6.3.1

[DevinTDHa]

📢 Spark NLP 6.3.1: LLM Backend Upgrade and Document Processing Improvements

Spark NLP 6.3.1 focuses on strengthening distributed local LLM inference by upgrading the jsl-llamacpp backend to a newer llama.cpp release, while also delivering important improvements in document structure handling, ONNX model compatibility, metadata consistency.

🔥 Highlights

• Upgraded jsl-llamacpp backend to llama.cpp tag b7247, bringing upstream performance improvements, stability fixes, and expanded model compatibility for local LLM inference.
• Structural position metadata is now preserved for tables and images across document readers.
• Reader2Image integration with AutoGGUFVisionModel

🚀 New Features & Enhancements

LLM Backend Upgrade (llama.cpp)

The jsl-llamacpp backend has been upgraded to llama.cpp tag b7247.

This upgrade brings:

• Performance and memory improvements, bug fixes from upstream llama.cpp
• Better support for newer GGUF/GGML model variants
• Improved reliability for local, offline LLM inference within Spark NLP pipelines

This is the most impactful change in this release for users running LLM workloads locally or in restricted environments.

Structural Metadata for Document Readers

All supported document readers now store structural position metadata for tables and images. Newly added metadata fields include:

• domPath
• orderTableIndex
• orderImageIndex

These additions enable layout-aware downstream processing and more precise document understanding, especially for HTML and rich document formats.

Reader2Image Integration with AutoGGUFVisionModel

Reader2Image now supports interoperability with AutoGGUFVisionModel by introducing flexible handling of encoded vs. decoded image bytes and optional prompt output.

• Added a new boolean parameter useEncodedImageBytes to control whether the image result stores:
  • true: Encoded (compressed) file bytes for models like AutoGGUFVisionModel
  • false: Decoded pixel matrix for models such as Qwen2VLTransformer
• outputPromptColumn parameter to optionally output a separate prompt column containing text prompts as Spark NLP Annotations.

Platform Setup Documentation

Added official documentation and instructions for setting up and running Spark NLP on Microsoft Fabric, simplifying configuration and improving developer onboarding on the platform. You can see them at Spark NLP - Installation

🐛 Bug Fixes

• Sentence metadata is now consistently included in DocumentAssembler outputs when using LightPipeline.
• Fixed an issue where resetting the cache in ResourceDownloader could fail under certain conditions.
• Fixed a document parsing bug where some HTML elements (such as section titles or diagnosis entries) could appear multiple times in the parsed output.
• Improved robustness when loading ONNX BertEmbeddings models with non-standard output tensor names.

❤️ Community Support

• Slack – real-time discussion with the Spark NLP community and team
• GitHub – issue tracking, feature requests, and contributions
• Discussions – community ideas and showcases
• Medium – latest Spark NLP articles and tutorials
• YouTube – educational videos and demos

💻 Installation

Python

pip install spark-nlp==6.3.1

Spark Packages

CPU

spark-shell --packages com.johnsnowlabs.nlp:spark-nlp_2.12:6.3.1
pyspark --packages com.johnsnowlabs.nlp:spark-nlp_2.12:6.3.1

GPU

spark-shell --packages com.johnsnowlabs.nlp:spark-nlp-gpu_2.12:6.3.1
pyspark --packages com.johnsnowlabs.nlp:spark-nlp-gpu_2.12:6.3.1

Apple Silicon

spark-shell --packages com.johnsnowlabs.nlp:spark-nlp-silicon_2.12:6.3.1
pyspark --packages com.johnsnowlabs.nlp:spark-nlp-silicon_2.12:6.3.1

AArch64

spark-shell --packages com.johnsnowlabs.nlp:spark-nlp-aarch64_2.12:6.3.1
pyspark --packages com.johnsnowlabs.nlp:spark-nlp-aarch64_2.12:6.3.1

Maven

Supported on on Apache Spark 3.x.

spark-nlp

<dependency>
    <groupId>com.johnsnowlabs.nlp</groupId>
    <artifactId>spark-nlp_2.12</artifactId>
    <version>6.3.1</version>
</dependency>

spark-nlp-gpu

<dependency>
    <groupId>com.johnsnowlabs.nlp</groupId>
    <artifactId>spark-nlp-gpu_2.12</artifactId>
    <version>6.3.1</version>
</dependency>

spark-nlp-silicon

<dependency>
    <groupId>com.johnsnowlabs.nlp</groupId>
    <artifactId>spark-nlp-silicon_2.12</artifactId>
    <version>6.3.1</version>
</dependency>

spark-nlp-aarch64

<dependency>
    <groupId>com.johnsnowlabs.nlp</groupId>
    <artifactId>spark-nlp-aarch64_2.12</artifactId>
    <version>6.3.1</version>
</dependency>

FAT JARs

• CPU: https://s3.amazonaws.com/auxdata.johnsnowlabs.com/public/jars/spark-nlp-assembly-6.3.1.jar
• GPU: https://s3.amazonaws.com/auxdata.johnsnowlabs.com/public/jars/spark-nlp-gpu-assembly-6.3.1.jar
• Apple Silicon: https://s3.amazonaws.com/auxdata.johnsnowlabs.com/public/jars/spark-nlp-silicon-assembly-6.3.1.jar
• AArch64: https://s3.amazonaws.com/auxdata.johnsnowlabs.com/public/jars/spark-nlp-aarch64-assembly-6.3.1.jar

What's Changed

• [SPARKNLP-1318] Reader2Image Integration with AutoGGUFVisionModel #14705 by @danilojsl
• [SPARKNLP-1323] Adding instructions to setup MS Frabric on spark-nlp #14710 by @danilojsl
• [SPARKNLP-1324] handle nonstandard output tensor names better when loading onnx bert models #14711 by @AbdullahMubeenAnwar
• Bugfix when resetting cache in ResourceDownloader #14712 by @mehmetbutgul
• SPARKNLP-1322 Store Structural Position for Tables and Images #14713 by @danilojsl
• Addition sentence info into DocumentAssembler output in LightPipeline #14714 by @mehmetbutgul
• [SPARKNLP-1304] Upgrade jsl-llamacpp to llama.cpp b7247 #14719 by @DevinTDHa

Full Changelog: 6.3.0...6.3.1

---

This discussion was created from the release 6.3.1.