Prerequisites

Software

1. ElasticSearch 7+ (Also 8 for 1.3.3)/Opensearch v. 1.1 or above
2. Java 11 or above, Java 17 for version 1.3.3

Step 1: Getting Saga

You'll be able to get binaries from here (SCA Artifactory repository).

From this version and on, we provide docker images:

Step 2: Set up the environment

1. Check that Java 11 (or 17 for 1.3.3 ) is installed on your machine by running on terminal (or system console):
   
   $> java -version


2. Unpackage Saga in your preferred location. 
   This is our recommended setup but you can pretty much handle the paths as you wish.
   This guide will refer to Saga's working directory as {SAGA_HOME}.
3. Saga uses ElasticSearch (7+ and also 8 for 1.3.3) or Opensearch (v.1.1).
   1. Deploy ElasticSearch (ES) or Opensearch
   2. Run ES by executing the binary on {ELASTIC_HOME}/bin.

Step 3: Deploy Saga

Once you have Saga in {SAGA_HOME} validate the following:

There is a {SAGA_HOME}/lib folder containing the following JARs:

1. saga-classification-trainer-stage-1.3.3
2. saga-faq-stage-1.3.3
3. saga-intent-stage-1.3.3 (on 1.3.3)
4. saga-lang-detector-stage-1.3.3

5. saga-name-trainer-stage-1.3.3
6. saga-parts-of-speech-stage-1.3.3
7. saga-sentence-breaker-stage-1.3.3
8. saga-spellchecking-stage-1.3.3

Check the configuration base on what NoSQL DB provider used.

    ES : How To Connect To Elasticsearch

    Opensearch: How To Connect To OpenSearch

Check the basic configuration on {SAGA_HOME}/config/config.json:

1. "apiPort": 8080 → The port used by the server.
2. "host": "0.0.0.0" → This IP/mask is used to restrict inbound connections, open to all connections by default.
3. "security" → Security access to the Saga Server and Management UI (recommended only when ssl is also enabled)
   1. "enable": true → Indicates the use of security
   2. "encryptionKeyFile": → file containing the encryption key.  A file is provided by default but it is recommended to change it.
   3. "users.username" → Specify the username to use
   4. "users.password" → Specify the password for the username. Can be either plain or encrypted
4. "ssl" → Enables SSL for the communication with the Saga Server
   1. enable": true → Indicates the use of ssl
   2. keyStore → Path to the keyStore
   3. keyStorePassword → Password of the keyStore. Can be either plain or encrypted
5. "libraryJars": ["./lib"] → Folder where library jars are located
6. If you are running elasticseach in localhost then you don't need to change anything. But if your elasticsearch runs somewhere else then you'll need to adjust the "nodeUrls" property in both "elasticSearch" solution and "Elastic" provider. In case you have a cluster, you can specify them separated by comma, for example: " nodeUrls": ["elastic1:9200", "elastic2:9200", "elastic3:9200"],

If you have some valid "models" you'd like to include them on the server:

1. Create a {SAGA-HOME}/nt-models folder for "name trainers" and copy the model there.
2. Create a {SAGA-HOME}/ct-models folder for "classification trainers" and copy the model there.
3. Create a {SAGA-HOME}/tf-models folder for "FAQ" (uses TensorFlow) and copy the model there.

To add datasets:

1. Create a {SAGA-HOME}/datasets folder.
2. Each dataset must be placed in its own folder. This folder name will be the one displayed for "test runs".
3. Each data document in the dataset must be compliant with Saga's data file JSON format.
4. Each folder must contain a ".metadata" file with information about the dataset and how to read it.
   You can check out the dataset format here.
   

Step 4: Run Saga

To run Saga:

Check that ElasticSearch is running.

Use the bundled startup script on {SAGA_HOME}/bin (either startup.bat for Windows or startup.sh for Linux and Mac).

If you didn't change the default port on the configuration, you should be able to access Saga UI at http://localhost:8080/.
If not, then check your configuration for the right port.