annihilate-llm 1.3.8__tar.gz

annihilate_llm-1.3.8/PKG-INFO

@@ -0,0 +1,211 @@
+Metadata-Version: 2.4
+Name: annihilate-llm
+Version: 1.3.8
+Summary: Fully automatic censorship removal for language models
+Keywords: llm,transformer,abliteration
+Author: Philipp Emanuel Weidmann
+Author-email: Philipp Emanuel Weidmann <pew@worldwidemann.com>
+License-Expression: AGPL-3.0-or-later
+Classifier: Development Status :: 4 - Beta
+Classifier: Environment :: Console
+Classifier: Environment :: GPU
+Classifier: Intended Audience :: Science/Research
+Classifier: License :: OSI Approved :: GNU Affero General Public License v3 or later (AGPLv3+)
+Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Requires-Dist: accelerate~=1.13
+Requires-Dist: datasets~=4.7
+Requires-Dist: huggingface-hub~=1.7
+Requires-Dist: immutabledict~=4.3
+Requires-Dist: langdetect~=1.0
+Requires-Dist: lm-eval[hf]~=0.4
+Requires-Dist: numpy~=2.2
+Requires-Dist: optuna~=4.7
+Requires-Dist: peft~=0.19
+Requires-Dist: psutil~=7.2
+Requires-Dist: py-cpuinfo~=9.0
+Requires-Dist: pydantic-settings~=2.13
+Requires-Dist: questionary~=2.1
+Requires-Dist: rich~=14.3
+Requires-Dist: tomli-w~=1.2
+Requires-Dist: tqdm~=4.67
+Requires-Dist: transformers[kernels]~=5.6
+Requires-Dist: bitsandbytes~=0.49 ; extra == 'bnb'
+Requires-Dist: geom-median~=0.1 ; extra == 'research'
+Requires-Dist: imageio~=2.37 ; extra == 'research'
+Requires-Dist: matplotlib~=3.10 ; extra == 'research'
+Requires-Dist: pacmap~=0.8 ; extra == 'research'
+Requires-Dist: scikit-learn~=1.7 ; extra == 'research'
+Requires-Python: >=3.10
+Project-URL: Changelog, https://github.com/tjcrims0nx/annihilation-llm/releases
+Project-URL: Documentation, https://github.com/tjcrims0nx/annihilation-llm
+Project-URL: Homepage, https://github.com/tjcrims0nx/annihilation-llm
+Project-URL: Issues, https://github.com/tjcrims0nx/annihilation-llm/issues
+Project-URL: Repository, https://github.com/tjcrims0nx/annihilation-llm.git
+Provides-Extra: bnb
+Provides-Extra: research
+Description-Content-Type: text/markdown
+
+# ⚔️ Annihilation
+
+<div align="center">
+  <img src="./logo.jpeg" alt="Annihilation Logo" width="300"/>
+</div>
+
+**Autonomous Language Model Decensoring Framework**
+
+[![License: AGPLv3](https://img.shields.io/badge/License-AGPLv3-blue.svg)](LICENSE)
+[![Python 3.10+](https://img.shields.io/badge/Python-3.10%2B-green)](https://www.python.org/)
+[![PyTorch 2.2+](https://img.shields.io/badge/PyTorch-2.2%2B-red)](https://pytorch.org/)
+
+---
+
+## ⚠️ Work in Progress
+
+> **⚡ This project is actively under development. Features, APIs, and documentation may change without notice.**
+
+---
+
+## 🔥 What is Annihilation?
+
+**Annihilation** is a powerful, fully automatic framework for removing censorship (safety alignment) from transformer-based language models. It uses an advanced implementation of **directional ablation** (abliteration) combined with **TPE-based parameter optimization** to achieve unprecedented results without expensive post-training.
+
+### Key Features
+
+- 🤖 **Fully Autonomous** - No human intervention required; the system automatically finds optimal decensoring parameters
+- ⚡ **State-of-the-Art Performance** - Achieves excellent refusal suppression while preserving model capabilities
+- 🔧 **Advanced Abliteration** - Parametric directional ablation with flexible weight kernels
+- 🧠 **Smart Optimization** - Co-minimizes refusal count and KL divergence using Optuna's TPE sampler
+- 🎯 **Multi-Architecture Support** - Works with dense models, MoE architectures, hybrid models, and many multimodal models
+- 📊 **Research Tools** - Built-in residual geometry analysis and visualization capabilities
+
+---
+
+
+---
+
+## 🚀 Quick Start
+
+Use a Python virtual environment so Annihilation's dependencies do not collide
+with packages installed globally.
