1. **Text encoder** converts words to numbers\n",
    "2. **Start with random noise** (pure static)\n",
    "3. **Neural network** gradually removes noise while being \"guided\" by the text\n",
    "4. **After 1000 steps:** You get an image of a cat in a hat!\n",
    "\n",
    "## Key Advantages\n",
    "\n",
    "- **High quality:** Produces very realistic results\n",
    "- **Stable training:** Doesn't have training problems like GANs\n",
    "- **Controllable:** Can guide generation with text, images, etc.\n",
    "- **Flexible:** Works for images, audio, 3D models, etc.\n",
    "\n",
    "## Simple Mental Model\n",
    "\n",
    "Think of diffusion models as:\n",
    "- **A student learning to clean dirty pictures**\n",
    "- **The teacher shows clean and dirty versions**\n",
    "- **Student practices removing dirt step by step**\n",
    "- **Eventually, student can clean any dirty picture**\n",
    "- **Even pictures they've never seen before!**\n",
    "\n",
    "This is exactly how diffusion models work - they learn to \"clean\" noisy data to generate new, realistic samples!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "d38384f9-8e10-4401-b7fb-9a109c19c408",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting tqdm\n",
      "  Downloading tqdm-4.67.1-py3-none-any.whl.metadata (57 kB)\n",
      "Downloading tqdm-4.67.1-py3-none-any.whl (78 kB)\n",
      "Installing collected packages: tqdm\n",
      "Successfully installed tqdm-4.67.1\n"
     ]
    }
   ],
   "source": [
    "!pip install tqdm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "1e207ab1-1cc3-40ae-9970-788c59d8221c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using device: mps\n",
      "Creating dataset...\n",
      "Created dataset with 200 points\n",
      "Data range: x[-2.97, 2.69], y[-2.59, 2.74]\n",
      "Initializing diffusion model...\n",
      "=== DIFFUSION MODEL INITIALIZED ===\n",
      "Device: mps\n",
      "Timesteps: 1000\n",
      "Beta range: 0.0001 to 0.02\n",
      "Parameters: 157698\n",
      "Starting training...\n",
      "\n",
      "=== STARTING TRAINING ===\n",
      "Dataset size: 200\n",
      "Epochs: 3000\n",
      "Epoch    0 | Loss: 0.845029 | LR: 0.001000\n",
      "Epoch  300 | Loss: 0.313191 | LR: 0.000945\n",
      "Epoch  600 | Loss: 0.365086 | LR: 0.000793\n",
      "Epoch  900 | Loss: 0.263697 | LR: 0.000577\n",
      "Epoch 1200 | Loss: 0.177150 | LR: 0.000345\n",
      "Epoch 1500 | Loss: 0.352184 | LR: 0.000146\n",
      "Epoch 1800 | Loss: 0.268311 | LR: 0.000024\n",
      "Epoch 2100 | Loss: 0.225336 | LR: 0.000006\n",
      "Epoch 2400 | Loss: 0.279304 | LR: 0.000096\n",
      "Epoch 2700 | Loss: 0.260822 | LR: 0.000274\n",
      "\n",
      "=== TRAINING SUMMARY ===\n",
      "Total time: 0:01:50.384296\n",
      "Training steps: 21000\n",
      "Sampling steps: 0\n",
      "Final loss: 0.295443\n",
      "Generating samples...\n",
      "\n",
      "=== SAMPLING 100 SAMPLES ===\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Sampling: 1000it [00:01, 685.00it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Visualizing results...\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1800x1200 with 6 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 2500x1000 with 10 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Plotting training metrics...\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x400 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Saving model...\n",
      "\n",
      "=== EXECUTION COMPLETE ===\n",
      "Generated files:\n",
      "- diffusion_results.png: Results visualization\n",
      "- diffusion_process.png: Sampling process steps\n",
      "- training_metrics.png: Training loss curves\n",
      "- diffusion_logs/: Detailed training logs\n",
      "- diffusion_model_complete.pth: Saved model\n",
      "\n",
      "Original Data Stats:\n",
      "Mean: [0.002, 0.012]\n",
      "Std:  [1.796, 1.808]\n",
      "\n",
      "Generated Data Stats:\n",
      "Mean: [-0.090, 0.145]\n",
      "Std:  [1.827, 1.726]\n"
     ]
    }
   ],
   "source": [
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from tqdm import tqdm\n",
    "import os\n",
    "import json\n",
    "from datetime import datetime\n",
    "\n",
    "torch.manual_seed(42)\n",
    "np.random.seed(42)\n",
    "\n",
    "device = torch.device(\"mps\" if torch.backends.mps.is_available() else \"cpu\")\n",
    "print(f\"Using device: {device}\")\n",
    "\n",
    "class DiffusionLogger:\n",
    "   def __init__(self, log_dir=\"diffusion_logs\"):\n",
    "       self.log_dir = log_dir\n",
    "       os.makedirs(log_dir, exist_ok=True)\n",
    "       self.training_log = []\n",
    "       self.sampling_log = []\n",
    "       self.start_time = datetime.now()\n",
    "       \n",
    "   def log_training_step(self, epoch, batch_idx, loss, lr, t_mean, epsilon_norm, pred_norm):\n",
    "       log_entry = {\n",
    "           'epoch': epoch,\n",
    "           'batch_idx': batch_idx,\n",
    "           'loss': loss,\n",
    "           'learning_rate': lr,\n",
    "           'timestep_mean': t_mean,\n",
    "           'epsilon_norm': epsilon_norm,\n",
    "           'prediction_norm': pred_norm,\n",
    "           'timestamp': datetime.now().isoformat()\n",
    "       }\n",
    "       self.training_log.append(log_entry)\n",
    "       \n",
    "   def log_sampling_step(self, step, x_mean, x_std, epsilon_pred_norm):\n",
    "       log_entry = {\n",
    "           'step': step,\n",
    "           'x_mean': x_mean,\n",
    "           'x_std': x_std,\n",
    "           'epsilon_pred_norm': epsilon_pred_norm,\n",
    "           'timestamp': datetime.now().isoformat()\n",
    "       }\n",
    "       self.sampling_log.append(log_entry)\n",
    "       \n",
    "   def save_logs(self):\n",
    "       with open(f\"{self.log_dir}/training_log.json\", 'w') as f:\n",
    "           json.dump(self.training_log, f, indent=2)\n",
    "       with open(f\"{self.log_dir}/sampling_log.json\", 'w') as f:\n",
    "           json.dump(self.sampling_log, f, indent=2)\n",
