{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 🔐 Ethical Hacking LLM – Qwen3-8B (Alternative for T4)\n",
    "\n",
    "**Model:** [Qwen3-8B](https://huggingface.co/Qwen/Qwen3-8B-Base) via Unsloth 4-bit  \n",
    "**Trade-off:** More parameters = better capacity for complex exploits, but tighter VRAM.\n",
    "\n",
    "> If this OOMs on your T4, switch to the [Qwen3-4B version](https://huggingface.co/asdf98/ethical-hacking-llm-colab/blob/main/EthicalHacking_Qwen3-4B_Ultimate_Colab.ipynb) instead."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%capture\n",
    "!pip install -q unsloth trl datasets accelerate transformers bitsandbytes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from unsloth import FastLanguageModel\n",
    "import torch\n",
    "\n",
    "MAX_SEQ_LENGTH = 2048\n",
    "LORA_R = 16\n",
    "LORA_ALPHA = 16\n",
    "BATCH_SIZE = 1\n",
    "GRAD_ACCUM = 4\n",
    "LEARNING_RATE = 2e-4\n",
    "NUM_EPOCHS = 1\n",
    "\n",
    "model, tokenizer = FastLanguageModel.from_pretrained(\n",
    "    model_name=\"unsloth/Qwen3-8B-unsloth-bnb-4bit\",\n",
    "    max_seq_length=MAX_SEQ_LENGTH,\n",
    "    dtype=None,\n",
    "    load_in_4bit=True,\n",
    ")\n",
    "\n",
    "model = FastLanguageModel.get_peft_model(\n",
    "    model,\n",
    "    r=LORA_R,\n",
    "    target_modules=[\"q_proj\", \"k_proj\", \"v_proj\", \"o_proj\",\n",
    "                   \"gate_proj\", \"up_proj\", \"down_proj\"],\n",
    "    lora_alpha=LORA_ALPHA,\n",
    "    lora_dropout=0,\n",
    "    bias=\"none\",\n",
    "    use_gradient_checkpointing=\"unsloth\",\n",
    "    random_state=3407,\n",
    "    use_rslora=False,\n",
    ")\n",
    "print(\"✅ Qwen3-8B loaded. Use Qwen3-4B notebook if this OOMs.\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from datasets import load_dataset, concatenate_datasets\n",
    "\n",
    "ds1 = load_dataset(\"AlicanKiraz0/Cybersecurity-Dataset-Fenrir-v2.1\", split=\"train\")\n",
    "ds2 = load_dataset(\"Trendyol/Trendyol-Cybersecurity-Instruction-Tuning-Dataset\", split=\"train\")\n",
    "\n",
    "def to_messages(example):\n",
    "    return {\"messages\": [\n",
    "        {\"role\": \"system\", \"content\": example[\"system\"]},\n",
    "        {\"role\": \"user\", \"content\": example[\"user\"]},\n",
    "        {\"role\": \"assistant\", \"content\": example[\"assistant\"]},\n",
    "    ]}\n",
    "\n",
    "ds1 = ds1.map(to_messages, remove_columns=ds1.column_names, batched=False)\n",
    "ds2 = ds2.map(to_messages, remove_columns=ds2.column_names, batched=False)\n",
    "train_dataset = concatenate_datasets([ds1, ds2])\n",
    "print(f\"✅ Messages dataset: {len(train_dataset)} rows\")\n",
    "\n",
    "# ========== PRE-PROCESS: messages → text with chat template ==========\n",
    "def convert_messages_to_text(examples):\n",
    "    texts = []\n",
    "    for msgs in examples[\"messages\"]:\n",
    "        text = tokenizer.apply_chat_template(\n",
    "            msgs,\n",
    "            tokenize=False,\n",
    "            add_generation_prompt=False,\n",
    "        )\n",
    "        texts.append(text)\n",
    "    return {\"text\": texts}\n",
    "\n",
    "print(\"🔄 Converting messages to text...\")\n",
    "train_dataset = train_dataset.map(\n",
    "    convert_messages_to_text,\n",
    "    batched=True,\n",
    "    remove_columns=[\"messages\"],\n",
    "    batch_size=100,\n",
    ")\n",
    "print(f\"✅ Dataset ready with columns: {train_dataset.column_names}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from trl import SFTTrainer\n",
    "from transformers import TrainingArguments\n",
    "\n",
    "trainer = SFTTrainer(\n",
    "    model=model,\n",
    "    tokenizer=tokenizer,\n",
    "    train_dataset=train_dataset,\n",
    "    dataset_text_field=\"text\",          # ← standard text format\n",
    "    max_seq_length=MAX_SEQ_LENGTH,\n",
    "    dataset_num_proc=2,\n",
    "    packing=False,\n",
    "    args=TrainingArguments(\n",
    "        per_device_train_batch_size=BATCH_SIZE,\n",
    "        gradient_accumulation_steps=GRAD_ACCUM,\n",
    "        warmup_steps=10,\n",
    "        num_train_epochs=NUM_EPOCHS,\n",
    "        learning_rate=LEARNING_RATE,\n",
    "        fp16=True,\n",
    "        logging_steps=5,\n",
    "        optim=\"adamw_8bit\",\n",
    "        weight_decay=0.01,\n",
    "        lr_scheduler_type=\"linear\",\n",
    "        seed=3407,\n",
    "        output_dir=\"./outputs\",\n",
    "        save_strategy=\"epoch\",\n",
    "        report_to=\"none\",\n",
    "    ),\n",
    ")\n",
    "\n",
    "trainer.train()\n",
    "model.save_pretrained(\"./cyber-lora-adapter\")\n",
    "tokenizer.save_pretrained(\"./cyber-lora-adapter\")\n",
    "print(\"✅ Training complete! Adapter saved to ./cyber-lora-adapter\")"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "name": "python",
   "version": "3.10.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}