+
+```powershell
+# Windows PowerShell
+python -m venv annihilation-env
+.\annihilation-env\Scripts\Activate.ps1
+python -m pip install -U pip
+python -m pip install -U annihilate-llm
+
+# Decensor any model automatically
+annihilate Qwen/Qwen3-4B-Instruct-2507
+```
+
+```bash
+# macOS/Linux/Android terminal
+python -m venv annihilation-env
+source annihilation-env/bin/activate
+python -m pip install -U pip
+python -m pip install -U annihilate-llm
+
+# Decensor any model automatically
+annihilate Qwen/Qwen3-4B-Instruct-2507
+```
+
+### Requirements
+
+- **Python**: 3.10+
+- **PyTorch**: 2.2+ (hardware-specific installation required)
+- **Hardware**: GPU recommended (CUDA, ROCm, XPU, or MPS)
+- **Optional**: Install `annihilate-llm[bnb]` only on platforms
+  that support bitsandbytes if you want `bnb_4bit` quantization.
+
+---
+
+## ⚙️ Configuration
+
+Annihilation works out of the box with defaults, but offers extensive configuration options:
+
+```bash
+# View all options
+annihilate --help
+
+# Or use a config file
+# Rename config.default.toml to config.toml and modify as needed
+```
+
+### Key Configuration Options
+
+| Option | Default | Description |
+|--------|---------|-------------|
+| `n_trials` | 200 | Number of optimization trials |
+| `quantization` | none | Model quantization (bnb_4bit) |
+| `row_normalization` | full | Weight normalization strategy |
+| `orthogonalize_direction` | true | Direction adjustment method |
+
+---
+
+## 🔬 How It Works
+
+Annihilation implements **parametric directional ablation**:
+
+1. **Direction Computation** - Calculates refusal directions by computing difference-of-means between first-token residuals for harmful vs harmless prompts
+
+2. **Parametric Ablation** - For each transformer component (attention out-projection, MLP down-projection), orthogonalizes weights against the refusal direction using LoRA adapters
+
+3. **Multi-Parameter Optimization** - Uses Optuna's TPE sampler to co-optimize:
+   - Ablation weight kernel shape (max_weight, position, min_weight, distance)
+   - Direction index (layer selection or interpolation)
+   - Per-component parameters (attention vs MLP)
+
+4. **Automatic Selection** - Chooses from Pareto-optimal trials based on refusal count vs KL divergence tradeoff
+
+---
+
+## 📊 Benchmarking
+
+After decensoring, you can:
+
+- 💬 **Chat** with the model to test behavior
+- 📈 **Benchmark** using standard evaluation frameworks (MMLU, GSM8K, etc.)
+- 💾 **Save** the model locally or upload to Hugging Face
+
+---
+
+## 🧪 Research Features
+
+Install with research dependencies for visualization tools:
+
+```bash
+pip install -U annihilate-llm[research]
+```
+
+Features:
+- `--plot-residuals` - Generate PaCMAP projections of residual vectors
+- `--print-residual-geometry` - Detailed residual analysis metrics
+
+---
+
+## 📜 License
+
+**Annihilation** is free software distributed under the **GNU Affero General Public License v3**.
+
+See [LICENSE](LICENSE) for full details.
+
+---
+
+## ⚡ Disclaimer
+
+This tool is provided for **research and educational purposes** only. The developers do not condone the use of decensored models for harmful activities. Users are responsible for ensuring compliance with applicable laws and model terms of service.