    "           \n",
    "   def print_summary(self):\n",
    "       total_time = datetime.now() - self.start_time\n",
    "       print(f\"\\n=== TRAINING SUMMARY ===\")\n",
    "       print(f\"Total time: {total_time}\")\n",
    "       print(f\"Training steps: {len(self.training_log)}\")\n",
    "       print(f\"Sampling steps: {len(self.sampling_log)}\")\n",
    "       if self.training_log:\n",
    "           final_loss = self.training_log[-1]['loss']\n",
    "           print(f\"Final loss: {final_loss:.6f}\")\n",
    "\n",
    "class NoisePredictor(nn.Module):\n",
    "   def __init__(self, data_dim=2, hidden_dim=256, time_embed_dim=64):\n",
    "       super(NoisePredictor, self).__init__()\n",
    "       self.data_dim = data_dim\n",
    "       self.time_embed_dim = time_embed_dim\n",
    "       \n",
    "       self.time_mlp = nn.Sequential(\n",
    "           nn.Linear(1, time_embed_dim),\n",
    "           nn.SiLU(),\n",
    "           nn.Linear(time_embed_dim, time_embed_dim),\n",
    "           nn.SiLU(),\n",
    "           nn.Linear(time_embed_dim, time_embed_dim)\n",
    "       )\n",
    "       \n",
    "       self.main_mlp = nn.Sequential(\n",
    "           nn.Linear(data_dim + time_embed_dim, hidden_dim),\n",
    "           nn.SiLU(),\n",
    "           nn.Dropout(0.1),\n",
    "           nn.Linear(hidden_dim, hidden_dim),\n",
    "           nn.SiLU(),\n",
    "           nn.Dropout(0.1),\n",
    "           nn.Linear(hidden_dim, hidden_dim),\n",
    "           nn.SiLU(),\n",
    "           nn.Linear(hidden_dim, data_dim)\n",
    "       )\n",
    "   \n",
    "   def forward(self, x, t):\n",
    "       batch_size = x.shape[0]\n",
    "       t_normalized = t.float() / 1000.0\n",
    "       t_embed = self.time_mlp(t_normalized.view(-1, 1))\n",
    "       \n",
    "       x_t_concat = torch.cat([x, t_embed], dim=1)\n",
    "       noise_pred = self.main_mlp(x_t_concat)\n",
    "       \n",
    "       return noise_pred\n",
    "\n",
    "class DiffusionModel:\n",
    "   def __init__(self, T=1000, beta_start=0.0001, beta_end=0.02, device='mps'):\n",
    "       self.T = T\n",
    "       self.device = device\n",
    "       \n",
    "       self.beta = torch.linspace(beta_start, beta_end, T).to(device)\n",
    "       self.alpha = 1. - self.beta\n",
    "       self.alpha_bar = torch.cumprod(self.alpha, dim=0)\n",
    "       \n",
    "       self.model = NoisePredictor().to(device)\n",
    "       self.optimizer = torch.optim.AdamW(self.model.parameters(), lr=0.001, weight_decay=0.01)\n",
    "       self.scheduler = torch.optim.lr_scheduler.CosineAnnealingLR(self.optimizer, T_max=2000)\n",
    "       self.logger = DiffusionLogger()\n",
    "       \n",
    "       print(f\"=== DIFFUSION MODEL INITIALIZED ===\")\n",
    "       print(f\"Device: {device}\")\n",
    "       print(f\"Timesteps: {T}\")\n",
    "       print(f\"Beta range: {beta_start} to {beta_end}\")\n",
    "       print(f\"Parameters: {sum(p.numel() for p in self.model.parameters())}\")\n",
    "       \n",
    "   def forward_process(self, x0, t):\n",
    "       batch_size = x0.shape[0]\n",
    "       \n",
    "       epsilon = torch.randn_like(x0)\n",
    "       \n",
    "       sqrt_alpha_bar = torch.sqrt(self.alpha_bar[t]).view(batch_size, 1)\n",
    "       sqrt_one_minus_alpha_bar = torch.sqrt(1 - self.alpha_bar[t]).view(batch_size, 1)\n",
    "       \n",
    "       xt = sqrt_alpha_bar * x0 + sqrt_one_minus_alpha_bar * epsilon\n",
    "       \n",
    "       return xt, epsilon\n",
    "   \n",
    "   def train_step(self, x0, epoch, batch_idx):\n",
    "       batch_size = x0.shape[0]\n",
    "       \n",
    "       t = torch.randint(0, self.T, (batch_size,)).to(self.device)\n",
    "       \n",
    "       xt, epsilon = self.forward_process(x0, t)\n",
    "       \n",
    "       epsilon_pred = self.model(xt, t)\n",
    "       \n",
    "       loss = F.mse_loss(epsilon_pred, epsilon)\n",
    "       \n",
    "       self.optimizer.zero_grad()\n",
    "       loss.backward()\n",
    "       torch.nn.utils.clip_grad_norm_(self.model.parameters(), max_norm=1.0)\n",
    "       self.optimizer.step()\n",
    "       \n",
    "       with torch.no_grad():\n",
    "           epsilon_norm = torch.norm(epsilon).item()\n",
    "           pred_norm = torch.norm(epsilon_pred).item()\n",
    "           t_mean = t.float().mean().item()\n",
    "           lr = self.optimizer.param_groups[0]['lr']\n",
    "           \n",
    "           self.logger.log_training_step(\n",
    "               epoch, batch_idx, loss.item(), lr, t_mean, epsilon_norm, pred_norm\n",
    "           )\n",
    "       \n",
    "       return loss.item()\n",
    "   \n",
    "   def sample(self, n_samples=64, save_intermediate=True):\n",
    "       self.model.eval()\n",
    "       \n",
    "       x = torch.randn(n_samples, 2).to(self.device)\n",
    "       \n",
    "       intermediate_samples = []\n",
    "       \n",
    "       print(f\"\\n=== SAMPLING {n_samples} SAMPLES ===\")\n",
    "       \n",
    "       with torch.no_grad():\n",
    "           for i in tqdm(reversed(range(self.T)), desc=\"Sampling\"):\n",
    "               t = torch.tensor([i]).repeat(n_samples).to(self.device)\n",
    "               \n",
    "               epsilon_pred = self.model(x, t)\n",
    "               \n",
    "               alpha_t = self.alpha[i]\n",
    "               beta_t = self.beta[i]\n",
    "               sqrt_one_minus_alpha_bar = torch.sqrt(1 - self.alpha_bar[i])\n",
    "               \n",
    "               mu = (x - (beta_t / sqrt_one_minus_alpha_bar) * epsilon_pred) / torch.sqrt(alpha_t)\n",
    "               \n",
    "               if i > 0:\n",
    "                   sigma = torch.sqrt(beta_t)\n",
    "                   x = mu + sigma * torch.randn_like(x)\n",
    "               else:\n",
    "                   x = mu\n",
    "               \n",
    "               if save_intermediate and i % 100 == 0:\n",
    "                   intermediate_samples.append(x.clone())\n",
