Class DeepCompressionStats

Namespace

AiDotNet.ModelCompression

Assembly

AiDotNet.dll

Statistics about Deep Compression performance.

public class DeepCompressionStats

Inheritance

object

DeepCompressionStats

Inherited Members

object.Equals(object)

object.Equals(object, object)

object.GetHashCode()

object.GetType()

object.MemberwiseClone()

object.ReferenceEquals(object, object)

object.ToString()

Remarks

For Beginners: These statistics help you understand how well compression worked: - CompressionRatio: How much smaller the model is (e.g., 35x means 35 times smaller) - Sparsity: What fraction of weights are zero (e.g., 0.9 = 90% zeros) - BitsPerWeight: How many bits are used per non-zero weight

Properties

BitsPerWeight

Effective bits per weight after quantization.

public double BitsPerWeight { get; set; }

Property Value

double

CompressedSizeBytes

Compressed size in bytes after all three stages.

public long CompressedSizeBytes { get; set; }

Property Value

long

CompressionRatio

Overall compression ratio (original / compressed).

public double CompressionRatio { get; set; }

Property Value

double

NumClusters

Number of clusters used for quantization.

public int NumClusters { get; set; }

Property Value

int

OriginalSizeBytes

Original size in bytes before compression.

public long OriginalSizeBytes { get; set; }

Property Value

long

PruningRatio

Compression ratio from pruning stage alone.

public double PruningRatio { get; }

Property Value

double

QuantizationRatio

Compression ratio from quantization stage alone.

public double QuantizationRatio { get; }

Property Value

double

Sparsity

Sparsity achieved by pruning (fraction of zeros).

public double Sparsity { get; set; }

Property Value

double