Uses of Enum Class
ai.beestream.jitnn.DType
Packages that use DType
Package
Description
The jitNN API:
Tensor (n-dimensional arrays, lazy on TornadoVM
devices), reverse-mode differentiation (Tensor.backward(), Grad),
the functional operators of F, linear algebra (Linalg),
Fourier transforms (Fft), sparse matrices (SparseMatrix),
mixed precision (Autocast) and user-defined operators
(CustomOp).The GGUF reader (llama.cpp's model format):
GgufFile gives a file's
metadata and its tensors, memory-mapped without a copy, in their quantised blocks or as F32, F16
and BF16.Checkpoints: named tensors and values (
StateDict) saved and loaded in
Hugging Face's safetensors format (SafeTensors).-
Uses of DType in ai.beestream.jitnn
Subclasses with type arguments of type DType in ai.beestream.jitnnModifier and TypeClassDescriptionenumElement type of a tensor: FP32 data, INT32 labels and indices, and the half-precision storage types for mixed precision, IEEE FP16 and bfloat16 (the top 16 bits of an FP32), and Q8_0, GGUF's 8-bit blocks (per 32 consecutive values an FP16 scale and 32 signed bytes, 34 bytes), and the OCP 8-bit floats FP8 E4M3 and E5M2, one byte per value, stored without their scale (seeFp8).Methods in ai.beestream.jitnn that return DTypeModifier and TypeMethodDescriptionstatic DTypeAutocast.current()The active autocast type, or null.Tensor.dtype()The element type of this tensor.static DTypeReturns the enum constant of this class with the specified name.static DType[]DType.values()Returns an array containing the constants of this enum class, in the order they are declared.Methods in ai.beestream.jitnn with parameters of type DTypeModifier and TypeMethodDescriptionstatic Grad.ScopeEnables autocast todtype(F16 or BF16) until the returned scope is closed.The inputs' element types, in order (null accepts any type).CustomOp.Builder.output(CustomOp.OutputShape shape, DType dtype) Adds an output of the given shape and element type.static TensorTensor.persistent(uk.ac.manchester.tornado.api.types.arrays.TornadoNativeArray array, DType dtype, int... dims) A persistent tensor (a weight that stays on the device) over an existing native array, without copying it: for example an array that wraps a memory-mapped file.static TensorA weight in one of GGUF's block-quantised formats (Q8_0, Q4_0, Q4_1, Q4_K, Q5_K, Q6_K, Q5_0, Q5_1, Q2_K, Q3_K) from its raw blocks, as a GGUF file stores them: persistent, so it stays on the device.static TensorAn output tensor for an operator; its storage is allocated when the operator runs.This tensor converted todtype: between FP32, FP16 and BF16 (round to nearest even; differentiable), or between FP32 and INT32 (truncating toward zero; no gradient). -
Uses of DType in ai.beestream.jitnn.gguf
Methods in ai.beestream.jitnn.gguf that return DType -
Uses of DType in ai.beestream.jitnn.io
Methods in ai.beestream.jitnn.io that return DTypeModifier and TypeMethodDescriptionSafeTensors.Entry.dtype()Returns the value of thedtyperecord component.SafeTensors.Source.dtype()Returns the value of thedtyperecord component.The dtype of tensorname.Methods in ai.beestream.jitnn.io with parameters of type DTypeModifier and TypeMethodDescriptionstatic StringThe safetensors name ofdtype; GGUF block formats have none.Constructors in ai.beestream.jitnn.io with parameters of type DType