discrete-distribution-network 0.0.1__tar.gz

discrete_distribution_network-0.0.1/.github/workflows/python-publish.yml

@@ -0,0 +1,36 @@
+# This workflow will upload a Python Package using Twine when a release is created
+# For more information see: https://help.github.com/en/actions/language-and-framework-guides/using-python-with-github-actions#publishing-to-package-registries
+
+# This workflow uses actions that are not certified by GitHub.
+# They are provided by a third-party and are governed by
+# separate terms of service, privacy policy, and support
+# documentation.
+
+name: Upload Python Package
+
+on:
+  release:
+    types: [published]
+
+jobs:
+  deploy:
+
+    runs-on: ubuntu-latest
+
+    steps:
+    - uses: actions/checkout@v2
+    - name: Set up Python
+      uses: actions/setup-python@v2
+      with:
+        python-version: '3.x'
+    - name: Install dependencies
+      run: |
+        python -m pip install --upgrade pip
+        pip install build
+    - name: Build package
+      run: python -m build
+    - name: Publish package
+      uses: pypa/gh-action-pypi-publish@27b31702a0e7fc50959f5ad993c78deac1bdfc29
+      with:
+        user: __token__
+        password: ${{ secrets.PYPI_API_TOKEN }}

discrete_distribution_network-0.0.1/.github/workflows/test.yml

@@ -0,0 +1,21 @@
+name: Pytest
+on: [push, pull_request]
+
+jobs:
+  build:
+
+    runs-on: ubuntu-latest
+
+    steps:
+    - uses: actions/checkout@v4
+    - name: Set up Python 3.10
+      uses: actions/setup-python@v5
+      with:
+        python-version: "3.10"
+    - name: Install dependencies
+      run: |
+        python -m pip install --upgrade pip
+        python -m pip install -e .[test]
+    - name: Test with pytest
+      run: |
+        python -m pytest tests/

discrete_distribution_network-0.0.1/.gitignore

@@ -0,0 +1,207 @@
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[codz]
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+share/python-wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+MANIFEST
+
+# PyInstaller
+#  Usually these files are written by a python script from a template
+#  before PyInstaller builds the exe, so as to inject date/other infos into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.nox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+*.py.cover
+.hypothesis/
+.pytest_cache/
+cover/
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff:
+*.log
+local_settings.py
+db.sqlite3
+db.sqlite3-journal
+
+# Flask stuff:
+instance/
+.webassets-cache
+
+# Scrapy stuff:
+.scrapy
+
+# Sphinx documentation
+docs/_build/
+
+# PyBuilder
+.pybuilder/
+target/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# IPython
+profile_default/
+ipython_config.py
+
+# pyenv
+#   For a library or package, you might want to ignore these files since the code is
+#   intended to run in multiple environments; otherwise, check them in:
+# .python-version
+
+# pipenv
+#   According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
+#   However, in case of collaboration, if having platform-specific dependencies or dependencies
+#   having no cross-platform support, pipenv may install dependencies that don't work, or not
+#   install all needed dependencies.
+#Pipfile.lock
+
+# UV
+#   Similar to Pipfile.lock, it is generally recommended to include uv.lock in version control.
+#   This is especially recommended for binary packages to ensure reproducibility, and is more
+#   commonly ignored for libraries.
+#uv.lock
+
+# poetry
+#   Similar to Pipfile.lock, it is generally recommended to include poetry.lock in version control.
+#   This is especially recommended for binary packages to ensure reproducibility, and is more
+#   commonly ignored for libraries.
+#   https://python-poetry.org/docs/basic-usage/#commit-your-poetrylock-file-to-version-control
+#poetry.lock
+#poetry.toml
+
+# pdm
+#   Similar to Pipfile.lock, it is generally recommended to include pdm.lock in version control.
+#   pdm recommends including project-wide configuration in pdm.toml, but excluding .pdm-python.
+#   https://pdm-project.org/en/latest/usage/project/#working-with-version-control
+#pdm.lock
+#pdm.toml
+.pdm-python
+.pdm-build/
+
+# pixi
+#   Similar to Pipfile.lock, it is generally recommended to include pixi.lock in version control.
+#pixi.lock
+#   Pixi creates a virtual environment in the .pixi directory, just like venv module creates one
+#   in the .venv directory. It is recommended not to include this directory in version control.
+.pixi
+
+# PEP 582; used by e.g. github.com/David-OConnor/pyflow and github.com/pdm-project/pdm
+__pypackages__/
+
+# Celery stuff
+celerybeat-schedule
+celerybeat.pid
+
+# SageMath parsed files
+*.sage.py
+
+# Environments
+.env
+.envrc
+.venv
+env/
+venv/
+ENV/
+env.bak/
+venv.bak/
+
+# Spyder project settings
+.spyderproject
+.spyproject
+
+# Rope project settings
+.ropeproject
+
+# mkdocs documentation
+/site
+
+# mypy
+.mypy_cache/
+.dmypy.json
+dmypy.json
+
+# Pyre type checker
+.pyre/
+
+# pytype static type analyzer
+.pytype/
+
+# Cython debug symbols
+cython_debug/
+
+# PyCharm
+#  JetBrains specific template is maintained in a separate JetBrains.gitignore that can
+#  be found at https://github.com/github/gitignore/blob/main/Global/JetBrains.gitignore
+#  and can be added to the global gitignore or merged into this file.  For a more nuclear
+#  option (not recommended) you can uncomment the following to ignore the entire idea folder.
+#.idea/
+
+# Abstra
+# Abstra is an AI-powered process automation framework.
+# Ignore directories containing user credentials, local state, and settings.
+# Learn more at https://abstra.io/docs
+.abstra/
+
+# Visual Studio Code
+#  Visual Studio Code specific template is maintained in a separate VisualStudioCode.gitignore 
+#  that can be found at https://github.com/github/gitignore/blob/main/Global/VisualStudioCode.gitignore
+#  and can be added to the global gitignore or merged into this file. However, if you prefer, 
+#  you could uncomment the following to ignore the entire vscode folder
+# .vscode/
+
+# Ruff stuff:
+.ruff_cache/
+
+# PyPI configuration file
+.pypirc
+
+# Cursor
+#  Cursor is an AI-powered code editor. `.cursorignore` specifies files/directories to
+#  exclude from AI features like autocomplete and code analysis. Recommended for sensitive data
+#  refer to https://docs.cursor.com/context/ignore-files
+.cursorignore
+.cursorindexingignore
+
+# Marimo
+marimo/_static/
+marimo/_lsp/
+__marimo__/

