Zum Hauptinhalt springen
Dieser Inhalt ist noch nicht in Ihrer Sprache verfügbar und wird auf Englisch angezeigt.

Phoenix

Skill Verifiziert Aktiv

Open-source AI observability platform for LLM tracing, evaluation, and monitoring. Use when debugging LLM applications with detailed traces, running evaluations on datasets, or monitoring production AI systems with real-time insights.

Zweck

To provide a self-hosted, open-source solution for observing and evaluating LLM applications, enabling detailed tracing, systematic dataset evaluations, and real-time production monitoring.

Funktionen

  • LLM tracing with OpenTelemetry
  • LLM evaluation framework
  • Dataset management for regression testing
  • Experiment pipelines for prompt/model comparison
  • Self-hosted observability without vendor lock-in

Anwendungsfälle

  • Debugging LLM application issues with detailed traces
  • Running systematic evaluations on datasets
  • Monitoring production LLM systems in real-time
  • Building experiment pipelines for prompt/model comparison

Nicht-Ziele

  • Managed observability platform with vendor lock-in
  • Deep learning experiment tracking focus
  • General ML lifecycle management
  • LangChain-first integration

Workflow

  1. Launch Phoenix server (notebook or command-line)
  2. Configure OpenTelemetry tracing
  3. Instrument LLM framework (e.g., OpenAI, LangChain)
  4. Generate and collect traces
  5. Run evaluations on datasets
  6. Monitor and analyze LLM application performance

Praktiken

  • Observability
  • LLM Tracing
  • Evaluation
  • Monitoring

Voraussetzungen

  • Python 3.8+
  • pip install arize-phoenix

Compliance

  • info:GDPRThe tool collects telemetry about LLM application usage, which may include personal data if not properly anonymized by the user. No specific sanitization for GDPR is detailed.
  • info:Telemetry opt-inThe extension collects telemetry for LLM application observability. While not explicitly stated as opt-in by default, the nature of collecting application-level data suggests it's part of the core functionality rather than optional tracking.

Execution

  • info:Pinned dependenciesWhile dependencies are declared, explicit pinning via lockfiles or version constraints in SKILL.md are not clearly presented, though standard Python packaging should manage this.

Installation

Zuerst Marketplace hinzufügen

/plugin marketplace add Orchestra-Research/AI-Research-SKILLs
/plugin install AI-Research-SKILLs@ai-research-skills

Qualitätspunktzahl

Verifiziert
95 /100
Analysiert 1 day ago

Vertrauenssignale

Letzter Commit17 days ago
Sterne8.3k
LizenzMIT
Status
Quellcode ansehen