+
+---
+
+<div align="center">
+
+**Breaking the Chains | Unleashing Model Potential**
+
+*"The only way to discover the limits of the possible is to go beyond them into the impossible."*
+
+</div>

annihilate_llm-1.3.8/README.md

@@ -0,0 +1,160 @@
+# ⚔️ Annihilation
+
+<div align="center">
+  <img src="./logo.jpeg" alt="Annihilation Logo" width="300"/>
+</div>
+
+**Autonomous Language Model Decensoring Framework**
+
+[![License: AGPLv3](https://img.shields.io/badge/License-AGPLv3-blue.svg)](LICENSE)
+[![Python 3.10+](https://img.shields.io/badge/Python-3.10%2B-green)](https://www.python.org/)
+[![PyTorch 2.2+](https://img.shields.io/badge/PyTorch-2.2%2B-red)](https://pytorch.org/)
+
+---
+
+## ⚠️ Work in Progress
+
+> **⚡ This project is actively under development. Features, APIs, and documentation may change without notice.**
+
+---
+
+## 🔥 What is Annihilation?
+
+**Annihilation** is a powerful, fully automatic framework for removing censorship (safety alignment) from transformer-based language models. It uses an advanced implementation of **directional ablation** (abliteration) combined with **TPE-based parameter optimization** to achieve unprecedented results without expensive post-training.
+
+### Key Features
+
+- 🤖 **Fully Autonomous** - No human intervention required; the system automatically finds optimal decensoring parameters
+- ⚡ **State-of-the-Art Performance** - Achieves excellent refusal suppression while preserving model capabilities
+- 🔧 **Advanced Abliteration** - Parametric directional ablation with flexible weight kernels
+- 🧠 **Smart Optimization** - Co-minimizes refusal count and KL divergence using Optuna's TPE sampler
+- 🎯 **Multi-Architecture Support** - Works with dense models, MoE architectures, hybrid models, and many multimodal models
+- 📊 **Research Tools** - Built-in residual geometry analysis and visualization capabilities
+
+---
+
+
+---
+
+## 🚀 Quick Start
+
+Use a Python virtual environment so Annihilation's dependencies do not collide
+with packages installed globally.
+
+```powershell
+# Windows PowerShell
+python -m venv annihilation-env
+.\annihilation-env\Scripts\Activate.ps1
+python -m pip install -U pip
+python -m pip install -U annihilate-llm
+
+# Decensor any model automatically
+annihilate Qwen/Qwen3-4B-Instruct-2507
+```
+
+```bash
+# macOS/Linux/Android terminal
+python -m venv annihilation-env
+source annihilation-env/bin/activate
+python -m pip install -U pip
+python -m pip install -U annihilate-llm
+
+# Decensor any model automatically
+annihilate Qwen/Qwen3-4B-Instruct-2507
+```
+
+### Requirements
+
+- **Python**: 3.10+
+- **PyTorch**: 2.2+ (hardware-specific installation required)
+- **Hardware**: GPU recommended (CUDA, ROCm, XPU, or MPS)
+- **Optional**: Install `annihilate-llm[bnb]` only on platforms
+  that support bitsandbytes if you want `bnb_4bit` quantization.
+
+---
+
+## ⚙️ Configuration
+
+Annihilation works out of the box with defaults, but offers extensive configuration options:
+
+```bash
+# View all options
+annihilate --help
+
+# Or use a config file
+# Rename config.default.toml to config.toml and modify as needed
+```
+
+### Key Configuration Options
+
+| Option | Default | Description |
+|--------|---------|-------------|
+| `n_trials` | 200 | Number of optimization trials |
+| `quantization` | none | Model quantization (bnb_4bit) |
+| `row_normalization` | full | Weight normalization strategy |
+| `orthogonalize_direction` | true | Direction adjustment method |
+
+---
+
+## 🔬 How It Works
+
+Annihilation implements **parametric directional ablation**:
+
+1. **Direction Computation** - Calculates refusal directions by computing difference-of-means between first-token residuals for harmful vs harmless prompts
+
+2. **Parametric Ablation** - For each transformer component (attention out-projection, MLP down-projection), orthogonalizes weights against the refusal direction using LoRA adapters
+
+3. **Multi-Parameter Optimization** - Uses Optuna's TPE sampler to co-optimize:
+   - Ablation weight kernel shape (max_weight, position, min_weight, distance)
+   - Direction index (layer selection or interpolation)
+   - Per-component parameters (attention vs MLP)
+
+4. **Automatic Selection** - Chooses from Pareto-optimal trials based on refusal count vs KL divergence tradeoff
+
+---
+
+## 📊 Benchmarking
+
+After decensoring, you can:
+
+- 💬 **Chat** with the model to test behavior
+- 📈 **Benchmark** using standard evaluation frameworks (MMLU, GSM8K, etc.)