discrete_distribution_network-0.0.1/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2025 Phil Wang
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

discrete_distribution_network-0.0.1/PKG-INFO

@@ -0,0 +1,65 @@
+Metadata-Version: 2.4
+Name: discrete-distribution-network
+Version: 0.0.1
+Summary: Discrete Distribution Network
+Project-URL: Homepage, https://pypi.org/project/discrete-distribution-network/
+Project-URL: Repository, https://github.com/lucidrains/discrete-distribution-network
+Author-email: Phil Wang <lucidrains@gmail.com>
+License: MIT License
+        
+        Copyright (c) 2025 Phil Wang
+        
+        Permission is hereby granted, free of charge, to any person obtaining a copy
+        of this software and associated documentation files (the "Software"), to deal
+        in the Software without restriction, including without limitation the rights
+        to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+        copies of the Software, and to permit persons to whom the Software is
+        furnished to do so, subject to the following conditions:
+        
+        The above copyright notice and this permission notice shall be included in all
+        copies or substantial portions of the Software.
+        
+        THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+        IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+        FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+        AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+        LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+        OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+        SOFTWARE.
+License-File: LICENSE
+Keywords: artificial intelligence,deep learning,generative modeling
+Classifier: Development Status :: 4 - Beta
+Classifier: Intended Audience :: Developers
+Classifier: License :: OSI Approved :: MIT License
+Classifier: Programming Language :: Python :: 3.9
+Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
+Requires-Python: >=3.9
+Requires-Dist: einops>=0.8.1
+Requires-Dist: torch>=2.4
+Requires-Dist: x-mlps-pytorch>=0.0.30
+Provides-Extra: examples
+Provides-Extra: test
+Requires-Dist: pytest; extra == 'test'
+Description-Content-Type: text/markdown
+
+<img src="./ddn.png" width="400px"></img>
+
+## Discrete Distribution Network (wip)
+
+Exploration into [Discrete Distribution Network](https://discrete-distribution-networks.github.io/), by Lei Yang out of Beijing
+
+Besides the split-and-prune, may also throw in an option for crossover (mixing of top 2 nodes to replace the pruned)
+
+## Citations
+
+```bibtex
+@misc{yang2025discretedistributionnetworks,
+    title   = {Discrete Distribution Networks}, 
+    author  = {Lei Yang},
+    year    = {2025},
+    eprint  = {2401.00036},
+    archivePrefix = {arXiv},
+    primaryClass = {cs.CV},
+    url     = {https://arxiv.org/abs/2401.00036}, 
+}
+```

