jRDF2Vec

steps to generate jar and save it locally and test locally

1. docker build --no-cache -t jrdf2vec . [/home/illusion/projects/docker/jRDF2Vec/jRDF2Vec-src]

2. docker-compose -f docker-compose.yml up [/home/illusion/projects/docker/jRDF2Vec/jRDF2Vec-docker]

3. docker exec -it jrdf2vec bash [check the /app/jrdf2vec.jar if it exist]

   for window:

    - ${PWD}
    - docker exec -it jrdf2vec sh -c "java -jar /app/jrdf2vec.jar -checkInstallation"
   

4. check installation status:

   • activate the conda env jrdf2vec_3_9_env and use the correct requirements.txt [requirements-correct.txt] java -jar /app/jrdf2vec.jar -checkInstallation [this is when inside the container] or docker exec -it jrdf2vec java -jar /app/jrdf2vec.jar -checkInstallation [this is to run when outside the container]

     Note: If error message is this: Missing Requirements: python==3.9.6 Installation is not ok! [❌]

       Ignore it as this was added manually to requirements.txt
     

5. docker cp jrdf2vec:/app/jrdf2vec.jar ${PWD} [this will copy the jar available inside the container to the host machine]

6. java -jar ../../jar/jrdf2vec.jar -graph sample_dbpedia_nt_file.nt [test the jar]

7. java -jar ../../jar/jrdf2vec.jar -light sample_dbpedia_entity_file.txt -graph sample_dbpedia_nt_file.nt [test the jar]

steps to generate jar and save it locally and test with docker

8. follow steps 1 & 2 as mentioned above.

9. docker exec -it jrdf2vec java -jar /app/jrdf2vec.jar -graph sample_dbpedia_nt_file.nt

10. docker command: docker run -it --rm
    -v ${PWD}/data:/data
    jrdf2vec-local
    -light /data/src/test/resources/sample_dbpedia_entity_file.txt
    -graph /data/src/test/resources/sample_dbpedia_nt_file.nt

Run Jar to generate vectors:

12. cd /home/illusion/projects/docker/jRDF2Vec/jRDF2Vec-docker/test
13. create test folders: test1 and copy the relevant graph files
14. cd /test1 and run following: java -jar ../../jar/jrdf2vec.jar -graph sample_dbpedia_nt_file.nt
15. check if the model.kv and vectors.txt file generated successfully.

Generating a Vocabulary Text File

16. java -jar jrdf2vec.jar -generateVocabularyFile walks/vectors.txt

Analyzing the Embedding Vocabulary

17. java -jar ../../jar/jrdf2vec.jar -analyzeVocab sample_dbpedia_nt_file.nt /walks/model.kv

18. run flask app: flask --app python-server run run in debug mode: flask --debug --app python_server run

Tests:

- command run from the following directory: /home/illusion/projects/docker/jRDF2Vec/jRDF2Vec-docker
- test-1: java -jar ../../jar/jrdf2vec.jar -graph sample_dbpedia_nt_file.nt
- test-2: java -Xmx10G -jar jar/jrdf2vec.jar -light test/test-2/orgs-iri-all.txt -graph test/test-2/orgs-full-graph-all.ttl
- test-3: java -Xmx10G -jar jar/jrdf2vec.jar -trainingMode cbow -threads 24 -light test/test-2/orgs-iri-all.txt -graph test/test-2/orgs-full-graph-all.ttl
- test-4: java -Xmx10G -jar jar/jrdf2vec.jar -trainingMode cbow -depth 10 -threads 30 -light test/test-2/orgs-iri-all.txt -graph test/test-2/orgs-full-graph-all.ttl

Ollama:

• sudo systemctl stop ollama

• ollama serve

  NAME ID SIZE MODIFIED
  codellama:13b 9f438cb9cd58 7.4 GB 33 seconds ago codellama:34b 685be00e1532 19 GB 5 minutes ago codellama:70b e59b580dfce7 38 GB 17 minutes ago codellama:7b 8fdf8f752f6e 3.8 GB 17 seconds ago codellama:latest 8fdf8f752f6e 3.8 GB 49 minutes ago