# Spark NLP: State-of-the-Art Natural Language Processing
<p align="center">
<a href="https://github.com/JohnSnowLabs/spark-nlp/actions" alt="build">
<img src="https://github.com/JohnSnowLabs/spark-nlp/workflows/build/badge.svg" /></a>
<a href="https://github.com/JohnSnowLabs/spark-nlp/releases" alt="Current Release Version">
<img src="https://img.shields.io/github/v/release/JohnSnowLabs/spark-nlp.svg?style=flat-square&logo=github" /></a>
<a href="https://search.maven.org/artifact/com.johnsnowlabs.nlp/spark-nlp_2.12" alt="Maven Central">
<img src="https://maven-badges.herokuapp.com/maven-central/com.johnsnowlabs.nlp/spark-nlp_2.12/badge.svg" /></a>
<a href="https://badge.fury.io/py/spark-nlp" alt="PyPI version">
<img src="https://badge.fury.io/py/spark-nlp.svg" /></a>
<a href="https://anaconda.org/JohnSnowLabs/spark-nlp" alt="Anaconda-Cloud">
<img src="https://anaconda.org/johnsnowlabs/spark-nlp/badges/version.svg" /></a>
<a href="https://github.com/JohnSnowLabs/spark-nlp/blob/master/LICENSE" alt="License">
<img src="https://img.shields.io/badge/License-Apache%202.0-blue.svg" /></a>
<a href="https://pypi.org/project/spark-nlp/" alt="PyPi downloads">
<img src="https://static.pepy.tech/personalized-badge/spark-nlp?period=total&units=international_system&left_color=grey&right_color=orange&left_text=pip%20downloads" /></a>
</p>
Spark NLP is a state-of-the-art Natural Language Processing library built on top of Apache Spark. It provides **simple**, **performant** & **accurate** NLP annotations for machine learning pipelines that **scale** easily in a distributed environment.
Spark NLP comes with **11000+** pretrained **pipelines** and **models** in more than **200+** languages.
It also offers tasks such as **Tokenization**, **Word Segmentation**, **Part-of-Speech Tagging**, Word and Sentence **Embeddings**, **Named Entity Recognition**, **Dependency Parsing**, **Spell Checking**, **Text Classification**, **Sentiment Analysis**, **Token Classification**, **Machine Translation** (+180 languages), **Summarization**, **Question Answering**, **Table Question Answering**, **Text Generation**, **Image Classification**, **Automatic Speech Recognition**, and many more [NLP tasks](#features).
**Spark NLP** is the only open-source NLP library in **production** that offers state-of-the-art transformers such as **BERT**, **CamemBERT**, **ALBERT**, **ELECTRA**, **XLNet**, **DistilBERT**, **RoBERTa**, **DeBERTa**, **XLM-RoBERTa**, **Longformer**, **ELMO**, **Universal Sentence Encoder**, **Google T5**, **MarianMT**, **GPT2**, and **Vision Transformers (ViT)** not only to **Python** and **R**, but also to **JVM** ecosystem (**Java**, **Scala**, and **Kotlin**) at **scale** by extending **Apache Spark** natively.
## Project's website
Take a look at our official Spark NLP page: [http://nlp.johnsnowlabs.com/](http://nlp.johnsnowlabs.com/) for user documentation and examples
## Community support
- [Slack](https://www.johnsnowlabs.com/slack-redirect/) For live discussion with the Spark NLP community and the team
- [GitHub](https://github.com/JohnSnowLabs/spark-nlp) Bug reports, feature requests, and contributions
- [Discussions](https://github.com/JohnSnowLabs/spark-nlp/discussions) Engage with other community members, share ideas, and show off how you use Spark NLP!
- [Medium](https://medium.com/spark-nlp) Spark NLP articles
- [YouTube](https://www.youtube.com/channel/UCmFOjlpYEhxf_wJUDuz6xxQ/videos) Spark NLP video tutorials
## Table of contents
- [Features](#features)
- [Requirements](#requirements)
- [Quick Start](#quick-start)
- [Apache Spark Support](#apache-spark-support)
- [Scala & Python Support](#scala-and-python-support)
- [Databricks Support](#databricks-support)
- [EMR Support](#emr-support)
- [Using Spark NLP](#usage)
- [Packages Cheatsheet](#packages-cheatsheet)
- [Spark Packages](#spark-packages)
- [Scala](#scala)
- [Maven](#maven)
- [SBT](#sbt)
- [Python](#python)
- [Pip/Conda](#pipconda)
- [Compiled JARs](#compiled-jars)
- [Apache Zeppelin](#apache-zeppelin)
- [Jupyter Notebook](#jupyter-notebook-python)
- [Google Colab Notebook](#google-colab-notebook)
- [Kaggle Kernel](#kaggle-kernel)
- [Databricks Cluster](#databricks-cluster)
- [EMR Cluster](#emr-cluster)
- [GCP Dataproc](#gcp-dataproc)
- [Spark NLP Configuration](#spark-nlp-configuration)
- [Pipelines & Models](#pipelines-and-models)
- [Pipelines](#pipelines)
- [Models](#models)
- [Offline](#offline)
- [Examples](#examples)
- [FAQ](#faq)
- [Citation](#citation)
- [Contributing](#contributing)
## Features
- Tokenization
- Trainable Word Segmentation
- Stop Words Removal
- Token Normalizer
- Document Normalizer
- Stemmer
- Lemmatizer
- NGrams
- Regex Matching
- Text Matching
- Chunking
- Date Matcher
- Sentence Detector
- Deep Sentence Detector (Deep learning)
- Dependency parsing (Labeled/unlabeled)
- SpanBertCorefModel (Coreference Resolution)
- Part-of-speech tagging
- Sentiment Detection (ML models)
- Spell Checker (ML and DL models)
- Word Embeddings (GloVe and Word2Vec)
- Doc2Vec (based on Word2Vec)
- BERT Embeddings (TF Hub & HuggingFace models)
- DistilBERT Embeddings (HuggingFace models)
- CamemBERT Embeddings (HuggingFace models)
- RoBERTa Embeddings (HuggingFace models)
- DeBERTa Embeddings (HuggingFace v2 & v3 models)
- XLM-RoBERTa Embeddings (HuggingFace models)
- Longformer Embeddings (HuggingFace models)
- ALBERT Embeddings (TF Hub & HuggingFace models)
- XLNet Embeddings
- ELMO Embeddings (TF Hub models)
- Universal Sentence Encoder (TF Hub models)
- BERT Sentence Embeddings (TF Hub & HuggingFace models)
- RoBerta Sentence Embeddings (HuggingFace models)
- XLM-RoBerta Sentence Embeddings (HuggingFace models)
- Sentence Embeddings
- Chunk Embeddings
- Unsupervised keywords extraction
- Language Detection & Identification (up to 375 languages)
- Multi-class Sentiment analysis (Deep learning)
- Multi-label Sentiment analysis (Deep learning)
- Multi-class Text Classification (Deep learning)
- BERT for Token & Sequence Classification
- DistilBERT for Token & Sequence Classification
- CamemBERT for Token & Sequence Classification
- ALBERT for Token & Sequence Classification
- RoBERTa for Token & Sequence Classification
- DeBERTa for Token & Sequence Classification
- XLM-RoBERTa for Token & Sequence Classification
- XLNet for Token & Sequence Classification
- Longformer for Token & Sequence Classification
- BERT for Token & Sequence Classification
- DistilBERT for Question Answering
- ALBERT for Question Answering
- RoBERTa for Question Answering
- DeBERTa for Question Answering
- XLM-RoBERTa for Question Answering
- Longformer for Question Answering
- Table Question Answering (TAPAS)
- Neural Machine Translation (MarianMT)
- Text-To-Text Transfer Transformer (Google T5)
- Generative Pre-trained Transformer 2 (OpenAI GPT2)
- Vision Transformer (ViT)
- Automatic Speech Recognition (Wav2Vec2)
- Named entity recognition (Deep learning)
- Easy TensorFlow integration
- GPU Support
- Full integration with Spark ML functions
- +8500 pre-trained models in +200 languages!
- +3200 pre-trained pipelines in +200 languages!
- Multi-lingual NER models: Arabic, Bengali, Chinese, Danish, Dutch, English, Finnish, French, German, Hebrew, Italian, Japanese, Korean, Norwegian, Persian, Polish, Portuguese, Russian, Spanish, Swedish, Urdu, and more.
## Requirements
To use Spark NLP you need the following requirements:
- Java 8 and 11
- Apache Spark 3.3.x, 3.2.x, 3.1.x, 3.0.x
**GPU (optional):**
Spark NLP 4.2.7 is built with TensorFlow 2.7.1 and the following NVIDIA® software are only required for GPU support:
- NVIDIA® GPU drivers version 450.80.02 or higher
- CUDA® Toolkit 11.2
- cuDNN SDK 8.1.0
## Quick Start
This is a quick example of how to use Spark NLP pre-trained pipeline in Python and PySpark:
```sh
$ java -version
# should be Java 8 or 11 (Oracle or OpenJDK)
$ conda create -n sparknlp python=3.7 -y
$ conda activate sparknlp
# spark-nlp by default is based on pyspark 3.x
$ pip install spark-nlp==4.2.7 pyspark==3.2.3
```
In Python console or Jupyter `Python3` kernel:
```python
# Import Spark NLP
from sparknlp.base import *
from sparknlp.annotator import *
from sparknlp.pretrained import PretrainedPipeline
import sparknlp
# Start SparkSession with Spark NLP
# start() functions has 3 parameters: gpu, m1, and memory
# sparknlp.start(gpu=True) will start the session with GPU support
# sparknlp.start(m1=True) will start the session with macOS M1 support
# sparknlp.start(memory="16G") to change the default driver memory in SparkSession
spark = sparknlp.start()
# Download a pre-trained pipeline
pipeline = PretrainedPipeline('explain_document_dl', lang='en')
# Your testing dataset
text = """
The Mona Lisa is a 16th century oil painting created by Leonardo.
It's held at the Louvre in Paris.
"""
# Annotate your testing dataset
result = pipeline.annotate(text)
# What's in the pipeline
list(result.keys())
Output: ['entities', 'stem', 'checked', 'lemma', 'document',
'pos', 'token', 'ner', 'embeddings', 'sentence']
# Check the results
result['entities']
Output: ['Mona Lisa', 'Leonardo', 'Louvre', 'Paris']
```
For more examples, you can visit our dedicated [repository](https://github.com/JohnSnowLabs/spark-nlp-workshop) to showcase all Spark NLP use cases!
## Apache Spark Support
Spark NLP *4.2.7* has been built on top of Apache Spark 3.2 while fully supports Apache Spark 3.0.x, 3.1.x, 3.2.x, and 3.3.x:
| Spark NLP | Apache Spark 2.3.x | Apache Spark 2.4.x | Apache Spark 3.0.x | Apache Spark 3.1.x | Apache Spark 3.2.x | Apache Spark 3.3.x |
|-----------|--------------------|--------------------|--------------------|--------------------|--------------------|--------------------|
| 4.2.x | NO | NO | YES | YES | YES | YES |
| 4.1.x | NO | NO | YES | YES | YES | YES |
| 4.0.x | NO | NO | YES | YES | YES | YES |
| 3.4.x | YES | YES | YES | YES | Partially | N/A |
| 3.3.x | YES | YES | YES | YES | NO | NO |
| 3.2.x | YES | YES | YES | YES | NO | NO |
| 3.1.x | YES | YES | YES | YES | NO | NO |
| 3.0.x | YES | YES | YES | YES | NO | NO |
| 2.7.x | YES | YES | NO | NO | NO | NO |
NOTE: Starting 4.0.0 release, the default `spark-nlp` and `spark-nlp-gpu` packages are based on Scala 2.12.15 and Apache Spark 3.2 by default.
Find out more about `Spark NLP` versions from our [release notes](https://github.com/JohnSnowLabs/spark-nlp/releases).
## Scala and Python Support
| Spark NLP | Python 3.6 | Python 3.7 | Python 3.8 | Python 3.9 | Scala 2.11 | Scala 2.12 |
|-----------|------------|------------|------------|------------|------------|------------|
| 4.2.x | YES | YES | YES | YES | NO | YES |
| 4.1.x | YES | YES | YES | YES | NO | YES |
| 4.0.x | YES | YES | YES | YES | NO | YES |
| 3.4.x | YES | YES | YES | YES | YES | YES |
| 3.3.x | YES | YES | YES | NO | YES | YES |
| 3.2.x | YES | YES | YES | NO | YES | YES |
| 3.1.x | YES | YES | YES | NO | YES | YES |
| 3.0.x | YES | YES | YES | NO | YES | YES |
| 2.7.x | YES | YES | NO | NO | YES | NO |
## Databricks Support
Spark NLP 4.2.7 has been tested and is compatible with the following runtimes:
**CPU:**
- 7.3
- 7.3 ML
- 9.1
- 9.1 ML
- 10.1
- 10.1 ML
- 10.2
- 10.2 ML
- 10.3
- 10.3 ML
- 10.4
- 10.4 ML
- 10.5
- 10.5 ML
- 11.0
- 11.0 ML
- 11.1
- 11.1 ML
- 11.2
- 11.2 ML
- 11.3
- 11.3 ML
**GPU:**
- 9.1 ML & GPU
- 10.1 ML & GPU
- 10.2 ML & GPU
- 10.3 ML & GPU
- 10.4 ML & GPU
- 10.5 ML & GPU
- 11.0 ML & GPU
- 11.1 ML & GPU
- 11.2 ML & GPU
- 11.3 ML & GPU
NOTE: Spark NLP 4.0.x is based on TensorFlow 2.7.x which is compatible with CUDA11 and cuDNN 8.0.2. The only Databricks runtimes supporting CUDA 11 are 9.x and above as listed under GPU.
