1
1
mirror of https://github.com/MarginaliaSearch/MarginaliaSearch.git synced 2025-10-05 21:22:39 +02:00

(setup) Pull POS tags from control svc on first boot

This commit also removes the old retrieval from setup.sh
This commit is contained in:
Viktor Lofgren
2025-09-15 10:05:17 +02:00
parent 4c6fdf6ebe
commit 360881fafd
3 changed files with 5 additions and 3 deletions

View File

@@ -32,6 +32,7 @@ dependencies {
implementation project(':code:common:config')
implementation project(':code:common:renderer')
implementation project(':code:libraries:message-queue')
implementation project(':code:libraries:language-processing')
implementation project(':code:functions:search-query:api')
implementation project(':code:execution:api')
implementation project(':code:index:api')

View File

@@ -4,6 +4,7 @@ import com.google.inject.Guice;
import com.google.inject.Inject;
import com.google.inject.Injector;
import nu.marginalia.WmsaHome;
import nu.marginalia.language.config.LanguageConfiguration;
import nu.marginalia.service.MainClass;
import nu.marginalia.service.ServiceId;
import nu.marginalia.service.discovery.ServiceRegistryIf;
@@ -42,6 +43,9 @@ public class ControlMain extends MainClass {
var registry = injector.getInstance(ServiceRegistryIf.class);
var configuration = injector.getInstance(ServiceConfiguration.class);
// Ensures we download POS tag models from control svc at first boot
var languageConfig = injector.getInstance(LanguageConfiguration.class);
// This must be run before orchestrateBoot, so that the other services don't
// start up until we're done
downloadAncillaryFiles(WmsaHome.getDataPath());

View File

@@ -64,9 +64,6 @@ else
mkdir -p {node-1,node-2}/{work,index,backup,samples/export,uploads}
fi
download_model model/English.DICT https://raw.githubusercontent.com/datquocnguyen/RDRPOSTagger/e0fa60db14eae90b66dc67691f0f519eb19e3e66/Models/POS/English.DICT 356d96a8832b62eb5e0ddac6f0301ada
download_model model/English.RDR https://raw.githubusercontent.com/datquocnguyen/RDRPOSTagger/e0fa60db14eae90b66dc67691f0f519eb19e3e66/Models/POS/English.RDR bec40a1160e12c33a1dd0563677104e4
download_model model/opennlp-sentence.bin https://archive.apache.org/dist/opennlp/models/ud-models-1.0/opennlp-en-ud-ewt-sentence-1.0-1.9.3.bin 5965ada99a2ca77beb8632bb47741b7a
download_model model/segments.bin https://huggingface.co/MarginaliaNu/MarginaliaModelData/resolve/c9339e4224f1dfad7f628809c32687e748198ae3/segments.bin?download=true a2650796c77968b1bd9db0d7c01e3150