Notus is a collection of fine-tuned LLMs using SFT, DPO, SFT+DPO, and/or any other RLHF techniques, while always keeping a data-first approach
-
Updated
Jan 15, 2024 - Python
#
preference-data
Here are 15 public repositories matching this topic...
Generate conversational, tool-calling, structured-output, and preference datasets — easily and at scale
conversational persona structured-output afterimage multi-turn dpo synthetic-dataset preference-data tool-calling
-
Updated
Aug 1, 2026 - Python
MCP server for human-in-the-loop surveys, A/B preference tests, ratings, and rankings. Get real human feedback inside Claude Code, Claude Desktop, Cursor, Windsurf, and any MCP client — powered by Datapoint AI.
annotation mcp cursor surveys datapoint human-in-the-loop ai-agents claude windsurf data-labeling ai-evaluation human-feedback rlhf anthropic llm-evaluation preference-data claude-desktop model-context-protocol mcp-server claude-code
-
Updated
Aug 31, 2026 - Python
Curated tools, papers, datasets, and practices for LLM training data engineering.
annotation data-quality synthetic-data training-data data-governance llm rlhf financial-ai rag-evaluation preference-data
-
Updated
Aug 12, 2026 - Python
synthetic-data research and engineering: generation, deterministic curation, validation, provenance, and release tooling using local and openrouter distill models
spiking-neural-networks neuromorphic multi-agent-systems synthetic-data huggingface dataset-curation preference-data agentic-coding
-
Updated
Sep 6, 2026 - Python
RLHF preference data curation pipeline: HH-RLHF + UltraFeedback + OASST1 → quality filter → MinHash dedup → DPO-ready JSONL
-
Updated
Jun 7, 2026 - Python
This repository contains all artifacts produced during my bachelor's thesis on data modeling for collective decision-making.
-
Updated
Apr 15, 2026 - Jupyter Notebook
Pairwise code evaluation: two responses rated on five dimensions, a 7-point preference, and a check that the preference matches the ratings.
-
Updated
Aug 15, 2026 - Python
Self-hostable labeling and LLM-evaluation platform: configurable task templates, gold questions, rater reputation, position-bias counterbalancing, and LLM judges that merge with human votes.
react active-learning annotation-tool model-evaluation human-in-the-loop fastapi rlhf preference-data llm-as-judge
-
Updated
Aug 19, 2026 - Python
Human evaluation workbench for scoring LLM responses, collecting pairwise preferences, detecting hallucinations, and analyzing model quality.
python annotation human-data human-evaluation instruction-following ai-evaluation generative-ai human-feedback rlhf llm-evaluation preference-data hallucination-detection
-
Updated
Aug 26, 2026 - Python
Capybara Arena — deployed platform collecting pairwise human preferences on AI-generated CAD models (Elo rating, blob storage, CI).
-
Updated
Aug 19, 2026 - TypeScript
A forkable example of the human-in-the-loop model-improvement loop: AI generates, humans judge via the Terac MCP, you improve the model. Built as an SVG illustration arena.
-
Updated
Jun 20, 2026 - TypeScript
Pairwise rating CLI for AI responses with per-axis scoring (helpfulness/harmlessness/accuracy/instruction-following), JSONL in/out, inter-rater Cohen's kappa
-
Updated
Aug 30, 2026 - Python
Measuring verbosity bias in UltraFeedback (61K GPT-4-annotated preference pairs): statistically unambiguous, but a small effect. Python + Streamlit.
-
Updated
Aug 11, 2026 - Python
Rapidata — independent third-party profile of a public API surface, by API Evangelist. Rapidata provides human feedback at scale — crowd-sourced labeling, model evaluation, ranking, and preference data (RLHF-style) collected from real people worldwide. Teams use the Rapidata API and Python SDK to run online RLHF, refine model behavior, and evaluate
machine-learning company ai annotation crowdsourcing model-evaluation data-labeling apis-json human-feedback rlhf preference-data
-
Updated
Sep 4, 2026
Add this topic to your repo
To associate your repository with the preference-data topic, visit your repo's landing page and select "manage topics."