AkaliKong/MiniOneRec

Health v2 · maintenance only

Flags: no_releases

Adoption not part of the score

1770 stars · 257 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

MiniOneRec is an open-source framework for generative recommendation built on large language models, covering the full pipeline of semantic ID (SID) construction, supervised fine-tuning, and recommendation-oriented reinforcement learning. It is a minimal reproduction of the OneRec system intended for research and experimentation with LLM-based recommenders.

Use cases

• reproduce generative recommendation with llms
• train a recommender system using sft and reinforcement learning
• build semantic ids for items with rq-kmeans
• experiment with scaling laws for generative recommenders
• fine-tune qwen models for recommendation tasks
• evaluate llm-based recommendation with constrained decoding

When to choose

• you want an end-to-end open-source generative recommendation pipeline
• you are researching llm-based or generative recommender systems
• you need reproducible OneRec-style training including SID construction and RL

When to avoid

• you need a production-ready recommendation service with low-latency serving
• you want a traditional collaborative filtering recommender without llms
• you lack GPU resources for fine-tuning large language models

Facets

framework · maturity active

machine-learning llm-training reinforcement-learning rag machine-learning large-language-models artificial-intelligence developer-tools python generative-recommendation recommender-system sid-construction sft scaling-laws llm-based-recommendation gpu linux

Member repositories

For agents

markdown · JSON · MCP: product_card(name="AkaliKong/MiniOneRec")