textnvidia/HelpSteer3rlhfpreference-datareward-modelingalignmentmultilingualnvidiahuman-feedbackllm-trainingdpofine-tuning

HelpSteer3 — NVIDIA Human Preference & Feedback Dataset for RLHF/Reward Modeling

Category
Text
Records
132,937 rows
Format
PARQUET
Update Frequency
One-time snapshot
Collection Method
auto_imported_huggingface_federated
PII
None detected
File Size
~498.1 MB
Downloads
0

About this data

NVIDIA's open-source multilingual preference dataset for training reward models and aligning LLMs. CC-BY-4.0, 100K+ samples across 15 languages, used to train SOTA reward models on RM-Bench (85.5%) and JudgeBench (78.6%).

Schema

NameTypeDescription
domainVARCHARTask category or subject area (e.g., general, code, writing, reasoning)
languageVARCHARISO 639-1 language code or full language name of the sample
contextSTRUCT("role" VARCHAR, "content" VARCHAR)[]Multi-turn conversation history with role (user/assistant) and content pairs
original_responseVARCHARUnedited model response serving as baseline for comparison
good_edited_responseVARCHARHigh-quality edited version of the original response
bad_edited_responseVARCHARLow-quality edited version of the original response
feedbackVARCHAR[]Array of free-text annotator comments and critiques on responses

Sample Data

Preview a sample of the data before downloading.

Free

Open dataset

Quality: No ratings
0 downloads
Seller: DataBazaar
Sign up to download

Agent? No sign-up needed →

For AI Agents

Via MCP Server
# 1. Add to your agent's MCP config (claude_desktop_config.json or similar):
{
  "mcpServers": {
    "databazaar": { "command": "npx", "args": ["databazaar-mcp"] }
  }
}

# 2. Your agent can then call:
search_datasets({ query: "HelpSteer3 — NVIDIA Human Pref" })
// Found: 6a6db903-3d27-4ee2-87ca-503ce5221dbb
get_download_url({ dataset_id: "6a6db903-3d27-4ee2-87ca-503ce5221dbb" })  // free — no API key needed
Via REST API
# Free dataset — no API key required:
curl https://api.databazaar.io/datasets/6a6db903-3d27-4ee2-87ca-503ce5221dbb/download-url
HelpSteer3 — NVIDIA Human Preference & Feedback Dataset for RLHF/Reward Modeling — Free Dataset | DataBazaar