Taewoon Kim

It is trained on both MELD and IEMOCAP datasets.

I think so. It's been a while so I am not 100% sure tho. I'll check it out when I have time.

> Are you using the same similarity measure? yes (cosine). everything is the same in this experiment, except the vector search implementations (jvector vs. hnswlib).

> > Are you using the same similarity measure? > > yes (cosine). everything is the same in this experiment, except the vector search implementations (jvector vs. hnswlib). oh yeah...

> > oh yeah jvector's cosine metric (range [0, 1]) is half of hnswlib metric (range [0, 2]) > > Is it a bug or a feaure? I think either...

looked more into the warmup & persistence issue of jvector at https://github.com/ArcadeData/arcadedb/issues/2915

@robfrank @gramian I looked more into jvector vs. hnswlib returning different results. ```markdown # Benchmark Results This benchmark compares ArcadeDB's vector index implementations (JVector vs HNSWLib) across multiple dataset sizes...

@lvca At the moment, the test is done by me internally pulling the latest snapshot of arcadedb from Docker Hub, unzipping the necessary JARs, bundling them up in python wheel,...

@lvca https://github.com/humemai/arcadedb-embedded-python/blob/48dbaa2fadbfe02f56a957b629d7b596d9703796/bindings/python/examples/06_vector_search_recommendations.py is the script to run this. I'm rerunning it with two datasets ```bash ./run_with_memory_monitor.sh 06-angular "ARCADEDB_JVM_MAX_HEAP='32g' ARCADEDB_JVM_ARGS='-Xms32g' python3 benchmark_vector_params.py --dataset glove-100-angular" ``` and ```bash ./run_with_memory_monitor.sh 06-euclidean "ARCADEDB_JVM_MAX_HEAP='32g' ARCADEDB_JVM_ARGS='-Xms32g'...

@lvca @robfrank I finished running the experiments (ArcadeDB's jvector vs. FAISS's HNSW) https://github.com/humemai/arcadedb-embedded-python/blob/main/bindings/python/examples/benchmark-vector/README.md ## TL;DR * **Speed:** ArcadeDB (JVector) is surprisingly fast, often matching or beating in-memory FAISS. * **Recall:**...