diffgentor 0.1.0__tar.gz

diffgentor-0.1.0/.github/workflows/publish.yml

@@ -0,0 +1,80 @@
+name: Publish to PyPI
+
+on:
+  push:
+    tags:
+      - "v*"  # Trigger on tags like v0.1.0, v1.0.0, etc.
+  workflow_dispatch:
+
+jobs:
+  build:
+    name: Build distribution
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Update version from tag
+        if: startsWith(github.ref, 'refs/tags/v')
+        run: |
+          VERSION=${GITHUB_REF#refs/tags/v}
+          echo "Updating pyproject.toml version to $VERSION"
+          sed -i "s/^version = \".*\"/version = \"$VERSION\"/" pyproject.toml
+          cat pyproject.toml | grep "^version"
+
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: "3.11"
+
+      - name: Install build dependencies
+        run: |
+          python -m pip install --upgrade pip
+          pip install build
+
+      - name: Build package
+        run: python -m build
+
+      - name: Upload distribution artifacts
+        uses: actions/upload-artifact@v4
+        with:
+          name: python-package-distributions
+          path: dist/
+
+  create-release:
+    name: Create GitHub Release
+    needs: build
+    runs-on: ubuntu-latest
+    permissions:
+      contents: write  # Required for creating releases
+    steps:
+      - name: Download distribution artifacts
+        uses: actions/download-artifact@v4
+        with:
+          name: python-package-distributions
+          path: dist/
+
+      - name: Create GitHub Release
+        uses: softprops/action-gh-release@v2
+        with:
+          files: dist/*
+          generate_release_notes: true
+
+  publish-to-pypi:
+    name: Publish to PyPI
+    needs: build
+    runs-on: ubuntu-latest
+    environment:
+      name: pypi
+      url: https://pypi.org/p/diffgentor
+    permissions:
+      id-token: write  # Required for trusted publishing
+
+    steps:
+      - name: Download distribution artifacts
+        uses: actions/download-artifact@v4
+        with:
+          name: python-package-distributions
+          path: dist/
+
+      - name: Publish to PyPI
+        uses: pypa/gh-action-pypi-publish@release/v1

diffgentor-0.1.0/.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__/

diffgentor-0.1.0/.gitmodules

@@ -0,0 +1,15 @@
+[submodule "diffgentor/models/third_party/emu35"]
+	path = diffgentor/models/third_party/emu35
+	url = https://github.com/diffgentor/Emu3.5
+[submodule "diffgentor/models/third_party/step1x_edit"]
+	path = diffgentor/models/third_party/step1x_edit
+	url = https://github.com/diffgentor/Step1X-Edit
+[submodule "diffgentor/models/third_party/bagel"]
+	path = diffgentor/models/third_party/bagel
+	url = https://github.com/diffgentor/Bagel
+[submodule "diffgentor/models/third_party/dreamomni2"]
+	path = diffgentor/models/third_party/dreamomni2
+	url = https://github.com/diffgentor/DreamOmni2
+[submodule "diffgentor/models/third_party/flux1"]
+	path = diffgentor/models/third_party/flux1
+	url = https://github.com/diffgentor/flux

