java.lang.Exception: empty response from the recommender #1547
Description
Description
I get an error when calling the recommend API.
Request
POST {{host}}/recommend/similar_reservations
Content-Type: application/json
{
"count": 10,
"user": "540851",
"items": ["2151"]
}
Config
models:
similar_reservations:
type: als
interactions:
- purchase
factors: 200
iterations: 200Error Log
08:05:52.638 ERROR ai.metarank.main.command.Serve$ - java.lang.Exception: empty response from the recommender
java.lang.Exception: empty response from the recommender
at ai.metarank.ml.recommend.MFRecommender$EmbeddingSimilarityModel.$anonfun$predict$6(MFRecommender.scala:74)
at map @ ai.metarank.ml.recommend.MFRecommender$EmbeddingSimilarityModel.$anonfun$predict$5(MFRecommender.scala:74)
at apply @ ai.metarank.ml.recommend.MFRecommender$EmbeddingSimilarityModel.$anonfun$predict$2(MFRecommender.scala:73)
at flatMap @ ai.metarank.ml.recommend.MFRecommender$EmbeddingSimilarityModel.$anonfun$predict$2(MFRecommender.scala:73)
at apply @ ai.metarank.ml.recommend.embedding.HnswJavaIndex$HnswIndexReader.lookup(HnswJavaIndex.scala:26)
at flatMap @ ai.metarank.ml.recommend.MFRecommender$EmbeddingSimilarityModel.$anonfun$predict$1(MFRecommender.scala:72)
at flatMap @ ai.metarank.ml.recommend.MFRecommender$EmbeddingSimilarityModel.predict(MFRecommender.scala:68)
at flatMap @ ai.metarank.ml.Recommender.$anonfun$recommend$4(Recommender.scala:36)
at apply @ ai.metarank.fstore.memory.MemModelStore.get(MemModelStore.scala:13)
at flatMap @ ai.metarank.fstore.cache.CachedModelStore.get(CachedModelStore.scala:21)
at flatMap @ ai.metarank.ml.Recommender.loadModel(Recommender.scala:48)
at flatMap @ ai.metarank.ml.Recommender.$anonfun$recommend$3(Recommender.scala:35)
at flatMap @ ai.metarank.ml.Recommender.$anonfun$recommend$2(Recommender.scala:20)
at apply @ ai.metarank.ml.Recommender.recommend(Recommender.scala:19)
at flatMap @ ai.metarank.ml.Recommender.recommend(Recommender.scala:19)
at map @ ai.metarank.api.routes.RecommendApi$$anonfun$1.$anonfun$applyOrElse$2(RecommendApi.scala:20)
at flatMap @ ai.metarank.api.routes.RecommendApi$$anonfun$1.$anonfun$applyOrElse$1(RecommendApi.scala:19)
08:05:52.639 INFO ai.metarank.main.command.Serve$ - HTTP/1.1 POST /recommend/similar_reservations
Some item IDs return results, but others cause this error.
Why does this happen and how to fix it? @shuttie