Metrics were measured on a subsample (100k products, 10k queries) with
all relevant documents included, so they are not comparable to official
Amazon ESCI benchmarks. Add inline notes at each metric table/result.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Log saturation (ReLU + log1p) is applied per-token before max pooling
across positions, not after. Fix text descriptions and pipeline diagram.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Update links from thierrypdamiba/finetune-ecommerce-search to
qdrant-labs/devrel-projects/tree/main/qdrant-sparse-finetune across
all 5 articles.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>