    "               \n",
    "               self.logger.log_sampling_step(\n",
    "                   i, x.mean().item(), x.std().item(), torch.norm(epsilon_pred).item()\n",
    "               )\n",
    "       \n",
    "       self.model.train()\n",
    "       return x, intermediate_samples\n",
    "   \n",
    "   def train(self, data, epochs=2000, log_interval=200):\n",
    "       print(f\"\\n=== STARTING TRAINING ===\")\n",
    "       print(f\"Dataset size: {len(data)}\")\n",
    "       print(f\"Epochs: {epochs}\")\n",
    "       \n",
    "       for epoch in range(epochs):\n",
    "           epoch_loss = 0\n",
    "           \n",
    "           np.random.shuffle(data)\n",
    "           \n",
    "           for batch_idx in range(0, len(data), 32):\n",
    "               batch_data = data[batch_idx:batch_idx+32]\n",
    "               x0 = torch.stack(batch_data).to(self.device)\n",
    "               loss = self.train_step(x0, epoch, batch_idx)\n",
    "               epoch_loss += loss\n",
    "           \n",
    "           self.scheduler.step()\n",
    "           avg_loss = epoch_loss / (len(data) // 32 + 1)\n",
    "           \n",
    "           if epoch % log_interval == 0:\n",
    "               lr = self.optimizer.param_groups[0]['lr']\n",
    "               print(f\"Epoch {epoch:4d} | Loss: {avg_loss:.6f} | LR: {lr:.6f}\")\n",
    "       \n",
    "       self.logger.save_logs()\n",
    "       self.logger.print_summary()\n",
    "\n",
    "def create_toy_dataset():\n",
    "   np.random.seed(42)\n",
    "   torch.manual_seed(42)\n",
    "   \n",
    "   centers = [\n",
    "       [2.0, 2.0],\n",
    "       [-2.0, 2.0],\n",
    "       [2.0, -2.0],\n",
    "       [-2.0, -2.0],\n",
    "       [0.0, 0.0]\n",
    "   ]\n",
    "   \n",
    "   data = []\n",
    "   for center in centers:\n",
    "       for _ in range(40):\n",
    "           noise = np.random.normal(0, 0.3, 2)\n",
    "           point = np.array(center) + noise\n",
    "           data.append(torch.tensor(point, dtype=torch.float32))\n",
    "   \n",
    "   print(f\"Created dataset with {len(data)} points\")\n",
    "   print(f\"Data range: x[{min(d[0] for d in data):.2f}, {max(d[0] for d in data):.2f}], y[{min(d[1] for d in data):.2f}, {max(d[1] for d in data):.2f}]\")\n",
    "   \n",
    "   return data\n",
    "\n",
    "def visualize_results(original_data, generated_samples, intermediate_samples=None):\n",
    "   fig, axes = plt.subplots(2, 3, figsize=(18, 12))\n",
    "   \n",
    "   original_np = torch.stack(original_data).numpy()\n",
    "   generated_np = generated_samples.cpu().numpy()\n",
    "   \n",
    "   axes[0, 0].scatter(original_np[:, 0], original_np[:, 1], c='blue', alpha=0.6, s=30)\n",
    "   axes[0, 0].set_title('Original Training Data', fontsize=14)\n",
    "   axes[0, 0].grid(True, alpha=0.3)\n",
    "   axes[0, 0].set_xlim(-4, 4)\n",
    "   axes[0, 0].set_ylim(-4, 4)\n",
    "   \n",
    "   axes[0, 1].scatter(generated_np[:, 0], generated_np[:, 1], c='red', alpha=0.6, s=30)\n",
    "   axes[0, 1].set_title('Generated Samples', fontsize=14)\n",
    "   axes[0, 1].grid(True, alpha=0.3)\n",
    "   axes[0, 1].set_xlim(-4, 4)\n",
    "   axes[0, 1].set_ylim(-4, 4)\n",
    "   \n",
    "   axes[0, 2].scatter(original_np[:, 0], original_np[:, 1], c='blue', alpha=0.4, s=20, label='Original')\n",
    "   axes[0, 2].scatter(generated_np[:, 0], generated_np[:, 1], c='red', alpha=0.6, s=30, label='Generated')\n",
    "   axes[0, 2].set_title('Comparison', fontsize=14)\n",
    "   axes[0, 2].legend()\n",
    "   axes[0, 2].grid(True, alpha=0.3)\n",
    "   axes[0, 2].set_xlim(-4, 4)\n",
    "   axes[0, 2].set_ylim(-4, 4)\n",
    "   \n",
    "   if intermediate_samples and len(intermediate_samples) >= 3:\n",
    "       for i, (samples, title) in enumerate(zip(intermediate_samples[-3:], ['Early Denoising', 'Mid Denoising', 'Late Denoising'])):\n",
    "           samples_np = samples.cpu().numpy()\n",
    "           axes[1, i].scatter(samples_np[:, 0], samples_np[:, 1], c='green', alpha=0.6, s=20)\n",
    "           axes[1, i].set_title(title, fontsize=14)\n",
    "           axes[1, i].grid(True, alpha=0.3)\n",
    "           axes[1, i].set_xlim(-4, 4)\n",
    "           axes[1, i].set_ylim(-4, 4)\n",
    "   \n",
    "   plt.tight_layout()\n",
    "   plt.savefig('diffusion_results.png', dpi=300, bbox_inches='tight')\n",
    "   plt.show()\n",
    "   \n",
    "   if intermediate_samples:\n",
    "       n_steps = min(10, len(intermediate_samples))\n",
    "       fig, axes = plt.subplots(2, 5, figsize=(25, 10))\n",
    "       axes = axes.flatten()\n",
    "       \n",
    "       for i in range(n_steps):\n",
    "           samples = intermediate_samples[i]\n",
    "           samples_np = samples.cpu().numpy()\n",
    "           \n",
    "           axes[i].scatter(samples_np[:, 0], samples_np[:, 1], c='purple', alpha=0.6, s=15)\n",
    "           step_num = 1000 - i * 100\n",
    "           axes[i].set_title(f'Step {step_num}', fontsize=12)\n",
    "           axes[i].grid(True, alpha=0.3)\n",
    "           axes[i].set_xlim(-4, 4)\n",
    "           axes[i].set_ylim(-4, 4)\n",
    "       \n",
    "       plt.tight_layout()\n",
    "       plt.savefig('diffusion_process.png', dpi=300, bbox_inches='tight')\n",
    "       plt.show()\n",
    "\n",
    "def plot_training_metrics(log_file=\"diffusion_logs/training_log.json\"):\n",
    "   try:\n",
    "       with open(log_file, 'r') as f:\n",
    "           logs = json.load(f)\n",
    "       \n",
    "       epochs = [log['epoch'] for log in logs if log['batch_idx'] == 0]\n",
    "       losses = [log['loss'] for log in logs if log['batch_idx'] == 0]\n",
    "       \n",
    "       plt.figure(figsize=(12, 4))\n",
    "       \n",
    "       plt.subplot(1, 2, 1)\n",
    "       plt.plot(epochs, losses)\n",
    "       plt.title('Training Loss Over Time')\n",
    "       plt.xlabel('Epoch')\n",
    "       plt.ylabel('Loss')\n",
    "       plt.grid(True, alpha=0.3)\n",
