TorchLean API

Docs HomeInstallationGuideExamplesGraphsTools

FloatLib.Floats.Formats.Posit.Configured.Conversion.Instances

Configured posit conversion instances #

Every configured posit supports exact-rational quantization under explicit infinity and exceptional-value policies.

source
@[instance_reducible]
instance FloatLib.Floats.ExecFloat.Posit.Conversion.quantizer {format : Formats.Posit.Format} {plan : Formats.Posit.Configured.StoragePlan format} {code : Type} [ModelCodec plan (Formats.Posit.Model format) code] :
Quantizer (ExecFloat (Formats.Posit.Configured.Family format code plan)) ℚ

Every configured posit supports exact-rational quantization under explicit special policy.

source
@[instance_reducible]
instance FloatLib.Floats.ExecFloat.Posit.Conversion.defaultQuantizer {format : Formats.Posit.Format} {plan : Formats.Posit.Configured.StoragePlan format} {code : Type} [ModelCodec plan (Formats.Posit.Model format) code] :
DefaultQuantizer (ExecFloat (Formats.Posit.Configured.Family format code plan)) ℚ

Context-free conversion uses the Posit Standard (2022), §6.5 infinity and NaN mapping to NaR.