Nemotron RL Instruction Following Structured Outputs V2 Direct Complete

A dataset of approximately 20,000 rows containing instruction-following examples for language models. It is derived from the NVIDIA Nemotron-RL-Instruction-Following-Structured-Outputs-v2 dataset, with added thinking traces and validated final outputs. The dataset was created by electroglyph and last updated on June 17, 2026.

Use Cases

• Training language models to produce structured outputs based on the 'prompt' and 'result' columns.
• Analyzing reasoning processes in language models using the 'thinking' traces.
• Benchmarking model performance on validated instruction-following tasks.
• Fine-tuning models for improved chain-of-thought reasoning.

Strengths

• Contains validated final outputs, suggesting a quality control step.
• Includes 'thinking' traces, which are not present in the original source dataset.
• Approximately 20,000 rows after validation removed around 8,000 failed rows.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is approximate; exact scale is unknown.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

Derived from NVIDIA's Nemotron-RL-Instruction-Following-Structured-Outputs-v2 dataset on Hugging Face.

Collection Method

Original data was processed to add thinking traces and validate final outputs.

Freshness

Last updated 2026-06-17 08:05:34; freshness should be verified.