    "       \n",
    "       recent_losses = losses[-50:] if len(losses) > 50 else losses\n",
    "       plt.subplot(1, 2, 2)\n",
    "       plt.plot(recent_losses)\n",
    "       plt.title('Recent Training Loss')\n",
    "       plt.xlabel('Recent Epochs')\n",
    "       plt.ylabel('Loss')\n",
    "       plt.grid(True, alpha=0.3)\n",
    "       \n",
    "       plt.tight_layout()\n",
    "       plt.savefig('training_metrics.png', dpi=300, bbox_inches='tight')\n",
    "       plt.show()\n",
    "       \n",
    "   except FileNotFoundError:\n",
    "       print(\"Training log file not found. Train the model first.\")\n",
    "\n",
    "print(\"Creating dataset...\")\n",
    "data = create_toy_dataset()\n",
    "\n",
    "print(\"Initializing diffusion model...\")\n",
    "diffusion = DiffusionModel(T=1000, device=device)\n",
    "\n",
    "print(\"Starting training...\")\n",
    "diffusion.train(data, epochs=3000, log_interval=300)\n",
    "\n",
    "print(\"Generating samples...\")\n",
    "generated_samples, intermediate_samples = diffusion.sample(n_samples=100)\n",
    "\n",
    "print(\"Visualizing results...\")\n",
    "visualize_results(data, generated_samples, intermediate_samples)\n",
    "\n",
    "print(\"Plotting training metrics...\")\n",
    "plot_training_metrics()\n",
    "\n",
    "print(\"Saving model...\")\n",
    "torch.save({\n",
    "   'model_state_dict': diffusion.model.state_dict(),\n",
    "   'optimizer_state_dict': diffusion.optimizer.state_dict(),\n",
    "   'beta': diffusion.beta,\n",
    "   'alpha': diffusion.alpha,\n",
    "   'alpha_bar': diffusion.alpha_bar,\n",
    "}, 'diffusion_model_complete.pth')\n",
    "\n",
    "print(\"\\n=== EXECUTION COMPLETE ===\")\n",
    "print(\"Generated files:\")\n",
    "print(\"- diffusion_results.png: Results visualization\")\n",
    "print(\"- diffusion_process.png: Sampling process steps\")\n",
    "print(\"- training_metrics.png: Training loss curves\")\n",
    "print(\"- diffusion_logs/: Detailed training logs\")\n",
    "print(\"- diffusion_model_complete.pth: Saved model\")\n",
    "\n",
    "generated_stats = {\n",
    "   'mean': generated_samples.mean(dim=0).cpu().numpy(),\n",
    "   'std': generated_samples.std(dim=0).cpu().numpy(),\n",
    "   'min': generated_samples.min(dim=0)[0].cpu().numpy(),\n",
    "   'max': generated_samples.max(dim=0)[0].cpu().numpy()\n",
    "}\n",
    "\n",
    "original_stats = {\n",
    "   'mean': torch.stack(data).mean(dim=0).numpy(),\n",
    "   'std': torch.stack(data).std(dim=0).numpy(),\n",
    "   'min': torch.stack(data).min(dim=0)[0].numpy(),\n",
    "   'max': torch.stack(data).max(dim=0)[0].numpy()\n",
    "}\n",
    "\n",
    "print(f\"\\nOriginal Data Stats:\")\n",
    "print(f\"Mean: [{original_stats['mean'][0]:.3f}, {original_stats['mean'][1]:.3f}]\")\n",
    "print(f\"Std:  [{original_stats['std'][0]:.3f}, {original_stats['std'][1]:.3f}]\")\n",
    "\n",
    "print(f\"\\nGenerated Data Stats:\")\n",
    "print(f\"Mean: [{generated_stats['mean'][0]:.3f}, {generated_stats['mean'][1]:.3f}]\")\n",
    "print(f\"Std:  [{generated_stats['std'][0]:.3f}, {generated_stats['std'][1]:.3f}]\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1e7dcffb-5f64-401c-8b82-cdf80dea8d8b",
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from tqdm import tqdm\n",
    "import os\n",
    "import json\n",
    "from datetime import datetime\n",
    "\n",
    "# Set random seeds for reproducible results across runs\n",
    "torch.manual_seed(42)\n",
    "np.random.seed(42)\n",
    "\n",
    "# Use MPS (Metal Performance Shaders) for M4 chip acceleration\n",
    "device = torch.device(\"mps\" if torch.backends.mps.is_available() else \"cpu\")\n",
    "print(f\"Using device: {device}\")\n",
    "\n",
    "class DiffusionLogger:\n",
    "    \"\"\"\n",
    "    Comprehensive logging system to track training and sampling metrics.\n",
    "    This helps us understand what's happening during the diffusion process.\n",
    "    \"\"\"\n",
    "    def __init__(self, log_dir=\"diffusion_logs\"):\n",
    "        self.log_dir = log_dir\n",
    "        os.makedirs(log_dir, exist_ok=True)  # Create directory if it doesn't exist\n",
    "        self.training_log = []  # Store training step information\n",
    "        self.sampling_log = []  # Store sampling step information\n",
    "        self.start_time = datetime.now()  # Track total training time\n",
    "        \n",
    "    def log_training_step(self, epoch, batch_idx, loss, lr, t_mean, epsilon_norm, pred_norm):\n",
    "        \"\"\"\n",
    "        Log metrics from each training step:\n",
    "        - epoch: Current training epoch\n",
    "        - batch_idx: Current batch index\n",
    "        - loss: MSE loss between predicted and actual noise\n",
    "        - lr: Current learning rate\n",
    "        - t_mean: Average timestep in this batch\n",
    "        - epsilon_norm: L2 norm of actual noise\n",
    "        - pred_norm: L2 norm of predicted noise\n",
    "        \"\"\"\n",
    "        log_entry = {\n",
    "            'epoch': epoch,\n",
    "            'batch_idx': batch_idx,\n",
    "            'loss': loss,\n",
    "            'learning_rate': lr,\n",
    "            'timestep_mean': t_mean,\n",
    "            'epsilon_norm': epsilon_norm,\n",
    "            'prediction_norm': pred_norm,\n",
    "            'timestamp': datetime.now().isoformat()\n",
    "        }\n",
    "        self.training_log.append(log_entry)\n",
    "        \n",
    "    def log_sampling_step(self, step, x_mean, x_std, epsilon_pred_norm):\n",
    "        \"\"\"\n",
    "        Log metrics from each sampling step:\n",
    "        - step: Current reverse diffusion step (1000 → 0)\n",
    "        - x_mean: Mean of current sample values\n",
    "        - x_std: Standard deviation of current sample values\n",
    "        - epsilon_pred_norm: L2 norm of predicted noise\n",
    "        \"\"\"\n",
    "        log_entry = {\n",
    "            'step': step,\n",
    "            'x_mean': x_mean,\n",
