mirror of
https://github.com/deepset-ai/haystack.git
synced 2026-01-08 13:06:29 +00:00
2a840ee248
* Make YAML files get the same version as Haystack and throw warning at load in case of mismatch * Update version of most YAMLs in the codebase (aesthethic chamge, only to avoid the warning). * Remove quotes from version in tests * Fix version in generate_json_schema.py Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
2590 lines
74 KiB
JSON
2590 lines
74 KiB
JSON
{
|
|
"$schema": "http://json-schema.org/draft-07/schema",
|
|
"$id": "https://haystack.deepset.ai/json-schemas/haystack-pipeline-1.1.0.schema.json",
|
|
"title": "Haystack Pipeline",
|
|
"description": "Haystack Pipeline YAML file describing the nodes of the pipelines. For more info read the docs at: https://haystack.deepset.ai/components/pipelines#yaml-file-definitions",
|
|
"type": "object",
|
|
"properties": {
|
|
"version": {
|
|
"title": "Version",
|
|
"description": "Version of the Haystack Pipeline file.",
|
|
"type": "string",
|
|
"const": "1.1.0"
|
|
},
|
|
"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": [
|
|
{
|
|
"$ref": "#/definitions/AzureConverterComponent"
|
|
},
|
|
{
|
|
"$ref": "#/definitions/CrawlerComponent"
|
|
},
|
|
{
|
|
"$ref": "#/definitions/DensePassageRetrieverComponent"
|
|
},
|
|
{
|
|
"$ref": "#/definitions/Docs2AnswersComponent"
|
|
},
|
|
{
|
|
"$ref": "#/definitions/DocxToTextConverterComponent"
|
|
},
|
|
{
|
|
"$ref": "#/definitions/ElasticsearchFilterOnlyRetrieverComponent"
|
|
},
|
|
{
|
|
"$ref": "#/definitions/ElasticsearchRetrieverComponent"
|
|
},
|
|
{
|
|
"$ref": "#/definitions/EmbeddingRetrieverComponent"
|
|
},
|
|
{
|
|
"$ref": "#/definitions/EntityExtractorComponent"
|
|
},
|
|
{
|
|
"$ref": "#/definitions/EvalAnswersComponent"
|
|
},
|
|
{
|
|
"$ref": "#/definitions/EvalDocumentsComponent"
|
|
},
|
|
{
|
|
"$ref": "#/definitions/FARMReaderComponent"
|
|
},
|
|
{
|
|
"$ref": "#/definitions/FileTypeClassifierComponent"
|
|
},
|
|
{
|
|
"$ref": "#/definitions/ImageToTextConverterComponent"
|
|
},
|
|
{
|
|
"$ref": "#/definitions/JoinDocumentsComponent"
|
|
},
|
|
{
|
|
"$ref": "#/definitions/MarkdownConverterComponent"
|
|
},
|
|
{
|
|
"$ref": "#/definitions/PDFToTextConverterComponent"
|
|
},
|
|
{
|
|
"$ref": "#/definitions/PDFToTextOCRConverterComponent"
|
|
},
|
|
{
|
|
"$ref": "#/definitions/ParsrConverterComponent"
|
|
},
|
|
{
|
|
"$ref": "#/definitions/PreProcessorComponent"
|
|
},
|
|
{
|
|
"$ref": "#/definitions/QuestionGeneratorComponent"
|
|
},
|
|
{
|
|
"$ref": "#/definitions/RAGeneratorComponent"
|
|
},
|
|
{
|
|
"$ref": "#/definitions/RCIReaderComponent"
|
|
},
|
|
{
|
|
"$ref": "#/definitions/SentenceTransformersRankerComponent"
|
|
},
|
|
{
|
|
"$ref": "#/definitions/Seq2SeqGeneratorComponent"
|
|
},
|
|
{
|
|
"$ref": "#/definitions/SklearnQueryClassifierComponent"
|
|
},
|
|
{
|
|
"$ref": "#/definitions/TableReaderComponent"
|
|
},
|
|
{
|
|
"$ref": "#/definitions/TableTextRetrieverComponent"
|
|
},
|
|
{
|
|
"$ref": "#/definitions/Text2SparqlRetrieverComponent"
|
|
},
|
|
{
|
|
"$ref": "#/definitions/TextConverterComponent"
|
|
},
|
|
{
|
|
"$ref": "#/definitions/TfidfRetrieverComponent"
|
|
},
|
|
{
|
|
"$ref": "#/definitions/TikaConverterComponent"
|
|
},
|
|
{
|
|
"$ref": "#/definitions/TransformersDocumentClassifierComponent"
|
|
},
|
|
{
|
|
"$ref": "#/definitions/TransformersQueryClassifierComponent"
|
|
},
|
|
{
|
|
"$ref": "#/definitions/TransformersReaderComponent"
|
|
},
|
|
{
|
|
"$ref": "#/definitions/TransformersSummarizerComponent"
|
|
},
|
|
{
|
|
"$ref": "#/definitions/TransformersTranslatorComponent"
|
|
}
|
|
]
|
|
},
|
|
"required": [
|
|
"type",
|
|
"name"
|
|
],
|
|
"additionalProperties": false
|
|
},
|
|
"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"
|
|
}
|
|
}
|
|
},
|
|
"additionalProperties": false
|
|
},
|
|
"required": [
|
|
"name",
|
|
"nodes"
|
|
],
|
|
"additionalProperties": false
|
|
}
|
|
},
|
|
"additionalProperties": false
|
|
}
|
|
}
|
|
},
|
|
"required": [
|
|
