Skip to content

Commit

Permalink
fix descriptions (#1920)
Browse files Browse the repository at this point in the history
  • Loading branch information
ebrehault authored Jan 21, 2025
1 parent a1e1506 commit f89146b
Show file tree
Hide file tree
Showing 5 changed files with 9 additions and 9 deletions.
2 changes: 1 addition & 1 deletion libs/common/common.babel
Original file line number Diff line number Diff line change
Expand Up @@ -25805,7 +25805,7 @@
</translations>
</concept_node>
<concept_node>
<name>generative_relation_ranking-help</name>
<name>generative_relation_ranking_help</name>
<description/>
<comment/>
<translations>
Expand Down
4 changes: 2 additions & 2 deletions libs/common/src/assets/i18n/ca.json
Original file line number Diff line number Diff line change
Expand Up @@ -1069,14 +1069,14 @@
"search.configuration.generative-answer.rag-strategies.graph.agentic_graph_only": "Només gràfo personalitzat",
"search.configuration.generative-answer.rag-strategies.graph.agentic_graph_only_help": "Si està activat, només es tenen en compte les relacions extretes d'un agent d'extracció de gràfes per a l'expansió del context.",
"search.configuration.generative-answer.rag-strategies.graph.generative_relation_ranking": "Rànquing amb model generatiu",
"search.configuration.generative-answer.rag-strategies.graph.generative_relation_ranking-help": "Després de la reclassificació regular, utilitzeu un model generatiu per classificar les relacions.\nÉs més lent i consumeix més fitxes, però pot proporcionar millors resultats.",
"search.configuration.generative-answer.rag-strategies.graph.generative_relation_ranking_help": "Després de la reclassificació regular, utilitzeu un model generatiu per classificar les relacions.\nÉs més lent i consumeix més fitxes, però pot proporcionar millors resultats.",
"search.configuration.generative-answer.rag-strategies.graph.hops": "Salts",
"search.configuration.generative-answer.rag-strategies.graph.hops-description": "Nombre de salts que cal fer en explorar el gràfic per a un context rellevant.\nPer exemple, hops=1 explorarà els veïns de les entitats inicials, hops=2 explorarà els veïns dels veïns de les entitats inicials, etc.\nEls valors més grans descobriran relacions més complexes, però també trigaran més temps a calcular.",
"search.configuration.generative-answer.rag-strategies.graph.relation_text_as_paragraphs": "Utilitzar les relacions com a context",
"search.configuration.generative-answer.rag-strategies.graph.relation_text_as_paragraphs_help": "Si està activat, el text de les relacions s'utilitza com a context en lloc de paràgrafs, això permet utilitzar un Top K més gran sense trobar-se amb els límits de context del model generatiu. Si està desactivat, els paràgrafs que contenen les relacions s'utilitzen com a context.",
"search.configuration.generative-answer.rag-strategies.graph.suggest_query_entity_detection": "Cerca difusa per detectar entitats",
"search.configuration.generative-answer.rag-strategies.graph.suggest_query_entity_detection_help": "Utilitza la cerca difusa per detectar entitats. És més ràpid i flexible, però pot tenir problemes per fer coincidir entitats compostes per diverses paraules (es tornarà a la detecció predeterminada si no es detecta cap entitat). La detecció predeterminada és més lenta i precisa, però requereix una coincidència de text exacta entre les entitats del Knowledge Box i les entitats de la consulta.",
"search.configuration.generative-answer.rag-strategies.graph.toggle-description": "Cerca entre les relacions amb els NERs esmentats a la pregunta.",
"search.configuration.generative-answer.rag-strategies.graph.toggle-description": "Recolle peces de context explorant el Knowledge Graph, a partir dels NER presents a la consulta. Funciona millor si el Knowledge Box té activat un agent d'extracció de gràf definit per l'usuari.",
"search.configuration.generative-answer.rag-strategies.graph.toggle-label": "Gràfo de coneixement",
"search.configuration.generative-answer.rag-strategies.graph.top_k": "Top K",
"search.configuration.generative-answer.rag-strategies.graph.top_k-description": "Quantitat màxima de resultats per recuperar.",
Expand Down
4 changes: 2 additions & 2 deletions libs/common/src/assets/i18n/en.json
Original file line number Diff line number Diff line change
Expand Up @@ -1069,14 +1069,14 @@
"search.configuration.generative-answer.rag-strategies.graph.agentic_graph_only": "Custom graph only",
"search.configuration.generative-answer.rag-strategies.graph.agentic_graph_only_help": "If enabled, only relationships extracted from a graph extraction agent are considered for context expansion.",
"search.configuration.generative-answer.rag-strategies.graph.generative_relation_ranking": "Ranking with generative model",
"search.configuration.generative-answer.rag-strategies.graph.generative_relation_ranking-help": "After regular reranking, use a generative model to rank relationships.\nIt is slower and consumes more tokens, but can provide better results.",