    "            'x_std': x_std,\n",
    "            'epsilon_pred_norm': epsilon_pred_norm,\n",
    "            'timestamp': datetime.now().isoformat()\n",
    "        }\n",
    "        self.sampling_log.append(log_entry)\n",
    "        \n",
    "    def save_logs(self):\n",
    "        \"\"\"Save all logs to JSON files for later analysis\"\"\"\n",
    "        with open(f\"{self.log_dir}/training_log.json\", 'w') as f:\n",
    "            json.dump(self.training_log, f, indent=2)\n",
    "        with open(f\"{self.log_dir}/sampling_log.json\", 'w') as f:\n",
    "            json.dump(self.sampling_log, f, indent=2)\n",
    "            \n",
    "    def print_summary(self):\n",
    "        \"\"\"Print a summary of the training process\"\"\"\n",
    "        total_time = datetime.now() - self.start_time\n",
    "        print(f\"\\n=== TRAINING SUMMARY ===\")\n",
    "        print(f\"Total time: {total_time}\")\n",
    "        print(f\"Training steps: {len(self.training_log)}\")\n",
    "        print(f\"Sampling steps: {len(self.sampling_log)}\")\n",
    "        if self.training_log:\n",
    "            final_loss = self.training_log[-1]['loss']\n",
    "            print(f\"Final loss: {final_loss:.6f}\")\n",
    "\n",
    "class NoisePredictor(nn.Module):\n",
    "    \"\"\"\n",
    "    Neural network that predicts noise ε_θ(x_t, t) given:\n",
    "    - x_t: noisy data at timestep t\n",
    "    - t: current timestep\n",
    "    \n",
    "    Architecture:\n",
    "    1. Time embedding: Convert timestep to rich representation\n",
    "    2. Concatenate: Combine data and time embeddings\n",
    "    3. MLP: Multi-layer perceptron to predict noise\n",
    "    \"\"\"\n",
    "    def __init__(self, data_dim=2, hidden_dim=256, time_embed_dim=64):\n",
    "        super(NoisePredictor, self).__init__()\n",
    "        self.data_dim = data_dim  # Dimension of input data (2D points)\n",
    "        self.time_embed_dim = time_embed_dim  # Dimension of time embedding\n",
    "        \n",
    "        # Time embedding network: converts timestep to rich representation\n",
    "        # This helps the model understand \"how much noise\" to expect\n",
    "        self.time_mlp = nn.Sequential(\n",
    "            nn.Linear(1, time_embed_dim),  # Input: single timestep value\n",
    "            nn.SiLU(),  # Swish activation (smooth, works well for diffusion)\n",
    "            nn.Linear(time_embed_dim, time_embed_dim),  # Hidden layer\n",
    "            nn.SiLU(),\n",
    "            nn.Linear(time_embed_dim, time_embed_dim)  # Output: time embedding\n",
    "        )\n",
    "        \n",
    "        # Main network: predicts noise from data + time embedding\n",
    "        self.main_mlp = nn.Sequential(\n",
    "            nn.Linear(data_dim + time_embed_dim, hidden_dim),  # Input: data + time\n",
    "            nn.SiLU(),  # Smooth activation\n",
    "            nn.Dropout(0.1),  # Prevent overfitting\n",
    "            nn.Linear(hidden_dim, hidden_dim),  # Hidden layer 1\n",
    "            nn.SiLU(),\n",
    "            nn.Dropout(0.1),\n",
    "            nn.Linear(hidden_dim, hidden_dim),  # Hidden layer 2\n",
    "            nn.SiLU(),\n",
    "            nn.Linear(hidden_dim, data_dim)  # Output: predicted noise (same dim as input)\n",
    "        )\n",
    "    \n",
    "    def forward(self, x, t):\n",
    "        \"\"\"\n",
    "        Forward pass:\n",
    "        x: [batch_size, data_dim] - noisy data\n",
    "        t: [batch_size] - timesteps\n",
    "        Returns: [batch_size, data_dim] - predicted noise\n",
    "        \"\"\"\n",
    "        batch_size = x.shape[0]\n",
    "        \n",
    "        # Normalize timestep to [0, 1] range for stable training\n",
    "        t_normalized = t.float() / 1000.0\n",
    "        \n",
    "        # Create time embedding: timestep → rich representation\n",
    "        t_embed = self.time_mlp(t_normalized.view(-1, 1))\n",
    "        \n",
    "        # Concatenate data and time embedding\n",
    "        x_t_concat = torch.cat([x, t_embed], dim=1)\n",
    "        \n",
    "        # Predict noise using main network\n",
    "        noise_pred = self.main_mlp(x_t_concat)\n",
    "        \n",
    "        return noise_pred\n",
    "\n",
    "class DiffusionModel:\n",
    "    \"\"\"\n",
    "    Complete Denoising Diffusion Probabilistic Model (DDPM) implementation.\n",
    "    \n",
    "    Key components:\n",
    "    1. Noise schedule: β_t, α_t, ᾱ_t\n",
    "    2. Forward process: q(x_t | x_0) - adds noise\n",
    "    3. Reverse process: p_θ(x_{t-1} | x_t) - removes noise\n",
    "    4. Training: minimize ||ε - ε_θ(x_t, t)||²\n",
    "    5. Sampling: start from noise, iteratively denoise\n",
    "    \"\"\"\n",
    "    def __init__(self, T=1000, beta_start=0.0001, beta_end=0.02, device='mps'):\n",
    "        self.T = T  # Number of diffusion timesteps\n",
    "        self.device = device\n",
    "        \n",
    "        # Noise schedule: β_t increases linearly from β_start to β_end\n",
    "        # β_t controls how much noise to add at each step\n",
    "        self.beta = torch.linspace(beta_start, beta_end, T).to(device)\n",
    "        \n",
    "        # α_t = 1 - β_t: signal retention at each step\n",
    "        self.alpha = 1. - self.beta\n",
    "        \n",
    "        # ᾱ_t = ∏_{s=1}^t α_s: cumulative signal retention\n",
    "        # This allows us to jump directly to any timestep t\n",
    "        self.alpha_bar = torch.cumprod(self.alpha, dim=0)\n",
    "        \n",
    "        # Initialize neural network and optimizer\n",
    "        self.model = NoisePredictor().to(device)\n",
    "        self.optimizer = torch.optim.AdamW(self.model.parameters(), lr=0.001, weight_decay=0.01)\n",
    "        \n",
    "        # Cosine annealing: gradually reduce learning rate\n",
    "        self.scheduler = torch.optim.lr_scheduler.CosineAnnealingLR(self.optimizer, T_max=2000)\n",
    "        \n",
    "        # Initialize logger\n",
