{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# LAMBA V1.0 — Quickstart (Colab GPU)\n",
    "\n",
    "A ~177M from-scratch **Mamba-3 + GQA** hybrid for **English + Turkish**.\n",
    "\n",
    "> ⚠️ LAMBA is small (177M). It **hallucinates** facts on its own — use it **with retrieval (RAG)**, which this notebook does by default.\n",
    "\n",
    "**Runtime → Change runtime type → GPU** (free T4 tier works). Then run the cells in order.\n",
    "\n",
    "💛 Support bigger open LAMBA models: https://www.patreon.com/c/kdirgul/membership"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 1) Setup — download the model + install the Mamba-3 GPU wheel (~1 min)\n",
    "!pip -q install einops sentencepiece \"huggingface_hub>=0.23\" sentence-transformers\n",
    "from huggingface_hub import snapshot_download\n",
    "REPO = \"kdirgul/LAMBA-V1.0-MAMBA3\"\n",
    "DIR = snapshot_download(REPO)            # public repo, no token needed\n",
    "!pip -q install --no-deps {DIR}/wheels/*.whl\n",
    "import os; print(\"LAMBA V1.0 ready at:\", DIR)\n",
    "print(\"files:\", os.listdir(DIR))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 2) RAG inference (recommended) — LAMBA answers from a built-in demo corpus\n",
    "!cd {DIR} && python lamba_rag.py --demo \\\n",
    "  --tokenizer tokenizer/tokenizer.model \\\n",
    "  --ckpt checkpoints/lamba_v1.pt \\\n",
    "  --temperature 0 --top_k 1 \\\n",
    "  --query \"Türkiye'nin başkenti neresi?\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 3) Your own documents — drop .txt/.md files in /content/my_docs, then ask\n",
    "import os; os.makedirs(\"/content/my_docs\", exist_ok=True)\n",
    "# (upload files into /content/my_docs via the Colab file panel first)\n",
    "!cd {DIR} && python lamba_rag.py --docs /content/my_docs \\\n",
    "  --tokenizer tokenizer/tokenizer.model \\\n",
    "  --ckpt checkpoints/lamba_v1.pt \\\n",
    "  --temperature 0 --top_k 1 \\\n",
    "  --query \"Write your question here\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Notes\n",
    "- **Use RAG** for any factual question. Without context, LAMBA may invent answers.\n",
    "- Works in **English and Turkish**; reasoning/CoT is strongest in English.\n",
    "- GPU is required for now (Mamba-3 Triton kernel). A CPU build is planned for **v1.1**.\n",
    "- Full details, eval scores, and limitations: see the **model card (README)**.\n",
    "\n",
    "Built by **Kadir Gül**. If LAMBA is useful to you, consider supporting compute for the next version 🙏"
   ]
  }
 ],
 "metadata": {
  "accelerator": "GPU",
  "colab": {"provenance": []},
  "kernelspec": {"display_name": "Python 3", "name": "python3"},
  "language_info": {"name": "python"}
 },
 "nbformat": 4,
 "nbformat_minor": 0
}