captix 1.0.0__tar.gz

captix-1.0.0/.gitignore

@@ -0,0 +1,23 @@
+# Python-generated files
+__pycache__/
+*.py[oc]
+build/
+dist/
+wheels/
+*.egg-info
+
+# Virtual environments
+.venv
+notebooks/data
+notebooks/data_merged
+notebooks/data_live
+notebooks/*.pt
+docs/notebooks/data
+docs/notebooks/data_merged
+docs/notebooks/data_live
+docs/notebooks/*.pt
+.claude/
+
+.env
+# Isolated agent worktrees
+.worktrees/

captix-1.0.0/CHANGELOG.md

@@ -0,0 +1,50 @@
+# Changelog
+
+All notable changes to `txtcaptcha` are documented here. The format follows
+[Keep a Changelog](https://keepachangelog.com/en/1.1.0/) and the project
+adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
+
+## [Unreleased]
+
+### Changed
+- Notebooks moved under `docs/notebooks/` and surfaced on the docs site.
+
+## [0.1.0] - 2026-04-11
+
+Initial public release.
+
+### Added
+- CRNN+CTC architecture (`txtcaptcha.CRNN`) with ResNet-style CNN backbone,
+  2-layer BiLSTM and CTC head. Handles variable-width input and variable
+  label lengths out of the box.
+- `read_captcha`, `Captcha`, `annotate` for loading/labeling images.
+- `CaptchaDataset`, `transform_image`, `encode_label`, `decode_indices`,
+  `pad_collate` for building training pipelines.
+- `fit_model` training loop with validation split, CTC loss and early
+  stopping.
+- `decrypt()` with three decoding modes:
+  - greedy CTC (default),
+  - fixed-length exact DP via `length=N`,
+  - decode-time masking via `mask=` (list, regex, or `re.Pattern`).
+- `save_model` / `load_model` with automatic CPU fallback when CUDA is
+  unavailable.
+- Dataset fetchers: `available_datasets`, `download_dataset`.
+- Live captcha fetchers via `download_captchas` CLI and the
+  `txtcaptcha.download` subpackage (10 Brazilian court / tax sources).
+- Hugging Face Hub integration:
+  - `save_pretrained`, `from_pretrained`, `push_to_hub`, `DEFAULT_REPO_ID`,
+  - safetensors + `config.json` + model card layout,
+  - `decrypt()` auto-downloads the unified checkpoint from
+    [`jtrecenti/txtcaptcha-crnn`](https://huggingface.co/jtrecenti/txtcaptcha-crnn)
+    on first call and caches it.
+- Unified pretrained checkpoint published on the Hub at
+  `jtrecenti/txtcaptcha-crnn` (~89% captcha-level accuracy across ten
+  Brazilian court datasets).
+- Notebooks: `train_unified_model.ipynb`, `eval_per_dataset.ipynb`,
+  `eval_per_dataset_live.ipynb`.
+- GitHub Actions:
+  - CI running pytest on push/PR,
+  - PyPI publishing via trusted publishing on GitHub releases,
+  - Quarto + quartodoc site deployed to GitHub Pages on pushes to `main`.
+
+[0.1.0]: https://github.com/jtrecenti/txtcaptcha/releases/tag/v0.1.0

captix-1.0.0/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 Julio Trecenti
+
+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.

