From b4da5703ff9ad8c61b9e8c9454abe6ef321a03d9 Mon Sep 17 00:00:00 2001 From: Vinta Chen Date: Fri, 2 Oct 2026 11:37:15 +0800 Subject: [PATCH 1/6] audit: sweep Deep Learning, re-tier frameworks The tensorflow line called it "the most popular Deep Learning framework", false by downloads (torch 59.8M/month vs tensorflow 11.4M on pypistats, 2026-10-02), and the Frameworks tiers no longer matched the evidence: tensorflow 2.21.0 still ships no cp314 wheels (tensorflow/tensorflow#102890) and its releases now focus on security and bug fixes, while keras's 11.76M is mostly tensorflow pulling it in (requires keras>=3.12.0). Both move to Second Tier; jax (8.0M, Google's own successor) and pytorch-lightning (10.1M) become obvious choices. Co-Authored-By: Claude --- README.md | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/README.md b/README.md index 33e8c691..549b8604 100644 --- a/README.md +++ b/README.md @@ -188,10 +188,10 @@ _Frameworks for Neural Networks and Deep Learning. Also see [awesome-deep-learni - Frameworks - [pytorch](https://github.com/pytorch/pytorch) - Tensors and Dynamic neural networks in Python with strong GPU acceleration. - - [tensorflow](https://github.com/tensorflow/tensorflow) - The most popular Deep Learning framework created by Google. - - [keras](https://github.com/keras-team/keras) - A high-level deep learning library with support for JAX, TensorFlow, and PyTorch backends. - - [jax](https://github.com/jax-ml/jax) - A library for high-performance numerical computing with automatic differentiation and JIT compilation. - [pytorch-lightning](https://github.com/Lightning-AI/pytorch-lightning) - Deep learning framework to train, deploy, and ship AI products Lightning fast. + - [jax](https://github.com/jax-ml/jax) - A library for high-performance numerical computing with automatic differentiation and JIT compilation. + - [keras](https://github.com/keras-team/keras) - A high-level deep learning library with support for JAX, TensorFlow, and PyTorch backends. + - [tensorflow](https://github.com/tensorflow/tensorflow) - An end-to-end machine learning platform from Google. - Reinforcement Learning - [gymnasium](https://github.com/Farama-Foundation/Gymnasium) - A standard API for reinforcement learning environments with popular reference environments ([gym](https://github.com/openai/gym) successor). - [stable-baselines3](https://github.com/DLR-RM/stable-baselines3) - PyTorch implementations of Stable Baselines (deep) reinforcement learning algorithms. From eee27857ab59069ee5955259d9a3c6855f5b54f5 Mon Sep 17 00:00:00 2001 From: Vinta Chen Date: Fri, 2 Oct 2026 11:37:37 +0800 Subject: [PATCH 2/6] feat: add prophet to Machine Learning > Time Series Forecasting Time Series Forecasting listed only timesfm, while prophet, the most-installed forecasting library, was missing. Co-Authored-By: Claude --- README.md | 1 + 1 file changed, 1 insertion(+) diff --git a/README.md b/README.md index 549b8604..a51a7a82 100644 --- a/README.md +++ b/README.md @@ -209,6 +209,7 @@ _Libraries for Machine Learning. Also see [awesome-machine-learning](https://git - [lightgbm](https://github.com/lightgbm-org/LightGBM) - A fast, distributed, high performance gradient boosting framework. - [catboost](https://github.com/catboost/catboost) - A fast, scalable, high performance gradient boosting on decision trees library. - Time Series Forecasting + - [prophet](https://github.com/facebook/prophet) - A tool for producing forecasts for time series with multiple seasonality and trend changes. - [timesfm](https://github.com/google-research/timesfm) - A pretrained foundation model from Google Research for time-series forecasting, with non-commercial default weights. ### Natural Language Processing From 3cee8ada5f53b9624c60cf8b458c672b64438f81 Mon Sep 17 00:00:00 2001 From: Vinta Chen Date: Fri, 2 Oct 2026 11:37:54 +0800 Subject: [PATCH 3/6] feat: add statsforecast to Machine Learning > Time Series Forecasting The use case had no fast classical statistical forecasting library (ARIMA, ETS, Theta). Co-Authored-By: Claude --- README.md | 1 + 1 file changed, 1 insertion(+) diff --git a/README.md b/README.md index a51a7a82..20f6eda5 100644 --- a/README.md +++ b/README.md @@ -210,6 +210,7 @@ _Libraries for Machine Learning. Also see [awesome-machine-learning](https://git - [catboost](https://github.com/catboost/catboost) - A fast, scalable, high performance gradient boosting on decision trees library. - Time Series Forecasting - [prophet](https://github.com/facebook/prophet) - A tool for producing forecasts for time series with multiple seasonality and trend changes. + - [statsforecast](https://github.com/Nixtla/statsforecast) - Fast statistical forecasting models such as ARIMA, ETS, and Theta, compiled with numba. - [timesfm](https://github.com/google-research/timesfm) - A pretrained foundation model from Google Research for time-series forecasting, with non-commercial default weights. ### Natural Language Processing From 823a69642a190da20d53055881836bdf70a0e429 Mon Sep 17 00:00:00 2001 From: Vinta Chen Date: Fri, 2 Oct 2026 11:38:02 +0800 Subject: [PATCH 4/6] feat: add sktime to Machine Learning > Time Series Forecasting The use case had no unified scikit-learn-style time-series framework. Co-Authored-By: Claude --- README.md | 1 + 1 file changed, 1 insertion(+) diff --git a/README.md b/README.md index 20f6eda5..6e452d2e 100644 --- a/README.md +++ b/README.md @@ -211,6 +211,7 @@ _Libraries for Machine Learning. Also see [awesome-machine-learning](https://git - Time Series Forecasting - [prophet](https://github.com/facebook/prophet) - A tool for producing forecasts for time series with multiple seasonality and trend changes. - [statsforecast](https://github.com/Nixtla/statsforecast) - Fast statistical forecasting models such as ARIMA, ETS, and Theta, compiled with numba. + - [sktime](https://github.com/sktime/sktime) - A unified scikit-learn-style framework for forecasting and other time-series learning tasks. - [timesfm](https://github.com/google-research/timesfm) - A pretrained foundation model from Google Research for time-series forecasting, with non-commercial default weights. ### Natural Language Processing From 041b055a0d31ac515df7f095e1a4ef0d190c888c Mon Sep 17 00:00:00 2001 From: Vinta Chen Date: Fri, 2 Oct 2026 11:38:13 +0800 Subject: [PATCH 5/6] feat: add paddleocr to Computer Vision > OCR The OCR list had pytesseract and easyocr but omitted PaddleOCR, which led 2026 open-source OCR benchmarks on multilingual text and complex layouts. Co-Authored-By: Claude --- README.md | 1 + 1 file changed, 1 insertion(+) diff --git a/README.md b/README.md index 6e452d2e..d6e18e31 100644 --- a/README.md +++ b/README.md @@ -240,6 +240,7 @@ _Libraries for image and video analysis, object detection, and OCR._ - OCR - [pytesseract](https://github.com/madmaze/pytesseract) - A wrapper for the [Tesseract OCR](https://github.com/tesseract-ocr) engine. - [easyocr](https://github.com/JaidedAI/EasyOCR) - Ready-to-use OCR with 80+ languages supported. + - [paddleocr](https://github.com/PaddlePaddle/PaddleOCR) - Multilingual OCR and document parsing toolkit based on PaddlePaddle. ### Recommender Systems From a7069559f2ccd9a837d4c9630d478aa788d73cea Mon Sep 17 00:00:00 2001 From: Vinta Chen Date: Fri, 2 Oct 2026 11:38:56 +0800 Subject: [PATCH 6/6] docs: record AI & ML audit overrides The 2026-10-02 AI & ML audit kept or admitted five Beta-classifier entries and kept annoy past the 12-month activity line, but docs/audit-logs.md did not record these maintainer decisions, so the next audit would drop them as rule failures. Co-Authored-By: Claude --- docs/audit-logs.md | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/docs/audit-logs.md b/docs/audit-logs.md index 088c1711..fb6b0a40 100644 --- a/docs/audit-logs.md +++ b/docs/audit-logs.md @@ -29,6 +29,11 @@ Entries admitted despite the Stable quality requirement, by maintainer decision: - zensical -- admitted 2026-08-23 at version 0.0.57, PyPI classifier `Development Status :: 3 - Alpha`, with no 1.0 target announced. Admitted as a challenger displacing mkdocs because the Documentation section was carrying a dying upstream: the Material for MkDocs team's announcement of 2025-11-05 put Material into maintenance mode for twelve months and called MkDocs itself unmaintained since 2024-08 and a supply chain risk. Re-check when 1.0 lands. - openviking -- admitted 2026-08-25 (PR #3302) at version 0.4.16, PyPI classifier `Development Status :: 3 - Alpha`, with a README that said "OpenViking is still in its early stages". Admitted as a Data Layer challenger because the adoption is real: 459,446 downloads/month on pypistats (mirror-excluded) on 2026-08-24, and releases every few days. Re-check when 1.0 lands. - claude-agent-sdk -- kept 2026-10-01 in the AI and Agents audit at version 0.2.163, PyPI classifier `Development Status :: 3 - Alpha`. Kept as a Vendor Agent SDKs obvious choice because the adoption is real: 30,033,689 downloads/month on ClickPy on 2026-10-01, up from 1.2M in March 2026 on pypistats. Re-check when 1.0 lands. +- keras -- kept 2026-10-02 in the AI & ML audit as a Deep Learning Frameworks Second Tier challenger at version 3.15.1, PyPI classifier `Development Status :: 4 - Beta`. Kept because Keras 3 is a production multi-backend API; the classifier lags the project. +- stanza -- kept 2026-10-02 in the AI & ML audit as an NLP General challenger at version 1.15.0, PyPI classifier `Development Status :: 4 - Beta`. Kept because it is Stanford NLP's official library with releases through 2026-10-01; the classifier lags the project. +- kornia -- kept 2026-10-02 in the AI & ML audit as a Computer Vision General challenger at version 0.8.3, PyPI classifier `Development Status :: 4 - Beta`. Kept on 2,181,803 downloads/month on pypistats; the classifier lags the project. +- implicit -- kept 2026-10-02 in the AI & ML audit as a Recommender Systems obvious choice at version 0.7.3, PyPI classifier `Development Status :: 4 - Beta`. Kept on 234,833 downloads/month on pypistats and cp314 wheels; the classifier lags the project. +- statsforecast -- admitted 2026-10-02 in the AI & ML audit as a Time Series Forecasting obvious choice at version 2.1.1, PyPI classifier `Development Status :: 4 - Beta`. Admitted on 1,339,344 downloads/month on pypistats; the classifier lags the project. ### Challenger Exceptions @@ -41,6 +46,7 @@ Entries admitted as challengers without adoption-trajectory evidence, by maintai These entries sit past the 12-month activity requirement without an override. Each one is kept by editorial judgment: mature, stable, and no successor exists. +- annoy -- kept 2026-10-02 in the AI & ML audit, last push 2025-10-29. Unlike the others it has a successor, Spotify's own Voyager (2023), but kept by maintainer decision on 911,913 downloads/month on pypistats. - ftfy - itsdangerous - jieba