{ "$schema": "https://json-schema.org/draft/2020-12/schema", "$id": "https://catalog.lintel.tools/schemas/schemastore/haystack-pipeline/_shared/latest--haystack-pipeline-1.0.0.schema.json", "title": "Haystack Pipeline", "description": "Haystack Pipeline YAML file describing the nodes of the pipelines. For more info read the docs at: ", "x-lintel": { "source": "https://raw.githubusercontent.com/deepset-ai/haystack-json-schema/main/json-schema/haystack-pipeline-1.0.0.schema.json", "sourceSha256": "f906aa2adfdfe1c442c69f1fe4765472681bbe180d1ec53858e367b2ae96423d" }, "type": "object", "properties": { "version": { "title": "Version", "description": "Version of the Haystack Pipeline file.", "type": "string", "const": "1.0.0" }, "components": { "title": "Components", "description": "Component nodes and their configurations, to later be used in the pipelines section. Define here all the building blocks for the pipelines.", "type": "array", "items": { "anyOf": [ { "$ref": "#/$defs/DeepsetCloudDocumentStoreComponent" }, { "$ref": "#/$defs/ElasticsearchDocumentStoreComponent" }, { "$ref": "#/$defs/FAISSDocumentStoreComponent" }, { "$ref": "#/$defs/GraphDBKnowledgeGraphComponent" }, { "$ref": "#/$defs/InMemoryDocumentStoreComponent" }, { "$ref": "#/$defs/Milvus2DocumentStoreComponent" }, { "$ref": "#/$defs/OpenDistroElasticsearchDocumentStoreComponent" }, { "$ref": "#/$defs/OpenSearchDocumentStoreComponent" }, { "$ref": "#/$defs/SQLDocumentStoreComponent" }, { "$ref": "#/$defs/WeaviateDocumentStoreComponent" }, { "$ref": "#/$defs/AzureConverterComponent" }, { "$ref": "#/$defs/CrawlerComponent" }, { "$ref": "#/$defs/DensePassageRetrieverComponent" }, { "$ref": "#/$defs/Docs2AnswersComponent" }, { "$ref": "#/$defs/DocxToTextConverterComponent" }, { "$ref": "#/$defs/ElasticsearchFilterOnlyRetrieverComponent" }, { "$ref": "#/$defs/ElasticsearchRetrieverComponent" }, { "$ref": "#/$defs/EmbeddingRetrieverComponent" }, { "$ref": "#/$defs/EntityExtractorComponent" }, { "$ref": "#/$defs/EvalAnswersComponent" }, { "$ref": "#/$defs/EvalDocumentsComponent" }, { "$ref": "#/$defs/FARMReaderComponent" }, { "$ref": "#/$defs/FileTypeClassifierComponent" }, { "$ref": "#/$defs/ImageToTextConverterComponent" }, { "$ref": "#/$defs/JoinAnswersComponent" }, { "$ref": "#/$defs/JoinDocumentsComponent" }, { "$ref": "#/$defs/MarkdownConverterComponent" }, { "$ref": "#/$defs/PDFToTextConverterComponent" }, { "$ref": "#/$defs/PDFToTextOCRConverterComponent" }, { "$ref": "#/$defs/ParsrConverterComponent" }, { "$ref": "#/$defs/PreProcessorComponent" }, { "$ref": "#/$defs/QuestionGeneratorComponent" }, { "$ref": "#/$defs/RAGeneratorComponent" }, { "$ref": "#/$defs/RCIReaderComponent" }, { "$ref": "#/$defs/RouteDocumentsComponent" }, { "$ref": "#/$defs/SentenceTransformersRankerComponent" }, { "$ref": "#/$defs/Seq2SeqGeneratorComponent" }, { "$ref": "#/$defs/SklearnQueryClassifierComponent" }, { "$ref": "#/$defs/TableReaderComponent" }, { "$ref": "#/$defs/TableTextRetrieverComponent" }, { "$ref": "#/$defs/Text2SparqlRetrieverComponent" }, { "$ref": "#/$defs/TextConverterComponent" }, { "$ref": "#/$defs/TfidfRetrieverComponent" }, { "$ref": "#/$defs/TikaConverterComponent" }, { "$ref": "#/$defs/TransformersDocumentClassifierComponent" }, { "$ref": "#/$defs/TransformersQueryClassifierComponent" }, { "$ref": "#/$defs/TransformersReaderComponent" }, { "$ref": "#/$defs/TransformersSummarizerComponent" }, { "$ref": "#/$defs/TransformersTranslatorComponent" } ] }, "required": [ "type", "name" ], "additionalProperties": true }, "pipelines": { "title": "Pipelines", "description": "Multiple pipelines can be defined using the components from the same YAML file.", "type": "array", "items": { "type": "object", "properties": { "name": { "title": "Name", "description": "Name of the pipeline.", "type": "string" }, "nodes": { "title": "Nodes", "description": "Nodes to be used by this particular pipeline", "type": "array", "items": { "type": "object", "properties": { "name": { "title": "Name", "description": "The name of this particular node in the pipeline. This should be one of the names from the components defined in the same file.", "type": "string" }, "inputs": { "title": "Inputs", "description": "Input parameters for this node.", "type": "array", "items": { "type": "string" } } }, "required": [ "name", "inputs" ], "additionalProperties": false }, "required": [ "name", "nodes" ], "additionalProperties": false }, "additionalProperties": false }, "additionalProperties": false } } }, "required": [ "version", "components", "pipelines" ], "additionalProperties": false, "$defs": { "DeepsetCloudDocumentStoreComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "DeepsetCloudDocumentStore" }, "params": { "title": "Parameters", "type": "object", "properties": { "api_key": { "title": "Api Key", "type": "string" }, "workspace": { "title": "Workspace", "default": "default", "type": "string" }, "index": { "title": "Index", "default": "default", "type": "string" }, "duplicate_documents": { "title": "Duplicate Documents", "default": "overwrite", "type": "string" }, "api_endpoint": { "title": "Api Endpoint", "type": "string" }, "similarity": { "title": "Similarity", "default": "dot_product", "type": "string" }, "return_embedding": { "title": "Return Embedding", "default": false, "type": "boolean" } }, "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "ElasticsearchDocumentStoreComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "ElasticsearchDocumentStore" }, "params": { "title": "Parameters", "type": "object", "properties": { "host": { "title": "Host", "default": "localhost", "anyOf": [ { "type": "string" }, { "type": "array", "items": { "type": "string" } } ] }, "port": { "title": "Port", "default": 9200, "anyOf": [ { "type": "integer" }, { "type": "array", "items": { "type": "integer" } } ] }, "username": { "title": "Username", "default": "", "type": "string" }, "password": { "title": "Password", "default": "", "type": "string" }, "api_key_id": { "title": "Api Key Id", "type": "string" }, "api_key": { "title": "Api Key", "type": "string" }, "aws4auth": { "title": "Aws4Auth" }, "index": { "title": "Index", "default": "document", "type": "string" }, "label_index": { "title": "Label Index", "default": "label", "type": "string" }, "search_fields": { "title": "Search Fields", "default": "content", "anyOf": [ { "type": "string" }, { "type": "array", "items": {} } ] }, "content_field": { "title": "Content Field", "default": "content", "type": "string" }, "name_field": { "title": "Name Field", "default": "name", "type": "string" }, "embedding_field": { "title": "Embedding Field", "default": "embedding", "type": "string" }, "embedding_dim": { "title": "Embedding Dim", "default": 768, "type": "integer" }, "custom_mapping": { "title": "Custom Mapping", "type": "object" }, "excluded_meta_data": { "title": "Excluded Meta Data", "type": "array", "items": {} }, "analyzer": { "title": "Analyzer", "default": "standard", "type": "string" }, "scheme": { "title": "Scheme", "default": "http", "type": "string" }, "ca_certs": { "title": "Ca Certs", "type": "string" }, "verify_certs": { "title": "Verify Certs", "default": true, "type": "boolean" }, "recreate_index": { "title": "Recreate Index", "default": false, "type": "boolean" }, "create_index": { "title": "Create Index", "default": true, "type": "boolean" }, "refresh_type": { "title": "Refresh Type", "default": "wait_for", "type": "string" }, "similarity": { "title": "Similarity", "default": "dot_product" }, "timeout": { "title": "Timeout", "default": 30 }, "return_embedding": { "title": "Return Embedding", "default": false, "type": "boolean" }, "duplicate_documents": { "title": "Duplicate Documents", "default": "overwrite", "type": "string" }, "index_type": { "title": "Index Type", "default": "flat", "type": "string" }, "scroll": { "title": "Scroll", "default": "1d", "type": "string" }, "skip_missing_embeddings": { "title": "Skip Missing Embeddings", "default": true, "type": "boolean" }, "synonyms": { "title": "Synonyms", "type": "array", "items": {} }, "synonym_type": { "title": "Synonym Type", "default": "synonym", "type": "string" }, "use_system_proxy": { "title": "Use System Proxy", "default": false, "type": "boolean" } }, "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "FAISSDocumentStoreComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "FAISSDocumentStore" }, "params": { "title": "Parameters", "type": "object", "properties": { "sql_url": { "title": "Sql Url", "default": "sqlite:///faiss_document_store.db", "type": "string" }, "vector_dim": { "title": "Vector Dim", "type": "integer" }, "embedding_dim": { "title": "Embedding Dim", "default": 768, "type": "integer" }, "faiss_index_factory_str": { "title": "Faiss Index Factory Str", "default": "Flat", "type": "string" }, "faiss_index": { "title": "Faiss Index", "type": "string", "default": null }, "return_embedding": { "title": "Return Embedding", "default": false, "type": "boolean" }, "index": { "title": "Index", "default": "document", "type": "string" }, "similarity": { "title": "Similarity", "default": "dot_product", "type": "string" }, "embedding_field": { "title": "Embedding Field", "default": "embedding", "type": "string" }, "progress_bar": { "title": "Progress Bar", "default": true, "type": "boolean" }, "duplicate_documents": { "title": "Duplicate Documents", "default": "overwrite", "type": "string" }, "faiss_index_path": { "title": "Faiss Index Path", "anyOf": [ { "type": "string" }, { "type": "string", "format": "path" } ] }, "faiss_config_path": { "title": "Faiss Config Path", "anyOf": [ { "type": "string" }, { "type": "string", "format": "path" } ] }, "isolation_level": { "title": "Isolation Level", "type": "string" } }, "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "GraphDBKnowledgeGraphComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "GraphDBKnowledgeGraph" }, "params": { "title": "Parameters", "type": "object", "properties": { "host": { "title": "Host", "default": "localhost", "type": "string" }, "port": { "title": "Port", "default": 7200, "type": "integer" }, "username": { "title": "Username", "default": "", "type": "string" }, "password": { "title": "Password", "default": "", "type": "string" }, "index": { "title": "Index", "type": "string" }, "prefixes": { "title": "Prefixes", "default": "", "type": "string" } }, "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "InMemoryDocumentStoreComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "InMemoryDocumentStore" }, "params": { "title": "Parameters", "type": "object", "properties": { "index": { "title": "Index", "default": "document", "type": "string" }, "label_index": { "title": "Label Index", "default": "label", "type": "string" }, "embedding_field": { "title": "Embedding Field", "default": "embedding", "type": "string" }, "embedding_dim": { "title": "Embedding Dim", "default": 768, "type": "integer" }, "return_embedding": { "title": "Return Embedding", "default": false, "type": "boolean" }, "similarity": { "title": "Similarity", "default": "dot_product", "type": "string" }, "progress_bar": { "title": "Progress Bar", "default": true, "type": "boolean" }, "duplicate_documents": { "title": "Duplicate Documents", "default": "overwrite", "type": "string" }, "use_gpu": { "title": "Use Gpu", "default": true, "type": "boolean" }, "scoring_batch_size": { "title": "Scoring Batch Size", "default": 500000, "type": "integer" } }, "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "Milvus2DocumentStoreComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "Milvus2DocumentStore" }, "params": { "title": "Parameters", "type": "object", "properties": { "sql_url": { "title": "Sql