<h1 align="center">dropout</h1>
Zero-config CLI to set up @dropout-ai/runtime in your project.
Invisible Node.js runtime for understanding behavioral impact of AI interactions.
Easy way of managing navigation in react based apps.
No description provided.
Graph Neural Network capabilities for Ruvector - Node.js bindings
ModelContextProtocol server for Ref
Download vwredql
WebAssembly bindings for ruvector-gnn - Graph Neural Network layers for browsers
HTML5 MPEG2-TS Stream Player
media player for_apocalypto
a easy to use fuzzy searcher
Generic drop/paste input and download/copy output component for web applications.
A hover-triggered popout dropdown component for React.
TypeScript + Rust Native machine learning library. Matrix ops, layers (Dense, Embedding, RNN, LSTM, GRU, MultiHeadAttention, etc), models (Sequential, Transformers), dan BPE tokenizer.
MCP server for Forge GPU kernel optimization — generate and optimize Triton/CUDA kernels on real H100/A100 GPUs from any AI coding agent
WebAssembly bindings for ruvector-gnn - Graph Neural Network layers for browsers
A functional library for JS
GPT model implemented with Tensorflow.js
WebAssembly bindings for RuVector GNN with tensor compression and differentiable search
Three.js-based 3D scene for geoscientific visualization: block models, sections, drillholes, surfaces, polylines, clip planes, raycasting. Expects THREE on globalThis or passed via opts.
ModelContextProtocol server for Ref
media player for_apocalypto
Ref MCP server enhanced with search optimization tips
Drop your objects out of main thread
A complete LSTM neural network library with training capabilities, multiple optimizers, and peephole variants.
A composable, deterministic text data pipeline for ML. Ingest, denoise, chunk, split, and sample multi-source corpora into reproducible training triplets.
Attention mechanisms for ruvector - geometric, graph, and sparse attention
Neural forecasting models implemented using ruv-FANN for time series forecasting
CNN feature extraction for image embeddings with SIMD acceleration
Vision Transformer models and building blocks for Rust using tch.
Neural network modules for Axonml ML framework
A simple and powreful configuration language, extending JSON with declarative and functional language features.
Comprehensive neural network library with dataset loading, batch normalization, 9 activation functions, 5 loss functions, multiple optimizers, regularization, and clean async-first API
Neural networks from scratch, in Rust.
Media file repair and recovery tools for OxiMedia
If your internet connection fails during a deploy, the deploy fails. And that means your site could stop functioning. Ouch. Deploy reliably and quickly from anywhere with slow or unreliable internet connections. It works by invoking the capistrano command from a remote server using `screen`. To ensure tasks run the same as if invoked locally, cap files are first copied to remote server via `scp`.
GRX brings PyTorch-style tensor operations to Ruby. Every arithmetic op, activation, and optimizer step runs through a native C library compiled with AVX2+FMA SIMD. Ruby is the interface — C does the work. Features: autograd, SGD/Adam optimizers, Linear/Sequential/Dropout/BatchNorm layers, MSE/BCE/CrossEntropy loss functions, Xavier and He weight init. Cross-platform: .so on Linux, .dylib on macOS, .dll on Windows.
ignis-dl is the deep-learning layer of the Ignis ecosystem: NN modules (Linear, Embedding, LayerNorm, RMSNorm, Dropout), optimizers (SGD/Adam/AdamW), losses, and a transformer stack (multi-head + grouped-query attention, RoPE, SwiGLU, KV cache) with HuggingFace weight loaders (GPT-2, Llama). Loads real GPT-2 and Llama-3.2 checkpoints and matches HuggingFace logits, and trains transformers from scratch — in Ruby, on native Windows. Installing this pulls the whole stack (ignis + ignis-autograd), so it also serves as the meta-gem.
GRYDRA v2.0 is a complete, modular Ruby library for building, training, and deploying neural networks. NEW in v2.0: - Complete modular architecture with 29 organized files - Keyword arguments API for better readability - Full implementations (no more "simplified" versions) - 8 loss functions (MSE, MAE, Huber, Cross-Entropy, Hinge, Log-Cosh, Quantile) - 5 optimizers (Adam, SGD, RMSprop, AdaGrad, AdamW) - 6 training callbacks (EarlyStopping, LearningRateScheduler, ReduceLROnPlateau, ModelCheckpoint, CSVLogger, ProgressBar) - Complete LSTM implementation with backpropagation - Complete 2D Convolutional layer with padding and stride - Real PCA with eigenvalue decomposition using Power Iteration - Multiple activation functions (Tanh, ReLU, Leaky ReLU, Sigmoid, Swish, GELU, Softmax) - Regularization (Dropout, L1, L2) - Weight initialization (Xavier, He) - Data normalization (Z-Score, Min-Max) - Comprehensive metrics (MSE, MAE, Accuracy, Precision, Recall, F1, Confusion Matrix, AUC-ROC) - Advanced training (mini-batch, early stopping, learning rate decay, validation split) - Cross-validation and hyperparameter search - Text processing (vocabulary, binary vectorization, TF-IDF) - Model persistence (save/load with Marshal) - Network visualization and gradient analysis - Simplified EasyNetwork interface - 100% backward compatibility with v1.x Perfect for machine learning projects, research, and education in Ruby.
No description provided.
No description provided.
No description provided.
No description provided.
No description provided.
No description provided.
No description provided.
No description provided.
No description provided.