memory/embedding-model.md

Embedding Model (2026-04-26)

Configuration

• Model: embeddinggemma-300M-Q8_0.gguf
• Service: systemd user service
• Endpoint: http://10.0.30.92:8080 (OpenAI-compatible, CLS pooling)

Status

• /health → ok
• /v1/embeddings → 200 OK

Usage

• Hindsight daemon needs hindsight_embed installed & configured