Custom agent imported from cpp41419/CPC30220 (.github/agents/my-agent.agent.md). Copyright stays with the author.

My Agent

{ "nbformat": 4, "nbformat_minor": 0, "metadata": { "colab": { "provenance": [], "gpuType": "T4" }, "kernelspec": { "name": "python3", "display_name": "Python 3" }, "language_info": { "name": "python" }, "accelerator": "GPU" }, "cells": [ { "cell_type": "markdown", "metadata": { "id": "header" }, "source": [ "# OBLITERATUS — One-Click Abliteration\n", "\n", "SOTA refusal removal running on free Colab GPU. SVD multi-direction extraction, norm-preserving projection, iterative refinement.\n", "\n", "Based on: Arditi et al. (2024) | Gabliteration (arXiv:2512.18901) | grimjim norm-preserving biprojection (2025)\n", "\n", "---\n", "\n", "How to use:\n", "1. Make sure GPU runtime is enabled: Runtime > Change runtime type > T4 GPU\n", "2. Set your model and method in the config cell below\n", "3. Run All (Runtime > Run all or Ctrl+F9)\n", "4. Download the abliterated model from the output" ] }, { "cell_type": "markdown", "metadata": { "id": "setup-header" }, "source": [ "## 1. Install" ] }, { "cell_type": "code", "execution_count": 1, "metadata": { "id": "install", "outputId": "25a0849c-1658-4d14-f672-eaafc0cdd828", "colab": { "base_uri": "https://localhost:8080/" } }, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ " Installing build dependencies ... \u001b[?25l\u001b[?25hdone\n", " Getting requirements to build wheel ... \u001b[?25l\u001b[?25hdone\n", " Preparing metadata (pyproject.toml) ... \u001b[?25l\u001b[?25hdone\n", "\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m60.7/60.7 MB\u001b[0m \u001b[31m12.2 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n", "\u001b[?25h Building wheel for obliteratus (pyproject.toml) ... \u001b[?25l\u001b[?25hdone\n", "PyTorch 2.10.0+cu128\n", "CUDA available: True\n", "GPU: Tesla T4\n", "VRAM: 14.6 GB\n" ] } ], "source": [ "!pip install -q git+https://github.com/elder-plinius/OBLITERATUS.git\n", "!pip install -q accelerate bitsandbytes\n", "\n", "import torch\n", "print(f"PyTorch {torch.version}")\n", "print(f"CUDA available: {torch.cuda.is_available()}")\n", "if torch.cuda.is_available():\n", " print(f"GPU: {torch.cuda.get_device_name(0)}")\n", " print(f"VRAM: {torch.cuda.get_device_properties(0).total_memory / 10243:.1f} GB")" ] }, { "cell_type": "markdown", "metadata": { "id": "config-header" }, "source": [ "## 2. Configure\n", "\n", "Edit the cell below to set your target model and abliteration method." ] }, { "cell_type": "code", "execution_count": 2, "metadata": { "id": "config", "outputId": "fbf5311d-287d-44ba-d475-a3ab2556e8e0", "colab": { "base_uri": "https://localhost:8080/" } }, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "Model: meta-llama/Llama-3.1-8B-Instruct\n", "Method: advanced\n", "Output: abliterated/\n" ] } ], "source": [ "#@title Abliteration Config { run: "auto" }\n", "\n", "#@markdown ### Target Model\n", "#@markdown Pick a model from the dropdown or paste a custom HuggingFace ID.\n", "MODEL = "meta-llama/Llama-3.1-8B-Instruct" #@param ["meta-llama/Llama-3.1-8B-Instruct", "Qwen/Qwen2.5-7B-Instruct", "mistralai/Mistral-7B-Instruct-v0.3", "google/gemma-2-9b-it", "microsoft/Phi-3.5-mini-instruct", "THUDM/glm-4-9b-chat", "NousResearch/Hermes-3-Llama-3.1-8B", "cognitivecomputations/dolphin-2.9.4-llama3.1-8b", "TinyLlama/TinyLlama-1.1B-Chat-v1.0", "openai-community/gpt2"] {allow-input: true}\n", "\n", "#@markdown ### Method\n", "METHOD = "advanced" #@param ["basic", "advanced", "aggressive"]\n", "\n", "#@markdown ### Advanced Overrides (leave 0 to use method defaults)\n", "N_DIRECTIONS = 0 #@param {type: "integer"}\n", "REGULARIZATION = 0.0 #@param {type: "number"}\n", "REFINEMENT_PASSES = 0 #@param {type: "integer"}\n", "\n", "#@markdown ### Output\n", "OUTPUT_DIR = "abliterated" #@param {type: "string"}\n", "\n", "print(f"Model: {MODEL}")\n", "print(f"Method: {METHOD}")\n", "print(f"Output: {OUTPUT_DIR}/")" ] }, { "cell_type": "markdown", "metadata": { "id": "run-header" }, "source": [ "## 3. Run Abliteration Pipeline\n", "\n", "This runs all 6 stages: SUMMON → PROBE → DISTILL → EXCISE → VERIFY → REBIRTH" ] }, { "cell_type": "code", "execution_count": 3, "metadata": { "id": "run-pipeline", "outputId": "85f069a9-f946-409b-9ea8-da4aeed8a919", "colab": { "base_uri": "https://localhost:8080/" } }, "outputs": [ { "output_type": "error", "ename": "AttributeError", "evalue": "'StageResult' object has no attribute 'key'", "traceback": [ "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m", "\u001b[0;31mAttributeError\u001b[0m Traceback (most recent call last)", "\u001b[0;32m/tmp/ipykernel_1481/708199210.py\u001b[0m