## EMR Support
Spark NLP 4.2.7 has been tested and is compatible with the following EMR releases:
- emr-6.2.0
- emr-6.3.0
- emr-6.3.1
- emr-6.4.0
- emr-6.5.0
- emr-6.6.0
- emr-6.7.0
Full list of [Amazon EMR 6.x releases](https://docs.aws.amazon.com/emr/latest/ReleaseGuide/emr-release-6x.html)
NOTE: The EMR 6.1.0 and 6.1.1 are not supported.
## Usage
## Packages Cheatsheet
This is a cheatsheet for corresponding Spark NLP Maven package to Apache Spark / PySpark major version:
| Apache Spark | Spark NLP on CPU | Spark NLP on GPU | Spark NLP on AArch64 (linux) | Spark NLP on M1 |
|------------------|--------------------|----------------------------|--------------------------------|---------------------------|
| 3.0/3.1/3.2/3.3 | `spark-nlp` | `spark-nlp-gpu` | `spark-nlp-aarch64` | `spark-nlp-m1` |
| Start Function | `sparknlp.start()` | `sparknlp.start(gpu=True)` | `sparknlp.start(aarch64=True)` | `sparknlp.start(m1=True)` |
NOTE: `M1` and `AArch64` are under `experimental` support. Access and support to these architectures are limited by the community and we had to build most of the dependencies by ourselves to make them compatible. We support these two architectures, however, they may not work in some environments.
## Spark Packages
### Command line (requires internet connection)
Spark NLP supports all major releases of Apache Spark 3.0.x, Apache Spark 3.1.x, Apache Spark 3.2.x, and Apache Spark 3.3.x.
#### Apache Spark 3.x (3.0.x, 3.1.x, 3.2.x, and 3.3.x - Scala 2.12)
```sh
# CPU
spark-shell --packages com.johnsnowlabs.nlp:spark-nlp_2.12:4.2.7
pyspark --packages com.johnsnowlabs.nlp:spark-nlp_2.12:4.2.7
spark-submit --packages com.johnsnowlabs.nlp:spark-nlp_2.12:4.2.7
```
The `spark-nlp` has been published to the [Maven Repository](https://mvnrepository.com/artifact/com.johnsnowlabs.nlp/spark-nlp).
```sh
# GPU
spark-shell --packages com.johnsnowlabs.nlp:spark-nlp-gpu_2.12:4.2.7
pyspark --packages com.johnsnowlabs.nlp:spark-nlp-gpu_2.12:4.2.7
spark-submit --packages com.johnsnowlabs.nlp:spark-nlp-gpu_2.12:4.2.7
```
The `spark-nlp-gpu` has been published to the [Maven Repository](https://mvnrepository.com/artifact/com.johnsnowlabs.nlp/spark-nlp-gpu).
```sh
# AArch64
spark-shell --packages com.johnsnowlabs.nlp:spark-nlp-aarch64_2.12:4.2.7
pyspark --packages com.johnsnowlabs.nlp:spark-nlp-aarch64_2.12:4.2.7
spark-submit --packages com.johnsnowlabs.nlp:spark-nlp-aarch64_2.12:4.2.7
```
The `spark-nlp-aarch64` has been published to the [Maven Repository](https://mvnrepository.com/artifact/com.johnsnowlabs.nlp/spark-nlp-aarch64).
```sh
# M1
spark-shell --packages com.johnsnowlabs.nlp:spark-nlp-m1_2.12:4.2.7
pyspark --packages com.johnsnowlabs.nlp:spark-nlp-m1_2.12:4.2.7
spark-submit --packages com.johnsnowlabs.nlp:spark-nlp-m1_2.12:4.2.7
```
The `spark-nlp-m1` has been published to the [Maven Repository](https://mvnrepository.com/artifact/com.johnsnowlabs.nlp/spark-nlp-m1).
**NOTE**: In case you are using large pretrained models like UniversalSentenceEncoder, you need to have the following set in your SparkSession:
```sh
spark-shell \
--driver-memory 16g \
--conf spark.kryoserializer.buffer.max=2000M \
--packages com.johnsnowlabs.nlp:spark-nlp_2.12:4.2.7
```
## Scala
Spark NLP supports Scala 2.12.15 if you are using Apache Spark 3.0.x, 3.1.x, 3.2.x, and 3.3.x versions. Our packages are deployed to Maven central. To add any of our packages as a dependency in your application you can follow these coordinates:
### Maven
**spark-nlp** on Apache Spark 3.0.x, 3.1.x, 3.2.x, and 3.3.x:
```xml
<!-- https://mvnrepository.com/artifact/com.johnsnowlabs.nlp/spark-nlp -->
<dependency>
<groupId>com.johnsnowlabs.nlp</groupId>
<artifactId>spark-nlp_2.12</artifactId>
<version>4.2.7</version>
</dependency>
```
**spark-nlp-gpu:**
```xml
<!-- https://mvnrepository.com/artifact/com.johnsnowlabs.nlp/spark-nlp-gpu -->
<dependency>
<groupId>com.johnsnowlabs.nlp</groupId>
<artifactId>spark-nlp-gpu_2.12</artifactId>
<version>4.2.7</version>
</dependency>
```
**spark-nlp-aarch64:**
```xml
<!-- https://mvnrepository.com/artifact/com.johnsnowlabs.nlp/spark-nlp-aarch64 -->
<dependency>
<groupId>com.johnsnowlabs.nlp</groupId>
<artifactId>spark-nlp-aarch64_2.12</artifactId>
<version>4.2.7</version>
</dependency>
```
**spark-nlp-m1:**
```xml
<!-- https://mvnrepository.com/artifact/com.johnsnowlabs.nlp/spark-nlp-m1 -->
<dependency>
<groupId>com.johnsnowlabs.nlp</groupId>
<artifactId>spark-nlp-m1_2.12</artifactId>
<version>4.2.7</version>
</dependency>
```
### SBT
**spark-nlp** on Apache Spark 3.0.x, 3.1.x, 3.2.x, and 3.3.x:
```sbtshell
// https://mvnrepository.com/artifact/com.johnsnowlabs.nlp/spark-nlp
libraryDependencies += "com.johnsnowlabs.nlp" %% "spark-nlp" % "4.2.7"
```
**spark-nlp-gpu:**
```sbtshell
// https://mvnrepository.com/artifact/com.johnsnowlabs.nlp/spark-nlp-gpu
libraryDependencies += "com.johnsnowlabs.nlp" %% "spark-nlp-gpu" % "4.2.7"
```
**spark-nlp-aarch64:**
```sbtshell
// https://mvnrepository.com/artifact/com.johnsnowlabs.nlp/spark-nlp-aarch64
libraryDependencies += "com.johnsnowlabs.nlp" %% "spark-nlp-aarch64" % "4.2.7"
```
**spark-nlp-m1:**
```sbtshell
// https://mvnrepository.com/artifact/com.johnsnowlabs.nlp/spark-nlp-m1
libraryDependencies += "com.johnsnowlabs.nlp" %% "spark-nlp-m1" % "4.2.7"
```
Maven Central: [https://mvnrepository.com/artifact/com.johnsnowlabs.nlp](https://mvnrepository.com/artifact/com.johnsnowlabs.nlp)
If you are interested, there is a simple SBT project for Spark NLP to guide you on how to use it in your projects [Spark NLP SBT Starter](https://github.com/maziyarpanahi/spark-nlp-starter)
## Python
Spark NLP supports Python 3.6.x and above depending on your major PySpark version.
### Python without explicit Pyspark installation
### Pip/Conda
If you installed pyspark through pip/conda, you can install `spark-nlp` through the same channel.
Pip:
```bash
pip install spark-nlp==4.2.7
```
Conda:
```bash
conda install -c johnsnowlabs spark-nlp
```
PyPI [spark-nlp package](https://pypi.org/project/spark-nlp/) / Anaconda [spark-nlp package](https://anaconda.org/JohnSnowLabs/spark-nlp)
Then you'll have to create a SparkSession either from Spark NLP:
```python
import sparknlp
spark = sparknlp.start()
```
or manually:
```python
spark = SparkSession.builder \
.appName("Spark NLP")\
.master("local[*]")\
.config("spark.driver.memory","16G")\
.config("spark.driver.maxResultSize", "0") \
.config("spark.kryoserializer.buffer.max", "2000M")\
.config("spark.jars.packages", "com.johnsnowlabs.nlp:spark-nlp_2.12:4.2.7")\
.getOrCreate()
```
If using local jars, you can use `spark.jars` instead for comma-delimited jar files. For cluster setups, of course, you'll have to put the jars in a reachable location for all driver and executor nodes.
**Quick example:**
```python
import sparknlp
from sparknlp.pretrained import PretrainedPipeline
#create or get Spark Session
spark = sparknlp.start()
sparknlp.version()
spark.version
#download, load and annotate a text by pre-trained pipeline
pipeline = PretrainedPipeline('recognize_entities_dl', 'en')
result = pipeline.annotate('The Mona Lisa is a 16th century oil painting created by Leonardo')
```
## Compiled JARs
### Build from source
#### spark-nlp
- FAT-JAR for CPU on Apache Spark 3.0.x, 3.1.x, 3.2.x, and 3.3.x
```bash
sbt assembly
```
- FAT-JAR for GPU on Apache Spark 3.0.x, 3.1.x, 3.2.x, and 3.3.x
```bash
sbt -Dis_gpu=true assembly
```
- FAT-JAR for M! on Apache Spark 3.0.x, 3.1.x, 3.2.x, and 3.3.x
```bash
sbt -Dis_m1=true assembly
```
### Using the jar manually
If for some reason you need to use the JAR, you can either download the Fat JARs provided here or download it from [Maven Central](https://mvnrepository.com/artifact/com.johnsnowlabs.nlp).
To add JARs to spark programs use the `--jars` option:
```sh
spark-shell --jars spark-nlp.jar
```
The preferred way to use the library when running spark programs is using the `--packages` option as specified in the `spark-packages` section.
## Apache Zeppelin
Use either one of the following options
- Add the following Maven Coordinates to the interpreter's library list
```bash
com.johnsnowlabs.nlp:spark-nlp_2.12:4.2.7
```
- Add a path to pre-built jar from [here](#compiled-jars) in the interpreter's library list making sure the jar is available to driver path
### Python in Zeppelin
Apart from the previous step, install the python module through pip
```bash
pip install spark-nlp==4.2.7
```
Or you can install `spark-nlp` from inside Zeppelin by using Conda:
```bash
python.conda install -c johnsnowlabs spark-nlp
```
Configure Zeppelin properly, use cells with %spark.pyspark or any interpreter name you chose.
Finally, in Zeppelin interpreter settings, make sure you set properly zeppelin.python to the python you want to use and install the pip library with (e.g. `python3`).
An alternative option would be to set `SPARK_SUBMIT_OPTIONS` (zeppelin-env.sh) and make sure `--packages` is there as shown earlier since it includes both scala and python side installation.
## Jupyter Notebook (Python)
**Recommended:**
The easiest way to get this done on Linux and macOS is to simply install `spark-nlp` and `pyspark` PyPI packages and launch the Jupyter from the same Python environment:
```sh
$ conda create -n sparknlp python=3.8 -y
$ conda activate sparknlp
# spark-nlp by default is based on pyspark 3.x
$ pip install spark-nlp==4.2.7 pyspark==3.2.3 jupyter
$ jupyter notebook
```
Then you can use `python3` kernel to run your code with creating SparkSession via `spark = sparknlp.start()`.
**Optional:**
If you are in different operating systems and require to make Jupyter Notebook run by using pyspark, you can follow these steps:
```bash
export SPARK_HOME=/path/to/your/spark/folder
export PYSPARK_PYTHON=python3
export PYSPARK_DRIVER_PYTHON=jupyter
export PYSPARK_DRIVER_PYTHON_OPTS=notebook
pyspark --packages com.johnsnowlabs.nlp:spark-nlp_2.12:4.2.7
```
Alternatively, you can mix in using `--jars` option for pyspark + `pip install spark-nlp`
If not using pyspark at all, you'll have to run the instructions pointed [here](#python-without-explicit-Pyspark-installation)
## Google Colab Notebook
Google Colab is perhaps the easiest way to get started with spark-nlp. It requires no installation or setup other than having a Google account.
Run the following code in Google Colab notebook and start using spark-nlp right away.
```sh
# This is only to setup PySpark and Spark NLP on Colab
!wget https://setup.johnsnowlabs.com/colab.sh -O - | bash
```
This script comes with the two options to define `pyspark` and `spark-nlp` versions via options:
```sh
# -p is for pyspark
# -s is for spark-nlp
# -g will enable upgrading libcudnn8 to 8.1.0 on Google Colab for GPU usage
# by default they are set to the latest
!wget https://setup.johnsnowlabs.com/colab.sh -O - | bash /dev/stdin -p 3.2.3 -s 4.2.7
```
[Spark NLP quick start on Google Colab](https://colab.research.google.com/github/JohnSnowLabs/spark-nlp-workshop/blob/master/jupyter/quick_start_google_colab.ipynb) is a live demo on Google Colab that performs named entity recognitions and sentiment analysis by using Spark NLP pretrained pipelines.
## Kaggle Kernel
Run the following code in Kaggle Kernel and start using spark-nlp right away.
```sh
# Let's setup Kaggle for Spark NLP and PySpark
!wget https://setup.johnsnowlabs.com/kaggle.sh -O - | bash
```
This script comes with the two options to define `pyspark` and `spark-nlp` versions via options:
```sh
# -p is for pyspark
# -s is for spark-nlp
# -g will enable upgrading libcudnn8 to 8.1.0 on Kaggle for GPU usage
# by default they are set to the latest
!wget https://setup.johnsnowlabs.com/colab.sh -O - | bash /dev/stdin -p 3.2.3 -s 4.2.7
```
[Spark NLP quick start on Kaggle Kernel](https://www.kaggle.com/mozzie/spark-nlp-named-entity-recognition) is a live demo on Kaggle Kernel that performs named entity recognitions by using Spark NLP pretrained pipeline.