captix-1.0.0/PKG-INFO

@@ -0,0 +1,261 @@
+Metadata-Version: 2.5
+Name: captix
+Version: 1.0.0
+Summary: Read, annotate, train and decrypt text captchas with a CRNN+CTC model.
+Project-URL: Homepage, https://github.com/Diaszano/captix
+Project-URL: Documentation, https://Diaszano.github.io/captix/
+Project-URL: Repository, https://github.com/Diaszano/captix
+Project-URL: Issues, https://github.com/Diaszano/captix/issues
+Project-URL: Changelog, https://github.com/Diaszano/captix/blob/main/CHANGELOG.md
+Project-URL: Model (HF Hub), https://huggingface.co/jtrecenti/txtcaptcha-crnn
+Author-email: Julio Trecenti <jtrecenti@gmail.com>
+Maintainer-email: Julio Trecenti <jtrecenti@gmail.com>
+License: MIT License
+        
+        Copyright (c) 2026 Julio Trecenti
+        
+        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: captcha,crnn,ctc,deep-learning,image-to-text,ocr,pytorch
+Classifier: Development Status :: 4 - Beta
+Classifier: Intended Audience :: Developers
+Classifier: Intended Audience :: Science/Research
+Classifier: License :: OSI Approved :: MIT License
+Classifier: Operating System :: OS Independent
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3 :: Only
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
+Classifier: Topic :: Scientific/Engineering :: Image Recognition
+Requires-Python: >=3.10
+Requires-Dist: huggingface-hub>=0.20
+Requires-Dist: matplotlib>=3.8
+Requires-Dist: numpy>=1.24
+Requires-Dist: pandas>=2.0
+Requires-Dist: pillow>=10.0
+Requires-Dist: python-dotenv>=1.0
+Requires-Dist: requests>=2.31
+Requires-Dist: safetensors>=0.4
+Requires-Dist: torch>=2.2
+Requires-Dist: torchvision>=0.17
+Requires-Dist: tqdm>=4.66
+Provides-Extra: dev
+Requires-Dist: build>=1.2; extra == 'dev'
+Requires-Dist: pre-commit<5,>=4.6; extra == 'dev'
+Requires-Dist: pytest-cov>=5.0; extra == 'dev'
+Requires-Dist: pytest>=8.0; extra == 'dev'
+Requires-Dist: python-semantic-release==10.7.0; extra == 'dev'
+Requires-Dist: twine>=5.0; extra == 'dev'
+Provides-Extra: docs
+Requires-Dist: griffe<1; extra == 'docs'
+Requires-Dist: quartodoc>=0.7; extra == 'docs'
+Provides-Extra: notebook
+Requires-Dist: ipykernel>=6.29; extra == 'notebook'
+Requires-Dist: jupyter>=1.0; extra == 'notebook'
+Description-Content-Type: text/markdown
+
+# Captix
+
+[![PyPI](https://img.shields.io/pypi/v/captix.svg)](https://pypi.org/project/captix/)
+[![Python](https://img.shields.io/pypi/pyversions/captix.svg)](https://pypi.org/project/captix/)
+[![License: MIT](https://img.shields.io/badge/license-MIT-blue.svg)](LICENSE)
+[![Model on HF](https://img.shields.io/badge/%F0%9F%A4%97%20HF-jtrecenti%2Ftxtcaptcha--crnn-yellow)](https://huggingface.co/jtrecenti/txtcaptcha-crnn)
+[![Docs](https://img.shields.io/badge/docs-online-brightgreen)](https://Diaszano.github.io/captix/)
+
+Read, annotate, train and decrypt **text captchas in images** with a modern
+CRNN + CTC pipeline in PyTorch.
+
+Captix ships:
+
+- a CRNN architecture that handles **arbitrary input sizes** and
+  **variable-length labels**,
+- the full alphanumeric vocabulary `0-9a-zA-Z` (62 classes + CTC blank),
+- **decode-time masking** so a single trained model can be restricted per
+  site (e.g. `mask="[0-9]"`),
+- **fixed-length decoding** via `length=N` for sites with a known length,
+- a pretrained unified model hosted on the
+  [Hugging Face Hub](https://huggingface.co/jtrecenti/txtcaptcha-crnn) with
+  ~89% captcha-level accuracy across ten Brazilian court captcha datasets.
+
+## Installation
+
+```bash
+pip install captix
+```
+
+Or from source with [`uv`](https://docs.astral.sh/uv/):
+
+```bash
+git clone https://github.com/Diaszano/captix
+cd captix
+uv sync --extra dev
+```
+
+The distribution installs as `captix`; Python imports use `captix`.
+
+## Quick start
+
+The first `decrypt` call downloads the pretrained model from the Hugging Face
+Hub into `~/.cache/huggingface/hub`; subsequent calls are free.
+
+```python
+from captix import read_captcha, decrypt
+
+cap = read_captcha("path/to/captcha.png")
+print(decrypt(cap))                          # greedy, variable length
+print(decrypt(cap, mask="[0-9]"))            # digits only
+print(decrypt(cap, length=5))                # force exactly 5 chars
+print(decrypt(cap, mask=list("abcdef0123"))) # explicit allowed set
+```
+
+Pin a specific release or load a different Hub repo explicitly:
+
+```python
+from captix import from_pretrained
+
+model = from_pretrained("jtrecenti/txtcaptcha-crnn", revision="v0.1.0")
+print(decrypt(cap, model=model))
+```
+
+### Training your own model
+
+```python
+from captix import fit_model, save_model, download_dataset
+
+data_dir = download_dataset("tjmg", "data")
+model, history = fit_model(
+    data_dir,
+    epochs=30,