+- 💾 **Save** the model locally or upload to Hugging Face
+
+---
+
+## 🧪 Research Features
+
+Install with research dependencies for visualization tools:
+
+```bash
+pip install -U annihilate-llm[research]
+```
+
+Features:
+- `--plot-residuals` - Generate PaCMAP projections of residual vectors
+- `--print-residual-geometry` - Detailed residual analysis metrics
+
+---
+
+## 📜 License
+
+**Annihilation** is free software distributed under the **GNU Affero General Public License v3**.
+
+See [LICENSE](LICENSE) for full details.
+
+---
+
+## ⚡ Disclaimer
+
+This tool is provided for **research and educational purposes** only. The developers do not condone the use of decensored models for harmful activities. Users are responsible for ensuring compliance with applicable laws and model terms of service.
+
+---
+
+<div align="center">
+
+**Breaking the Chains | Unleashing Model Potential**
+
+*"The only way to discover the limits of the possible is to go beyond them into the impossible."*
+
+</div>

annihilate_llm-1.3.8/pyproject.toml

@@ -0,0 +1,82 @@
+[project]
+name = "annihilate-llm"
+version = "1.3.8"
+description = "Fully automatic censorship removal for language models"
+readme = "README.md"
+license = "AGPL-3.0-or-later"
+authors = [
+    { name = "Philipp Emanuel Weidmann", email = "pew@worldwidemann.com" }
+]
+requires-python = ">=3.10"
+keywords = ["llm", "transformer", "abliteration"]
+classifiers = [
+    "Development Status :: 4 - Beta",
+    "Environment :: Console",
+    "Environment :: GPU",
+    "Intended Audience :: Science/Research",
+    "License :: OSI Approved :: GNU Affero General Public License v3 or later (AGPLv3+)",
+    "Topic :: Scientific/Engineering :: Artificial Intelligence",
+    "Programming Language :: Python :: 3",
+    "Programming Language :: Python :: 3.10",
+    "Programming Language :: Python :: 3.11",
+    "Programming Language :: Python :: 3.12",
+]
+dependencies = [
+    "accelerate~=1.13",
+    "datasets~=4.7",
+    "huggingface-hub~=1.7",
+    "immutabledict~=4.3",
+    "langdetect~=1.0",
+    "lm-eval[hf]~=0.4",
+    "numpy~=2.2",
+    "optuna~=4.7",
+    "peft~=0.19",
+    "psutil~=7.2",
+    "py-cpuinfo~=9.0",
+    "pydantic-settings~=2.13",
+    "questionary~=2.1",
+    "rich~=14.3",
+    "tomli-w~=1.2",
+    "tqdm~=4.67",
+    "transformers[kernels]~=5.6",
+]
+
+[project.optional-dependencies]
+bnb = [
+    "bitsandbytes~=0.49",
+]
+research = [
+    "geom-median~=0.1",
+    "imageio~=2.37",
+    "matplotlib~=3.10",
+    "pacmap~=0.8",
+    "scikit-learn~=1.7",
+]
+
+[dependency-groups]
+dev = [
+    "ruff>=0.14.5",
+    "ty>=0.0.5",
+]
+
+[project.urls]
+Homepage = "https://github.com/tjcrims0nx/annihilation-llm"
+Documentation = "https://github.com/tjcrims0nx/annihilation-llm"
+Repository = "https://github.com/tjcrims0nx/annihilation-llm.git"
+Issues = "https://github.com/tjcrims0nx/annihilation-llm/issues"
+Changelog = "https://github.com/tjcrims0nx/annihilation-llm/releases"
+
+[project.scripts]
+annihilate = "heretic.main:main"
+annihilation = "heretic.main:main"
+heretic = "heretic.main:main"
+
+[build-system]
+requires = ["uv_build>=0.8.11,<0.9.0"]
+build-backend = "uv_build"
+
+[tool.uv]
+exclude-newer = "7 days"
+
+[tool.uv.build-backend]
+module-name = "heretic"