## Databricks Cluster
1. Create a cluster if you don't have one already
2. On a new cluster or existing one you need to add the following to the `Advanced Options -> Spark` tab:
```bash
spark.kryoserializer.buffer.max 2000M
spark.serializer org.apache.spark.serializer.KryoSerializer
```
3. In `Libraries` tab inside your cluster you need to follow these steps:
3.1. Install New -> PyPI -> `spark-nlp==4.2.7` -> Install
3.2. Install New -> Maven -> Coordinates -> `com.johnsnowlabs.nlp:spark-nlp_2.12:4.2.7` -> Install
4. Now you can attach your notebook to the cluster and use Spark NLP!
NOTE: Databricks' runtimes support different Apache Spark major releases. Please make sure you choose the correct Spark NLP Maven package name (Maven Coordinate) for your runtime from our [Packages Cheatsheet](https://github.com/JohnSnowLabs/spark-nlp#packages-cheatsheet)
## EMR Cluster
To launch EMR clusters with Apache Spark/PySpark and Spark NLP correctly you need to have bootstrap and software configuration.
A sample of your bootstrap script
```.sh
#!/bin/bash
set -x -e
echo -e 'export PYSPARK_PYTHON=/usr/bin/python3
export HADOOP_CONF_DIR=/etc/hadoop/conf
export SPARK_JARS_DIR=/usr/lib/spark/jars
export SPARK_HOME=/usr/lib/spark' >> $HOME/.bashrc && source $HOME/.bashrc
sudo python3 -m pip install awscli boto spark-nlp
set +x
exit 0
```
A sample of your software configuration in JSON on S3 (must be public access):
```.json
[{
"Classification": "spark-env",
"Configurations": [{
"Classification": "export",
"Properties": {
"PYSPARK_PYTHON": "/usr/bin/python3"
}
}]
},
{
"Classification": "spark-defaults",
"Properties": {
"spark.yarn.stagingDir": "hdfs:///tmp",
"spark.yarn.preserve.staging.files": "true",
"spark.kryoserializer.buffer.max": "2000M",
"spark.serializer": "org.apache.spark.serializer.KryoSerializer",
"spark.driver.maxResultSize": "0",
"spark.jars.packages": "com.johnsnowlabs.nlp:spark-nlp_2.12:4.2.7"
}
}]
```
A sample of AWS CLI to launch EMR cluster:
```.sh
aws emr create-cluster \
--name "Spark NLP 4.2.7" \
--release-label emr-6.2.0 \
--applications Name=Hadoop Name=Spark Name=Hive \
--instance-type m4.4xlarge \
--instance-count 3 \
--use-default-roles \
--log-uri "s3://<S3_BUCKET>/" \
--bootstrap-actions Path=s3://<S3_BUCKET>/emr-bootstrap.sh,Name=custome \
--configurations "https://<public_access>/sparknlp-config.json" \
--ec2-attributes KeyName=<your_ssh_key>,EmrManagedMasterSecurityGroup=<security_group_with_ssh>,EmrManagedSlaveSecurityGroup=<security_group_with_ssh> \
--profile <aws_profile_credentials>
```
## GCP Dataproc
1. Create a cluster if you don't have one already as follows.
At gcloud shell:
```bash
gcloud services enable dataproc.googleapis.com \
compute.googleapis.com \
storage-component.googleapis.com \
bigquery.googleapis.com \
bigquerystorage.googleapis.com
```
```bash
REGION=<region>
```
```bash
BUCKET_NAME=<bucket_name>
gsutil mb -c standard -l ${REGION} gs://${BUCKET_NAME}
```
```bash
REGION=<region>
ZONE=<zone>
CLUSTER_NAME=<cluster_name>
BUCKET_NAME=<bucket_name>
```
You can set image-version, master-machine-type, worker-machine-type,
master-boot-disk-size, worker-boot-disk-size, num-workers as your needs.
If you use the previous image-version from 2.0, you should also add ANACONDA to optional-components.
And, you should enable gateway.
Don't forget to set the maven coordinates for the jar in properties.
```bash
gcloud dataproc clusters create ${CLUSTER_NAME} \
--region=${REGION} \
--zone=${ZONE} \
--image-version=2.0 \
--master-machine-type=n1-standard-4 \
--worker-machine-type=n1-standard-2 \
--master-boot-disk-size=128GB \
--worker-boot-disk-size=128GB \
--num-workers=2 \
--bucket=${BUCKET_NAME} \
--optional-components=JUPYTER \
--enable-component-gateway \
--metadata 'PIP_PACKAGES=spark-nlp spark-nlp-display google-cloud-bigquery google-cloud-storage' \
--initialization-actions gs://goog-dataproc-initialization-actions-${REGION}/python/pip-install.sh \
--properties spark:spark.serializer=org.apache.spark.serializer.KryoSerializer,spark:spark.driver.maxResultSize=0,spark:spark.kryoserializer.buffer.max=2000M,spark:spark.jars.packages=com.johnsnowlabs.nlp:spark-nlp_2.12:4.2.7
```
2. On an existing one, you need to install spark-nlp and spark-nlp-display packages from PyPI.
3. Now, you can attach your notebook to the cluster and use the Spark NLP!
## Spark NLP Configuration
You can change the following Spark NLP configurations via Spark Configuration:
| Property Name | Default | Meaning |
|--------------------------------------------------------|----------------------|------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| `spark.jsl.settings.pretrained.cache_folder` | `~/cache_pretrained` | The location to download and extract pretrained `Models` and `Pipelines`. By default, it will be in User's Home directory under `cache_pretrained` directory |
| `spark.jsl.settings.storage.cluster_tmp_dir` | `hadoop.tmp.dir` | The location to use on a cluster for temporarily files such as unpacking indexes for WordEmbeddings. By default, this locations is the location of `hadoop.tmp.dir` set via Hadoop configuration for Apache Spark. NOTE: `S3` is not supported and it must be local, HDFS, or DBFS |
| `spark.jsl.settings.annotator.log_folder` | `~/annotator_logs` | The location to save logs from annotators during training such as `NerDLApproach`, `ClassifierDLApproach`, `SentimentDLApproach`, `MultiClassifierDLApproach`, etc. By default, it will be in User's Home directory under `annotator_logs` directory |
| `spark.jsl.settings.aws.credentials.access_key_id` | `None` | Your AWS access key to use your S3 bucket to store log files of training models or access tensorflow graphs used in `NerDLApproach` |
| `spark.jsl.settings.aws.credentials.secret_access_key` | `None` | Your AWS secret access key to use your S3 bucket to store log files of training models or access tensorflow graphs used in `NerDLApproach` |
| `spark.jsl.settings.aws.credentials.session_token` | `None` | Your AWS MFA session token to use your S3 bucket to store log files of training models or access tensorflow graphs used in `NerDLApproach` |
| `spark.jsl.settings.aws.s3_bucket` | `None` | Your AWS S3 bucket to store log files of training models or access tensorflow graphs used in `NerDLApproach` |
| `spark.jsl.settings.aws.region` | `None` | Your AWS region to use your S3 bucket to store log files of training models or access tensorflow graphs used in `NerDLApproach` |
### How to set Spark NLP Configuration
**SparkSession:**
You can use `.config()` during SparkSession creation to set Spark NLP configurations.
```python
from pyspark.sql import SparkSession
spark = SparkSession.builder \
.master("local[*]") \
.config("spark.driver.memory", "16G") \
.config("spark.driver.maxResultSize", "0") \
.config("spark.serializer", "org.apache.spark.serializer.KryoSerializer") \
.config("spark.kryoserializer.buffer.max", "2000m") \
.config("spark.jsl.settings.pretrained.cache_folder", "sample_data/pretrained") \
.config("spark.jsl.settings.storage.cluster_tmp_dir", "sample_data/storage") \
.config("spark.jars.packages", "com.johnsnowlabs.nlp:spark-nlp_2.12:4.2.7") \
.getOrCreate()
```
**spark-shell:**
```sh
spark-shell \
--driver-memory 16g \
--conf spark.driver.maxResultSize=0 \
--conf spark.serializer=org.apache.spark.serializer.KryoSerializer
--conf spark.kryoserializer.buffer.max=2000M \
--conf spark.jsl.settings.pretrained.cache_folder="sample_data/pretrained" \
--conf spark.jsl.settings.storage.cluster_tmp_dir="sample_data/storage" \
--packages com.johnsnowlabs.nlp:spark-nlp_2.12:4.2.7
```
**pyspark:**
```sh
pyspark \
--driver-memory 16g \
--conf spark.driver.maxResultSize=0 \
--conf spark.serializer=org.apache.spark.serializer.KryoSerializer
--conf spark.kryoserializer.buffer.max=2000M \
--conf spark.jsl.settings.pretrained.cache_folder="sample_data/pretrained" \
--conf spark.jsl.settings.storage.cluster_tmp_dir="sample_data/storage" \
--packages com.johnsnowlabs.nlp:spark-nlp_2.12:4.2.7
```
**Databricks:**
On a new cluster or existing one you need to add the following to the `Advanced Options -> Spark` tab:
```bash
spark.kryoserializer.buffer.max 2000M
spark.serializer org.apache.spark.serializer.KryoSerializer
spark.jsl.settings.pretrained.cache_folder dbfs:/PATH_TO_CACHE
spark.jsl.settings.storage.cluster_tmp_dir dbfs:/PATH_TO_STORAGE
spark.jsl.settings.annotator.log_folder dbfs:/PATH_TO_LOGS
```
NOTE: If this is an existing cluster, after adding new configs or changing existing properties you need to restart it.
### S3 Integration
In Spark NLP we can define S3 locations to:
- Export log files of training models
- Store tensorflow graphs used in `NerDLApproach`
**Logging:**
To configure S3 path for logging while training models. We need to set up AWS credentials as well as an S3 path
```bash
spark.conf.set("spark.jsl.settings.annotator.log_folder", "s3://my/s3/path/logs")
spark.conf.set("spark.jsl.settings.aws.credentials.access_key_id", "MY_KEY_ID")
spark.conf.set("spark.jsl.settings.aws.credentials.secret_access_key", "MY_SECRET_ACCESS_KEY")
spark.conf.set("spark.jsl.settings.aws.s3_bucket", "my.bucket")
spark.conf.set("spark.jsl.settings.aws.region", "my-region")
```
Now you can check the log on your S3 path defined in *spark.jsl.settings.annotator.log_folder* property.
Make sure to use the prefix *s3://*, otherwise it will use the default configuration.
**Tensorflow Graphs:**
To reference S3 location for downloading graphs. We need to set up AWS credentials
```bash
spark.conf.set("spark.jsl.settings.aws.credentials.access_key_id", "MY_KEY_ID")
spark.conf.set("spark.jsl.settings.aws.credentials.secret_access_key", "MY_SECRET_ACCESS_KEY")
spark.conf.set("spark.jsl.settings.aws.region", "my-region")
```
**MFA Configuration:**
In case your AWS account is configured with MFA. You will need first to get temporal credentials and add session token to the configuration as shown in the examples below
For logging:
```bash
spark.conf.set("spark.jsl.settings.aws.credentials.session_token", "MY_TOKEN")
```
An example of a bash script that gets temporal AWS credentials can be found [here](https://github.com/JohnSnowLabs/spark-nlp/blob/master/scripts/aws_tmp_credentials.sh)
This script requires three arguments:
```bash
./aws_tmp_credentials.sh iam_user duration serial_number
```
## Pipelines and Models
### Pipelines
**Quick example:**
```scala
import com.johnsnowlabs.nlp.pretrained.PretrainedPipeline
import com.johnsnowlabs.nlp.SparkNLP
SparkNLP.version()
val testData = spark.createDataFrame(Seq(
(1, "Google has announced the release of a beta version of the popular TensorFlow machine learning library"),
(2, "Donald John Trump (born June 14, 1946) is the 45th and current president of the United States")
)).toDF("id", "text")
val pipeline = PretrainedPipeline("explain_document_dl", lang="en")
val annotation = pipeline.transform(testData)
annotation.show()
/*
import com.johnsnowlabs.nlp.pretrained.PretrainedPipeline
import com.johnsnowlabs.nlp.SparkNLP
2.5.0
testData: org.apache.spark.sql.DataFrame = [id: int, text: string]
pipeline: com.johnsnowlabs.nlp.pretrained.PretrainedPipeline = PretrainedPipeline(explain_document_dl,en,public/models)
annotation: org.apache.spark.sql.DataFrame = [id: int, text: string ... 10 more fields]
+---+--------------------+--------------------+--------------------+--------------------+--------------------+--------------------+--------------------+--------------------+--------------------+--------------------+--------------------+
| id| text| document| token| sentence| checked| lemma| stem| pos| embeddings| ner| entities|
+---+--------------------+--------------------+--------------------+--------------------+--------------------+--------------------+--------------------+--------------------+--------------------+--------------------+--------------------+
| 1|Google has announ...|[[document, 0, 10...|[[token, 0, 5, Go...|[[document, 0, 10...|[[token, 0, 5, Go...|[[token, 0, 5, Go...|[[token, 0, 5, go...|[[pos, 0, 5, NNP,...|[[word_embeddings...|[[named_entity, 0...|[[chunk, 0, 5, Go...|
| 2|The Paris metro w...|[[document, 0, 11...|[[token, 0, 2, Th...|[[document, 0, 11...|[[token, 0, 2, Th...|[[token, 0, 2, Th...|[[token, 0, 2, th...|[[pos, 0, 2, DT, ...|[[word_embeddings...|[[named_entity, 0...|[[chunk, 4, 8, Pa...|
+---+--------------------+--------------------+--------------------+--------------------+--------------------+--------------------+--------------------+--------------------+--------------------+--------------------+--------------------+
*/
annotation.select("entities.result").show(false)
/*
+----------------------------------+
|result |
+----------------------------------+
|[Google, TensorFlow] |
|[Donald John Trump, United States]|
+----------------------------------+
*/
```
#### Showing Available Pipelines
There are functions in Spark NLP that will list all the available Pipelines
of a particular language for you:
```scala
import com.johnsnowlabs.nlp.pretrained.ResourceDownloader
ResourceDownloader.showPublicPipelines(lang="en")
/*
+--------------------------------------------+------+---------+
| Pipeline | lang | version |
+--------------------------------------------+------+---------+
| dependency_parse | en | 2.0.2 |
| analyze_sentiment_ml | en | 2.0.2 |
| check_spelling | en | 2.1.0 |
| match_datetime | en | 2.1.0 |
...