    "        self.logger = DiffusionLogger()\n",
    "        \n",
    "        print(f\"=== DIFFUSION MODEL INITIALIZED ===\")\n",
    "        print(f\"Device: {device}\")\n",
    "        print(f\"Timesteps: {T}\")\n",
    "        print(f\"Beta range: {beta_start} to {beta_end}\")\n",
    "        print(f\"Parameters: {sum(p.numel() for p in self.model.parameters())}\")\n",
    "        \n",
    "    def forward_process(self, x0, t):\n",
    "        \"\"\"\n",
    "        Forward process: q(x_t | x_0) = N(x_t; √ᾱ_t x_0, (1-ᾱ_t)I)\n",
    "        \n",
    "        This is the reparameterization trick:\n",
    "        x_t = √ᾱ_t x_0 + √(1-ᾱ_t) ε, where ε ~ N(0, I)\n",
    "        \n",
    "        Args:\n",
    "            x0: [batch_size, data_dim] - clean data\n",
    "            t: [batch_size] - timesteps\n",
    "        \n",
    "        Returns:\n",
    "            xt: [batch_size, data_dim] - noisy data\n",
    "            epsilon: [batch_size, data_dim] - noise that was added\n",
    "        \"\"\"\n",
    "        batch_size = x0.shape[0]\n",
    "        \n",
    "        # Sample random noise from standard normal distribution\n",
    "        epsilon = torch.randn_like(x0)\n",
    "        \n",
    "        # Get noise schedule values for timesteps t\n",
    "        sqrt_alpha_bar = torch.sqrt(self.alpha_bar[t]).view(batch_size, 1)\n",
    "        sqrt_one_minus_alpha_bar = torch.sqrt(1 - self.alpha_bar[t]).view(batch_size, 1)\n",
    "        \n",
    "        # Apply reparameterization trick: x_t = √ᾱ_t x_0 + √(1-ᾱ_t) ε\n",
    "        xt = sqrt_alpha_bar * x0 + sqrt_one_minus_alpha_bar * epsilon\n",
    "        \n",
    "        return xt, epsilon\n",
    "    \n",
    "    def train_step(self, x0, epoch, batch_idx):\n",
    "        \"\"\"\n",
    "        Single training step:\n",
    "        1. Sample random timestep t\n",
    "        2. Add noise to get x_t\n",
    "        3. Predict noise: ε_θ(x_t, t)\n",
    "        4. Compute loss: ||ε - ε_θ(x_t, t)||²\n",
    "        5. Backpropagate and update parameters\n",
    "        \"\"\"\n",
    "        batch_size = x0.shape[0]\n",
    "        \n",
    "        # Sample random timestep for each data point\n",
    "        t = torch.randint(0, self.T, (batch_size,)).to(self.device)\n",
    "        \n",
    "        # Forward process: add noise to clean data\n",
    "        xt, epsilon = self.forward_process(x0, t)\n",
    "        \n",
    "        # Predict noise using neural network\n",
    "        epsilon_pred = self.model(xt, t)\n",
    "        \n",
    "        # Compute MSE loss between actual and predicted noise\n",
    "        loss = F.mse_loss(epsilon_pred, epsilon)\n",
    "        \n",
    "        # Backpropagation\n",
    "        self.optimizer.zero_grad()  # Clear gradients\n",
    "        loss.backward()  # Compute gradients\n",
    "        \n",
    "        # Gradient clipping to prevent exploding gradients\n",
    "        torch.nn.utils.clip_grad_norm_(self.model.parameters(), max_norm=1.0)\n",
    "        \n",
    "        self.optimizer.step()  # Update parameters\n",
    "        \n",
    "        # Log training metrics (without gradients to save memory)\n",
    "        with torch.no_grad():\n",
    "            epsilon_norm = torch.norm(epsilon).item()\n",
    "            pred_norm = torch.norm(epsilon_pred).item()\n",
    "            t_mean = t.float().mean().item()\n",
    "            lr = self.optimizer.param_groups[0]['lr']\n",
    "            \n",
    "            self.logger.log_training_step(\n",
    "                epoch, batch_idx, loss.item(), lr, t_mean, epsilon_norm, pred_norm\n",
    "            )\n",
    "        \n",
    "        return loss.item()\n",
    "    \n",
    "    def sample(self, n_samples=64, save_intermediate=True):\n",
    "        \"\"\"\n",
    "        Sampling process: p_θ(x_{t-1} | x_t)\n",
    "        \n",
    "        Algorithm:\n",
    "        1. Start with x_T ~ N(0, I) (pure noise)\n",
    "        2. For t = T, T-1, ..., 1:\n",
    "           - Predict noise: ε_θ(x_t, t)\n",
    "           - Compute mean: μ_θ(x_t, t)\n",
    "           - Sample: x_{t-1} ~ N(μ_θ(x_t, t), σ_t² I)\n",
    "        3. Return x_0 (clean sample)\n",
    "        \"\"\"\n",
    "        self.model.eval()  # Set to evaluation mode\n",
    "        \n",
    "        # Start with pure noise: x_T ~ N(0, I)\n",
    "        x = torch.randn(n_samples, 2).to(self.device)\n",
    "        \n",
    "        intermediate_samples = []  # Store intermediate results\n",
    "        \n",
    "        print(f\"\\n=== SAMPLING {n_samples} SAMPLES ===\")\n",
    "        \n",
    "        # Disable gradient computation for efficiency\n",
    "        with torch.no_grad():\n",
    "            # Reverse diffusion: T → T-1 → ... → 1 → 0\n",
    "            for i in tqdm(reversed(range(self.T)), desc=\"Sampling\"):\n",
    "                # Current timestep for all samples\n",
    "                t = torch.tensor([i]).repeat(n_samples).to(self.device)\n",
    "                \n",
    "                # Predict noise at current timestep\n",
    "                epsilon_pred = self.model(x, t)\n",
    "                \n",
    "                # Get noise schedule values\n",
    "                alpha_t = self.alpha[i]\n",
    "                beta_t = self.beta[i]\n",
    "                sqrt_one_minus_alpha_bar = torch.sqrt(1 - self.alpha_bar[i])\n",
    "                \n",
    "                # Compute mean of reverse process:\n",
    "                # μ_θ(x_t, t) = (1/√α_t) * (x_t - (β_t/√(1-ᾱ_t)) * ε_θ(x_t, t))\n",
    "                mu = (x - (beta_t / sqrt_one_minus_alpha_bar) * epsilon_pred) / torch.sqrt(alpha_t)\n",
    "                \n",
    "                if i > 0:\n",
    "                    # Add noise for all steps except the last one\n",
    "                    # σ_t = √β_t\n",
    "                    sigma = torch.sqrt(beta_t)\n",
    "                    x = mu + sigma * torch.randn_like(x)\n",
    "                else:\n",