Url", "default": "sqlite:///", "type": "string" }, "host": { "title": "Host", "default": "localhost", "type": "string" }, "port": { "title": "Port", "default": "19530", "type": "string" }, "connection_pool": { "title": "Connection Pool", "default": "SingletonThread", "type": "string" }, "index": { "title": "Index", "default": "document", "type": "string" }, "vector_dim": { "title": "Vector Dim", "type": "integer" }, "embedding_dim": { "title": "Embedding Dim", "default": 768, "type": "integer" }, "index_file_size": { "title": "Index File Size", "default": 1024, "type": "integer" }, "similarity": { "title": "Similarity", "default": "dot_product", "type": "string" }, "index_type": { "title": "Index Type", "default": "IVF_FLAT", "type": "string" }, "index_param": { "title": "Index Param", "type": "object" }, "search_param": { "title": "Search Param", "type": "object" }, "return_embedding": { "title": "Return Embedding", "default": false, "type": "boolean" }, "embedding_field": { "title": "Embedding Field", "default": "embedding", "type": "string" }, "id_field": { "title": "Id Field", "default": "id", "type": "string" }, "custom_fields": { "title": "Custom Fields", "type": "array", "items": {} }, "progress_bar": { "title": "Progress Bar", "default": true, "type": "boolean" }, "duplicate_documents": { "title": "Duplicate Documents", "default": "overwrite", "type": "string" }, "isolation_level": { "title": "Isolation Level", "type": "string" }, "consistency_level": { "title": "Consistency Level", "default": 0, "type": "integer" } }, "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "OpenDistroElasticsearchDocumentStoreComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "OpenDistroElasticsearchDocumentStore" }, "params": { "title": "Parameters", "type": "object", "properties": { "host": { "title": "Host", "default": "https://admin:admin@localhost:9200/" }, "similarity": { "title": "Similarity", "default": "cosine" } }, "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "OpenSearchDocumentStoreComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "OpenSearchDocumentStore" }, "params": { "title": "Parameters", "type": "object", "properties": { "verify_certs": { "title": "Verify Certs", "default": false }, "scheme": { "title": "Scheme", "default": "https" }, "username": { "title": "Username", "default": "admin" }, "password": { "title": "Password", "default": "admin" }, "port": { "title": "Port", "default": 9200 } }, "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "SQLDocumentStoreComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "SQLDocumentStore" }, "params": { "title": "Parameters", "type": "object", "properties": { "url": { "title": "Url", "default": "sqlite://", "type": "string" }, "index": { "title": "Index", "default": "document", "type": "string" }, "label_index": { "title": "Label Index", "default": "label", "type": "string" }, "duplicate_documents": { "title": "Duplicate Documents", "default": "overwrite", "type": "string" }, "check_same_thread": { "title": "Check Same Thread", "default": false, "type": "boolean" }, "isolation_level": { "title": "Isolation Level", "type": "string" } }, "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "WeaviateDocumentStoreComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "WeaviateDocumentStore" }, "params": { "title": "Parameters", "type": "object", "properties": { "host": { "title": "Host", "default": "http://localhost", "anyOf": [ { "type": "string" }, { "type": "array", "items": { "type": "string" } } ] }, "port": { "title": "Port", "default": 8080, "anyOf": [ { "type": "integer" }, { "type": "array", "items": { "type": "integer" } } ] }, "timeout_config": { "title": "Timeout Config", "default": [ 5, 15 ], "type": "array", "items": {} }, "username": { "title": "Username", "type": "string" }, "password": { "title": "Password", "type": "string" }, "index": { "title": "Index", "default": "Document", "type": "string" }, "embedding_dim": { "title": "Embedding Dim", "default": 768, "type": "integer" }, "content_field": { "title": "Content Field", "default": "content", "type": "string" }, "name_field": { "title": "Name Field", "default": "name", "type": "string" }, "similarity": { "title": "Similarity", "default": "cosine", "type": "string" }, "index_type": { "title": "Index Type", "default": "hnsw", "type": "string" }, "custom_schema": { "title": "Custom Schema", "type": "object" }, "return_embedding": { "title": "Return Embedding", "default": false, "type": "boolean" }, "embedding_field": { "title": "Embedding Field", "default": "embedding", "type": "string" }, "progress_bar": { "title": "Progress Bar", "default": true, "type": "boolean" }, "duplicate_documents": { "title": "Duplicate Documents", "default": "overwrite", "type": "string" } }, "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "AzureConverterComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "AzureConverter" }, "params": { "title": "Parameters", "type": "object", "properties": { "endpoint": { "title": "Endpoint", "type": "string" }, "credential_key": { "title": "Credential Key", "type": "string" }, "model_id": { "title": "Model Id", "default": "prebuilt-document", "type": "string" }, "valid_languages": { "title": "Valid Languages", "type": "array", "items": { "type": "string" } }, "save_json": { "title": "Save Json", "default": false, "type": "boolean" }, "preceding_context_len": { "title": "Preceding Context Len", "default": 3, "type": "integer" }, "following_context_len": { "title": "Following Context Len", "default": 3, "type": "integer" }, "merge_multiple_column_headers": { "title": "Merge Multiple Column Headers", "default": true, "type": "boolean" } }, "required": [ "endpoint", "credential_key" ], "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "CrawlerComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "Crawler" }, "params": { "title": "Parameters", "type": "object", "properties": { "output_dir": { "title": "Output Dir", "type": "string" }, "urls": { "title": "Urls", "type": "array", "items": { "type": "string" } }, "crawler_depth": { "title": "Crawler Depth", "default": 1, "type": "integer" }, "filter_urls": { "title": "Filter Urls", "type": "array", "items": {} }, "overwrite_existing_files": { "title": "Overwrite Existing Files", "default": true } }, "required": [ "output_dir" ], "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "DensePassageRetrieverComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "DensePassageRetriever" }, "params": { "title": "Parameters", "type": "object", "properties": { "document_store": { "title": "Document Store", "type": "string" }, "query_embedding_model": { "title": "Query Embedding Model", "default": "facebook/dpr-question_encoder-single-nq-base", "anyOf": [ { "type": "string", "format": "path" }, { "type": "string" } ] }, "passage_embedding_model": { "title": "Passage Embedding Model", "default": "facebook/dpr-ctx_encoder-single-nq-base", "anyOf": [ { "type": "string", "format": "path" }, { "type": "string" } ] }, "model_version": { "title": "Model Version", "type": "string" }, "max_seq_len_query": { "title": "Max Seq Len Query", "default": 64, "type": "integer" }, "max_seq_len_passage": { "title": "Max Seq Len Passage", "default": 256, "type": "integer" }, "top_k": { "title": "Top K", "default": 10, "type": "integer" }, "use_gpu": { "title": "Use Gpu", "default": true, "type": "boolean" }, "batch_size": { "title": "Batch Size", "default": 16, "type": "integer" }, "embed_title": { "title": "Embed Title", "default": true, "type": "boolean" }, "use_fast_tokenizers": { "title": "Use Fast Tokenizers", "default": true, "type": "boolean" }, "infer_tokenizer_classes": { "title": "Infer Tokenizer Classes", "default": false, "type": "boolean" }, "similarity_function": { "title": "Similarity Function", "default": "dot_product", "type": "string" }, "global_loss_buffer_size": { "title": "Global Loss Buffer Size", "default": 150000, "type": "integer" }, "progress_bar": { "title": "Progress Bar", "default": true, "type": "boolean" }, "devices": { "title": "Devices", "type": "array", "items": { "anyOf": [ { "type": "integer" }, { "type": "string" }, { "type": "string" } ] } }, "use_auth_token": { "title": "Use Auth Token", "anyOf": [ { "type": "boolean" }, { "type": "string" } ] } }, "required": [ "document_store" ], "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "Docs2AnswersComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "Docs2Answers" }, "params": { "title": "Parameters", "type": "object", "properties": {}, "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "DocxToTextConverterComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "DocxToTextConverter" }, "params": { "title": "Parameters", "type": "object", "properties": { "remove_numeric_tables": { "title": "Remove Numeric Tables", "default": false, "type": "boolean" }, "valid_languages": { "title": "Valid Languages", "type": "array", "items": { "type": "string" } } }, "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "ElasticsearchFilterOnlyRetrieverComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "ElasticsearchFilterOnlyRetriever" }, "params": { "title": "Parameters", "type": "object", "properties": { "document_store": { "title": "Document Store", "type": "string" }, "top_k": { "title": "Top K", "default": 10, "type": "integer" }, "custom_query": { "title": "Custom Query", "type": "string" } }, "required": [ "document_store" ], "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "ElasticsearchRetrieverComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "ElasticsearchRetriever" }, "params": { "title": "Parameters", "type": "object", "properties": { "document_store": { "title": "Document Store", "type": "string" }, "top_k": { "title": "Top K", "default": 10, "type": "integer" }, "custom_query": { "title": "Custom Query", "type": "string" } }, "required": [ "document_store" ], "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "EmbeddingRetrieverComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "EmbeddingRetriever" }, "params": { "title": "Parameters", "type": "object", "properties": { "document_store": { "title": "Document Store", "type": "string" }, "embedding_model": { "title": "Embedding Model", "type": "string" }, "model_version": { "title": "Model Version", "type": "string" }, "use_gpu": { "title": "Use Gpu", "default": true, "type": "boolean" }, "batch_size": { "title": "Batch Size", "default": 32, "type": "integer" }, "max_seq_len": { "title": "Max Seq Len", "default": 512, "type": "integer" }, "model_format": { "title": "Model Format", "default": "farm", "type": "string" }, "pooling_strategy": { "title": "Pooling Strategy", "default": "reduce_mean", "type": "string" }, "emb_extraction_layer": { "title": "Emb Extraction Layer", "default": -1, "type": "integer" }, "top_k": { "title": "Top K", "default": 10, "type": "integer" }, "progress_bar": { "title": "Progress Bar", "default": true, "type": "boolean" }, "devices": { "title": "Devices", "type": "array", "items": { "anyOf": [ { "type": "integer" }, { "type": "string" }, { "type": "string" } ] } }, "use_auth_token": { "title": "Use Auth Token", "anyOf": [ { "type": "boolean" }, { "type": "string" } ] } }, "required": [ "document_store", "embedding_model" ], "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "EntityExtractorComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "EntityExtractor" }, "params": { "title": "Parameters", "type": "object", "properties": { "model_name_or_path": { "title": "Model Name Or Path", "default": "dslim/bert-base-NER", "type": "string" }, "use_gpu": { "title": "Use