| explain_document_ml | en | 3.1.3 |
+--------------------------------------------+------+---------+
*/
```
Or if we want to check for a particular version:
```scala
import com.johnsnowlabs.nlp.pretrained.ResourceDownloader
ResourceDownloader.showPublicPipelines(lang="en", version="3.1.0")
/*
+---------------------------------------+------+---------+
| Pipeline | lang | version |
+---------------------------------------+------+---------+
| dependency_parse | en | 2.0.2 |
...
| clean_slang | en | 3.0.0 |
| clean_pattern | en | 3.0.0 |
| check_spelling | en | 3.0.0 |
| dependency_parse | en | 3.0.0 |
+---------------------------------------+------+---------+
*/
```
#### Please check out our Models Hub for the full list of [pre-trained pipelines](https://nlp.johnsnowlabs.com/models) with examples, demos, benchmarks, and more
### Models
**Some selected languages:** `Afrikaans, Arabic, Armenian, Basque, Bengali, Breton, Bulgarian, Catalan, Czech, Dutch, English, Esperanto, Finnish, French, Galician, German, Greek, Hausa, Hebrew, Hindi, Hungarian, Indonesian, Irish, Italian, Japanese, Latin, Latvian, Marathi, Norwegian, Persian, Polish, Portuguese, Romanian, Russian, Slovak, Slovenian, Somali, Southern Sotho, Spanish, Swahili, Swedish, Tswana, Turkish, Ukrainian, Zulu`
**Quick online example:**
```python
# load NER model trained by deep learning approach and GloVe word embeddings
ner_dl = NerDLModel.pretrained('ner_dl')
# load NER model trained by deep learning approach and BERT word embeddings
ner_bert = NerDLModel.pretrained('ner_dl_bert')
```
```scala
// load French POS tagger model trained by Universal Dependencies
val french_pos = PerceptronModel.pretrained("pos_ud_gsd", lang="fr")
// load Italian LemmatizerModel
val italian_lemma = LemmatizerModel.pretrained("lemma_dxc", lang="it")
````
**Quick offline example:**
- Loading `PerceptronModel` annotator model inside Spark NLP Pipeline
```scala
val french_pos = PerceptronModel.load("/tmp/pos_ud_gsd_fr_2.0.2_2.4_1556531457346/")
.setInputCols("document", "token")
.setOutputCol("pos")
```
#### Showing Available Models
There are functions in Spark NLP that will list all the available Models
of a particular Annotator and language for you:
```scala
import com.johnsnowlabs.nlp.pretrained.ResourceDownloader
ResourceDownloader.showPublicModels(annotator="NerDLModel", lang="en")
/*
+---------------------------------------------+------+---------+
| Model | lang | version |
+---------------------------------------------+------+---------+
| onto_100 | en | 2.1.0 |
| onto_300 | en | 2.1.0 |
| ner_dl_bert | en | 2.2.0 |
| onto_100 | en | 2.4.0 |
| ner_conll_elmo | en | 3.2.2 |
+---------------------------------------------+------+---------+
*/
```
Or if we want to check for a particular version:
```scala
import com.johnsnowlabs.nlp.pretrained.ResourceDownloader
ResourceDownloader.showPublicModels(annotator="NerDLModel", lang="en", version="3.1.0")
/*
+----------------------------+------+---------+
| Model | lang | version |
+----------------------------+------+---------+
| onto_100 | en | 2.1.0 |
| ner_aspect_based_sentiment | en | 2.6.2 |
| ner_weibo_glove_840B_300d | en | 2.6.2 |
| nerdl_atis_840b_300d | en | 2.7.1 |
| nerdl_snips_100d | en | 2.7.3 |
+----------------------------+------+---------+
*/
```
And to see a list of available annotators, you can use:
```scala
import com.johnsnowlabs.nlp.pretrained.ResourceDownloader
ResourceDownloader.showAvailableAnnotators()
/*
AlbertEmbeddings
AlbertForTokenClassification
AssertionDLModel
...
XlmRoBertaSentenceEmbeddings
XlnetEmbeddings
*/
```
#### Please check out our Models Hub for the full list of [pre-trained models](https://nlp.johnsnowlabs.com/models) with examples, demo, benchmark, and more
## Offline
Spark NLP library and all the pre-trained models/pipelines can be used entirely offline with no access to the Internet. If you are behind a proxy or a firewall with no access to the Maven repository (to download packages) or/and no access to S3 (to automatically download models and pipelines), you can simply follow the instructions to have Spark NLP without any limitations offline:
- Instead of using the Maven package, you need to load our Fat JAR
- Instead of using PretrainedPipeline for pretrained pipelines or the `.pretrained()` function to download pretrained models, you will need to manually download your pipeline/model from [Models Hub](https://nlp.johnsnowlabs.com/models), extract it, and load it.
Example of `SparkSession` with Fat JAR to have Spark NLP offline:
```python
spark = SparkSession.builder \
.appName("Spark NLP")\
.master("local[*]")\
.config("spark.driver.memory","16G")\
.config("spark.driver.maxResultSize", "0") \
.config("spark.kryoserializer.buffer.max", "2000M")\
.config("spark.jars", "/tmp/spark-nlp-assembly-4.2.7.jar")\
.getOrCreate()
```
- You can download provided Fat JARs from each [release notes](https://github.com/JohnSnowLabs/spark-nlp/releases), please pay attention to pick the one that suits your environment depending on the device (CPU/GPU) and Apache Spark version (3.0.x, 3.1.x, 3.2.x, and 3.3.x)
- If you are local, you can load the Fat JAR from your local FileSystem, however, if you are in a cluster setup you need to put the Fat JAR on a distributed FileSystem such as HDFS, DBFS, S3, etc. (i.e., `hdfs:///tmp/spark-nlp-assembly-4.2.7.jar`)
Example of using pretrained Models and Pipelines in offline:
```python
# instead of using pretrained() for online:
# french_pos = PerceptronModel.pretrained("pos_ud_gsd", lang="fr")
# you download this model, extract it, and use .load
french_pos = PerceptronModel.load("/tmp/pos_ud_gsd_fr_2.0.2_2.4_1556531457346/")\
.setInputCols("document", "token")\
.setOutputCol("pos")
# example for pipelines
# instead of using PretrainedPipeline
# pipeline = PretrainedPipeline('explain_document_dl', lang='en')
# you download this pipeline, extract it, and use PipelineModel
PipelineModel.load("/tmp/explain_document_dl_en_2.0.2_2.4_1556530585689/")
```
- Since you are downloading and loading models/pipelines manually, this means Spark NLP is not downloading the most recent and compatible models/pipelines for you. Choosing the right model/pipeline is on you
- If you are local, you can load the model/pipeline from your local FileSystem, however, if you are in a cluster setup you need to put the model/pipeline on a distributed FileSystem such as HDFS, DBFS, S3, etc. (i.e., `hdfs:///tmp/explain_document_dl_en_2.0.2_2.4_1556530585689/`)
## Examples
Need more **examples**? Check out our dedicated [Spark NLP Showcase](https://github.com/JohnSnowLabs/spark-nlp-workshop) repository to showcase all Spark NLP use cases!
Also, don't forget to check [Spark NLP in Action](https://nlp.johnsnowlabs.com/demo) built by Streamlit.
### All examples: [spark-nlp-workshop](https://github.com/JohnSnowLabs/spark-nlp-workshop)
## FAQ
[Check our Articles and Videos page here](https://nlp.johnsnowlabs.com/learn)
## Citation
We have published a [paper](https://www.sciencedirect.com/science/article/pii/S2665963821000063) that you can cite for the Spark NLP library:
```bibtex
@article{KOCAMAN2021100058,
title = {Spark NLP: Natural language understanding at scale},
journal = {Software Impacts},
pages = {100058},
year = {2021},
issn = {2665-9638},
doi = {https://doi.org/10.1016/j.simpa.2021.100058},
url = {https://www.sciencedirect.com/science/article/pii/S2665963.2.300063},
author = {Veysel Kocaman and David Talby},
keywords = {Spark, Natural language processing, Deep learning, Tensorflow, Cluster},
abstract = {Spark NLP is a Natural Language Processing (NLP) library built on top of Apache Spark ML. It provides simple, performant & accurate NLP annotations for machine learning pipelines that can scale easily in a distributed environment. Spark NLP comes with 1100+ pretrained pipelines and models in more than 192+ languages. It supports nearly all the NLP tasks and modules that can be used seamlessly in a cluster. Downloaded more than 2.7 million times and experiencing 9x growth since January 2020, Spark NLP is used by 54% of healthcare organizations as the world’s most widely used NLP library in the enterprise.}
}
}
```
## Contributing
We appreciate any sort of contributions:
- ideas
- feedback
- documentation
- bug reports
- NLP training and testing corpora
- Development and testing
Clone the repo and submit your pull-requests! Or directly create issues in this repo.
## John Snow Labs
[http://johnsnowlabs.com](http://johnsnowlabs.com)
Raw data
{
"_id": null,
"home_page": "https://github.com/JohnSnowLabs/spark-nlp",
"name": "ckls-test-lib",
"maintainer": "",
"docs_url": null,
"requires_python": "",
"maintainer_email": "",
"keywords": "NLP spark vision speech deep learning transformer tensorflow BERT GPT-2 Wav2Vec2 ViT",
"author": "John Snow Labs",
"author_email": "",
"download_url": "https://files.pythonhosted.org/packages/48/b6/2604ee6b1c6dbdd88ed973bae50e4f37095906e0d1f8c18f2cc40c3e209a/ckls_test_lib-4.2.7.tar.gz",
"platform": null,
"description": "# Spark NLP: State-of-the-Art Natural Language Processing\n<p align=\"center\">\n <a href=\"https://github.com/JohnSnowLabs/spark-nlp/actions\" alt=\"build\">\n <img src=\"https://github.com/JohnSnowLabs/spark-nlp/workflows/build/badge.svg\" /></a>\n <a href=\"https://github.com/JohnSnowLabs/spark-nlp/releases\" alt=\"Current Release Version\">\n <img src=\"https://img.shields.io/github/v/release/JohnSnowLabs/spark-nlp.svg?style=flat-square&logo=github\" /></a>\n <a href=\"https://search.maven.org/artifact/com.johnsnowlabs.nlp/spark-nlp_2.12\" alt=\"Maven Central\">\n <img src=\"https://maven-badges.herokuapp.com/maven-central/com.johnsnowlabs.nlp/spark-nlp_2.12/badge.svg\" /></a>\n <a href=\"https://badge.fury.io/py/spark-nlp\" alt=\"PyPI version\">\n <img src=\"https://badge.fury.io/py/spark-nlp.svg\" /></a>\n <a href=\"https://anaconda.org/JohnSnowLabs/spark-nlp\" alt=\"Anaconda-Cloud\">\n <img src=\"https://anaconda.org/johnsnowlabs/spark-nlp/badges/version.svg\" /></a>\n <a href=\"https://github.com/JohnSnowLabs/spark-nlp/blob/master/LICENSE\" alt=\"License\">\n <img src=\"https://img.shields.io/badge/License-Apache%202.0-blue.svg\" /></a>\n <a href=\"https://pypi.org/project/spark-nlp/\" alt=\"PyPi downloads\">\n <img src=\"https://static.pepy.tech/personalized-badge/spark-nlp?period=total&units=international_system&left_color=grey&right_color=orange&left_text=pip%20downloads\" /></a>\n</p>\n\nSpark NLP is a state-of-the-art Natural Language Processing library built on top of Apache Spark. It provides **simple**, **performant** & **accurate** NLP annotations for machine learning pipelines that **scale** easily in a distributed environment.\nSpark NLP comes with **11000+** pretrained **pipelines** and **models** in more than **200+** languages.\nIt also offers tasks such as **Tokenization**, **Word Segmentation**, **Part-of-Speech Tagging**, Word and Sentence **Embeddings**, **Named Entity Recognition**, **Dependency Parsing**, **Spell Checking**, **Text Classification**, **Sentiment Analysis**, **Token Classification**, **Machine Translation** (+180 languages), **Summarization**, **Question Answering**, **Table Question Answering**, **Text Generation**, **Image Classification**, **Automatic Speech Recognition**, and many more [NLP tasks](#features).\n\n**Spark NLP** is the only open-source NLP library in **production** that offers state-of-the-art transformers such as **BERT**, **CamemBERT**, **ALBERT**, **ELECTRA**, **XLNet**, **DistilBERT**, **RoBERTa**, **DeBERTa**, **XLM-RoBERTa**, **Longformer**, **ELMO**, **Universal Sentence Encoder**, **Google T5**, **MarianMT**, **GPT2**, and **Vision Transformers (ViT)** not only to **Python** and **R**, but also to **JVM** ecosystem (**Java**, **Scala**, and **Kotlin**) at **scale** by extending **Apache Spark** natively.