core concepts
Inference
The process of using a trained model to make predictions or generate outputs on new, unseen data. Inference is the production use of a model, as opposed to training, and its speed and cost are critical factors for deployment.
In practice
When you ask ChatGPT a question, the model performs inference to generate your response.
In the index
Tools that mention Inference
- Hugging FaceTrust 88
- SeldonTrust 87
- GroqTrust 84
- BentoMLTrust 76
- FeastTrust 74
- Fireworks AITrust 70
- Together AITrust 70
- BeamTrust 63
Matched on each tool’s own description and feature list, highest trust score first.