    "                    # Final step: no noise added\n",
    "                    x = mu\n",
    "                \n",
    "                # Save intermediate results every 100 steps\n",
    "                if save_intermediate and i % 100 == 0:\n",
    "                    intermediate_samples.append(x.clone())\n",
    "                \n",
    "                # Log sampling metrics\n",
    "                self.logger.log_sampling_step(\n",
    "                    i, x.mean().item(), x.std().item(), torch.norm(epsilon_pred).item()\n",
    "                )\n",
    "        \n",
    "        self.model.train()  # Return to training mode\n",
    "        return x, intermediate_samples\n",
    "    \n",
    "    def train(self, data, epochs=2000, log_interval=200):\n",
    "        \"\"\"\n",
    "        Complete training loop:\n",
    "        1. For each epoch:\n",
    "           - Shuffle data\n",
    "           - Process in batches\n",
    "           - Update learning rate\n",
    "        2. Save logs and print summary\n",
    "        \"\"\"\n",
    "        print(f\"\\n=== STARTING TRAINING ===\")\n",
    "        print(f\"Dataset size: {len(data)}\")\n",
    "        print(f\"Epochs: {epochs}\")\n",
    "        \n",
    "        for epoch in range(epochs):\n",
    "            epoch_loss = 0\n",
    "            \n",
    "            # Shuffle data each epoch for better training\n",
    "            np.random.shuffle(data)\n",
    "            \n",
    "            # Process data in batches of 32\n",
    "            for batch_idx in range(0, len(data), 32):\n",
    "                batch_data = data[batch_idx:batch_idx+32]\n",
    "                x0 = torch.stack(batch_data).to(self.device)\n",
    "                loss = self.train_step(x0, epoch, batch_idx)\n",
    "                epoch_loss += loss\n",
    "            \n",
    "            # Update learning rate schedule\n",
    "            self.scheduler.step()\n",
    "            avg_loss = epoch_loss / (len(data) // 32 + 1)\n",
    "            \n",
    "            # Print progress\n",
    "            if epoch % log_interval == 0:\n",
    "                lr = self.optimizer.param_groups[0]['lr']\n",
    "                print(f\"Epoch {epoch:4d} | Loss: {avg_loss:.6f} | LR: {lr:.6f}\")\n",
    "        \n",
    "        # Save training logs\n",
    "        self.logger.save_logs()\n",
    "        self.logger.print_summary()\n",
    "\n",
    "def create_toy_dataset():\n",
    "    \"\"\"\n",
    "    Create a 2D toy dataset with 5 clusters:\n",
    "    - 4 clusters at corners: (±2, ±2)\n",
    "    - 1 cluster at center: (0, 0)\n",
    "    \n",
    "    Each cluster has 40 points with Gaussian noise (σ=0.3)\n",
    "    Total: 200 points\n",
    "    \"\"\"\n",
    "    np.random.seed(42)\n",
    "    torch.manual_seed(42)\n",
    "    \n",
    "    # Define cluster centers\n",
    "    centers = [\n",
    "        [2.0, 2.0],   # Top-right\n",
    "        [-2.0, 2.0],  # Top-left\n",
    "        [2.0, -2.0],  # Bottom-right\n",
    "        [-2.0, -2.0], # Bottom-left\n",
    "        [0.0, 0.0]    # Center\n",
    "    ]\n",
    "    \n",
    "    data = []\n",
    "    # Generate points around each center\n",
    "    for center in centers:\n",
    "        for _ in range(40):  # 40 points per cluster\n",
    "            # Add Gaussian noise to center position\n",
    "            noise = np.random.normal(0, 0.3, 2)  # σ=0.3 for tight clusters\n",
    "            point = np.array(center) + noise\n",
    "            data.append(torch.tensor(point, dtype=torch.float32))\n",
    "    \n",
    "    print(f\"Created dataset with {len(data)} points\")\n",
    "    print(f\"Data range: x[{min(d[0] for d in data):.2f}, {max(d[0] for d in data):.2f}], y[{min(d[1] for d in data):.2f}, {max(d[1] for d in data):.2f}]\")\n",
    "    \n",
    "    return data\n",
    "\n",
    "def visualize_results(original_data, generated_samples, intermediate_samples=None):\n",
    "    \"\"\"\n",
    "    Create comprehensive visualizations:\n",
    "    1. Original training data\n",
    "    2. Generated samples\n",
    "    3. Comparison overlay\n",
    "    4. Intermediate sampling steps (if available)\n",
    "    \"\"\"\n",
    "    fig, axes = plt.subplots(2, 3, figsize=(18, 12))\n",
    "    \n",
    "    # Convert to numpy for plotting\n",
    "    original_np = torch.stack(original_data).numpy()\n",
    "    generated_np = generated_samples.cpu().numpy()\n",
    "    \n",
    "    # Plot 1: Original training data\n",
    "    axes[0, 0].scatter(original_np[:, 0], original_np[:, 1], c='blue', alpha=0.6, s=30)\n",
    "    axes[0, 0].set_title('Original Training Data', fontsize=14)\n",
    "    axes[0, 0].grid(True, alpha=0.3)\n",
    "    axes[0, 0].set_xlim(-4, 4)\n",
    "    axes[0, 0].set_ylim(-4, 4)\n",
    "    \n",
    "    # Plot 2: Generated samples\n",
    "    axes[0, 1].scatter(generated_np[:, 0], generated_np[:, 1], c='red', alpha=0.6, s=30)\n",
    "    axes[0, 1].set_title('Generated Samples', fontsize=14)\n",
    "    axes[0, 1].grid(True, alpha=0.3)\n",
    "    axes[0, 1].set_xlim(-4, 4)\n",
    "    axes[0, 1].set_ylim(-4, 4)\n",
    "    \n",
    "    # Plot 3: Comparison overlay\n",
    "    axes[0, 2].scatter(original_np[:, 0], original_np[:, 1], c='blue', alpha=0.4, s=20, label='Original')\n",
    "    axes[0, 2].scatter(generated_np[:, 0], generated_np[:, 1], c='red', alpha=0.6, s=30, label='Generated')\n",
    "    axes[0, 2].set_title('Comparison', fontsize=14)\n",
    "    axes[0, 2].legend()\n",
    "    axes[0, 2].grid(True, alpha=0.3)\n",
    "    axes[0, 2].set_xlim(-4, 4)\n",
    "    axes[0, 2].set_ylim(-4, 4)\n",
    "    \n",
    "    # Plot 4-6: Intermediate sampling steps\n",
    "    if intermediate_samples and len(intermediate_samples) >= 3:\n",
    "        for i, (samples, title) in enumerate(zip(intermediate_samples[-3:], ['Early Denoising', 'Mid Denoising', 'Late Denoising'])):\n",
    "            samples_np = samples.cpu().numpy()\n",
