Package org.apache.tika.inference
package org.apache.tika.inference
-
ClassDescriptionDeprecated.A content chunk with multimodal locators and an optional embedding vector.Serializes and deserializes a list of
Chunkobjects to/from JSON.Where the chunks produced for a unit land and how they name where they came from.An engine that turns images into vectors, many per request.Theembedtask.Deprecated.since 4.1.0, removed in 4.2.0 withOpenAIImageEmbeddingParser.Runtime-only config that prevents modification of security-sensitive and cost-sensitive fields (baseUrl,apiKey,model) at parse time.Configuration for the inference metadata filters.Runtime-only config that prevents modification of security-sensitive and cost-sensitive fields (baseUrl,apiKey,model) at parse time.Deprecated.since 4.1.0, removed in 4.2.0; seeAbstractEmbeddingFilter.Splits markdown text into chunks that respect structural boundaries.One audio or video item for an engine: the bytes, their type, and which channel they are.An OpenAI-compatible embeddings endpoint (POST /v1/embeddings), named in"engines".Deprecated.since 4.1.0, removed in 4.2.0; seeAbstractEmbeddingFilter.Deprecated.since 4.1.0, removed in 4.2.0.Serializes and deserializes float vectors as base64-encoded big-endian float32 byte arrays.
"engines"and bind it with aTEXT"inference"binding, which embeds the whole document tree in batched requests.