Waiting for the features to be added . It will allow to benchmark on multiple data sets as well as add XPERT retrieval candidates into reranking pipelines.
Add Amazon-10M dataset also. ( Can be done later as 1M if reproducible would suffice to follow)
Add text format for both Amazon-1M and Amazon-10M dataset. ( only for 1M )
Add dataset creation scripts. ( This would be needed)
Add base embedding extractions scripts and model. ( Ngame embeddings are part of NGAME repo in extreme repository , for now normal sentence transformers etc can be done for reproducibility)
Add global interest creation (clustering) code. ( This would be needed for reproducibility)
Add training scripts for morph operators. ( This would be needed for reproducibility)
Hi Team ,
Waiting for the features to be added . It will allow to benchmark on multiple data sets as well as add XPERT retrieval candidates into reranking pipelines.