    "            axes[1, i].scatter(samples_np[:, 0], samples_np[:, 1], c='green', alpha=0.6, s=20)\n",
    "            axes[1, i].set_title(title, fontsize=14)\n",
    "            axes[1, i].grid(True, alpha=0.3)\n",
    "            axes[1, i].set_xlim(-4, 4)\n",
    "            axes[1, i].set_ylim(-4, 4)\n",
    "    \n",
    "    plt.tight_layout()\n",
    "    plt.savefig('diffusion_results.png', dpi=300, bbox_inches='tight')\n",
    "    plt.show()\n",
    "    \n",
    "    # Additional plot: Complete sampling process\n",
    "    if intermediate_samples:\n",
    "        n_steps = min(10, len(intermediate_samples))\n",
    "        fig, axes = plt.subplots(2, 5, figsize=(25, 10))\n",
    "        axes = axes.flatten()\n",
    "        \n",
    "        for i in range(n_steps):\n",
    "            samples = intermediate_samples[i]\n",
    "            samples_np = samples.cpu().numpy()\n",
    "            \n",
    "            axes[i].scatter(samples_np[:, 0], samples_np[:, 1], c='purple', alpha=0.6, s=15)\n",
    "            step_num = 1000 - i * 100\n",
    "            axes[i].set_title(f'Step {step_num}', fontsize=12)\n",
    "            axes[i].grid(True, alpha=0.3)\n",
    "            axes[i].set_xlim(-4, 4)\n",
    "            axes[i].set_ylim(-4, 4)\n",
    "        \n",
    "        plt.tight_layout()\n",
    "        plt.savefig('diffusion_process.png', dpi=300, bbox_inches='tight')\n",
    "        plt.show()\n",
    "\n",
    "def plot_training_metrics(log_file=\"diffusion_logs/training_log.json\"):\n",
    "    \"\"\"\n",
    "    Plot training metrics from saved logs:\n",
    "    1. Complete training loss curve\n",
    "    2. Recent training loss (last 50 epochs)\n",
    "    \"\"\"\n",
    "    try:\n",
    "        with open(log_file, 'r') as f:\n",
    "            logs = json.load(f)\n",
    "        \n",
    "        # Extract loss values for first batch of each epoch\n",
    "        epochs = [log['epoch'] for log in logs if log['batch_idx'] == 0]\n",
    "        losses = [log['loss'] for log in logs if log['batch_idx'] == 0]\n",
    "        \n",
    "        plt.figure(figsize=(12, 4))\n",
    "        \n",
    "        # Plot 1: Complete training curve\n",
    "        plt.subplot(1, 2, 1)\n",
    "        plt.plot(epochs, losses)\n",
    "        plt.title('Training Loss Over Time')\n",
    "        plt.xlabel('Epoch')\n",
    "        plt.ylabel('Loss')\n",
    "        plt.grid(True, alpha=0.3)\n",
    "        \n",
    "        # Plot 2: Recent training (last 50 epochs)\n",
    "        recent_losses = losses[-50:] if len(losses) > 50 else losses\n",
    "        plt.subplot(1, 2, 2)\n",
    "        plt.plot(recent_losses)\n",
    "        plt.title('Recent Training Loss')\n",
    "        plt.xlabel('Recent Epochs')\n",
    "        plt.ylabel('Loss')\n",
    "        plt.grid(True, alpha=0.3)\n",
    "        \n",
    "        plt.tight_layout()\n",
    "        plt.savefig('training_metrics.png', dpi=300, bbox_inches='tight')\n",
    "        plt.show()\n",
    "        \n",
    "    except FileNotFoundError:\n",
    "        print(\"Training log file not found. Train the model first.\")\n",
    "\n",
    "# === MAIN EXECUTION ===\n",
    "\n",
    "print(\"Creating dataset...\")\n",
    "data = create_toy_dataset()  # Create 5-cluster 2D dataset\n",
    "\n",
    "print(\"Initializing diffusion model...\")\n",
    "diffusion = DiffusionModel(T=1000, device=device)  # Initialize DDPM\n",
    "\n",
    "print(\"Starting training...\")\n",
    "diffusion.train(data, epochs=3000, log_interval=300)  # Train for 3000 epochs\n",
    "\n",
    "print(\"Generating samples...\")\n",
    "generated_samples, intermediate_samples = diffusion.sample(n_samples=100)  # Generate 100 new samples\n",
    "\n",
    "print(\"Visualizing results...\")\n",
    "visualize_results(data, generated_samples, intermediate_samples)  # Create plots\n",
    "\n",
    "print(\"Plotting training metrics...\")\n",
    "plot_training_metrics()  # Plot loss curves\n",
    "\n",
    "print(\"Saving model...\")\n",
    "# Save complete model state for later use\n",
    "torch.save({\n",
    "    'model_state_dict': diffusion.model.state_dict(),\n",
    "    'optimizer_state_dict': diffusion.optimizer.state_dict(),\n",
    "    'beta': diffusion.beta,\n",
    "    'alpha': diffusion.alpha,\n",
    "    'alpha_bar': diffusion.alpha_bar,\n",
    "}, 'diffusion_model_complete.pth')\n",
    "\n",
    "print(\"\\n=== EXECUTION COMPLETE ===\")\n",
    "print(\"Generated files:\")\n",
    "print(\"- diffusion_results.png: Results visualization\")\n",
    "print(\"- diffusion_process.png: Sampling process steps\")\n",
    "print(\"- training_metrics.png: Training loss curves\")\n",
    "print(\"- diffusion_logs/: Detailed training logs\")\n",
    "print(\"- diffusion_model_complete.pth: Saved model\")\n",
    "\n",
    "# Calculate and display statistics\n",
    "generated_stats = {\n",
    "    'mean': generated_samples.mean(dim=0).cpu().numpy(),\n",
    "    'std': generated_samples.std(dim=0).cpu().numpy(),\n",
    "    'min': generated_samples.min(dim=0)[0].cpu().numpy(),\n",
    "    'max': generated_samples.max(dim=0)[0].cpu().numpy()\n",
    "}\n",
    "\n",
    "original_stats = {\n",
    "    'mean': torch.stack(data).mean(dim=0).numpy(),\n",
    "    'std': torch.stack(data).std(dim=0).numpy(),\n",
    "    'min': torch.stack(data).min(dim=0)[0].numpy(),\n",
    "    'max': torch.stack(data).max(dim=0)[0].numpy()\n",
    "}\n",
    "\n",
    "print(f\"\\nOriginal Data Stats:\")\n",
    "print(f\"Mean: [{original_stats['mean'][0]:.3f}, {original_stats['mean'][1]:.3f}]\")\n",
    "print(f\"Std:  [{original_stats['std'][0]:.3f}, {original_stats['std'][1]:.3f}]\")\n",
    "\n",
    "print(f\"\\nGenerated Data Stats:\")\n",
    "print(f\"Mean: [{generated_stats['mean'][0]:.3f}, {generated_stats['mean'][1]:.3f}]\")\n",
    "print(f\"Std:  [{generated_stats['std'][0]:.3f}, {generated_stats['std'][1]:.3f}]\")"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.13.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}