All Tools
P
OtherFreeOpen Source
PADDLEREC
Large-scale recommendation algorithm library on PaddlePaddle
Apache-2.0
ABOUT
Recommendation research splits CTR, ranking, and sequential models across paper repos. PaddleRec puts DeepFM, DIN, DCN, MMOE, and Bert4Rec on PaddlePaddle with shared datasets so teams train and compare recsys models without rebuilding each paper's pipeline.
INSTALL
pip install paddle-recINTEGRATION GUIDE
1. Train DeepFM or DIN click-through models on Criteo-style datasets
2. Benchmark multi-task recommenders such as MMOE and PLE
3. Run sequential recommenders like Bert4Rec and GRU4Rec
4. Prototype industrial ranking pipelines on PaddlePaddle
TAGS
pythonrecommendationctrrankingpaddlepaddledeepfmrecsys