Show HN: Spinf-serve – inference server for bulk scoring and decisions
An inference server was released that reads content once on a single GPU and scores all related questions in the same pass for bulk classification and decisions.
Key points
- •Performs bulk classification and decision inference on a single GPU.
- •Reads content once and scores every question about it in the same pass.
- •It is OpenAI-compatible and provided under the Apache 2.0 license.