Gpu", "default": true, "type": "boolean" } }, "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "EvalAnswersComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "EvalAnswers" }, "params": { "title": "Parameters", "type": "object", "properties": { "skip_incorrect_retrieval": { "title": "Skip Incorrect Retrieval", "default": true, "type": "boolean" }, "open_domain": { "title": "Open Domain", "default": true, "type": "boolean" }, "sas_model": { "title": "Sas Model", "type": "string" }, "debug": { "title": "Debug", "default": false, "type": "boolean" } }, "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "EvalDocumentsComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "EvalDocuments" }, "params": { "title": "Parameters", "type": "object", "properties": { "debug": { "title": "Debug", "default": false, "type": "boolean" }, "open_domain": { "title": "Open Domain", "default": true, "type": "boolean" }, "top_k": { "title": "Top K", "default": 10, "type": "integer" } }, "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "FARMReaderComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "FARMReader" }, "params": { "title": "Parameters", "type": "object", "properties": { "model_name_or_path": { "title": "Model Name Or Path", "type": "string" }, "model_version": { "title": "Model Version", "type": "string" }, "context_window_size": { "title": "Context Window Size", "default": 150, "type": "integer" }, "batch_size": { "title": "Batch Size", "default": 50, "type": "integer" }, "use_gpu": { "title": "Use Gpu", "default": true, "type": "boolean" }, "no_ans_boost": { "title": "No Ans Boost", "default": 0.0, "type": "number" }, "return_no_answer": { "title": "Return No Answer", "default": false, "type": "boolean" }, "top_k": { "title": "Top K", "default": 10, "type": "integer" }, "top_k_per_candidate": { "title": "Top K Per Candidate", "default": 3, "type": "integer" }, "top_k_per_sample": { "title": "Top K Per Sample", "default": 1, "type": "integer" }, "num_processes": { "title": "Num Processes", "type": "integer" }, "max_seq_len": { "title": "Max Seq Len", "default": 256, "type": "integer" }, "doc_stride": { "title": "Doc Stride", "default": 128, "type": "integer" }, "progress_bar": { "title": "Progress Bar", "default": true, "type": "boolean" }, "duplicate_filtering": { "title": "Duplicate Filtering", "default": 0, "type": "integer" }, "use_confidence_scores": { "title": "Use Confidence Scores", "default": true, "type": "boolean" }, "proxies": { "title": "Proxies", "type": "object", "additionalProperties": { "type": "string" } }, "local_files_only": { "title": "Local Files Only", "default": false }, "force_download": { "title": "Force Download", "default": false }, "use_auth_token": { "title": "Use Auth Token", "anyOf": [ { "type": "boolean" }, { "type": "string" } ] } }, "required": [ "model_name_or_path" ], "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "FileTypeClassifierComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "FileTypeClassifier" }, "params": { "title": "Parameters", "type": "object", "properties": { "supported_types": { "title": "Supported Types", "default": [ "txt", "pdf", "md", "docx", "html" ], "type": "array", "items": { "type": "string" } } }, "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "ImageToTextConverterComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "ImageToTextConverter" }, "params": { "title": "Parameters", "type": "object", "properties": { "remove_numeric_tables": { "title": "Remove Numeric Tables", "default": false, "type": "boolean" }, "valid_languages": { "title": "Valid Languages", "default": [ "eng" ], "type": "array", "items": { "type": "string" } } }, "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "JoinAnswersComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "JoinAnswers" }, "params": { "title": "Parameters", "type": "object", "properties": { "join_mode": { "title": "Join Mode", "default": "concatenate", "type": "string" }, "weights": { "title": "Weights", "type": "array", "items": { "type": "number" } }, "top_k_join": { "title": "Top K Join", "type": "integer" } }, "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "JoinDocumentsComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "JoinDocuments" }, "params": { "title": "Parameters", "type": "object", "properties": { "join_mode": { "title": "Join Mode", "default": "concatenate", "type": "string" }, "weights": { "title": "Weights", "type": "array", "items": { "type": "number" } }, "top_k_join": { "title": "Top K Join", "type": "integer" } }, "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "MarkdownConverterComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "MarkdownConverter" }, "params": { "title": "Parameters", "type": "object", "properties": { "remove_numeric_tables": { "title": "Remove Numeric Tables", "default": false, "type": "boolean" }, "valid_languages": { "title": "Valid Languages", "type": "array", "items": { "type": "string" } } }, "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "PDFToTextConverterComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "PDFToTextConverter" }, "params": { "title": "Parameters", "type": "object", "properties": { "remove_numeric_tables": { "title": "Remove Numeric Tables", "default": false, "type": "boolean" }, "valid_languages": { "title": "Valid Languages", "type": "array", "items": { "type": "string" } } }, "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "PDFToTextOCRConverterComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "PDFToTextOCRConverter" }, "params": { "title": "Parameters", "type": "object", "properties": { "remove_numeric_tables": { "title": "Remove Numeric Tables", "default": false, "type": "boolean" }, "valid_languages": { "title": "Valid Languages", "default": [ "eng" ], "type": "array", "items": { "type": "string" } } }, "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "ParsrConverterComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "ParsrConverter" }, "params": { "title": "Parameters", "type": "object", "properties": { "parsr_url": { "title": "Parsr Url", "default": "http://localhost:3001", "type": "string" }, "extractor": { "title": "Extractor", "default": "pdfminer", "enum": [ "pdfminer", "pdfjs" ], "type": "string" }, "table_detection_mode": { "title": "Table Detection Mode", "default": "lattice", "enum": [ "lattice", "stream" ], "type": "string" }, "preceding_context_len": { "title": "Preceding Context Len", "default": 3, "type": "integer" }, "following_context_len": { "title": "Following Context Len", "default": 3, "type": "integer" }, "remove_page_headers": { "title": "Remove Page Headers", "default": false, "type": "boolean" }, "remove_page_footers": { "title": "Remove Page Footers", "default": false, "type": "boolean" }, "remove_table_of_contents": { "title": "Remove Table Of Contents", "default": false, "type": "boolean" }, "valid_languages": { "title": "Valid Languages", "type": "array", "items": { "type": "string" } } }, "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "PreProcessorComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "PreProcessor" }, "params": { "title": "Parameters", "type": "object", "properties": { "clean_whitespace": { "title": "Clean Whitespace", "default": true, "type": "boolean" }, "clean_header_footer": { "title": "Clean Header Footer", "default": false, "type": "boolean" }, "clean_empty_lines": { "title": "Clean Empty Lines", "default": true, "type": "boolean" }, "remove_substrings": { "title": "Remove Substrings", "default": [], "type": "array", "items": { "type": "string" } }, "split_by": { "title": "Split By", "default": "word", "type": "string" }, "split_length": { "title": "Split Length", "default": 200, "type": "integer" }, "split_overlap": { "title": "Split Overlap", "default": 0, "type": "integer" }, "split_respect_sentence_boundary": { "title": "Split Respect Sentence Boundary", "default": true, "type": "boolean" }, "language": { "title": "Language", "default": "en", "type": "string" } }, "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "QuestionGeneratorComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "QuestionGenerator" }, "params": { "title": "Parameters", "type": "object", "properties": { "model_name_or_path": { "title": "Model Name Or Path", "default": "valhalla/t5-base-e2e-qg" }, "model_version": { "title": "Model Version" }, "num_beams": { "title": "Num Beams", "default": 4 }, "max_length": { "title": "Max Length", "default": 256 }, "no_repeat_ngram_size": { "title": "No Repeat Ngram Size", "default": 3 }, "length_penalty": { "title": "Length Penalty", "default": 1.5 }, "early_stopping": { "title": "Early Stopping", "default": true }, "split_length": { "title": "Split Length", "default": 50 }, "split_overlap": { "title": "Split Overlap", "default": 10 }, "use_gpu": { "title": "Use Gpu", "default": true }, "prompt": { "title": "Prompt", "default": "generate questions:" } }, "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "RAGeneratorComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "RAGenerator" }, "params": { "title": "Parameters", "type": "object", "properties": { "model_name_or_path": { "title": "Model Name Or Path", "default": "facebook/rag-token-nq", "type": "string" }, "model_version": { "title": "Model Version", "type": "string" }, "retriever": { "title": "Retriever", "type": "string", "default": null }, "generator_type": { "default": [ 1 ], "allOf": [ { "$ref": "#/$defs/RAGeneratorType" } ] }, "top_k": { "title": "Top K", "default": 2, "type": "integer" }, "max_length": { "title": "Max Length", "default": 200, "type": "integer" }, "min_length": { "title": "Min Length", "default": 2, "type": "integer" }, "num_beams": { "title": "Num Beams", "default": 2, "type": "integer" }, "embed_title": { "title": "Embed Title", "default": true, "type": "boolean" }, "prefix": { "title": "Prefix", "type": "string" }, "use_gpu": { "title": "Use Gpu", "default": true, "type": "boolean" } }, "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "RAGeneratorType": { "title": "RAGeneratorType", "description": "An enumeration.", "enum": [ [ 1 ], 2 ] }, "RCIReaderComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "RCIReader" }, "params": { "title": "Parameters", "type": "object", "properties": { "row_model_name_or_path": { "title": "Row Model Name Or Path", "default": "michaelrglass/albert-base-rci-wikisql-row", "type": "string" }, "column_model_name_or_path": { "title": "Column Model Name Or Path", "default": "michaelrglass/albert-base-rci-wikisql-col", "type": "string" }, "row_model_version": { "title": "Row Model Version", "type": "string" }, "column_model_version": { "title": "Column Model Version", "type": "string" }, "row_tokenizer": { "title": "Row Tokenizer", "type": "string" }, "column_tokenizer": { "title": "Column Tokenizer", "type": "string" }, "use_gpu": { "title": "Use Gpu", "default": true, "type": "boolean" }, "top_k": { "title": "Top K", "default": 10, "type": "integer" }, "max_seq_len": { "title": "Max Seq Len", "default": 256, "type": "integer" } }, "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "RouteDocumentsComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "RouteDocuments" }, "params": { "title": "Parameters", "type": "object", "properties": { "split_by": { "title": "Split By", "default": "content_type", "type": "string" }, "metadata_values": { "title": "Metadata Values", "type": "array", "items": { "type": "string" } } }, "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "SentenceTransformersRankerComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "SentenceTransformersRanker" }, "params": { "title": "Parameters", "type": "object", "properties": { "model_name_or_path": { "title": "Model