# Post-training quantization

> Post-training quantization: Quantizing a model after it has been fully trained. It is the most common approach, and the one application developers use.

Canonical: https://llmdojo.dev/glossary/ptq

Also called ptq.

Quantizing a model after it has been fully trained. It is the most common approach, and the one application developers use.

Learn it in the lesson: [What to quantize, and when](https://llmdojo.dev/learn/inference-optimization/precision-and-quantization/what-and-when) (Inference and Model Optimization · Precision and Quantization: Fewer Bytes per Weight).

## Interview questions about Post-training quantization

- [You must serve an open 13B model on one 80 GB accelerator and want more concurrent conversations. Which quantization would you use, and what does it actually buy?](https://llmdojo.dev/interview/q/you-must-serve-an-open-13b-model-on-one-80-gb-accelerator)
- [Explain quantization-aware training, knowledge distillation, eval mode and contrastive learning. How would you verify that each one improved a production model?](https://llmdojo.dev/interview/q/explain-quantization-aware-training-knowledge-distillation-eval-mode-and-contrastive-learning-how)

## Related
- [Quantization](https://llmdojo.dev/glossary/quantization)
- [BF16 (bfloat16)](https://llmdojo.dev/glossary/bf16)
- [Quantization-aware training](https://llmdojo.dev/glossary/qat)
- [Glossary of every term](https://llmdojo.dev/glossary)