Knowledgator SDKs
Knowledgator provides official client libraries in Python and TypeScript/JavaScript. Both SDKs provide complete access to all inference models, dataset management, and fine-tuning pipelines with async/await support, retry policies, and type definitions.
Available SDKs
| Language | Package Name | Installation | Source |
|---|---|---|---|
| Python | knowledgator | pip install knowledgator | GitHub |
| TypeScript / Node.js | @knowledgator/typescript-sdk | npm install @knowledgator/typescript-sdk | GitHub |
Services & Modules Available Across All SDKs
Both SDKs mirror the exact same service structure as the REST API:
- Comprehend-it: Zero-shot and few-shot text classification with confidence scores.
- Extract-it: Zero-shot named entity recognition (NER), relation extraction, and summarization.
- Instruct-it: Structured extraction of information conforming to a schema.
- Datasets: Dataset CRUD, file uploads, direct & streaming downloads, dataset planning, and async generation.
- Fine-tuning: Start, monitor, evaluate, and stop custom training jobs.
- Open Models: Query finetuned or custom models using dynamic IDs.
- Raw Data: Upload and manage unannotated raw corpora.
Getting Started
Select your language to view installation instructions, configuration parameters, and code examples: