You are viewing a single comment's thread from:

RE: LeoThread 2024-10-28 03:27

Quantization can be done using various algorithms, including:

  1. K-means quantization: Grouping the weights and activations into k clusters and assigning each cluster to a lower precision data type.
  2. Hierarchical quantization: Quantizing the weights and activations in a hierarchical manner, starting with the most important weights and activations.
  3. Nearest-neighbor quantization: Finding the nearest neighbor in a quantization table and assigning the weight or activation to that neighbor.