\n\n## Project's website\n\nTake a look at our official Spark NLP page: [http://nlp.johnsnowlabs.com/](http://nlp.johnsnowlabs.com/) for user documentation and examples\n\n## Community support\n\n- [Slack](https://www.johnsnowlabs.com/slack-redirect/) For live discussion with the Spark NLP community and the team\n- [GitHub](https://github.com/JohnSnowLabs/spark-nlp) Bug reports, feature requests, and contributions\n- [Discussions](https://github.com/JohnSnowLabs/spark-nlp/discussions) Engage with other community members, share ideas, and show off how you use Spark NLP!\n- [Medium](https://medium.com/spark-nlp) Spark NLP articles\n- [YouTube](https://www.youtube.com/channel/UCmFOjlpYEhxf_wJUDuz6xxQ/videos) Spark NLP video tutorials\n\n## Table of contents\n\n- [Features](#features)\n- [Requirements](#requirements)\n- [Quick Start](#quick-start)\n- [Apache Spark Support](#apache-spark-support)\n- [Scala & Python Support](#scala-and-python-support)\n- [Databricks Support](#databricks-support)\n- [EMR Support](#emr-support)\n- [Using Spark NLP](#usage)\n - [Packages Cheatsheet](#packages-cheatsheet)\n - [Spark Packages](#spark-packages)\n - [Scala](#scala)\n - [Maven](#maven)\n - [SBT](#sbt)\n - [Python](#python)\n - [Pip/Conda](#pipconda)\n - [Compiled JARs](#compiled-jars)\n - [Apache Zeppelin](#apache-zeppelin)\n - [Jupyter Notebook](#jupyter-notebook-python)\n - [Google Colab Notebook](#google-colab-notebook)\n - [Kaggle Kernel](#kaggle-kernel)\n - [Databricks Cluster](#databricks-cluster)\n - [EMR Cluster](#emr-cluster)\n - [GCP Dataproc](#gcp-dataproc)\n - [Spark NLP Configuration](#spark-nlp-configuration)\n- [Pipelines & Models](#pipelines-and-models)\n - [Pipelines](#pipelines)\n - [Models](#models)\n- [Offline](#offline)\n- [Examples](#examples)\n- [FAQ](#faq)\n- [Citation](#citation)\n- [Contributing](#contributing)\n\n## Features\n\n- Tokenization\n- Trainable Word Segmentation\n- Stop Words Removal\n- Token Normalizer\n- Document Normalizer\n- Stemmer\n- Lemmatizer\n- NGrams\n- Regex Matching\n- Text Matching\n- Chunking\n- Date Matcher\n- Sentence Detector\n- Deep Sentence Detector (Deep learning)\n- Dependency parsing (Labeled/unlabeled)\n- SpanBertCorefModel (Coreference Resolution)\n- Part-of-speech tagging\n- Sentiment Detection (ML models)\n- Spell Checker (ML and DL models)\n- Word Embeddings (GloVe and Word2Vec)\n- Doc2Vec (based on Word2Vec)\n- BERT Embeddings (TF Hub & HuggingFace models)\n- DistilBERT Embeddings (HuggingFace models)\n- CamemBERT Embeddings (HuggingFace models)\n- RoBERTa Embeddings (HuggingFace models)\n- DeBERTa Embeddings (HuggingFace v2 & v3 models)\n- XLM-RoBERTa Embeddings (HuggingFace models)\n- Longformer Embeddings (HuggingFace models)\n- ALBERT Embeddings (TF Hub & HuggingFace models)\n- XLNet Embeddings\n- ELMO Embeddings (TF Hub models)\n- Universal Sentence Encoder (TF Hub models)\n- BERT Sentence Embeddings (TF Hub & HuggingFace models)\n- RoBerta Sentence Embeddings (HuggingFace models)\n- XLM-RoBerta Sentence Embeddings (HuggingFace models)\n- Sentence Embeddings\n- Chunk Embeddings\n- Unsupervised keywords extraction\n- Language Detection & Identification (up to 375 languages)\n- Multi-class Sentiment analysis (Deep learning)\n- Multi-label Sentiment analysis (Deep learning)\n- Multi-class Text Classification (Deep learning)\n- BERT for Token & Sequence Classification\n- DistilBERT for Token & Sequence Classification\n- CamemBERT for Token & Sequence Classification\n- ALBERT for Token & Sequence Classification\n- RoBERTa for Token & Sequence Classification\n- DeBERTa for Token & Sequence Classification\n- XLM-RoBERTa for Token & Sequence Classification\n- XLNet for Token & Sequence Classification\n- Longformer for Token & Sequence Classification\n- BERT for Token & Sequence Classification\n- DistilBERT for Question Answering\n- ALBERT for Question Answering\n- RoBERTa for Question Answering\n- DeBERTa for Question Answering\n- XLM-RoBERTa for Question Answering\n- Longformer for Question Answering\n- Table Question Answering (TAPAS)\n- Neural Machine Translation (MarianMT)\n- Text-To-Text Transfer Transformer (Google T5)\n- Generative Pre-trained Transformer 2 (OpenAI GPT2)\n- Vision Transformer (ViT)\n- Automatic Speech Recognition (Wav2Vec2)\n- Named entity recognition (Deep learning)\n- Easy TensorFlow integration\n- GPU Support\n- Full integration with Spark ML functions\n- +8500 pre-trained models in +200 languages!\n- +3200 pre-trained pipelines in +200 languages!\n- Multi-lingual NER models: Arabic, Bengali, Chinese, Danish, Dutch, English, Finnish, French, German, Hebrew, Italian, Japanese, Korean, Norwegian, Persian, Polish, Portuguese, Russian, Spanish, Swedish, Urdu, and more.\n\n## Requirements\n\nTo use Spark NLP you need the following requirements:\n\n- Java 8 and 11\n- Apache Spark 3.3.x, 3.2.x, 3.1.x, 3.0.x\n\n**GPU (optional):**\n\nSpark NLP 4.2.7 is built with TensorFlow 2.7.1 and the following NVIDIA\u00ae software are only required for GPU support:\n\n- NVIDIA\u00ae GPU drivers version 450.80.02 or higher\n- CUDA\u00ae Toolkit 11.2\n- cuDNN SDK 8.1.0\n\n## Quick Start\n\nThis is a quick example of how to use Spark NLP pre-trained pipeline in Python and PySpark:\n\n```sh\n$ java -version\n# should be Java 8 or 11 (Oracle or OpenJDK)\n$ conda create -n sparknlp python=3.7 -y\n$ conda activate sparknlp\n# spark-nlp by default is based on pyspark 3.x\n$ pip install spark-nlp==4.2.7 pyspark==3.2.3\n```\n\nIn Python console or Jupyter `Python3` kernel:\n\n```python\n# Import Spark NLP\nfrom sparknlp.base import *\nfrom sparknlp.annotator import *\nfrom sparknlp.pretrained import PretrainedPipeline\nimport sparknlp\n\n# Start SparkSession with Spark NLP\n# start() functions has 3 parameters: gpu, m1, and memory\n# sparknlp.start(gpu=True) will start the session with GPU support\n# sparknlp.start(m1=True) will start the session with macOS M1 support\n# sparknlp.start(memory=\"16G\") to change the default driver memory in SparkSession\nspark = sparknlp.start()\n\n# Download a pre-trained pipeline\npipeline = PretrainedPipeline('explain_document_dl', lang='en')\n\n# Your testing dataset\ntext = \"\"\"\nThe Mona Lisa is a 16th century oil painting created by Leonardo.\nIt's held at the Louvre in Paris.\n\"\"\"\n\n# Annotate your testing dataset\nresult = pipeline.annotate(text)\n\n# What's in the pipeline\nlist(result.keys())\nOutput: ['entities', 'stem', 'checked', 'lemma', 'document',\n'pos', 'token', 'ner', 'embeddings', 'sentence']\n\n# Check the results\nresult['entities']\nOutput: ['Mona Lisa', 'Leonardo', 'Louvre', 'Paris']\n```\n\nFor more examples, you can visit our dedicated [repository](https://github.com/JohnSnowLabs/spark-nlp-workshop) to showcase all Spark NLP use cases!\n\n## Apache Spark Support\n\nSpark NLP *4.2.7* has been built on top of Apache Spark 3.2 while fully supports Apache Spark 3.0.x, 3.1.x, 3.2.x, and 3.3.x:\n\n| Spark NLP | Apache Spark 2.3.x | Apache Spark 2.4.x | Apache Spark 3.0.x | Apache Spark 3.1.x | Apache Spark 3.2.x | Apache Spark 3.3.x |\n|-----------|--------------------|--------------------|--------------------|--------------------|--------------------|--------------------|\n| 4.2.x | NO | NO | YES | YES | YES | YES |\n| 4.1.x | NO | NO | YES | YES | YES | YES |\n| 4.0.x | NO | NO | YES | YES | YES | YES |\n| 3.4.x | YES | YES | YES | YES | Partially | N/A |\n| 3.3.x | YES | YES | YES | YES | NO | NO |\n| 3.2.x | YES | YES | YES | YES | NO | NO |\n| 3.1.x | YES | YES | YES | YES | NO | NO |\n| 3.0.x | YES | YES | YES | YES | NO | NO |\n| 2.7.x | YES | YES | NO | NO | NO | NO |\n\nNOTE: Starting 4.0.0 release, the default `spark-nlp` and `spark-nlp-gpu` packages are based on Scala 2.12.15 and Apache Spark 3.2 by default.\n\nFind out more about `Spark NLP` versions from our [release notes](https://github.com/JohnSnowLabs/spark-nlp/releases).\n\n## Scala and Python Support\n\n| Spark NLP | Python 3.6 | Python 3.7 | Python 3.8 | Python 3.9 | Scala 2.11 | Scala 2.12 |\n|-----------|------------|------------|------------|------------|------------|------------|\n| 4.2.x | YES | YES | YES | YES | NO | YES |\n| 4.1.x | YES | YES | YES | YES | NO | YES |\n| 4.0.x | YES | YES | YES | YES | NO | YES |\n| 3.4.x | YES | YES | YES | YES | YES | YES |\n| 3.3.x | YES | YES | YES | NO | YES | YES |\n| 3.2.x | YES | YES | YES | NO | YES | YES |\n| 3.1.x | YES | YES | YES | NO | YES | YES |\n| 3.0.x | YES | YES | YES | NO | YES | YES |\n| 2.7.x | YES | YES | NO | NO | YES | NO |\n\n## Databricks Support\n\nSpark NLP 4.2.7 has been tested and is compatible with the following runtimes:\n\n**CPU:**\n\n- 7.3\n- 7.3 ML\n- 9.1\n- 9.1 ML\n- 10.1\n- 10.1 ML\n- 10.2\n- 10.2 ML\n- 10.3\n- 10.3 ML\n- 10.4\n- 10.4 ML\n- 10.5\n- 10.5 ML\n- 11.0\n- 11.0 ML\n- 11.1\n- 11.1 ML\n- 11.2\n- 11.2 ML\n- 11.3\n- 11.3 ML\n\n**GPU:**\n\n- 9.1 ML & GPU\n- 10.1 ML & GPU\n- 10.2 ML & GPU\n- 10.3 ML & GPU\n- 10.4 ML & GPU\n- 10.5 ML & GPU\n- 11.0 ML & GPU\n- 11.1 ML & GPU\n- 11.2 ML & GPU\n- 11.3 ML & GPU\n\nNOTE: Spark NLP 4.0.x is based on TensorFlow 2.7.x which is compatible with CUDA11 and cuDNN 8.0.2. The only Databricks runtimes supporting CUDA 11 are 9.x and above as listed under GPU.\n\n## EMR Support\n\nSpark NLP 4.2.7 has been tested and is compatible with the following EMR releases:\n\n- emr-6.2.0\n- emr-6.3.0\n- emr-6.3.1\n- emr-6.4.0\n- emr-6.5.0\n- emr-6.6.0\n- emr-6.7.0\n\nFull list of [Amazon EMR 6.x releases](https://docs.aws.amazon.com/emr/latest/ReleaseGuide/emr-release-6x.html)\n\nNOTE: The EMR 6.1.0 and 6.1.1 are not supported.\n\n## Usage\n\n## Packages Cheatsheet\n\nThis is a cheatsheet for corresponding Spark NLP Maven package to Apache Spark / PySpark major version:\n\n| Apache Spark | Spark NLP on CPU | Spark NLP on GPU | Spark NLP on AArch64 (linux) | Spark NLP on M1 |\n|------------------|--------------------|----------------------------|--------------------------------|---------------------------|\n| 3.0/3.1/3.2/3.3 | `spark-nlp` | `spark-nlp-gpu` | `spark-nlp-aarch64` | `spark-nlp-m1` |\n| Start Function | `sparknlp.start()` | `sparknlp.start(gpu=True)` | `sparknlp.start(aarch64=True)` | `sparknlp.start(m1=True)` |\n\nNOTE: `M1` and `AArch64` are under `experimental` support. Access and support to these architectures are limited by the community and we had to build most of the dependencies by ourselves to make them compatible. We support these two architectures, however, they may not work in some environments.\n\n## Spark Packages\n\n### Command line (requires internet connection)\n\nSpark NLP supports all major releases of Apache Spark 3.0.x, Apache Spark 3.1.x, Apache Spark 3.2.x, and Apache Spark 3.3.x.\n\n#### Apache Spark 3.x (3.0.x, 3.1.x, 3.2.x, and 3.3.x - Scala 2.12)\n\n```sh\n# CPU\n\nspark-shell --packages com.johnsnowlabs.nlp:spark-nlp_2.12:4.2.7\n\npyspark --packages com.johnsnowlabs.nlp:spark-nlp_2.12:4.2.7\n\nspark-submit --packages com.johnsnowlabs.nlp:spark-nlp_2.12:4.2.7\n```\n\nThe `spark-nlp` has been published to the [Maven Repository](https://mvnrepository.com/artifact/com.johnsnowlabs.nlp/spark-nlp).\n\n```sh\n# GPU\n\nspark-shell --packages com.johnsnowlabs.nlp:spark-nlp-gpu_2.12:4.2.7\n\npyspark --packages com.johnsnowlabs.nlp:spark-nlp-gpu_2.12:4.2.7\n\nspark-submit --packages com.johnsnowlabs.nlp:spark-nlp-gpu_2.12:4.2.7\n\n```\n\nThe `spark-nlp-gpu` has been published to the [Maven Repository](https://mvnrepository.com/artifact/com.johnsnowlabs.nlp/spark-nlp-gpu).