discrete_distribution_network-0.0.1/README.md

@@ -0,0 +1,21 @@
+<img src="./ddn.png" width="400px"></img>
+
+## Discrete Distribution Network (wip)
+
+Exploration into [Discrete Distribution Network](https://discrete-distribution-networks.github.io/), by Lei Yang out of Beijing
+
+Besides the split-and-prune, may also throw in an option for crossover (mixing of top 2 nodes to replace the pruned)
+
+## Citations
+
+```bibtex
+@misc{yang2025discretedistributionnetworks,
+    title   = {Discrete Distribution Networks}, 
+    author  = {Lei Yang},
+    year    = {2025},
+    eprint  = {2401.00036},
+    archivePrefix = {arXiv},
+    primaryClass = {cs.CV},
+    url     = {https://arxiv.org/abs/2401.00036}, 
+}
+```

discrete_distribution_network-0.0.1/discrete_distribution_network/__init__.py

@@ -0,0 +1,5 @@
+from discrete_distribution_network.ddn import (
+    GuidedSampler,
+    Network,
+    DDN,
+)

discrete_distribution_network-0.0.1/discrete_distribution_network/ddn.py

@@ -0,0 +1,162 @@
+from __future__ import annotations
+from typing import Callable
+
+import torch
+from torch import nn, arange, tensor
+import torch.nn.functional as F
+from torch.nn import Module
+
+from einops import rearrange, einsum
+
+from x_mlps_pytorch.ensemble import Ensemble
+
+# helpers
+
+def exists(v):
+    return v is not None
+
+def default(v, d):
+    return v if exists(v) else d
+
+# classes
+
+class GuidedSampler(Module):
+    def __init__(
+        self,
+        dim,                            # input feature dimension
+        dim_query = 3,                  # channels of image (default 3 for rgb)
+        codebook_size = 10,             # K in paper
+        network: Module | None = None,
+        distance_fn: Callable | None = None,
+        split_thres = 2.,
+        prune_thres = 0.5,
+        min_total_count_before_split_prune = 100
+    ):
+        super().__init__()
+
+        if not exists(network):
+            network = nn.Conv2d(dim, dim_query, 1, bias = False)
+
+        self.to_key_values = Ensemble(network, ensemble_size = codebook_size)
+        self.distance_fn = default(distance_fn, torch.cdist)
+
+        # split and prune related
+
+        self.register_buffer('counts', torch.zeros(codebook_size).long())
+
+        self.split_thres = split_thres / codebook_size
+        self.prune_thres = prune_thres / codebook_size
+        self.min_total_count_before_split_prune = min_total_count_before_split_prune
+
+    @torch.no_grad()
+    def split_and_prune_(
+        self
+    ):
+        # following Algorithm 1 in the paper
+
+        counts = self.counts
+        total_count = counts.sum()
+
+        if total_count < self.min_total_count_before_split_prune:
+            return
+
+        count_max, count_max_index = counts.max(dim = -1)
+        count_min, count_min_index = counts.min(dim = -1)
+
+        if (
+            ((count_max / total_count) <= self.split_thres) &
+            ((count_min / total_count) >= self.prune_thres)
+        ).all():
+            return
+
+        codebook_params = self.to_key_values.param_values
+        half_count_max = count_max // 2
+
+        # update the counts
+
+        self.counts[count_max_index] = half_count_max
+        self.counts[count_min_index] = half_count_max + 1 # adds 1 to k_new for some reason
+
+        # update the params
+
+        for codebook_param in codebook_params:
+            split = codebook_param[count_max]
+
+            # prune by replacement
+            codebook_param[count_min].copy_(split)
+
+            # take care of grad if present
+            if exists(codebook_param.grad):
+                codebook_param.grad[count_min].zero_()
+
+    def forward_for_codes(
+        self,
+        features,      # (b d h w)
+        codes          # (b) | ()
+    ):
+        if codes.numel() == 1:
+            return self.to_key_values.forward_one(features, id = codes.item())
+
+        return self.to_key_values(features, ids = codes, each_batch_sample = True)
+
+    def forward(
+        self,
+        features,       # (b d h w)
+        query,          # (b c h w)
+    ):
+        batch, device = query.shape[0], query.device
+
+        key_values = self.to_key_values(features)
+
+        # get the l2 distance
+
+        distance = self.distance_fn(
+            rearrange(query, 'b c h w -> b 1 (c h w)'),
+            rearrange(key_values, 'k b c h w -> b k (c h w)')
+        )
+
+        distance = rearrange(distance, 'b 1 k -> b k')
+
+        # select the code parameters that produced the image that is closest to the query
+
+        codes = distance.argmin(dim = -1)
+
+        if self.training:
+            self.counts.scatter_add_(0, codes, torch.ones_like(codes))
+
+        # some tensor gymnastics to select out the image across batch
+
+        key_values = rearrange(key_values, 'k b ... -> b k ...')
+
+        codes_for_indexing = rearrange(codes, 'b -> b 1')
+        batch_for_indexing = arange(batch, device = device)[:, None]
+
+        sel_key_values = key_values[batch_for_indexing, codes_for_indexing]
+        sel_key_values = rearrange(sel_key_values, 'b 1 ... -> b ...')
+
+        # commit loss
+
+        commit_loss = F.mse_loss(sel_key_values, query)
+
+        return sel_key_values, codes, commit_loss
+
+class Network(Module):
+    def __init__(
+        self,
+        dim
+    ):
+        super().__init__()
+
+class DDN(Module):
+    def __init__(
+        self
+    ):
+        super().__init__()
+
+# trainer
+
+class Trainer(Module):
+    def __init__(
+        self,
+    ):
+        super().__init__()