"search.configuration.generative-answer.rag-strategies.graph.generative_relation_ranking_help": "After regular reranking, use a generative model to rank relationships.\nIt is slower and consumes more tokens, but can provide better results.",
"search.configuration.generative-answer.rag-strategies.graph.hops": "Hops",
"search.configuration.generative-answer.rag-strategies.graph.hops-description": "Number of hops to take when exploring the graph for relevant context.\nFor example, hops=1 will explore the neighbors of the starting entities, hops=2 will explore the neighbors of the neighbors of the starting entities, and so on.\nBigger values will discover more intricate relationships but will also take more time to compute.",
"search.configuration.generative-answer.rag-strategies.graph.relation_text_as_paragraphs": "Use relations as context",
"search.configuration.generative-answer.rag-strategies.graph.relation_text_as_paragraphs_help": "If enabled, the text of the relationships is used as context instead of paragraphs, this enables to use bigger top K values without running into the generative model's context limits. If disabled, the paragraphs that contain the relationships are used as context.",
"search.configuration.generative-answer.rag-strategies.graph.suggest_query_entity_detection": "Fuzzy search to detect entities",
"search.configuration.generative-answer.rag-strategies.graph.suggest_query_entity_detection_help": "Uses fuzzy search to detect entities. It's faster and more flexible but might have trouble matching entities composed of multiple words (it will fallback to default detection if no entities are detected). The default detection is slower and more accurate but requires an exact text match between Knowledge Box entities and entities in the query.",
"search.configuration.generative-answer.rag-strategies.graph.toggle-description": "Search accross relations with NERs mentioned in the question.",
"search.configuration.generative-answer.rag-strategies.graph.toggle-description": "Collect context pieces by exploring the Knowledge Graph, starting from the NERs present in the query. It works best if the Knowledge Box has a user-defined Graph Extraction agent enabled.",
"search.configuration.generative-answer.rag-strategies.graph.toggle-label": "Knowledge graph",
"search.configuration.generative-answer.rag-strategies.graph.top_k": "Top K",
"search.configuration.generative-answer.rag-strategies.graph.top_k-description": "Maximum amount of results to retrieve.",
Expand Down
4 changes: 2 additions & 2 deletions libs/common/src/assets/i18n/es.json
Original file line number Diff line number Diff line change
Expand Up @@ -1069,14 +1069,14 @@
"search.configuration.generative-answer.rag-strategies.graph.agentic_graph_only": "Solo gráfo personalizado",
"search.configuration.generative-answer.rag-strategies.graph.agentic_graph_only_help": "Si está habilitado, solo se consideran las relaciones extraídas de un agente de extracción de gráfos para la expansión del contexto.",
"search.configuration.generative-answer.rag-strategies.graph.generative_relation_ranking": "Ranking con modelo generativo",
"search.configuration.generative-answer.rag-strategies.graph.generative_relation_ranking-help": "Después de la reclasificación periódica, utilice un modelo generativo para clasificar las relaciones.\nEs más lento y consume más tokens, pero puede brindar mejores resultados.",
"search.configuration.generative-answer.rag-strategies.graph.generative_relation_ranking_help": "Después de la reclasificación periódica, utilice un modelo generativo para clasificar las relaciones.\nEs más lento y consume más tokens, pero puede brindar mejores resultados.",
"search.configuration.generative-answer.rag-strategies.graph.hops": "Saltos",
"search.configuration.generative-answer.rag-strategies.graph.hops-description": "Número de saltos que se deben realizar al explorar el gráfico para obtener el contexto relevante.\nPor ejemplo, hops=1 explorará los vecinos de las entidades iniciales, hops=2 explorará los vecinos de los vecinos de las entidades iniciales, y así sucesivamente.\nLos valores más grandes descubrirán relaciones más complejas, pero también llevarán más tiempo de cálculo.",
"search.configuration.generative-answer.rag-strategies.graph.relation_text_as_paragraphs": "Utilice las relaciones como contexto",
"search.configuration.generative-answer.rag-strategies.graph.relation_text_as_paragraphs_help": "Si está habilitado, el texto de las relaciones se usa como contexto en lugar de párrafos, lo que permite un Top K más alto sin entrar en los límites de contexto del modelo generativo. Si está deshabilitado, los párrafos que contienen las relaciones se usan como contexto.",
"search.configuration.generative-answer.rag-strategies.graph.suggest_query_entity_detection": "Búsqueda difusa para detectar entidades",