"version",
|
|
"components",
|
|
"pipelines"
|
|
],
|
|
"additionalProperties": false,
|
|
"definitions": {
|
|
"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"
|
|
],
|
|
"additionalProperties": false,
|
|
"description": "Each parameter can reference other components defined in the same YAML file."
|
|
}
|
|
},
|
|
"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"
|
|
],
|
|
"additionalProperties": false,
|
|
"description": "Each parameter can reference other components defined in the same YAML file."
|
|
}
|
|
},
|
|
"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"
|
|
],
|
|
"additionalProperties": false,
|
|
"description": "Each parameter can reference other components defined in the same YAML file."
|
|
}
|
|
},
|
|
"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": {},
|
|
"additionalProperties": false,
|
|
"description": "Each parameter can reference other components defined in the same YAML file."
|
|
}
|
|
},
|
|
"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"
|
|
}
|
|
}
|
|
},
|
|
"additionalProperties": false,
|
|
"description": "Each parameter can reference other components defined in the same YAML file."
|
|
}
|
|
},
|
|
"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"
|
|
],
|
|
"additionalProperties": false,
|
|
"description": "Each parameter can reference other components defined in the same YAML file."
|
|
}
|
|
},
|
|
"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"
|
|
],
|
|
"additionalProperties": false,
|
|
"description": "Each parameter can reference other components defined in the same YAML file."
|
|
}
|
|
},
|
|
"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"
|
|
],
|
|
"additionalProperties": false,
|
|
"description": "Each parameter can reference other components defined in the same YAML file."
|
|
}
|
|
},
|
|
"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"
|
|
}
|
|
},
|
|
"additionalProperties": false,
|
|
"description": "Each parameter can reference other components defined in the same YAML file."
|
|
}
|
|
},
|
|
"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"
|
|
}
|
|
},
|
|
"additionalProperties": false,
|
|
"description": "Each parameter can reference other components defined in the same YAML file."
|
|
}
|
|
},
|
|
"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"
|
|
}
|
|
},
|
|
"additionalProperties": false,
|
|
"description": "Each parameter can reference other components defined in the same YAML file."
|
|
}
|
|
},
|
|
"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"
|
|
],
|
|
"additionalProperties": false,
|
|
"description": "Each parameter can reference other components defined in the same YAML file."
|
|
}
|
|
},
|
|
"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"
|
|
}
|
|
}
|
|
},
|
|
"additionalProperties": false,
|
|
"description": "Each parameter can reference other components defined in the same YAML file."
|
|
}
|
|
},
|
|
"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"
|
|
}
|
|
}
|
|
},
|
|
"additionalProperties": false,
|
|
"description": "Each parameter can reference other components defined in the same YAML file."
|
|
}
|
|
},
|
|
"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"
|
|
}
|
|
},
|
|
"additionalProperties": false,
|
|
"description": "Each parameter can reference other components defined in the same YAML file."
|
|
}
|
|
},
|
|
"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"
|
|
}
|
|
}
|
|
},
|
|
"additionalProperties": false,
|
|
"description": "Each parameter can reference other components defined in the same YAML file."
|
|
}
|
|
},
|
|
"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"
|
|
}
|
|
}
|
|
},
|
|
"additionalProperties": false,
|
|
"description": "Each parameter can reference other components defined in the same YAML file."
|
|
}
|
|
},
|
|
"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"
|
|
}
|
|
}
|
|
},
|
|
"additionalProperties": false,
|
|
"description": "Each parameter can reference other components defined in the same YAML file."