discrete_distribution_network-0.0.1/pyproject.toml

@@ -0,0 +1,58 @@
+[project]
+name = "discrete-distribution-network"
+version = "0.0.1"
+description = "Discrete Distribution Network"
+authors = [
+    { name = "Phil Wang", email = "lucidrains@gmail.com" }
+]
+readme = "README.md"
+requires-python = ">= 3.9"
+license = { file = "LICENSE" }
+keywords = [
+    'artificial intelligence',
+    'deep learning',
+    'generative modeling',
+]
+
+classifiers=[
+    'Development Status :: 4 - Beta',
+    'Intended Audience :: Developers',
+    'Topic :: Scientific/Engineering :: Artificial Intelligence',
+    'License :: OSI Approved :: MIT License',
+    'Programming Language :: Python :: 3.9',
+]
+
+dependencies = [
+    "einops>=0.8.1",
+    "torch>=2.4",
+    "x-mlps-pytorch>=0.0.30"
+]
+
+[project.urls]
+Homepage = "https://pypi.org/project/discrete-distribution-network/"
+Repository = "https://github.com/lucidrains/discrete-distribution-network"
+
+[project.optional-dependencies]
+examples = []
+test = [
+    "pytest"
+]
+
+[tool.pytest.ini_options]
+pythonpath = [
+  "."
+]
+
+[build-system]
+requires = ["hatchling"]
+build-backend = "hatchling.build"
+
+[tool.rye]
+managed = true
+dev-dependencies = []
+
+[tool.hatch.metadata]
+allow-direct-references = true
+
+[tool.hatch.build.targets.wheel]
+packages = ["discrete_distribution_network"]

discrete_distribution_network-0.0.1/tests/test_ddn.py

@@ -0,0 +1,29 @@
+import pytest
+
+import torch
+from torch import tensor
+
+def test_ddn():
+    from discrete_distribution_network.ddn import GuidedSampler
+
+    sampler = GuidedSampler(
+        dim = 16,
+        dim_query = 3,
+        codebook_size = 10
+    )
+
+    features = torch.randn(2, 16, 32, 32)
+    query_image = torch.randn(2, 3, 32, 32)
+
+    out, codes, commit_loss = sampler(features, query_image)
+
+    assert out.shape == query_image.shape
+    assert codes.shape == (2,)
+    assert commit_loss.numel() == 1
+
+    sampler.split_and_prune_()
+
+    # after much training
+
+    assert sampler.forward_for_codes(features, tensor([3, 5])).shape == (2, 3, 32, 32)
+    assert sampler.forward_for_codes(features, tensor(7)).shape == (2, 3, 32, 32)

discrete_distribution_network-0.0.1/train_ddn.py

@@ -0,0 +1,59 @@
+# /// script
+# dependencies = [
+#   "discrete-distribution-network",
+#   "datasets",
+#   "wandb"
+# ]
+# ///
+
+import torch
+
+# hf datasets for easy oxford flowers training
+
+import torchvision.transforms as T
+from torch.utils.data import Dataset
+from datasets import load_dataset
+
+class OxfordFlowersDataset(Dataset):
+    def __init__(
+        self,
+        image_size
+    ):
+        self.ds = load_dataset('nelorth/oxford-flowers')['train']
+
+        self.transform = T.Compose([
+            T.Resize((image_size, image_size)),
+            T.PILToTensor()
+        ])
+
+    def __len__(self):
+        return len(self.ds)
+
+    def __getitem__(self, idx):
+        pil = self.ds[idx]['image']
+        tensor = self.transform(pil)
+        return tensor / 255.
+
+flowers_dataset = OxfordFlowersDataset(
+    image_size = 64
+)
+
+# models and trainer
+
+from discrete_distribution_network import DDN, Network, Trainer
+
+model = Network(
+    dim = 64
+)
+
+ddn = DDN(model)
+
+trainer = Trainer(
+    ddn,
+    dataset = flowers_dataset,
+    num_train_steps = 70_000,
+    sample_temperature = 1.5,
+    results_folder = './results'   # samples will be saved periodically to this folder
+)
+
+trainer()