Name Or Path", "anyOf": [ { "type": "string" }, { "type": "string", "format": "path" } ] }, "model_version": { "title": "Model Version", "type": "string" }, "top_k": { "title": "Top K", "default": 10, "type": "integer" }, "use_gpu": { "title": "Use Gpu", "default": true, "type": "boolean" }, "devices": { "title": "Devices", "type": "array", "items": { "anyOf": [ { "type": "integer" }, { "type": "string" }, { "type": "string" } ] } } }, "required": [ "model_name_or_path" ], "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "Seq2SeqGeneratorComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "Seq2SeqGenerator" }, "params": { "title": "Parameters", "type": "object", "properties": { "model_name_or_path": { "title": "Model Name Or Path", "type": "string" }, "input_converter": { "title": "Input Converter", "type": "string", "default": null }, "top_k": { "title": "Top K", "default": 1, "type": "integer" }, "max_length": { "title": "Max Length", "default": 200, "type": "integer" }, "min_length": { "title": "Min Length", "default": 2, "type": "integer" }, "num_beams": { "title": "Num Beams", "default": 8, "type": "integer" }, "use_gpu": { "title": "Use Gpu", "default": true, "type": "boolean" } }, "required": [ "model_name_or_path" ], "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "SklearnQueryClassifierComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "SklearnQueryClassifier" }, "params": { "title": "Parameters", "type": "object", "properties": { "model_name_or_path": { "title": "Model Name Or Path", "default": "https://ext-models-haystack.s3.eu-central-1.amazonaws.com/gradboost_query_classifier/model.pickle", "anyOf": [ { "type": "string" }, {} ] }, "vectorizer_name_or_path": { "title": "Vectorizer Name Or Path", "default": "https://ext-models-haystack.s3.eu-central-1.amazonaws.com/gradboost_query_classifier/vectorizer.pickle", "anyOf": [ { "type": "string" }, {} ] } }, "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "TableReaderComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "TableReader" }, "params": { "title": "Parameters", "type": "object", "properties": { "model_name_or_path": { "title": "Model Name Or Path", "default": "google/tapas-base-finetuned-wtq", "type": "string" }, "model_version": { "title": "Model Version", "type": "string" }, "tokenizer": { "title": "Tokenizer", "type": "string" }, "use_gpu": { "title": "Use Gpu", "default": true, "type": "boolean" }, "top_k": { "title": "Top K", "default": 10, "type": "integer" }, "top_k_per_candidate": { "title": "Top K Per Candidate", "default": 3, "type": "integer" }, "return_no_answer": { "title": "Return No Answer", "default": false, "type": "boolean" }, "max_seq_len": { "title": "Max Seq Len", "default": 256, "type": "integer" } }, "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "TableTextRetrieverComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "TableTextRetriever" }, "params": { "title": "Parameters", "type": "object", "properties": { "document_store": { "title": "Document Store", "type": "string" }, "query_embedding_model": { "title": "Query Embedding Model", "default": "deepset/bert-small-mm_retrieval-question_encoder", "anyOf": [ { "type": "string", "format": "path" }, { "type": "string" } ] }, "passage_embedding_model": { "title": "Passage Embedding Model", "default": "deepset/bert-small-mm_retrieval-passage_encoder", "anyOf": [ { "type": "string", "format": "path" }, { "type": "string" } ] }, "table_embedding_model": { "title": "Table Embedding Model", "default": "deepset/bert-small-mm_retrieval-table_encoder", "anyOf": [ { "type": "string", "format": "path" }, { "type": "string" } ] }, "model_version": { "title": "Model Version", "type": "string" }, "max_seq_len_query": { "title": "Max Seq Len Query", "default": 64, "type": "integer" }, "max_seq_len_passage": { "title": "Max Seq Len Passage", "default": 256, "type": "integer" }, "max_seq_len_table": { "title": "Max Seq Len Table", "default": 256, "type": "integer" }, "top_k": { "title": "Top K", "default": 10, "type": "integer" }, "use_gpu": { "title": "Use Gpu", "default": true, "type": "boolean" }, "batch_size": { "title": "Batch Size", "default": 16, "type": "integer" }, "embed_meta_fields": { "title": "Embed Meta Fields", "default": [ "name", "section_title", "caption" ], "type": "array", "items": { "type": "string" } }, "use_fast_tokenizers": { "title": "Use Fast Tokenizers", "default": true, "type": "boolean" }, "infer_tokenizer_classes": { "title": "Infer Tokenizer Classes", "default": false, "type": "boolean" }, "similarity_function": { "title": "Similarity Function", "default": "dot_product", "type": "string" }, "global_loss_buffer_size": { "title": "Global Loss Buffer Size", "default": 150000, "type": "integer" }, "progress_bar": { "title": "Progress Bar", "default": true, "type": "boolean" }, "devices": { "title": "Devices", "type": "array", "items": { "anyOf": [ { "type": "integer" }, { "type": "string" }, { "type": "string" } ] } }, "use_auth_token": { "title": "Use Auth Token", "anyOf": [ { "type": "boolean" }, { "type": "string" } ] } }, "required": [ "document_store" ], "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "Text2SparqlRetrieverComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "Text2SparqlRetriever" }, "params": { "title": "Parameters", "type": "object", "properties": { "knowledge_graph": { "title": "Knowledge Graph" }, "model_name_or_path": { "title": "Model Name Or Path" }, "top_k": { "title": "Top K", "default": 1, "type": "integer" } }, "required": [ "knowledge_graph", "model_name_or_path" ], "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "TextConverterComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "TextConverter" }, "params": { "title": "Parameters", "type": "object", "properties": { "remove_numeric_tables": { "title": "Remove