{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/duoan/TorchCode/blob/master/templates/32_topk_sampling.ipynb)\n",
    "\n",
    "# 🟠 Medium: Top-k / Top-p (Nucleus) Sampling\n",
    "\n",
    "Implement **sampling with top-k and top-p filtering** — the standard LLM decoding strategy.\n",
    "\n",
    "### Signature\n",
    "```python\n",
    "def sample_top_k_top_p(logits, top_k=0, top_p=1.0, temperature=1.0) -> int:\n",
    "    # logits: (V,) unnormalized log-probabilities\n",
    "    # Returns: sampled token index\n",
    "```\n",
    "\n",
    "### Algorithm\n",
    "1. Scale by temperature: `logits /= temperature`\n",
    "2. Top-k: keep only top-k logits, set rest to `-inf`\n",
    "3. Top-p: sort by prob, mask tokens where cumulative prob exceeds p\n",
    "4. Sample from filtered distribution"
   ],
   "outputs": []
  },
  {
   "cell_type": "code",
   "metadata": {},
   "source": [
    "# Install torch-judge in Colab (no-op in JupyterLab/Docker)\n",
    "try:\n",
    "    import google.colab\n",
    "    get_ipython().run_line_magic('pip', 'install -q torch-judge')\n",
    "except ImportError:\n",
    "    pass\n"
   ],
   "outputs": [],
   "execution_count": null
  },
  {
   "cell_type": "code",
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch"
   ],
   "execution_count": null
  },
  {
   "cell_type": "code",
   "metadata": {},
   "outputs": [],
   "source": [
    "# ✏️ YOUR IMPLEMENTATION HERE\n",
    "\n",
    "def sample_top_k_top_p(logits, top_k=0, top_p=1.0, temperature=1.0):\n",
    "    pass  # temperature, top-k filter, top-p filter, sample"
   ],
   "execution_count": null
  },
  {
   "cell_type": "code",
   "metadata": {},
   "outputs": [],
   "source": [
    "# 🧪 Debug\n",
    "logits = torch.tensor([1.0, 5.0, 2.0, 0.5])\n",
    "print('top_k=1:', sample_top_k_top_p(logits.clone(), top_k=1))\n",
    "print('top_p=0.5:', sample_top_k_top_p(logits.clone(), top_p=0.5))\n",
    "print('temp=0.01:', sample_top_k_top_p(logits.clone(), temperature=0.01))"
   ],
   "execution_count": null
  },
  {
   "cell_type": "code",
   "metadata": {},
   "outputs": [],
   "source": [
    "# ✅ SUBMIT\n",
    "from torch_judge import check\n",
    "check('topk_sampling')"
   ],
   "execution_count": null
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "name": "python",
   "version": "3.11.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}