Distillation and model compression · MLD-0885
Per-tensor quantization
Using one quantization parameter set for an entire tensor.
Original reference page · detailed expansion pending. Original definitions, examples, notes and citations are preserved below. New examples are illustrative, not benchmark results.
Original example
Apply a simple shared weight scale.
Original caveat
Outlier channels can reduce precision for other values.
Context and aliases
Level: Intermediate. Stages: Training · Adaptation · Deployment.
Tags: DL · SLM · LLM.
Aliases: None recorded.
Related concepts
Concepts linking here
Original references
- Distilling the Knowledge in a Neural Network ↗ · Consulted
- PEFT: LoRA conceptual guide ↗ · Consulted
- PEFT: quantization guide ↗ · Consulted
- Original reference URL ↗