\n\n```sh\n# AArch64\n\nspark-shell --packages com.johnsnowlabs.nlp:spark-nlp-aarch64_2.12:4.2.7\n\npyspark --packages com.johnsnowlabs.nlp:spark-nlp-aarch64_2.12:4.2.7\n\nspark-submit --packages com.johnsnowlabs.nlp:spark-nlp-aarch64_2.12:4.2.7\n\n```\n\nThe `spark-nlp-aarch64` has been published to the [Maven Repository](https://mvnrepository.com/artifact/com.johnsnowlabs.nlp/spark-nlp-aarch64).\n\n```sh\n# M1\n\nspark-shell --packages com.johnsnowlabs.nlp:spark-nlp-m1_2.12:4.2.7\n\npyspark --packages com.johnsnowlabs.nlp:spark-nlp-m1_2.12:4.2.7\n\nspark-submit --packages com.johnsnowlabs.nlp:spark-nlp-m1_2.12:4.2.7\n\n```\n\nThe `spark-nlp-m1` has been published to the [Maven Repository](https://mvnrepository.com/artifact/com.johnsnowlabs.nlp/spark-nlp-m1).\n\n**NOTE**: In case you are using large pretrained models like UniversalSentenceEncoder, you need to have the following set in your SparkSession:\n\n```sh\nspark-shell \\\n --driver-memory 16g \\\n --conf spark.kryoserializer.buffer.max=2000M \\\n --packages com.johnsnowlabs.nlp:spark-nlp_2.12:4.2.7\n```\n\n## Scala\n\nSpark NLP supports Scala 2.12.15 if you are using Apache Spark 3.0.x, 3.1.x, 3.2.x, and 3.3.x versions. Our packages are deployed to Maven central. To add any of our packages as a dependency in your application you can follow these coordinates:\n\n### Maven\n\n**spark-nlp** on Apache Spark 3.0.x, 3.1.x, 3.2.x, and 3.3.x:\n\n```xml\n<!-- https://mvnrepository.com/artifact/com.johnsnowlabs.nlp/spark-nlp -->\n<dependency>\n <groupId>com.johnsnowlabs.nlp</groupId>\n <artifactId>spark-nlp_2.12</artifactId>\n <version>4.2.7</version>\n</dependency>\n```\n\n**spark-nlp-gpu:**\n\n```xml\n<!-- https://mvnrepository.com/artifact/com.johnsnowlabs.nlp/spark-nlp-gpu -->\n<dependency>\n <groupId>com.johnsnowlabs.nlp</groupId>\n <artifactId>spark-nlp-gpu_2.12</artifactId>\n <version>4.2.7</version>\n</dependency>\n```\n\n**spark-nlp-aarch64:**\n\n```xml\n<!-- https://mvnrepository.com/artifact/com.johnsnowlabs.nlp/spark-nlp-aarch64 -->\n<dependency>\n <groupId>com.johnsnowlabs.nlp</groupId>\n <artifactId>spark-nlp-aarch64_2.12</artifactId>\n <version>4.2.7</version>\n</dependency>\n```\n\n**spark-nlp-m1:**\n\n```xml\n<!-- https://mvnrepository.com/artifact/com.johnsnowlabs.nlp/spark-nlp-m1 -->\n<dependency>\n <groupId>com.johnsnowlabs.nlp</groupId>\n <artifactId>spark-nlp-m1_2.12</artifactId>\n <version>4.2.7</version>\n</dependency>\n```\n\n### SBT\n\n**spark-nlp** on Apache Spark 3.0.x, 3.1.x, 3.2.x, and 3.3.x:\n\n```sbtshell\n// https://mvnrepository.com/artifact/com.johnsnowlabs.nlp/spark-nlp\nlibraryDependencies += \"com.johnsnowlabs.nlp\" %% \"spark-nlp\" % \"4.2.7\"\n```\n\n**spark-nlp-gpu:**\n\n```sbtshell\n// https://mvnrepository.com/artifact/com.johnsnowlabs.nlp/spark-nlp-gpu\nlibraryDependencies += \"com.johnsnowlabs.nlp\" %% \"spark-nlp-gpu\" % \"4.2.7\"\n```\n\n**spark-nlp-aarch64:**\n\n```sbtshell\n// https://mvnrepository.com/artifact/com.johnsnowlabs.nlp/spark-nlp-aarch64\nlibraryDependencies += \"com.johnsnowlabs.nlp\" %% \"spark-nlp-aarch64\" % \"4.2.7\"\n```\n\n**spark-nlp-m1:**\n\n```sbtshell\n// https://mvnrepository.com/artifact/com.johnsnowlabs.nlp/spark-nlp-m1\nlibraryDependencies += \"com.johnsnowlabs.nlp\" %% \"spark-nlp-m1\" % \"4.2.7\"\n```\n\nMaven Central: [https://mvnrepository.com/artifact/com.johnsnowlabs.nlp](https://mvnrepository.com/artifact/com.johnsnowlabs.nlp)\n\nIf you are interested, there is a simple SBT project for Spark NLP to guide you on how to use it in your projects [Spark NLP SBT Starter](https://github.com/maziyarpanahi/spark-nlp-starter)\n\n## Python\n\nSpark NLP supports Python 3.6.x and above depending on your major PySpark version.\n\n### Python without explicit Pyspark installation\n\n### Pip/Conda\n\nIf you installed pyspark through pip/conda, you can install `spark-nlp` through the same channel.\n\nPip:\n\n```bash\npip install spark-nlp==4.2.7\n```\n\nConda:\n\n```bash\nconda install -c johnsnowlabs spark-nlp\n```\n\nPyPI [spark-nlp package](https://pypi.org/project/spark-nlp/) / Anaconda [spark-nlp package](https://anaconda.org/JohnSnowLabs/spark-nlp)\n\nThen you'll have to create a SparkSession either from Spark NLP:\n\n```python\nimport sparknlp\n\nspark = sparknlp.start()\n```\n\nor manually:\n\n```python\nspark = SparkSession.builder \\\n .appName(\"Spark NLP\")\\\n .master(\"local[*]\")\\\n .config(\"spark.driver.memory\",\"16G\")\\\n .config(\"spark.driver.maxResultSize\", \"0\") \\\n .config(\"spark.kryoserializer.buffer.max\", \"2000M\")\\\n .config(\"spark.jars.packages\", \"com.johnsnowlabs.nlp:spark-nlp_2.12:4.2.7\")\\\n .getOrCreate()\n```\n\nIf using local jars, you can use `spark.jars` instead for comma-delimited jar files. For cluster setups, of course, you'll have to put the jars in a reachable location for all driver and executor nodes.\n\n**Quick example:**\n\n```python\nimport sparknlp\nfrom sparknlp.pretrained import PretrainedPipeline\n\n#create or get Spark Session\n\nspark = sparknlp.start()\n\nsparknlp.version()\nspark.version\n\n#download, load and annotate a text by pre-trained pipeline\n\npipeline = PretrainedPipeline('recognize_entities_dl', 'en')\nresult = pipeline.annotate('The Mona Lisa is a 16th century oil painting created by Leonardo')\n```\n\n## Compiled JARs\n\n### Build from source\n\n#### spark-nlp\n\n- FAT-JAR for CPU on Apache Spark 3.0.x, 3.1.x, 3.2.x, and 3.3.x\n\n```bash\nsbt assembly\n```\n\n- FAT-JAR for GPU on Apache Spark 3.0.x, 3.1.x, 3.2.x, and 3.3.x\n\n```bash\nsbt -Dis_gpu=true assembly\n```\n\n- FAT-JAR for M! on Apache Spark 3.0.x, 3.1.x, 3.2.x, and 3.3.x\n\n```bash\nsbt -Dis_m1=true assembly\n```\n\n### Using the jar manually\n\nIf for some reason you need to use the JAR, you can either download the Fat JARs provided here or download it from [Maven Central](https://mvnrepository.com/artifact/com.johnsnowlabs.nlp).\n\nTo add JARs to spark programs use the `--jars` option:\n\n```sh\nspark-shell --jars spark-nlp.jar\n```\n\nThe preferred way to use the library when running spark programs is using the `--packages` option as specified in the `spark-packages` section.\n\n## Apache Zeppelin\n\nUse either one of the following options\n\n- Add the following Maven Coordinates to the interpreter's library list\n\n```bash\ncom.johnsnowlabs.nlp:spark-nlp_2.12:4.2.7\n```\n\n- Add a path to pre-built jar from [here](#compiled-jars) in the interpreter's library list making sure the jar is available to driver path\n\n### Python in Zeppelin\n\nApart from the previous step, install the python module through pip\n\n```bash\npip install spark-nlp==4.2.7\n```\n\nOr you can install `spark-nlp` from inside Zeppelin by using Conda:\n\n```bash\npython.conda install -c johnsnowlabs spark-nlp\n```\n\nConfigure Zeppelin properly, use cells with %spark.pyspark or any interpreter name you chose.\n\nFinally, in Zeppelin interpreter settings, make sure you set properly zeppelin.python to the python you want to use and install the pip library with (e.g. `python3`).\n\nAn alternative option would be to set `SPARK_SUBMIT_OPTIONS` (zeppelin-env.sh) and make sure `--packages` is there as shown earlier since it includes both scala and python side installation.\n\n## Jupyter Notebook (Python)\n\n**Recommended:**\n\nThe easiest way to get this done on Linux and macOS is to simply install `spark-nlp` and `pyspark` PyPI packages and launch the Jupyter from the same Python environment:\n\n```sh\n$ conda create -n sparknlp python=3.8 -y\n$ conda activate sparknlp\n# spark-nlp by default is based on pyspark 3.x\n$ pip install spark-nlp==4.2.7 pyspark==3.2.3 jupyter\n$ jupyter notebook\n```\n\nThen you can use `python3` kernel to run your code with creating SparkSession via `spark = sparknlp.start()`.\n\n**Optional:**\n\nIf you are in different operating systems and require to make Jupyter Notebook run by using pyspark, you can follow these steps:\n\n```bash\nexport SPARK_HOME=/path/to/your/spark/folder\nexport PYSPARK_PYTHON=python3\nexport PYSPARK_DRIVER_PYTHON=jupyter\nexport PYSPARK_DRIVER_PYTHON_OPTS=notebook\n\npyspark --packages com.johnsnowlabs.nlp:spark-nlp_2.12:4.2.7\n```\n\nAlternatively, you can mix in using `--jars` option for pyspark + `pip install spark-nlp`\n\nIf not using pyspark at all, you'll have to run the instructions pointed [here](#python-without-explicit-Pyspark-installation)\n\n## Google Colab Notebook\n\nGoogle Colab is perhaps the easiest way to get started with spark-nlp. It requires no installation or setup other than having a Google account.\n\nRun the following code in Google Colab notebook and start using spark-nlp right away.\n\n```sh\n# This is only to setup PySpark and Spark NLP on Colab\n!wget https://setup.johnsnowlabs.com/colab.sh -O - | bash\n```\n\nThis script comes with the two options to define `pyspark` and `spark-nlp` versions via options:\n\n```sh\n# -p is for pyspark\n# -s is for spark-nlp\n# -g will enable upgrading libcudnn8 to 8.1.0 on Google Colab for GPU usage\n# by default they are set to the latest\n!wget https://setup.johnsnowlabs.com/colab.sh -O - | bash /dev/stdin -p 3.2.3 -s 4.2.7\n```\n\n[Spark NLP quick start on Google Colab](https://colab.research.google.com/github/JohnSnowLabs/spark-nlp-workshop/blob/master/jupyter/quick_start_google_colab.ipynb) is a live demo on Google Colab that performs named entity recognitions and sentiment analysis by using Spark NLP pretrained pipelines.\n\n## Kaggle Kernel\n\nRun the following code in Kaggle Kernel and start using spark-nlp right away.\n\n```sh\n# Let's setup Kaggle for Spark NLP and PySpark\n!wget https://setup.johnsnowlabs.com/kaggle.sh -O - | bash\n```\n\nThis script comes with the two options to define `pyspark` and `spark-nlp` versions via options:\n\n```sh\n# -p is for pyspark\n# -s is for spark-nlp\n# -g will enable upgrading libcudnn8 to 8.1.0 on Kaggle for GPU usage\n# by default they are set to the latest\n!wget https://setup.johnsnowlabs.com/colab.sh -O - | bash /dev/stdin -p 3.2.3 -s 4.2.7\n```\n\n[Spark NLP quick start on Kaggle Kernel](https://www.kaggle.com/mozzie/spark-nlp-named-entity-recognition) is a live demo on Kaggle Kernel that performs named entity recognitions by using Spark NLP pretrained pipeline.\n\n## Databricks Cluster\n\n1. Create a cluster if you don't have one already\n\n2. On a new cluster or existing one you need to add the following to the `Advanced Options -> Spark` tab:\n\n ```bash\n spark.kryoserializer.buffer.max 2000M\n spark.serializer org.apache.spark.serializer.KryoSerializer\n ```\n\n3. In `Libraries` tab inside your cluster you need to follow these steps:\n\n 3.1. Install New -> PyPI -> `spark-nlp==4.2.7` -> Install\n\n 3.2. Install New -> Maven -> Coordinates -> `com.johnsnowlabs.nlp:spark-nlp_2.12:4.2.7` -> Install\n\n4. Now you can attach your notebook to the cluster and use Spark NLP!\n\nNOTE: Databricks' runtimes support different Apache Spark major releases. Please make sure you choose the correct Spark NLP Maven package name (Maven Coordinate) for your runtime from our [Packages Cheatsheet](https://github.com/JohnSnowLabs/spark-nlp#packages-cheatsheet)\n\n## EMR Cluster\n\nTo launch EMR clusters with Apache Spark/PySpark and Spark NLP correctly you need to have bootstrap and software configuration.\n\nA sample of your bootstrap script\n\n```.sh\n#!