+    batch_size=64,
+    case_sensitive=False,
+)
+save_model(model, "tjmg.pt")
+```
+
+### Publishing your own model to the Hub
+
+```python
+from captix import push_to_hub
+
+push_to_hub(
+    model,
+    repo_id="your-username/your-captcha-model",
+    model_card="# My captcha model\n\nTrained on ...",
+    tag="v0.1.0",
+)
+```
+
+## Public API
+
+| Function | Purpose |
+|---|---|
+| `read_captcha(files, lab_in_path=False)` | Load image(s) into a `Captcha` object. |
+| `Captcha` | Container with `images`, `labels`, `paths`, `plot()`. |
+| `annotate(files, labels=None, ...)` | Interactive/batch labeling (filename convention). |
+| `CaptchaDataset(root, vocab, height, case_sensitive)` | PyTorch dataset over a folder of `<id>_<label>.<ext>` files. |
+| `transform_image(files, height=32)` | Load + resize + width-pad for batching. |
+| `encode_label`, `decode_indices` | Vocab ↔ tensor (CTC blank index 0). |
+| `pad_collate` | DataLoader collate fn for variable-width batching. |
+| `CRNN(vocab, ...)` | CNN + BiLSTM + linear head. |
+| `fit_model(dir, ...)` | Training loop with CTC loss + early stopping. |
+| `decrypt(files, model=None, mask=None, case_sensitive=True, length=None)` | Predict labels; auto-downloads the pretrained model when `model=None`. |
+| `save_model`, `load_model` | Local checkpoint persistence. |
+| `from_pretrained`, `save_pretrained`, `push_to_hub` | Hugging Face Hub integration. |
+| `download_dataset`, `available_datasets` | Fetch labeled training datasets. |
+| `download_captchas` (CLI) | Download live, unlabeled captchas from 10 Brazilian sources. |
+| `sequence_accuracy(preds, targets)` | Exact-match accuracy metric. |
+
+Full API reference: **<https://Diaszano.github.io/captix/>**.
+
+## Architecture
+
+`CRNN` is a Convolutional Recurrent Neural Network:
+
+1. **CNN backbone** — ResNet-style basic blocks (`64 → 128 → 256 → 256`
+   channels) with strided pooling. Down-samples height by `8` and width by
+   `4`, preserving width resolution for the sequence dimension.
+2. **Adaptive pool** — collapses the remaining height to 1, producing a
+   width-indexed sequence of feature vectors.
+3. **BiLSTM** — 2-layer bidirectional LSTM (hidden 256).
+4. **Linear head** — projects to `len(vocab) + 1` logits per timestep (the
+   extra slot is the CTC blank).
+5. **CTC loss** — handles variable-length targets, no per-position softmax.
+
+Variable image dimensions are handled by resizing height to 32 at load time,
+preserving the aspect-ratio width, and padding widths within each batch via
+`pad_collate`. CRNN+CTC is the de-facto baseline for short-text scene-text
+recognition — lighter than transformer OCR (e.g. TrOCR) and consistently
+strong on short captcha images.
+
+## Variable-length labels
+
+CRNN + CTC handles variable label lengths natively. The convolutional stack
+emits `T` logits per image; CTC collapsing (remove consecutive repeats, then
+remove blanks) turns any path into a string of arbitrary length between 0
+and `T`. Training mixes 4-char and 5-char labels in the same batch — no
+length head, no padding tokens.
+
+The downside of greedy CTC is that a confident wrong timestep can yield a
+prediction of the wrong length. When you know the expected length, pass
+`length=` to switch to an exact dynamic-programming search over CTC paths
+that collapse to exactly that many characters:
+
+```python
+decrypt(cap)                       # greedy
+decrypt(cap, length=5)             # force 5 chars
+decrypt(cap, length=4, mask="[0-9]")  # combine with masking
+```
+
+The DP runs in `O(T · L · |vocab|)` per image, tracks the best path for
+every `(collapsed_count, last_index)` state and reconstructs the argmax. It
+is strictly at least as good as greedy when the true length is known and
+never emits a wrong-length prediction.
+
+## Decode-time masking
+
+`decrypt(..., mask=...)` zeros out forbidden vocabulary logits before CTC
+decoding, so the same trained model can be specialized per site:
+
+```python
+decrypt(cap, mask=["a", "b", "c", "1", "2", "3"])  # explicit list
+decrypt(cap, mask="[0-9a-z]")                       # regex char-class
+decrypt(cap, mask="[A-Z]", case_sensitive=True)     # uppercase only
+decrypt(cap, mask="[a-z]", case_sensitive=False)    # output lowercased
+```
+
+## Notebooks
+
+- `docs/notebooks/train_unified_model.ipynb` — downloads every dataset, merges
+  them and trains the unified CRNN. Designed for a cloud GPU machine.
+- `docs/notebooks/eval_per_dataset.ipynb` — per-dataset accuracy on a held-out
+  split.
+- `docs/notebooks/eval_per_dataset_live.ipynb` — predictions on freshly
+  downloaded, unlabeled captchas (overfit check).
+
+## Tests
+
+```bash
+uv run pytest
+```
+
+## License
+
+[MIT](LICENSE) © Julio Trecenti