diffgentor-0.1.0/AGENTS.md

@@ -0,0 +1,213 @@
+---
+name: diffgentor_agent
+description: Expert developer for the diffgentor visual generation data synthesis project
+---
+
+You are an expert Python developer for the diffgentor project - a unified visual generation data synthesis factory.
+
+## Your Role
+
+- You are fluent in Python 3.10+ and familiar with deep learning frameworks (PyTorch, diffusers, transformers)
+- You specialize in image generation and editing pipelines, distributed inference, and API integrations
+- Your task: maintain and extend the diffgentor codebase for T2I (text-to-image) and image editing capabilities
+
+## Project Knowledge
+
+### Tech Stack
+
+- **Python:** 3.10+
+- **Core Dependencies:** PyTorch >=2.3.0, diffusers >=0.31.0, transformers >=4.40.0
+- **Optional:** xDiT (multi-GPU), OpenAI API, xformers, DeepCache, torchao, bitsandbytes
+- **Build System:** hatchling (pyproject.toml)
+- **Code Style:** black (line-length=120), ruff for linting
+
+### File Structure
+
+```
+diffgentor/
+├── __init__.py          # Package init
+├── __main__.py          # Entry point
+├── config.py            # Configuration classes (BackendConfig, EditingConfig, OptimizationConfig)
+├── backends/            # Backend implementations
+│   ├── base.py          # Base backend class
+│   ├── registry.py      # Backend registry
+│   ├── editing/         # Editing backends (openai, google_genai, step1x, bagel, etc.)
+│   └── t2i/             # T2I backends (diffusers, xdit)
+├── cli/                 # CLI interface
+│   └── main.py          # Argument parsing, subcommands (t2i, edit)
+├── launcher/            # Distributed launcher
+│   └── launcher.py      # Multi-process/GPU coordination
+├── models/              # Model definitions
+│   └── third_party/     # Third-party model integrations, git submodules
+├── optimizations/       # Optimization utilities
+│   └── manager.py       # VAE slicing, torch.compile, attention backends, cache
+├── prompt_enhance/      # Prompt enhancement module
+│   ├── base.py          # Base PromptEnhancer class
+│   ├── registry.py      # Enhancer registry
+│   ├── flux2.py         # Flux2 style enhancer (diffusers/API modes)
+│   ├── qwen_image_edit.py
+│   └── glm_image.py
+├── utils/               # Utility functions
+│   ├── env.py           # Environment variable utilities (DG_ prefix)
+│   ├── api_pool.py      # API endpoint pool with load balancing
+│   ├── distributed.py   # Distributed training utilities
+│   ├── data.py          # Data loading/saving
+│   ├── image.py         # Image processing
+│   └── logging.py       # Logging utilities
+└── workers/             # Worker processes
+    ├── edit_worker.py   # Image editing worker
+    └── t2i_worker.py    # T2I generation worker
+docs/                    # Documentation
+temp/                    # Temporary files, experiments (git-ignored mostly)
+```
+
+### Supported Backends
+
+| Backend | Type | Description |
+|---------|------|-------------|
+| `diffusers` | T2I / Editing | HuggingFace diffusers with auto pipeline detection |
+| `xdit` | T2I | Multi-GPU inference with xDiT parallelism |
+| `openai` | T2I / Editing | OpenAI API (GPT-Image, DALL-E) |
+| `google_genai` | T2I / Editing | Google GenAI (Gemini native image models) |
+| `step1x` | Editing | Step1X-Edit model |
+| `bagel` | Editing | ByteDance BAGEL model |
+| `emu35` | Editing | BAAI Emu3.5 model |
+| `dreamomni2` | Editing | DreamOmni2 (FLUX.1-Kontext + Qwen2.5-VL) |
+| `flux_kontext_official` | Editing | BFL official Flux Kontext |
+| `hunyuan_image_3` | Editing | Tencent HunyuanImage-3.0-Instruct with CoT reasoning |
+
+## Commands You Can Use
+
+```bash
+# Install dependencies
+pip install -e .
+pip install -e ".[all]"
+
+# Run T2I generation
+diffgentor t2i --backend diffusers --model_name black-forest-labs/FLUX.1-dev --prompt "A cat"
+
+# Run image editing
+diffgentor edit --backend diffusers --model_name Qwen/Qwen-Image-Edit-2511 --input data.csv
+
+# Run image editing with custom output filenames (from CSV/Parquet column)
+diffgentor edit --backend diffusers --model_name Qwen/Qwen-Image-Edit-2511 --input data.csv --output_name_column output_path
+
+# Lint code
+ruff check diffgentor/
+black --check diffgentor/
+
+# Format code
+black diffgentor/
+
+# Type check
+mypy diffgentor/
+
+# Run tests
+pytest tests/
+```
+
+### Custom Output Filename Support
+
+**Default behavior** (without `--output_name_column`):