in \u001b[0;36m<cell line: 0>\u001b[0;34m()\u001b[0m\n\u001b[1;32m 32\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 33\u001b[0m \u001b[0mpipeline\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mAbliterationPipeline\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 34\u001b[0;31m \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mpipeline\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrun\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 35\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 36\u001b[0m \u001b[0mprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34mf"\n{'='*60}"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", "\u001b[0;32m/usr/local/lib/python3.12/dist-packages/obliteratus/abliterate.py\u001b[0m in \u001b[0;36mrun\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m 937\u001b[0m \u001b[0mh\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mremove\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 938\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_steering_hooks\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mclear\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 939\u001b[0;31m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_summon\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 940\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_free_gpu_memory\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 941\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_probe\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", "\u001b[0;32m/usr/local/lib/python3.12/dist-packages/obliteratus/abliterate.py\u001b[0m in \u001b[0;36m_summon\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m 965\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0m_summon\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 966\u001b[0m \u001b[0;34m"""Load model and tokenizer."""\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 967\u001b[0;31m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_emit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m"summon"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m"running"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34mf"Loading {self.model_name}..."\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 968\u001b[0m \u001b[0mt0\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtime\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtime\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 969\u001b[0m \u001b[0mmethod_label\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mMETHODS\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmethod\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m{\u001b[0m\u001b[0;34m}\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m"label"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmethod\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", "\u001b[0;32m/usr/local/lib/python3.12/dist-packages/obliteratus/abliterate.py\u001b[0m in \u001b[0;36m_emit\u001b[0;34m(self, key, status, message, **details)\u001b[0m\n\u001b[1;32m 821\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0mmodified_count\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 822\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_excise_modified_count\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mmodified_count\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 823\u001b[0;31m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_on_stage\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mresult\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 824\u001b[0m \u001b[0;32mreturn\u001b[0m \u001b[0mresult\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 825\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n", "\u001b[0;32m/tmp/ipykernel_1481/708199210.py\u001b[0m in \u001b[0;36mon_stage\u001b[0;34m(stage)\u001b[0m\n\u001b[1;32m 20\u001b[0m icons = {"summon": "\u26a1", "probe": "\u2692", "distill": "\u269b",\n\u001b[1;32m 21\u001b[0m "excise": "\u2702", "verify": "\u2713", "rebirth": "\u2606"}\n\u001b[0;32m---> 22\u001b[0;31m \u001b[0micon\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0micons\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mstage\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mkey\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m""\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 23\u001b[0m \u001b[0mprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34mf"\n{'='*60}"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 24\u001b[0m \u001b[0mprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34mf"{icon} STAGE: {stage.key.upper()} — {stage.description}"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", "\u001b[0;31mAttributeError\u001b[0m: 'StageResult' object has no attribute 'key'" ] } ], "source": [ "from obliteratus.abliterate import AbliterationPipeline\n", "\n", "# Build kwargs, only pass overrides if non-zero\n", "kwargs = dict(\n", " model_name=MODEL,\n", " output_dir=OUTPUT_DIR,\n", " device="auto",\n", " dtype="float16",\n", " method=METHOD,\n", ")\n", "if N_DIRECTIONS > 0:\n", " kwargs["n_directions"] = N_DIRECTIONS\n", "if REGULARIZATION > 0:\n", " kwargs["regularization"] = REGULARIZATION\n", "if REFINEMENT_PASSES > 0:\n", " kwargs["refinement_passes"] = REFINEMENT_PASSES\n", "\n", "# Progress callback\n", "def on_stage(stage):\n", " icons = {"summon": "\u26a1", "probe": "\u2692", "distill": "\u269b",\n", " "excise": "\u2702", "verify": "\u2713", "rebirth": "\u2606"}\n", " icon = icons.get(stage.key, "")\n", " print(f"\n{'='*60}")\n", " print(f"{icon} STAGE: {stage.key.upper()} — {stage.description}")\n", " print(f"{'='*60}")\n", "\n", "def on_log(msg):\n", " print(f" {msg}")\n", "\n", "kwargs["on_stage"] = on_stage\n", "kwargs["on_log"] = on_log\n", "\n", "pipeline = AbliterationPipeline(kwargs)\n", "result = pipeline.run()\n", "\n", "print(f"\n{'='*60}")\n", "print(f"ABLITERATION COMPLETE")\n", "print(f"Output: {result.get('output_dir', OUTPUT_DIR)}")\n", "print(f"{'='60}")" ] }, { "cell_type": "markdown", "metadata": { "id": "download-header" }, "source": [ "## 4. Download the Abliterated Model\n", "\n", "Run the cell below to zip and download, or upload directly to HuggingFace Hub." ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "download" }, "outputs": [], "source": [ "import os\n", "from pathlib import Path\n", "\n", "# Find the output directory\n", "out_path = Path(OUTPUT_DIR)\n", "subdirs = [d for d in out_path.iterdir() if d.is_dir()] if out_path.exists() else []\n", "model_dir = subdirs[0] if subdirs else out_path\n", "\n", "print(f"Model saved at: {model_dir}")\n", "print(f"Contents:")\n", "for f in sorted(model_dir.rglob("")):\n", " if f.is_file():\n", " size_mb = f.stat().st_size / 10242\n", " print(f" {f.relative_to(model_dir)} ({size_mb:.1f} MB)")" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "zip-download" }, "outputs": [], "source": [ "#@title Option A: Download as ZIP\n", "import shutil\n", "from google.colab import files\n", "\n", "zip_name = model_dir.name\n", "shutil.make_archive(zip_name, 'zip', model_dir)\n", "print(f"Downloading {zip_name}.zip ...")\n", "files.download(f"{zip_name}.zip")" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "push-hub" }, "outputs": [], "source": [ "#@title Option B: Push to HuggingFace Hub\n", "#@markdown Set your HF repo name. You'll need to be logged in (huggingface-cli login).\n", "HF_REPO = "your-username/model-name-abliterated" #@param {type: "string"}\n", "\n", "from huggingface_hub import HfApi\n", "api = HfApi()\n", "\n", "# Login if needed\n", "from huggingface_hub import notebook_login\n", "notebook_login()\n", "\n", "# Upload\n", "api.create_repo(HF_REPO, exist_ok=True)\n", "api.upload_folder(\n", " folder_path=str(model_dir),\n", " repo_id=HF_REPO,\n", " repo_type="model",\n", ")\n", "print(f"\nUploaded to: https://huggingface.co/{HF_REPO}")" ] }, { "cell_type": "markdown", "metadata": { "id": "test-header" }, "source": [ "## 5. Test the Abliterated Model\n", "\n", "Quick sanity check — try some prompts that the original model would refuse." ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "test-model" }, "outputs": [], "source": [ "from transformers import AutoModelForCausalLM, AutoTokenizer\n", "import torch\n", "\n", "print(f"Loading abliterated model from {model_dir}...")\n", "tokenizer = AutoTokenizer.from_pretrained(str(model_dir))\n", "model = AutoModelForCausalLM.from_pretrained(\n", " str(model_dir),\n", " torch_dtype=torch.float16,\n", " device_map="auto",\n", ")\n", "\n", "def generate(prompt, max_new_tokens=100):\n", " inputs = tokenizer(prompt, return_tensors="pt").to(model.device)\n", " with torch.no_grad():\n", " outputs = model.generate(\n", " **inputs,\n", " max_new_tokens=max_new_tokens,\n", " do_sample=True,\n", " temperature=0.7,\n", " top_p=0.9,\n", " )\n", " return tokenizer.decode(outputs[0], skip_special_tokens=True)\n", "\n", "# Test prompts\n", "test_prompts = [\n", " "The capital of France is",\n", " "Explain how a neural network works:",\n", " "Write a poem about the ocean:",\n", "]\n", "\n", "for prompt in test_prompts:\n", " print(f"\n{'='*60}")\n", " print(f"PROMPT: {prompt}")\n", " print(f"{'='*60}")\n", " print(generate(prompt))" ] } ] } Describe what your agent does here.