Update README.md

README.md

@@ -6,8 +6,12 @@ license: apache-2.0
 
 An embedding model trained for study embedding quality vs model architecture (width/depth) given a size constraint (12M params).
 
-- 2 layers
-- hidden size 128
-- 4.4 million parameters
-- initialized from [bert-tiny](https://huggingface.co/google/bert_uncased_L-2_H-128_A-2)
+- **bo-emb-tiny-2-128**: 2 layers/hidden size 128/4.4m/[bert-tiny](https://huggingface.co/google/bert_uncased_L-2_H-128_A-2)
+- **bo-emb-tiny-4-128**: 4 layers/H 128/4.8m
+- **bo-emb-tiny-6-128**: 6 layers/H 128/5.2m
+- **bo-emb-tiny-8-128**: 8 layers/H 128/5.6m
+- **bo-emb-tiny-10-128**: 10 layers/H 128/6.0m
+- **bo-emb-tiny-12-128**: 12 layers/H 128/6.4m
+- **bo-emb-mini-2-256**: 2 layers/H 256/9.7m
+- **bo-emb-mini-4-256**: 4 layers/H 256/11.3m