# ustcllm/RecFM

Comprehensive tools and frameworks for developing foundation models tailored to recommendation systems.

Repository: https://github.com/ustcllm/RecFM
Canonical: https://ross.abutalabs.com/products/recfm
License: Apache-2.0
License Family: permissive
Last push: 2025-09-16T01:38:26+00:00

## Health v2 (maintenance only)
Score: 40/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 42, release rhythm 35, longevity 46
- inputs: {"age_days": 645, "days_push": 352, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1033, forks 667 (observed 2026-08-28T04:03:18.363321+00:00)

## What it is
RecFM is a collection of tools and frameworks from USTCLLM for developing foundation models tailored to recommendation systems. It bundles modular training libraries (RecStudio, RecStudio4Industry, Nexus) and text embedding alignment models (CELA, GRE) for building and deploying recommender models.

## Use cases
- build and train recommendation system models in PyTorch
- deploy multi-stage recommendation services in industry settings
- convert text embedding models for recommendation use
- extract universal text representations for multi-domain recommendation
- build document retrieval and information retrieval pipelines
- visualize and debug recommendation model training

## When to choose
- you need a modular library for rapid recommender model prototyping and training
- you want industrial-scale recommendation training with HDFS-friendly data interfaces
- you need text embeddings aligned for recommendation tasks
- you are building multi-stage recommendation or retrieval services

## When to avoid
- you need a turnkey hosted recommendation service rather than a toolkit
- your project is unrelated to recommendation or retrieval
- you require a single unified package rather than multiple subprojects

## Facets
- artifact type: framework
- maturity: active
- function: machine-learning, deep-learning, search-engine, data-science, sdk
- domain: machine-learning, large-language-models, data-science
- platform: python
- tags: recommendation-systems, foundation-models, retrieval, text-embedding, recsys, retrieval-augmented-generation, search

## Member repositories
- ustcllm/RecFM (main) score 40

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:18.363321+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T07:06:42.439519+00:00, confidence not recorded.
  - readme: https://github.com/ustcllm/RecFM (fetched 2026-08-28T04:03:18.363321+00:00, sha 6be789c7f5c1)
- Data as of 2026-08-30T08:39:29.467469+00:00.