"search.configuration.generative-answer.rag-strategies.graph.suggest_query_entity_detection_help": "Utiliza la búsqueda difusa para detectar entidades. Es más rápida y más flexible, pero puede tener problemas para encontrar entidades compuestas por varias palabras (recurrirá a la detección predeterminada si no se detectan entidades). La detección predeterminada es más lenta y precisa, pero requiere una coincidencia exacta de texto entre las entidades de Knowledge Box y las entidades de la consulta.",
"search.configuration.generative-answer.rag-strategies.graph.toggle-description": "Busque entre las relaciones con los NERs mencionados en la pregunta.",
"search.configuration.generative-answer.rag-strategies.graph.toggle-description": "Recopile fragmentos de contexto explorando el gráfico de conocimiento, comenzando por los NER presentes en la consulta. Funciona mejor si el cuadro de conocimiento tiene habilitado un agente de extracción de gráfos definido por el usuario.",
"search.configuration.generative-answer.rag-strategies.graph.toggle-label": "Gráfo de conocimiento",
"search.configuration.generative-answer.rag-strategies.graph.top_k": "Top K",
"search.configuration.generative-answer.rag-strategies.graph.top_k-description": "Cantidad máxima de resultados a recuperar.",
Expand Down
4 changes: 2 additions & 2 deletions libs/common/src/assets/i18n/fr.json
Original file line number Diff line number Diff line change
Expand Up @@ -1069,14 +1069,14 @@
"search.configuration.generative-answer.rag-strategies.graph.agentic_graph_only": "Graphe personnalisé uniquement",
"search.configuration.generative-answer.rag-strategies.graph.agentic_graph_only_help": "Si cette option est activée, seules les relations extraites d'un agent d'extraction de graphe sont prises en compte pour l'extension de contexte.",
"search.configuration.generative-answer.rag-strategies.graph.generative_relation_ranking": "Classement avec modèle génératif",
"search.configuration.generative-answer.rag-strategies.graph.generative_relation_ranking-help": "Après un reclassement régulier, utilisez un modèle génératif pour classer les relations.\nIl est plus lent et consomme plus de jetons, mais peut fournir de meilleurs résultats.",
"search.configuration.generative-answer.rag-strategies.graph.generative_relation_ranking_help": "Après un reclassement régulier, utilisez un modèle génératif pour classer les relations.\nIl est plus lent et consomme plus de jetons, mais peut fournir de meilleurs résultats.",
"search.configuration.generative-answer.rag-strategies.graph.hops": "Sauts",
"search.configuration.generative-answer.rag-strategies.graph.hops-description": "Nombre de sauts à effectuer lors de l'exploration du graphique pour un contexte pertinent.\nPar exemple, hops=1 explorera les voisins des entités de départ, hops=2 explorera les voisins des voisins des entités de départ, et ainsi de suite.\nDes valeurs plus élevées découvriront des relations plus complexes mais prendront également plus de temps à calculer.",
"search.configuration.generative-answer.rag-strategies.graph.relation_text_as_paragraphs": "Utiliser les relations comme contexte",
"search.configuration.generative-answer.rag-strategies.graph.relation_text_as_paragraphs_help": "Si cette option est activée, le texte des relations est utilisé comme contexte au lieu des paragraphes, ce qui permet d'utiliser un Top K plus grand sans atteindre les limites de contexte du modèle génératif. Si elle est désactivée, les paragraphes qui contiennent les relations sont utilisés comme contexte.",
"search.configuration.generative-answer.rag-strategies.graph.suggest_query_entity_detection": "Recherche floue pour détecter des entités",
"search.configuration.generative-answer.rag-strategies.graph.suggest_query_entity_detection_help": "Utilise la recherche floue pour détecter les entités. Elle est plus rapide et plus flexible, mais peut rencontrer des difficultés pour faire correspondre les entités composées de plusieurs mots (elle reviendra à la détection par défaut si aucune entité n'est détectée). La détection par défaut est plus lente et plus précise, mais nécessite une correspondance exacte du texte entre les entités de Knowledge Box et les entités de la requête.",
"search.configuration.generative-answer.rag-strategies.graph.toggle-description": "Rechercher parmi les relations avec les NERs mentionnés dans la question.",
"search.configuration.generative-answer.rag-strategies.graph.toggle-description": "Collecte des éléments de contexte en explorant le Knowledge Graph, à partir des NER présents dans la requête. Cela fonctionne mieux si la Knowledge Box dispose d'un agent d'extraction de graphes défini par l'utilisateur activé.",
"search.configuration.generative-answer.rag-strategies.graph.toggle-label": "Graphe de connaissances",
"search.configuration.generative-answer.rag-strategies.graph.top_k": "Top K",
"search.configuration.generative-answer.rag-strategies.graph.top_k-description": "Quantité maximale de résultats à récupérer.",
Expand Down

0 comments on commit f89146b

Please sign in to comment.