captix-1.0.0/README.md

@@ -0,0 +1,187 @@
+# Captix
+
+[![PyPI](https://img.shields.io/pypi/v/captix.svg)](https://pypi.org/project/captix/)
+[![Python](https://img.shields.io/pypi/pyversions/captix.svg)](https://pypi.org/project/captix/)
+[![License: MIT](https://img.shields.io/badge/license-MIT-blue.svg)](LICENSE)
+[![Model on HF](https://img.shields.io/badge/%F0%9F%A4%97%20HF-jtrecenti%2Ftxtcaptcha--crnn-yellow)](https://huggingface.co/jtrecenti/txtcaptcha-crnn)
+[![Docs](https://img.shields.io/badge/docs-online-brightgreen)](https://Diaszano.github.io/captix/)
+
+Read, annotate, train and decrypt **text captchas in images** with a modern
+CRNN + CTC pipeline in PyTorch.
+
+Captix ships:
+
+- a CRNN architecture that handles **arbitrary input sizes** and
+  **variable-length labels**,
+- the full alphanumeric vocabulary `0-9a-zA-Z` (62 classes + CTC blank),
+- **decode-time masking** so a single trained model can be restricted per
+  site (e.g. `mask="[0-9]"`),
+- **fixed-length decoding** via `length=N` for sites with a known length,
+- a pretrained unified model hosted on the
+  [Hugging Face Hub](https://huggingface.co/jtrecenti/txtcaptcha-crnn) with
+  ~89% captcha-level accuracy across ten Brazilian court captcha datasets.
+
+## Installation
+
+```bash
+pip install captix
+```
+
+Or from source with [`uv`](https://docs.astral.sh/uv/):
+
+```bash
+git clone https://github.com/Diaszano/captix
+cd captix
+uv sync --extra dev
+```
+
+The distribution installs as `captix`; Python imports use `captix`.
+
+## Quick start
+
+The first `decrypt` call downloads the pretrained model from the Hugging Face
+Hub into `~/.cache/huggingface/hub`; subsequent calls are free.
+
+```python
+from captix import read_captcha, decrypt
+
+cap = read_captcha("path/to/captcha.png")
+print(decrypt(cap))                          # greedy, variable length
+print(decrypt(cap, mask="[0-9]"))            # digits only
+print(decrypt(cap, length=5))                # force exactly 5 chars
+print(decrypt(cap, mask=list("abcdef0123"))) # explicit allowed set
+```
+
+Pin a specific release or load a different Hub repo explicitly:
+
+```python
+from captix import from_pretrained
+
+model = from_pretrained("jtrecenti/txtcaptcha-crnn", revision="v0.1.0")
+print(decrypt(cap, model=model))
+```
+
+### Training your own model
+
+```python
+from captix import fit_model, save_model, download_dataset
+
+data_dir = download_dataset("tjmg", "data")
+model, history = fit_model(
+    data_dir,
+    epochs=30,
+    batch_size=64,
+    case_sensitive=False,
+)
+save_model(model, "tjmg.pt")
+```
+
+### Publishing your own model to the Hub
+
+```python
+from captix import push_to_hub
+
+push_to_hub(
+    model,
+    repo_id="your-username/your-captcha-model",
+    model_card="# My captcha model\n\nTrained on ...",
+    tag="v0.1.0",
+)
+```
+
+## Public API
+
+| Function | Purpose |
+|---|---|
+| `read_captcha(files, lab_in_path=False)` | Load image(s) into a `Captcha` object. |
+| `Captcha` | Container with `images`, `labels`, `paths`, `plot()`. |
+| `annotate(files, labels=None, ...)` | Interactive/batch labeling (filename convention). |
+| `CaptchaDataset(root, vocab, height, case_sensitive)` | PyTorch dataset over a folder of `<id>_<label>.<ext>` files. |
+| `transform_image(files, height=32)` | Load + resize + width-pad for batching. |
+| `encode_label`, `decode_indices` | Vocab ↔ tensor (CTC blank index 0). |
+| `pad_collate` | DataLoader collate fn for variable-width batching. |
+| `CRNN(vocab, ...)` | CNN + BiLSTM + linear head. |
+| `fit_model(dir, ...)` | Training loop with CTC loss + early stopping. |