Numeric Tables", "default": false, "type": "boolean" }, "valid_languages": { "title": "Valid Languages", "type": "array", "items": { "type": "string" } } }, "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "TfidfRetrieverComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "TfidfRetriever" }, "params": { "title": "Parameters", "type": "object", "properties": { "document_store": { "title": "Document Store", "type": "string" }, "top_k": { "title": "Top K", "default": 10, "type": "integer" }, "auto_fit": { "title": "Auto Fit", "default": true } }, "required": [ "document_store" ], "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "TikaConverterComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "TikaConverter" }, "params": { "title": "Parameters", "type": "object", "properties": { "tika_url": { "title": "Tika Url", "default": "http://localhost:9998/tika", "type": "string" }, "remove_numeric_tables": { "title": "Remove Numeric Tables", "default": false, "type": "boolean" }, "valid_languages": { "title": "Valid Languages", "type": "array", "items": { "type": "string" } } }, "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "TransformersDocumentClassifierComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "TransformersDocumentClassifier" }, "params": { "title": "Parameters", "type": "object", "properties": { "model_name_or_path": { "title": "Model Name Or Path", "default": "bhadresh-savani/distilbert-base-uncased-emotion", "type": "string" }, "model_version": { "title": "Model Version", "type": "string" }, "tokenizer": { "title": "Tokenizer", "type": "string" }, "use_gpu": { "title": "Use Gpu", "default": true, "type": "boolean" }, "return_all_scores": { "title": "Return All Scores", "default": false, "type": "boolean" }, "task": { "title": "Task", "default": "text-classification", "type": "string" }, "labels": { "title": "Labels", "type": "array", "items": { "type": "string" } }, "batch_size": { "title": "Batch Size", "default": -1, "type": "integer" }, "classification_field": { "title": "Classification Field", "type": "string" } }, "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "TransformersQueryClassifierComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "TransformersQueryClassifier" }, "params": { "title": "Parameters", "type": "object", "properties": { "model_name_or_path": { "title": "Model Name Or Path", "default": "shahrukhx01/bert-mini-finetune-question-detection", "anyOf": [ { "type": "string", "format": "path" }, { "type": "string" } ] }, "use_gpu": { "title": "Use Gpu", "default": true, "type": "boolean" } }, "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "TransformersReaderComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "TransformersReader" }, "params": { "title": "Parameters", "type": "object", "properties": { "model_name_or_path": { "title": "Model Name Or Path", "default": "distilbert-base-uncased-distilled-squad", "type": "string" }, "model_version": { "title": "Model Version", "type": "string" }, "tokenizer": { "title": "Tokenizer", "type": "string" }, "context_window_size": { "title": "Context Window Size", "default": 70, "type": "integer" }, "use_gpu": { "title": "Use Gpu", "default": true, "type": "boolean" }, "top_k": { "title": "Top K", "default": 10, "type": "integer" }, "top_k_per_candidate": { "title": "Top K Per Candidate", "default": 4, "type": "integer" }, "return_no_answers": { "title": "Return No Answers", "default": true, "type": "boolean" }, "max_seq_len": { "title": "Max Seq Len", "default": 256, "type": "integer" }, "doc_stride": { "title": "Doc Stride", "default": 128, "type": "integer" } }, "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "TransformersSummarizerComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "TransformersSummarizer" }, "params": { "title": "Parameters", "type": "object", "properties": { "model_name_or_path": { "title": "Model Name Or Path", "default": "google/pegasus-xsum", "type": "string" }, "model_version": { "title": "Model Version", "type": "string" }, "tokenizer": { "title": "Tokenizer", "type": "string" }, "max_length": { "title": "Max Length", "default": 200, "type": "integer" }, "min_length": { "title": "Min Length", "default": 5, "type": "integer" }, "use_gpu": { "title": "Use Gpu", "default": true, "type": "boolean" }, "clean_up_tokenization_spaces": { "title": "Clean Up Tokenization Spaces", "default": true, "type": "boolean" }, "separator_for_single_summary": { "title": "Separator For Single Summary", "default": " ", "type": "string" }, "generate_single_summary": { "title": "Generate Single Summary", "default": false, "type": "boolean" } }, "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false }, "TransformersTranslatorComponent": { "type": "object", "properties": { "name": { "title": "Name", "description": "Custom name for the component. Helpful for visualization and debugging.", "type": "string" }, "type": { "title": "Type", "description": "Haystack Class name for the component.", "type": "string", "const": "TransformersTranslator" }, "params": { "title": "Parameters", "type": "object", "properties": { "model_name_or_path": { "title": "Model Name Or Path", "type": "string" }, "tokenizer_name": { "title": "Tokenizer Name", "type": "string" }, "max_seq_len": { "title": "Max Seq Len", "type": "integer" }, "clean_up_tokenization_spaces": { "title": "Clean Up Tokenization Spaces", "default": true, "type": "boolean" }, "use_gpu": { "title": "Use Gpu", "default": true, "type": "boolean" } }, "required": [ "model_name_or_path" ], "description": "Each parameter can reference other components defined in the same YAML file.", "additionalProperties": false } }, "required": [ "type", "name" ], "additionalProperties": false } } }