LlamaEmbeddingInput class final
One embedding request: ordered, interleaved text, image and audio parts.
Uses the same media sources as chat content. The list is copied; callers must keep media buffers unchanged until the embedding future settles. Embedding PCM input uses mono 16 kHz Float32 samples. Source/codec availability depends on the selected runtime. No chat template, task prefix, truncation or frame extraction is applied.
Video and chat-only parts throw LlamaUnsupportedException: a portable video sampling/timestamp contract has not been defined. An empty list throws LlamaArgumentException; a single empty text part retains raw text semantics.
Constructors
-
LlamaEmbeddingInput(List<
LlamaContentPart> parts) -
Creates an embedding input from existing public content parts.
factory
Properties
- hashCode → int
-
The hash code for this object.
no setterinherited
-
partKinds
→ List<
LlamaEmbeddingModality> -
The ordered part kinds, without text, paths, URLs or media buffers.
final
-
parts
→ List<
LlamaContentPart> -
An unmodifiable snapshot of the ordered content parts.
final
- runtimeType → Type
-
A representation of the runtime type of the object.
no setterinherited
Methods
-
noSuchMethod(
Invocation invocation) → dynamic -
Invoked when a nonexistent method or property is accessed.
inherited
-
toString(
) → String -
A string representation of this object.
inherited
Operators
-
operator ==(
Object other) → bool -
The equality operator.
inherited