CompressionType

Enums · AiDotNet.Enums

Defines the types of model compression strategies available in the AiDotNet library.

For Beginners

Model compression reduces the size of AI models while trying to maintain their accuracy. Think of it like compressing a photo - you want a smaller file size but still a recognizable image. Different compression techniques work better for different scenarios and model types.

Fields

Field Summary
CompressionType DeepCompression Deep Compression combines pruning, quantization, and Huffman coding (Han et al.
CompressionType HuffmanEncoding Huffman encoding uses variable-length codes where frequent values get shorter codes.
CompressionType HybridClusteringPruning Combines weight clustering with pruning (removing unimportant weights).
CompressionType HybridClusteringQuantization Combines weight clustering with quantization for improved compression.
CompressionType HybridHuffmanClustering Combines Huffman encoding with weight clustering for maximum compression.
CompressionType LowRankFactorization Low-rank matrix factorization approximates weight matrices with lower-rank representations.
CompressionType None No compression applied to the model.
CompressionType ProductQuantization Product quantization divides weight vectors into sub-vectors and quantizes each separately.
CompressionType SparsePruning Sparse pruning removes small-magnitude weights, setting them to zero.
CompressionType WeightClustering Weight clustering groups similar weight values together and replaces them with cluster representatives.
int value__