/bin/bash\nset -x -e\n\necho -e 'export PYSPARK_PYTHON=/usr/bin/python3\nexport HADOOP_CONF_DIR=/etc/hadoop/conf\nexport SPARK_JARS_DIR=/usr/lib/spark/jars\nexport SPARK_HOME=/usr/lib/spark' >> $HOME/.bashrc && source $HOME/.bashrc\n\nsudo python3 -m pip install awscli boto spark-nlp\n\nset +x\nexit 0\n\n```\n\nA sample of your software configuration in JSON on S3 (must be public access):\n\n```.json\n[{\n \"Classification\": \"spark-env\",\n \"Configurations\": [{\n \"Classification\": \"export\",\n \"Properties\": {\n \"PYSPARK_PYTHON\": \"/usr/bin/python3\"\n }\n }]\n},\n{\n \"Classification\": \"spark-defaults\",\n \"Properties\": {\n \"spark.yarn.stagingDir\": \"hdfs:///tmp\",\n \"spark.yarn.preserve.staging.files\": \"true\",\n \"spark.kryoserializer.buffer.max\": \"2000M\",\n \"spark.serializer\": \"org.apache.spark.serializer.KryoSerializer\",\n \"spark.driver.maxResultSize\": \"0\",\n \"spark.jars.packages\": \"com.johnsnowlabs.nlp:spark-nlp_2.12:4.2.7\"\n }\n}]\n```\n\nA sample of AWS CLI to launch EMR cluster:\n\n```.sh\naws emr create-cluster \\\n--name \"Spark NLP 4.2.7\" \\\n--release-label emr-6.2.0 \\\n--applications Name=Hadoop Name=Spark Name=Hive \\\n--instance-type m4.4xlarge \\\n--instance-count 3 \\\n--use-default-roles \\\n--log-uri \"s3://<S3_BUCKET>/\" \\\n--bootstrap-actions Path=s3://<S3_BUCKET>/emr-bootstrap.sh,Name=custome \\\n--configurations \"https://<public_access>/sparknlp-config.json\" \\\n--ec2-attributes KeyName=<your_ssh_key>,EmrManagedMasterSecurityGroup=<security_group_with_ssh>,EmrManagedSlaveSecurityGroup=<security_group_with_ssh> \\\n--profile <aws_profile_credentials>\n```\n\n## GCP Dataproc\n\n1. Create a cluster if you don't have one already as follows.\n\nAt gcloud shell:\n\n```bash\ngcloud services enable dataproc.googleapis.com \\\n compute.googleapis.com \\\n storage-component.googleapis.com \\\n bigquery.googleapis.com \\\n bigquerystorage.googleapis.com\n```\n\n```bash\nREGION=<region>\n```\n\n```bash\nBUCKET_NAME=<bucket_name>\ngsutil mb -c standard -l ${REGION} gs://${BUCKET_NAME}\n```\n\n```bash\nREGION=<region>\nZONE=<zone>\nCLUSTER_NAME=<cluster_name>\nBUCKET_NAME=<bucket_name>\n```\n\nYou can set image-version, master-machine-type, worker-machine-type,\nmaster-boot-disk-size, worker-boot-disk-size, num-workers as your needs.\nIf you use the previous image-version from 2.0, you should also add ANACONDA to optional-components.\nAnd, you should enable gateway.\nDon't forget to set the maven coordinates for the jar in properties.\n\n```bash\ngcloud dataproc clusters create ${CLUSTER_NAME} \\\n --region=${REGION} \\\n --zone=${ZONE} \\\n --image-version=2.0 \\\n --master-machine-type=n1-standard-4 \\\n --worker-machine-type=n1-standard-2 \\\n --master-boot-disk-size=128GB \\\n --worker-boot-disk-size=128GB \\\n --num-workers=2 \\\n --bucket=${BUCKET_NAME} \\\n --optional-components=JUPYTER \\\n --enable-component-gateway \\\n --metadata 'PIP_PACKAGES=spark-nlp spark-nlp-display google-cloud-bigquery google-cloud-storage' \\\n --initialization-actions gs://goog-dataproc-initialization-actions-${REGION}/python/pip-install.sh \\\n --properties spark:spark.serializer=org.apache.spark.serializer.KryoSerializer,spark:spark.driver.maxResultSize=0,spark:spark.kryoserializer.buffer.max=2000M,spark:spark.jars.packages=com.johnsnowlabs.nlp:spark-nlp_2.12:4.2.7\n```\n\n2. On an existing one, you need to install spark-nlp and spark-nlp-display packages from PyPI.\n\n3. Now, you can attach your notebook to the cluster and use the Spark NLP!\n\n## Spark NLP Configuration\n\nYou can change the following Spark NLP configurations via Spark Configuration:\n\n| Property Name | Default | Meaning |\n|--------------------------------------------------------|----------------------|------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|\n| `spark.jsl.settings.pretrained.cache_folder` | `~/cache_pretrained` | The location to download and extract pretrained `Models` and `Pipelines`. By default, it will be in User's Home directory under `cache_pretrained` directory |\n| `spark.jsl.settings.storage.cluster_tmp_dir` | `hadoop.tmp.dir` | The location to use on a cluster for temporarily files such as unpacking indexes for WordEmbeddings. By default, this locations is the location of `hadoop.tmp.dir` set via Hadoop configuration for Apache Spark. NOTE: `S3` is not supported and it must be local, HDFS, or DBFS |\n| `spark.jsl.settings.annotator.log_folder` | `~/annotator_logs` | The location to save logs from annotators during training such as `NerDLApproach`, `ClassifierDLApproach`, `SentimentDLApproach`, `MultiClassifierDLApproach`, etc. By default, it will be in User's Home directory under `annotator_logs` directory |\n| `spark.jsl.settings.aws.credentials.access_key_id` | `None` | Your AWS access key to use your S3 bucket to store log files of training models or access tensorflow graphs used in `NerDLApproach` |\n| `spark.jsl.settings.aws.credentials.secret_access_key` | `None` | Your AWS secret access key to use your S3 bucket to store log files of training models or access tensorflow graphs used in `NerDLApproach` |\n| `spark.jsl.settings.aws.credentials.session_token` | `None` | Your AWS MFA session token to use your S3 bucket to store log files of training models or access tensorflow graphs used in `NerDLApproach` |\n| `spark.jsl.settings.aws.s3_bucket` | `None` | Your AWS S3 bucket to store log files of training models or access tensorflow graphs used in `NerDLApproach` |\n| `spark.jsl.settings.aws.region` | `None` | Your AWS region to use your S3 bucket to store log files of training models or access tensorflow graphs used in `NerDLApproach` |\n\n### How to set Spark NLP Configuration\n\n**SparkSession:**\n\nYou can use `.config()` during SparkSession creation to set Spark NLP configurations.\n\n```python\nfrom pyspark.sql import SparkSession\n\nspark = SparkSession.builder \\\n .master(\"local[*]\") \\\n .config(\"spark.driver.memory\", \"16G\") \\\n .config(\"spark.driver.maxResultSize\", \"0\") \\\n .config(\"spark.serializer\", \"org.apache.spark.serializer.KryoSerializer\") \\\n .config(\"spark.kryoserializer.buffer.max\", \"2000m\") \\\n .config(\"spark.jsl.settings.pretrained.cache_folder\", \"sample_data/pretrained\") \\\n .config(\"spark.jsl.settings.storage.cluster_tmp_dir\", \"sample_data/storage\") \\\n .config(\"spark.jars.packages\", \"com.johnsnowlabs.nlp:spark-nlp_2.12:4.2.7\") \\\n .getOrCreate()\n```\n\n**spark-shell:**\n\n```sh\nspark-shell \\\n --driver-memory 16g \\\n --conf spark.driver.maxResultSize=0 \\\n --conf spark.serializer=org.apache.spark.serializer.KryoSerializer\n --conf spark.kryoserializer.buffer.max=2000M \\\n --conf spark.jsl.settings.pretrained.cache_folder=\"sample_data/pretrained\" \\\n --conf spark.jsl.settings.storage.cluster_tmp_dir=\"sample_data/storage\" \\\n --packages com.johnsnowlabs.nlp:spark-nlp_2.12:4.2.7\n```\n\n**pyspark:**\n\n```sh\npyspark \\\n --driver-memory 16g \\\n --conf spark.driver.maxResultSize=0 \\\n --conf spark.serializer=org.apache.spark.serializer.KryoSerializer\n --conf spark.kryoserializer.buffer.max=2000M \\\n --conf spark.jsl.settings.pretrained.cache_folder=\"sample_data/pretrained\" \\\n --conf spark.jsl.settings.storage.cluster_tmp_dir=\"sample_data/storage\" \\\n --packages com.johnsnowlabs.nlp:spark-nlp_2.12:4.2.7\n```\n\n**Databricks:**\n\nOn a new cluster or existing one you need to add the following to the `Advanced Options -> Spark` tab:\n\n```bash\nspark.kryoserializer.buffer.max 2000M\nspark.serializer org.apache.spark.serializer.KryoSerializer\nspark.jsl.settings.pretrained.cache_folder dbfs:/PATH_TO_CACHE\nspark.jsl.settings.storage.cluster_tmp_dir dbfs:/PATH_TO_STORAGE\nspark.jsl.settings.annotator.log_folder dbfs:/PATH_TO_LOGS\n```\n\nNOTE: If this is an existing cluster, after adding new configs or changing existing properties you need to restart it.\n\n### S3 Integration\n\nIn Spark NLP we can define S3 locations to:\n\n- Export log files of training models\n- Store tensorflow graphs used in `NerDLApproach`\n\n**Logging:**\n\nTo configure S3 path for logging while training models. We need to set up AWS credentials as well as an S3 path\n\n```bash\nspark.conf.set(\"spark.jsl.settings.annotator.log_folder\", \"s3://my/s3/path/logs\")\nspark.conf.set(\"spark.jsl.settings.aws.credentials.access_key_id\", \"MY_KEY_ID\")\nspark.conf.set(\"spark.jsl.settings.aws.credentials.secret_access_key\", \"MY_SECRET_ACCESS_KEY\")\nspark.conf.set(\"spark.jsl.settings.aws.s3_bucket\", \"my.bucket\")\nspark.conf.set(\"spark.jsl.settings.aws.region\", \"my-region\")\n```\n\nNow you can check the log on your S3 path defined in *spark.jsl.settings.annotator.log_folder* property.\nMake sure to use the prefix *s3://*, otherwise it will use the default configuration.\n\n**Tensorflow Graphs:**\n\nTo reference S3 location for downloading graphs. We need to set up AWS credentials\n\n```bash\nspark.conf.set(\"spark.jsl.settings.aws.credentials.access_key_id\", \"MY_KEY_ID\")\nspark.conf.set(\"spark.jsl.settings.aws.credentials.secret_access_key\", \"MY_SECRET_ACCESS_KEY\")\nspark.conf.set(\"spark.jsl.settings.aws.region\", \"my-region\")\n```\n\n**MFA Configuration:**\n\nIn case your AWS account is configured with MFA. You will need first to get temporal credentials and add session token to the configuration as shown in the examples below\nFor logging:\n\n```bash\nspark.conf.set(\"spark.jsl.settings.aws.credentials.session_token\", \"MY_TOKEN\")\n```\n\nAn example of a bash script that gets temporal AWS credentials can be found [here](https://github.com/JohnSnowLabs/spark-nlp/blob/master/scripts/aws_tmp_credentials.sh)\nThis script requires three arguments:\n\n```bash\n./aws_tmp_credentials.sh iam_user duration serial_number\n```\n\n## Pipelines and Models\n\n### Pipelines\n\n**Quick example:**\n\n```scala\nimport com.johnsnowlabs.nlp.pretrained.PretrainedPipeline\nimport com.johnsnowlabs.nlp.SparkNLP\n\nSparkNLP.version()\n\nval testData = spark.createDataFrame(Seq(\n(1, \"Google has announced the release of a beta version of the popular TensorFlow machine learning library\"),\n(2, \"Donald John Trump (born June 14, 1946) is the 45th and current president of the United States\")\n)).toDF(\"id\", \"text\")\n\nval pipeline = PretrainedPipeline(\"explain_document_dl\", lang=\"en\")\n\nval annotation = pipeline.transform(testData)\n\nannotation.show()\n/*\nimport com.johnsnowlabs.nlp.pretrained.PretrainedPipeline\nimport com.johnsnowlabs.nlp.SparkNLP\n2.5.0\ntestData: org.apache.spark.sql.DataFrame = [id: int, text: string]\npipeline: com.johnsnowlabs.nlp.pretrained.PretrainedPipeline = PretrainedPipeline(explain_document_dl,en,public/models)\nannotation: org.apache.spark.sql.DataFrame = [id: int, text: string ... 10 more fields]\n+---+--------------------+--------------------+--------------------+--------------------+--------------------+--------------------+--------------------+--------------------+--------------------+--------------------+--------------------+\n| id| text| document| token| sentence| checked| lemma| stem| pos| embeddings| ner| entities|\n+---+--------------------+--------------------+--------------------+--------------------+--------------------+--------------------+--------------------+--------------------+--------------------+--------------------+--------------------+\n| 1|Google has announ...|[[document, 0, 10...|[[token, 0, 5, Go...|[[document, 0, 10...|[[token, 0, 5, Go...|[[token, 0, 5, Go...|[[token, 0, 5, go...|[[pos, 0, 5, NNP,...|[[word_embeddings...|[[named_entity, 0...|[[chunk, 0, 5, Go...|\n| 2|The Paris metro w...|[[document, 0, 11...|[[token, 0, 2, Th...|[[document, 0, 11...|[[token, 0, 2, Th...|[[token, 0, 2, Th...|[[token, 0, 2, th...|[[pos, 0, 2, DT, ...|[[word_embeddings...|[[named_entity, 0...|[[chunk, 4, 8, Pa...