cross-encoder/ms-marco-MiniLM-L6-v2
Full fine-tune (fewer layers) vs cross-encoder/ms-marco-MiniLM-L12-v2. cross-encoder/ms-marco-MiniLM-L6-v2 matches part of cross-encoder/ms-marco-MiniLM-L12-v2 (106 of its 202 tensors; the rest of cross-encoder/ms-marco-MiniLM-L12-v2 has no counterpart). On the shared part: Fine-tune: 100% of the weights changed, across 6 of 6 layers.
Tensors changed: Embeddings 5/6, Norms 24/24, Attention 48/48, other 27/28
4 derivatives checked
- 2013khansohail/cartographer-ecommerce-reranker-MiniLM-L6-v2: Full fine-tune
- sinjab/ms-marco-MiniLM-L6-v2-Q8_0-GGUF: Not comparable yet
- cstr/ms-marco-MiniLM-L-6-v2-GGUF: Not comparable yet
- SamuelMauli/parity-reranker-juridico-br-v1: Full fine-tune
7 other repos hold byte-for-byte identical weight files.