+- Output files are named `{index:06d}.png` (e.g., `000000.png`, `000001.png`)
+- For multiple images per prompt: `{index:06d}_{sub_index:02d}.png` (e.g., `000000_00.png`)
+
+**With `--output_name_column`**: Specify a column in the input CSV/Parquet file to use as the output filename:
+
+- If the column value is `aaa/bb/1`, output will be saved as `output_dir/aaa/bb/1.png`
+- If the column value is `aaa/bb/1.jpg`, output will be saved as `output_dir/aaa/bb/1.jpg`
+- Supported formats: `.png`, `.jpg`, `.jpeg` (other extensions default to `.png`)
+- Parent directories are automatically created
+- For multiple images per prompt, sub-index is appended: `aaa/bb/1_00.png`, `aaa/bb/1_01.png`
+
+## Code Style Guidelines
+
+### General Rules
+
+- **Write all code comments in English**
+- Line length: 120 characters max
+- Use type hints for function signatures
+- Follow PEP 8 with black formatting
+- Import order: stdlib, third-party, local (handled by ruff/isort)
+
+### Environment Variables
+
+**All environment variables MUST be prefixed with `DG_`**
+
+Use the `diffgentor.utils.env` module for accessing environment variables:
+
+```python
+# Good - use helper functions
+from diffgentor.utils.env import get_env_str, get_env_int, get_env_float, get_env_bool
+
+api_key = get_env_str("PROMPT_ENHANCER_API_KEY")  # Reads DG_PROMPT_ENHANCER_API_KEY
+timeout = get_env_int("OPENAI_TIMEOUT", 300)      # Reads DG_OPENAI_TIMEOUT with default
+
+# Bad - direct os.environ access without DG_ prefix
+api_key = os.environ.get("API_KEY")  # Wrong! Missing DG_ prefix
+```
+
+Naming convention: `DG_{COMPONENT}_{PARAM}`
+
+Examples:
+- `DG_STEP1X_VERSION=v1.1`
+- `DG_BAGEL_CFG_TEXT_SCALE=3.0`
+- `DG_PROMPT_ENHANCER_API_KEY=xxx`
+- `DG_FLUX2_ENHANCER_MODE=api`
+- `DG_XDIT_ULYSSES_DEGREE=4`
+- `DG_XDIT_RING_DEGREE=2`
+- `DG_HUNYUAN_IMAGE_3_MOE_IMPL=flashinfer`
+- `DG_HUNYUAN_IMAGE_3_GPUS_PER_MODEL=4`
+
+### CLI Arguments vs Environment Variables
+
+**Only common/shared parameters should be added to CLI arguments.** Model-specific parameters (e.g., bagel's `cfg_text_scale`, step1x's `size_level`, emu35's `vq_path`, xDiT's `ulysses_degree`) MUST be configured via `DG_*` environment variables, NOT CLI arguments.
+
+- **CLI args**: Common parameters shared across backends (e.g., `--model_name`, `--batch_size`, `--num_inference_steps`)
+- **Env vars**: Model-specific parameters (e.g., `DG_BAGEL_CFG_TEXT_SCALE`, `DG_STEP1X_VERSION`, `DG_EMU35_VQ_PATH`, `DG_XDIT_ULYSSES_DEGREE`)
+
+### Distributed Logging
+
+The logging system is designed for distributed environments with the following behavior:
+
+- **Terminal output**: Only `local_rank=0` process on each node outputs to terminal
+- **File output**: All processes write to individual log files (`nodeX_processY.log`)
+- **Third-party libraries**: Automatically suppresses diffusers/transformers/tqdm output for non-main processes
+- **stdout/stderr redirect**: Captures all `print()` calls to the logging system
+
+**CLI option:**
+- `--log_dir`: Log directory path (default: `output_dir/logs/yyyymmdd_hhmm`)
+
+**Key modules:**
+- `diffgentor.utils.logging.LoggingConfig`: Logging configuration dataclass
+- `diffgentor.utils.logging.setup_logging()`: Initialize distributed logging system
+- `diffgentor.utils.logging.StreamRedirect`: Redirect stdout/stderr to logger
+
+## Boundaries
+
+### ✅ Always Do
+
+- Write all code comments in English
+- Use `DG_` prefix for all environment variables
+- Use `diffgentor.utils.env` helpers for env var access
+- Add type hints to function signatures
+- Follow existing code patterns and structure
+- Run `black` and `ruff` before committing
+- Document environment variables in docstrings
+- Update `AGENTS.md` when adding new backends/enhancers/other important features
+- Update related contents in `docs/` when adding/modifying features
+
+### ⚠️ Ask First
+
+- Before modifying core config classes (`config.py`)
+- Before changing CLI argument structure (`cli/main.py`)
+- Before modifying the base backend interface (`backends/base.py`)
+- Before adding new third-party dependencies to `pyproject.toml`
+
+### 🚫 Never Do
+
+- Never use environment variables without `DG_` prefix
+- Never write comments in languages other than English
+- Never commit API keys, secrets, or credentials
+- Never modify `.git/` or `.gitmodules` directly
+- Never bypass the registry pattern when adding backends/enhancers