Minimal Policy Search Toolbox

Contextual Relative Entropy Policy Search for Reinforcement Learning in Python

Simple and self-contained TensorFlow implementation of reinforcement learning algorithms for continuous control, integrated with OpenAI Gym and other physics engines.

High-performance Rust metrics library — counters, gauges, timers, histograms, rate meters, labels with bounded cardinality, async/adaptive controls, system health, and built-in exporters for Prometheus, OpenMetrics, JSON, StatsD UDP, and OTLP/HTTP+JSON. Cross-platform, minimal deps.

Relative entropy policy search (REPS) using neural networks in OpenAI Gym environments

Error Forge is a flexible, high-performance Rust framework for rich, structured errors. Define typed errors with macros/derives, add contextual metadata & severities, wire sync/async hooks for logging/telemetry, and use recovery helpers (retry/fallback/backoff) that keep your code clean and resilient.

High-performance embedded vector database for Rust. HNSW, IVF, and flat indexes with first-class metadata filtering. In-process, persistent, zero-network similarity search.

Vector quantization (scalar, product, binary) for memory-efficient vector search - part of the iQDB family.

The JG Rust Collection. Libraries and tools built to REPS standards. High-performance, zero-cost abstractions, minimal allocations.

Do the important things daily, and increase incrementally over time.

Rust library for NOML (Nested Object Markup Language) - blazing-fast config parser with variable interpolation, imports, and schema validation.

Backend-agnostic Raft consensus as a pure state machine - bring your own network and storage.

Parallel index construction, incremental updates, and merging - part of the iQDB family.

MVCC transaction primitive for Rust databases - snapshot isolation, SSI, deadlock detection.

LSM-tree storage engine library for Rust - memtable, WAL, SSTables, compaction, bloom filters.

An embedded, document-oriented database for Rust - schemaless documents, secondary indexes, and ACID single-file storage, with zero network and zero external services.

Week 1 Javascript homework using arrays and objects with mocha/chai tests [js, basics, intro, arrays, objects, fundamentals, tests, mocha]

Disk persistence, WAL, and crash recovery for iQDB indexes - part of the iQDB family.

Outbound throttling and resilience for Rust - rate limiters, backoff, circuit breakers, adaptive control.

IVF clustered nearest-neighbor index for very large vector datasets - part of the iQDB family.