+| `decrypt(files, model=None, mask=None, case_sensitive=True, length=None)` | Predict labels; auto-downloads the pretrained model when `model=None`. |
+| `save_model`, `load_model` | Local checkpoint persistence. |
+| `from_pretrained`, `save_pretrained`, `push_to_hub` | Hugging Face Hub integration. |
+| `download_dataset`, `available_datasets` | Fetch labeled training datasets. |
+| `download_captchas` (CLI) | Download live, unlabeled captchas from 10 Brazilian sources. |
+| `sequence_accuracy(preds, targets)` | Exact-match accuracy metric. |
+
+Full API reference: **<https://Diaszano.github.io/captix/>**.
+
+## Architecture
+
+`CRNN` is a Convolutional Recurrent Neural Network:
+
+1. **CNN backbone** — ResNet-style basic blocks (`64 → 128 → 256 → 256`
+   channels) with strided pooling. Down-samples height by `8` and width by
+   `4`, preserving width resolution for the sequence dimension.
+2. **Adaptive pool** — collapses the remaining height to 1, producing a
+   width-indexed sequence of feature vectors.
+3. **BiLSTM** — 2-layer bidirectional LSTM (hidden 256).
+4. **Linear head** — projects to `len(vocab) + 1` logits per timestep (the
+   extra slot is the CTC blank).
+5. **CTC loss** — handles variable-length targets, no per-position softmax.
+
+Variable image dimensions are handled by resizing height to 32 at load time,
+preserving the aspect-ratio width, and padding widths within each batch via
+`pad_collate`. CRNN+CTC is the de-facto baseline for short-text scene-text
+recognition — lighter than transformer OCR (e.g. TrOCR) and consistently
+strong on short captcha images.
+
+## Variable-length labels
+
+CRNN + CTC handles variable label lengths natively. The convolutional stack
+emits `T` logits per image; CTC collapsing (remove consecutive repeats, then
+remove blanks) turns any path into a string of arbitrary length between 0
+and `T`. Training mixes 4-char and 5-char labels in the same batch — no
+length head, no padding tokens.
+
+The downside of greedy CTC is that a confident wrong timestep can yield a
+prediction of the wrong length. When you know the expected length, pass
+`length=` to switch to an exact dynamic-programming search over CTC paths
+that collapse to exactly that many characters:
+
+```python
+decrypt(cap)                       # greedy
+decrypt(cap, length=5)             # force 5 chars
+decrypt(cap, length=4, mask="[0-9]")  # combine with masking
+```
+
+The DP runs in `O(T · L · |vocab|)` per image, tracks the best path for
+every `(collapsed_count, last_index)` state and reconstructs the argmax. It
+is strictly at least as good as greedy when the true length is known and
+never emits a wrong-length prediction.
+
+## Decode-time masking
+
+`decrypt(..., mask=...)` zeros out forbidden vocabulary logits before CTC
+decoding, so the same trained model can be specialized per site:
+
+```python
+decrypt(cap, mask=["a", "b", "c", "1", "2", "3"])  # explicit list
+decrypt(cap, mask="[0-9a-z]")                       # regex char-class
+decrypt(cap, mask="[A-Z]", case_sensitive=True)     # uppercase only
+decrypt(cap, mask="[a-z]", case_sensitive=False)    # output lowercased
+```
+
+## Notebooks
+
+- `docs/notebooks/train_unified_model.ipynb` — downloads every dataset, merges
+  them and trains the unified CRNN. Designed for a cloud GPU machine.
+- `docs/notebooks/eval_per_dataset.ipynb` — per-dataset accuracy on a held-out
+  split.
+- `docs/notebooks/eval_per_dataset_live.ipynb` — predictions on freshly
+  downloaded, unlabeled captchas (overfit check).
+
+## Tests
+
+```bash
+uv run pytest
+```
+
+## License
+
+[MIT](LICENSE) © Julio Trecenti

