Nemotron Instruction Following Chat V1: Multi-Turn Dialogue Data for LLM Training

Nemotron-Instruction-Following-Chat-v1 is designed to strengthen model capabilities in open-ended chat, precise instruction following, and structured output generation. It combines refreshed multi-turn chat data with synthetic dialogues generated by frontier models like GPT-OSS-120B and Qwen3-235B variants. The dataset was created by NVIDIA and last updated on December 15, 2025.

Use Cases

• Fine-tuning models for open-ended conversational AI based on the multi-turn chat data.
• Training models for precise instruction execution based on the instruction-following focus.
• Improving model reliability in generating structured outputs like JSON or code.
• Augmenting training data with synthetic dialogues from high-parameter models like Qwen3-235B.

Strengths

• Data is designed for multiple capabilities: open-ended chat, instruction following, and structured output generation.
• Includes synthetic dialogues generated by frontier models such as GPT-OSS-120B and Qwen3-235B variants.
• Dataset was last updated on 2025-12-15, suggesting recent maintenance.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Data may reflect source bias inherent to the models and chat data used for synthesis.

Provenance

Source

NVIDIA

Collection Method

Combines refreshed chat data from Nemotron-Post-Training-Dataset-v2 with synthetic dialogues produced by frontier models.

Freshness

Last updated 2025-12-15 05:27:15