|
|
}
|
|
},
|
|
"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": 1,
|
|
"type": "integer"
|
|
},
|
|
"following_context_len": {
|
|
"title": "Following Context Len",
|
|
"default": 1,
|
|
"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"
|
|
}
|
|
}
|
|
},
|
|
"additionalProperties": false,
|
|
"description": "Each parameter can reference other components defined in the same YAML file."
|
|
}
|
|
},
|
|
"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"
|
|
},
|
|
"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"
|
|
}
|
|
},
|
|
"additionalProperties": false,
|
|
"description": "Each parameter can reference other components defined in the same YAML file."
|
|
}
|
|
},
|
|
"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:"
|
|
}
|
|
},
|
|
"additionalProperties": false,
|
|
"description": "Each parameter can reference other components defined in the same YAML file."
|
|
}
|
|
},
|
|
"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": "#/definitions/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"
|
|
}
|
|
},
|
|
"additionalProperties": false,
|
|
"description": "Each parameter can reference other components defined in the same YAML file."
|
|
}
|
|
},
|
|
"required": [
|
|
"type",
|
|
"name"
|
|
],
|
|
"additionalProperties": false
|
|
},
|
|
"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"
|
|
}
|
|
},
|
|
"additionalProperties": false,
|
|
"description": "Each parameter can reference other components defined in the same YAML file."
|
|
}
|
|
},
|
|
"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"
|
|
],
|
|
"additionalProperties": false,
|
|
"description": "Each parameter can reference other components defined in the same YAML file."
|
|
}
|
|
},
|
|
"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"
|
|
],
|
|
"additionalProperties": false,
|
|
"description": "Each parameter can reference other components defined in the same YAML file."
|
|
}
|
|
},
|
|
"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"
|
|
},
|
|
{}
|
|
]
|
|
}
|
|
},
|
|
"additionalProperties": false,
|
|
"description": "Each parameter can reference other components defined in the same YAML file."
|
|
}
|
|
},
|
|
"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"
|
|
}
|
|
},
|
|
"additionalProperties": false,
|
|
"description": "Each parameter can reference other components defined in the same YAML file."
|
|
}
|
|
},
|
|
"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"
|
|
],
|
|
"additionalProperties": false,
|
|
"description": "Each parameter can reference other components defined in the same YAML file."
|
|
}
|
|
},
|
|
"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"
|
|
],
|
|
"additionalProperties": false,
|
|
"description": "Each parameter can reference other components defined in the same YAML file."
|
|
}
|
|
},
|
|
"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"
|
|
}
|
|
}
|
|
},
|
|
"additionalProperties": false,
|
|
"description": "Each parameter can reference other components defined in the same YAML file."
|
|
}
|
|
},
|
|
"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"
|
|
],
|
|
"additionalProperties": false,
|
|
"description": "Each parameter can reference other components defined in the same YAML file."
|
|
}
|
|
},
|
|
"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"
|
|
}
|
|
}
|
|
},
|
|
"additionalProperties": false,
|
|
"description": "Each parameter can reference other components defined in the same YAML file."
|
|
}
|
|
},
|
|
"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"
|
|
}
|
|
},
|
|
"additionalProperties": false,
|
|
"description": "Each parameter can reference other components defined in the same YAML file."
|
|
}
|
|
},
|
|
"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"
|
|
}
|
|
},
|
|
"additionalProperties": false,
|
|
"description": "Each parameter can reference other components defined in the same YAML file."
|
|
}
|
|
},
|
|
"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"
|
|
}
|
|
},
|
|
"additionalProperties": false,
|
|
"description": "Each parameter can reference other components defined in the same YAML file."
|
|
}
|
|
},
|
|
"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"
|
|
}
|
|
},
|
|
"additionalProperties": false,
|
|
"description": "Each parameter can reference other components defined in the same YAML file."
|
|
}
|
|
},
|
|
"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"
|
|
],
|
|
"additionalProperties": false,
|
|
"description": "Each parameter can reference other components defined in the same YAML file."
|
|
}
|
|
},
|
|
"required": [
|
|
"type",
|
|
"name"
|
|
],
|
|
"additionalProperties": false
|
|
},
|
|
"RAGeneratorType": {
|
|
"title": "RAGeneratorType",
|
|
"description": "An enumeration.",
|
|
"enum": [
|
|
[
|
|
1
|
|
],
|
|
2
|
|
]
|
|
}
|
|
}
|
|
} |