機械学習 · GitHub
モデルの学習・評価・推論に使うフレームワークを探し、注目度を比較できます。
便利なGitHubプロジェクト
MLX-DLSS
iamwavecut
aihwkit
IBM
IBM Analog Hardware Acceleration Kit
- A library of device presets that are calibrated to real hardware data and based on models in the literature, along with a configuration that specifies a particular device and optimizer choice.
- A module for executing high-level use cases ("experiments"), such as neural network training with minimal code overhead.
detrex
IDEA-Research
frouros
IFCA-Advanced-Computing
Frouros: an open-source Python library for drift detection in machine learning systems.
AlphaGPT
imbue-bit
HEPML-LivingReview
iml-wg
Living Review of Machine Learning for Particle Physics
machine-learning-research
imteekay
✨ AI/ML Research
ml-quant-trading
initial-d
PyTorch ML quant research: 213 factors, bias correction, portfolio optimization, backtesting, and System-One/Jev-style decisions.
open-jev-typed-decision-engine
intikhab49
torchTT
ion-g-ion
Tensor-Train decomposition in pytorch
LiquidNetworks
Ipsedo
Liquid Time-constant Networks implementation with PyTorch
ClashRoyaleAi
itzik123

A fast, deterministic Clash Royale battle simulator in C++ with Python bindings, a recurrent PPO agent, lookahead search, and a computer-vision bridge to the real game.
- 132 cards and 41 Evolutions: troops, buildings, spells, spawners, Champions and their abilities, death effects, knockback, slows, stuns and rolling spells. Every card's stats are in docs/CARDS.md, generated from the engine.
- Real-game timing: 10 ticks per second, deploy time, King Tower activation, and 1×/2×/3× elixir phases on the real schedule.
EasyOCR
JaidedAI
bonsai
jax-ml
Minimal, lightweight JAX implementations of popular models.
NABLA-SciML
jdtoscano94
Physics Informed Machine Learning Tutorials (Pytorch and Jax)
YetAnother-WhosAtMyFeeder
Jellman86
🐦 AI-powered bird identification for Frigate NVR. Classifies birds from your feeder camera using TensorFlow Lite and displays them in a real-time dashboard.
- Identify locally: choose a model and validate it on your CPU, Intel GPU/NPU or NVIDIA GPU. RoPE ViT-B14 remains the default; coverage and hardware support vary by model.
- Review the evidence: explore seen visits and heard calls, inspect individual captures, correct species and count birds in a photograph. See Explorer.
DNABERT
jerryji1993
DNABERT: pre-trained Bidirectional Encoder Representations from Transformers model for DNA-language in genome
snntorch
jeshraghian
Deep and online learning with spiking neural networks in Python
Classify Kaggle San Francisco Crime Description into 39 classes. Build the model with CNN, RNN (GRU and LSTM) and Word Embeddings on Tensorflow.
Deep-learning-activity-recognition
jindongwang
A tutorial for using deep learning for activity recognition (Pytorch and Tensorflow)
Multimodal-Recommendation-Library
Jinfeng-Xu
A Continuously Updated Library for Advanced Models for Multimodal Recommendation
Any3R
Jingong-Chen

Any3R: Feedforward 3D Reconstruction from Any Camera Type — pinhole, fisheye beyond 180°, 360° panoramas, or any mix, in one forward pass
- One model for every central camera. Thirteen camera models are built in (pinhole, Brown–Conrady, Kannala–Brandt, equidistant, equisolid, stereographic, FOV, MEI, double sphere, EUCM, cylindrical, equirectangular and equirectangular windows), and any other central camera works through its ray field.
- Mixed cameras in one scene. Views from different cameras, with different sizes and fields of view, are reconstructed together into one consistent point cloud.




