Skip to main content
Version: 25.6

snorkelai.sdk.develop

(Beta) Object-oriented SDK for Snorkel Flow

Classes

Batch(name, uid, dataset_uid, label_schemas, ...)The Batch object represents an annotation batch in Snorkel Flow.
Benchmark(*args, **kwargs)A benchmark is the collection of characteristics that you care about for a particular GenAI application, and the measurements you use to assess the performance against those characteristics.
BenchmarkExecution(*args, **kwargs)Represents a single execution run of a benchmark for a dataset.
BenchmarkExecutionExportFormat(value)An enumeration.
BenchmarkExportFormat(value)An enumeration.
CodeEvaluator(*args, **kwargs)An evaluator that uses custom Python code to assess an AI application's responses.
Criteria(*args, **kwargs)A criteria represents a specific characteristic or feature being evaluated as part of a benchmark.
CsvExportConfig([sep, quotechar, escapechar])
Dataset(name, uid, mta_enabled)The Dataset object represents a dataset in Snorkel Flow.
Evaluator(*args, **kwargs)Base class for all evaluators.
JsonExportConfig()
LabelSchema(name, uid, dataset_uid, ...[, ...])The LabelSchema object represents a label schema in Snorkel Flow.
ModelNode(uid, application_uid, config)ModelNode class represents a model node.
Node(uid, application_uid, config)The Node object represents atomic data processing units in Snorkel Flow.
OperatorNode(uid, application_uid, config)OperatorNode class represents a non-model, operator node.
PromptEvaluator(*args, **kwargs)An evaluator that uses LLM prompts to assess model outputs.
Slice(dataset, slice_uid, name[, ...])