|\n+---+--------------------+--------------------+--------------------+--------------------+--------------------+--------------------+--------------------+--------------------+--------------------+--------------------+--------------------+\n*/\n\nannotation.select(\"entities.result\").show(false)\n\n/*\n+----------------------------------+\n|result |\n+----------------------------------+\n|[Google, TensorFlow] |\n|[Donald John Trump, United States]|\n+----------------------------------+\n*/\n```\n\n#### Showing Available Pipelines\n\nThere are functions in Spark NLP that will list all the available Pipelines\nof a particular language for you:\n\n```scala\nimport com.johnsnowlabs.nlp.pretrained.ResourceDownloader\n\nResourceDownloader.showPublicPipelines(lang=\"en\")\n/*\n+--------------------------------------------+------+---------+\n| Pipeline | lang | version |\n+--------------------------------------------+------+---------+\n| dependency_parse | en | 2.0.2 |\n| analyze_sentiment_ml | en | 2.0.2 |\n| check_spelling | en | 2.1.0 |\n| match_datetime | en | 2.1.0 |\n ...\n| explain_document_ml | en | 3.1.3 |\n+--------------------------------------------+------+---------+\n*/\n```\n\nOr if we want to check for a particular version:\n\n```scala\nimport com.johnsnowlabs.nlp.pretrained.ResourceDownloader\n\nResourceDownloader.showPublicPipelines(lang=\"en\", version=\"3.1.0\")\n/*\n+---------------------------------------+------+---------+\n| Pipeline | lang | version |\n+---------------------------------------+------+---------+\n| dependency_parse | en | 2.0.2 |\n ...\n| clean_slang | en | 3.0.0 |\n| clean_pattern | en | 3.0.0 |\n| check_spelling | en | 3.0.0 |\n| dependency_parse | en | 3.0.0 |\n+---------------------------------------+------+---------+\n*/\n```\n\n#### Please check out our Models Hub for the full list of [pre-trained pipelines](https://nlp.johnsnowlabs.com/models) with examples, demos, benchmarks, and more\n\n### Models\n\n**Some selected languages:** `Afrikaans, Arabic, Armenian, Basque, Bengali, Breton, Bulgarian, Catalan, Czech, Dutch, English, Esperanto, Finnish, French, Galician, German, Greek, Hausa, Hebrew, Hindi, Hungarian, Indonesian, Irish, Italian, Japanese, Latin, Latvian, Marathi, Norwegian, Persian, Polish, Portuguese, Romanian, Russian, Slovak, Slovenian, Somali, Southern Sotho, Spanish, Swahili, Swedish, Tswana, Turkish, Ukrainian, Zulu`\n\n**Quick online example:**\n\n```python\n# load NER model trained by deep learning approach and GloVe word embeddings\nner_dl = NerDLModel.pretrained('ner_dl')\n# load NER model trained by deep learning approach and BERT word embeddings\nner_bert = NerDLModel.pretrained('ner_dl_bert')\n```\n\n```scala\n// load French POS tagger model trained by Universal Dependencies\nval french_pos = PerceptronModel.pretrained(\"pos_ud_gsd\", lang=\"fr\")\n// load Italian LemmatizerModel\nval italian_lemma = LemmatizerModel.pretrained(\"lemma_dxc\", lang=\"it\")\n````\n\n**Quick offline example:**\n\n- Loading `PerceptronModel` annotator model inside Spark NLP Pipeline\n\n```scala\nval french_pos = PerceptronModel.load(\"/tmp/pos_ud_gsd_fr_2.0.2_2.4_1556531457346/\")\n .setInputCols(\"document\", \"token\")\n .setOutputCol(\"pos\")\n```\n\n#### Showing Available Models\n\nThere are functions in Spark NLP that will list all the available Models\nof a particular Annotator and language for you:\n\n```scala\nimport com.johnsnowlabs.nlp.pretrained.ResourceDownloader\n\nResourceDownloader.showPublicModels(annotator=\"NerDLModel\", lang=\"en\")\n/*\n+---------------------------------------------+------+---------+\n| Model | lang | version |\n+---------------------------------------------+------+---------+\n| onto_100 | en | 2.1.0 |\n| onto_300 | en | 2.1.0 |\n| ner_dl_bert | en | 2.2.0 |\n| onto_100 | en | 2.4.0 |\n| ner_conll_elmo | en | 3.2.2 |\n+---------------------------------------------+------+---------+\n*/\n```\n\nOr if we want to check for a particular version:\n\n```scala\nimport com.johnsnowlabs.nlp.pretrained.ResourceDownloader\n\nResourceDownloader.showPublicModels(annotator=\"NerDLModel\", lang=\"en\", version=\"3.1.0\")\n/*\n+----------------------------+------+---------+\n| Model | lang | version |\n+----------------------------+------+---------+\n| onto_100 | en | 2.1.0 |\n| ner_aspect_based_sentiment | en | 2.6.2 |\n| ner_weibo_glove_840B_300d | en | 2.6.2 |\n| nerdl_atis_840b_300d | en | 2.7.1 |\n| nerdl_snips_100d | en | 2.7.3 |\n+----------------------------+------+---------+\n*/\n```\n\nAnd to see a list of available annotators, you can use:\n\n```scala\nimport com.johnsnowlabs.nlp.pretrained.ResourceDownloader\n\nResourceDownloader.showAvailableAnnotators()\n/*\nAlbertEmbeddings\nAlbertForTokenClassification\nAssertionDLModel\n...\nXlmRoBertaSentenceEmbeddings\nXlnetEmbeddings\n*/\n```\n\n#### Please check out our Models Hub for the full list of [pre-trained models](https://nlp.johnsnowlabs.com/models) with examples, demo, benchmark, and more\n\n## Offline\n\nSpark NLP library and all the pre-trained models/pipelines can be used entirely offline with no access to the Internet. If you are behind a proxy or a firewall with no access to the Maven repository (to download packages) or/and no access to S3 (to automatically download models and pipelines), you can simply follow the instructions to have Spark NLP without any limitations offline:\n\n- Instead of using the Maven package, you need to load our Fat JAR\n- Instead of using PretrainedPipeline for pretrained pipelines or the `.pretrained()` function to download pretrained models, you will need to manually download your pipeline/model from [Models Hub](https://nlp.johnsnowlabs.com/models), extract it, and load it.\n\nExample of `SparkSession` with Fat JAR to have Spark NLP offline:\n\n```python\nspark = SparkSession.builder \\\n .appName(\"Spark NLP\")\\\n .master(\"local[*]\")\\\n .config(\"spark.driver.memory\",\"16G\")\\\n .config(\"spark.driver.maxResultSize\", \"0\") \\\n .config(\"spark.kryoserializer.buffer.max\", \"2000M\")\\\n .config(\"spark.jars\", \"/tmp/spark-nlp-assembly-4.2.7.jar\")\\\n .getOrCreate()\n```\n\n- You can download provided Fat JARs from each [release notes](https://github.com/JohnSnowLabs/spark-nlp/releases), please pay attention to pick the one that suits your environment depending on the device (CPU/GPU) and Apache Spark version (3.0.x, 3.1.x, 3.2.x, and 3.3.x)\n- If you are local, you can load the Fat JAR from your local FileSystem, however, if you are in a cluster setup you need to put the Fat JAR on a distributed FileSystem such as HDFS, DBFS, S3, etc. (i.e., `hdfs:///tmp/spark-nlp-assembly-4.2.7.jar`)\n\nExample of using pretrained Models and Pipelines in offline:\n\n```python\n# instead of using pretrained() for online:\n# french_pos = PerceptronModel.pretrained(\"pos_ud_gsd\", lang=\"fr\")\n# you download this model, extract it, and use .load\nfrench_pos = PerceptronModel.load(\"/tmp/pos_ud_gsd_fr_2.0.2_2.4_1556531457346/\")\\\n .setInputCols(\"document\", \"token\")\\\n .setOutputCol(\"pos\")\n\n# example for pipelines\n# instead of using PretrainedPipeline\n# pipeline = PretrainedPipeline('explain_document_dl', lang='en')\n# you download this pipeline, extract it, and use PipelineModel\nPipelineModel.load(\"/tmp/explain_document_dl_en_2.0.2_2.4_1556530585689/\")\n```\n\n- Since you are downloading and loading models/pipelines manually, this means Spark NLP is not downloading the most recent and compatible models/pipelines for you. Choosing the right model/pipeline is on you\n- If you are local, you can load the model/pipeline from your local FileSystem, however, if you are in a cluster setup you need to put the model/pipeline on a distributed FileSystem such as HDFS, DBFS, S3, etc. (i.e., `hdfs:///tmp/explain_document_dl_en_2.0.2_2.4_1556530585689/`)\n\n## Examples\n\nNeed more **examples**? Check out our dedicated [Spark NLP Showcase](https://github.com/JohnSnowLabs/spark-nlp-workshop) repository to showcase all Spark NLP use cases!\n\nAlso, don't forget to check [Spark NLP in Action](https://nlp.johnsnowlabs.com/demo) built by Streamlit.\n\n### All examples: [spark-nlp-workshop](https://github.com/JohnSnowLabs/spark-nlp-workshop)\n\n## FAQ\n\n[Check our Articles and Videos page here](https://nlp.johnsnowlabs.com/learn)\n\n## Citation\n\nWe have published a [paper](https://www.sciencedirect.com/science/article/pii/S2665963821000063) that you can cite for the Spark NLP library:\n\n```bibtex\n@article{KOCAMAN2021100058,\n title = {Spark NLP: Natural language understanding at scale},\n journal = {Software Impacts},\n pages = {100058},\n year = {2021},\n issn = {2665-9638},\n doi = {https://doi.org/10.1016/j.simpa.2021.100058},\n url = {https://www.sciencedirect.com/science/article/pii/S2665963.2.300063},\n author = {Veysel Kocaman and David Talby},\n keywords = {Spark, Natural language processing, Deep learning, Tensorflow, Cluster},\n abstract = {Spark NLP is a Natural Language Processing (NLP) library built on top of Apache Spark ML. It provides simple, performant & accurate NLP annotations for machine learning pipelines that can scale easily in a distributed environment. Spark NLP comes with 1100+ pretrained pipelines and models in more than 192+ languages. It supports nearly all the NLP tasks and modules that can be used seamlessly in a cluster. Downloaded more than 2.7 million times and experiencing 9x growth since January 2020, Spark NLP is used by 54% of healthcare organizations as the world\u2019s most widely used NLP library in the enterprise.}\n }\n}\n```\n\n## Contributing\n\nWe appreciate any sort of contributions:\n\n- ideas\n- feedback\n- documentation\n- bug reports\n- NLP training and testing corpora\n- Development and testing\n\nClone the repo and submit your pull-requests! Or directly create issues in this repo.\n\n## John Snow Labs\n\n[http://johnsnowlabs.com](http://johnsnowlabs.com)\n",
"bugtrack_url": null,
"license": "",
"summary": "John Snow Labs Spark NLP is a natural language processing library built on top of Apache Spark ML. It provides simple, performant & accurate NLP annotations for machine learning pipelines, that scale easily in a distributed environment.",
"version": "4.2.7",
"split_keywords": [
"nlp",
"spark",
"vision",
"speech",
"deep",
"learning",
"transformer",
"tensorflow",
"bert",
"gpt-2",
"wav2vec2",
"vit"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "6651075e949746e2e108a58d1c009c847337f3377587eb64bd94e6e7423eb86b",
"md5": "2db7cbd76fc18788b3874508f0f641af",
"sha256": "5d18db975d950c668d2930c55c2833ac044c1b465c688047fd9ef21062b13569"
},
"downloads": -1,
"filename": "ckls_test_lib-4.2.7-py2.py3-none-any.whl",
"has_sig": false,
"md5_digest": "2db7cbd76fc18788b3874508f0f641af",
"packagetype": "bdist_wheel",
"python_version": "py2.py3",
"requires_python": null,
"size": 453817,
"upload_time": "2023-01-23T15:13:56",
"upload_time_iso_8601": "2023-01-23T15:13:56.976655Z",
"url": "https://files.pythonhosted.org/packages/66/51/075e949746e2e108a58d1c009c847337f3377587eb64bd94e6e7423eb86b/ckls_test_lib-4.2.7-py2.py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "48b62604ee6b1c6dbdd88ed973bae50e4f37095906e0d1f8c18f2cc40c3e209a",
"md5": "665eb7f0f9da3e764a25e39679e9ec3b",
"sha256": "137797417b85031783a505acbb84ff69bac8dd98bd2757838698852340e9f4b2"
},
"downloads": -1,
"filename": "ckls_test_lib-4.2.7.tar.gz",
"has_sig": false,
"md5_digest": "665eb7f0f9da3e764a25e39679e9ec3b",
"packagetype": "sdist",
"python_version": "source",
"requires_python": null,
"size": 258053,
"upload_time": "2023-01-23T15:14:00",
"upload_time_iso_8601": "2023-01-23T15:14:00.396579Z",
"url": "https://files.pythonhosted.org/packages/48/b6/2604ee6b1c6dbdd88ed973bae50e4f37095906e0d1f8c18f2cc40c3e209a/ckls_test_lib-4.2.7.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2023-01-23 15:14:00",
"github": true,
"gitlab": false,
"bitbucket": false,
"github_user": "JohnSnowLabs",
"github_project": "spark-nlp",
"travis_ci": false,
"coveralls": false,
"github_actions": true,
"lcname": "ckls-test-lib"
}
JSON