captix-1.0.0/pyproject.toml

@@ -0,0 +1,112 @@
+[project]
+name = "captix"
+version = "1.0.0"
+description = "Read, annotate, train and decrypt text captchas with a CRNN+CTC model."
+readme = "README.md"
+requires-python = ">=3.10"
+license = { file = "LICENSE" }
+authors = [
+    { name = "Julio Trecenti", email = "jtrecenti@gmail.com" },
+]
+maintainers = [
+    { name = "Julio Trecenti", email = "jtrecenti@gmail.com" },
+]
+keywords = [
+    "captcha",
+    "ocr",
+    "crnn",
+    "ctc",
+    "pytorch",
+    "deep-learning",
+    "image-to-text",
+]
+classifiers = [
+    "Development Status :: 4 - Beta",
+    "Intended Audience :: Developers",
+    "Intended Audience :: Science/Research",
+    "License :: OSI Approved :: MIT License",
+    "Operating System :: OS Independent",
+    "Programming Language :: Python :: 3",
+    "Programming Language :: Python :: 3 :: Only",
+    "Programming Language :: Python :: 3.10",
+    "Programming Language :: Python :: 3.11",
+    "Programming Language :: Python :: 3.12",
+    "Topic :: Scientific/Engineering :: Artificial Intelligence",
+    "Topic :: Scientific/Engineering :: Image Recognition",
+]
+dependencies = [
+    "torch>=2.2",
+    "torchvision>=0.17",
+    "Pillow>=10.0",
+    "numpy>=1.24",
+    "requests>=2.31",
+    "tqdm>=4.66",
+    "matplotlib>=3.8",
+    "pandas>=2.0",
+    "huggingface-hub>=0.20",
+    "safetensors>=0.4",
+    "python-dotenv>=1.0",
+]
+
+[project.optional-dependencies]
+dev = [
+    "pytest>=8.0",
+    "pytest-cov>=5.0",
+    "build>=1.2",
+    "twine>=5.0",
+    "pre-commit>=4.6,<5",
+    "python-semantic-release==10.7.0",
+]
+notebook = [
+    "jupyter>=1.0",
+    "ipykernel>=6.29",
+]
+docs = [
+    "quartodoc>=0.7",
+    "griffe<1",
+]
+
+[project.urls]
+Homepage = "https://github.com/Diaszano/captix"
+Documentation = "https://Diaszano.github.io/captix/"
+Repository = "https://github.com/Diaszano/captix"
+Issues = "https://github.com/Diaszano/captix/issues"
+Changelog = "https://github.com/Diaszano/captix/blob/main/CHANGELOG.md"
+"Model (HF Hub)" = "https://huggingface.co/jtrecenti/txtcaptcha-crnn"
+
+[project.scripts]
+download_captchas = "captix.download.cli:main"
+
+[build-system]
+requires = ["hatchling"]
+build-backend = "hatchling.build"
+
+[tool.hatch.build.targets.wheel]
+packages = ["src/captix"]
+
+[tool.hatch.build.targets.sdist]
+include = [
+    "src/captix",
+    "tests",
+    "README.md",
+    "LICENSE",
+    "CHANGELOG.md",
+    "pyproject.toml",
+]
+
+[tool.semantic_release]
+version_toml = ["pyproject.toml:project.version"]
+version_variables = ["src/captix/__init__.py:__version__"]
+tag_format = "v{version}"
+commit_parser = "conventional"
+allow_zero_version = true
+major_on_zero = true
+commit_message = "chore(release): {version}"
+
+[tool.semantic_release.branches.main]
+match = "main"
+
+[tool.pytest.ini_options]
+testpaths = ["tests"]
+addopts = "-q"
+pythonpath = ["."]