tinker-types

Explain Tinker SDK types Datum, ModelInput, and TensorData for training pipelines.

4.0k|507|Updated Jul 14, 2025

Search Tags:#sdk #training #types #datum #tinker #model-input #tensordata

Ready to run

npx skills add https://github.com/thinking-machines-lab/tinker-cookbook --skill tinker-types

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: tinker-types
Source: https://github.com/thinking-machines-lab/tinker-cookbook/tree/main/.claude/skills/tinker-types
Command: npx skills add https://github.com/thinking-machines-lab/tinker-cookbook --skill tinker-types

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill helps developers understand and correctly use core Tinker SDK types such as Datum, ModelInput, and TensorData, which are essential for building training data and structuring model inputs.

Core Features & Use Cases

Core data types: Datum, ModelInput, TensorData, and related helpers.
Guidance on constructing model inputs and interpreting responses.
Use cases include preparing training data, validating type schemas, and debugging data pipelines.

Quick Start

Refer to the type reference to begin constructing Datum and ModelInput for your training data.

Frequently Asked Questions about tinker-types

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I construct Datum and ModelInput for Tinker SDK training data?▼

To construct Datum and ModelInput for Tinker SDK training data, instantiate the core type schemas using the provided type references and examples to ensure your training pipelines receive structurally valid inputs.

What is the difference between Datum and TensorData in Tinker SDK workflows?▼

Datum and TensorData are core Tinker SDK types used to build training data and structure model inputs. TensorData specifically handles tensor representations, while Datum serves as a foundational data wrapper for training pipelines.

Why does my Tinker SDK data pipeline throw type validation errors?▼

Type validation errors in Tinker SDK data pipelines occur when instantiated types like Datum, ModelInput, or TensorData do not match the expected schemas. Use the type reference to validate schemas and debug data pipelines.

How do I interpret API responses and error types in Tinker workflows?▼

Interpreting API responses and error types in Tinker workflows involves mapping the returned data against the core SDK types. The skill provides guidance on interpreting responses and handling common error types encountered during data processing.

Can I use Tinker SDK types to validate training data schemas before model execution?▼

Yes, you can use Tinker SDK types to validate training data schemas before model execution. Constructing ModelInput and Datum instances according to the type reference ensures your data conforms to the required structure for training pipelines.