waris-gill/langcache-embed-v2
Partial fine-tune vs redis/langcache-embed-v1. Partial change: 74% of the weights changed (attention, embeddings, mlp, norms), in 22 of 22 layers.
Tensors changed: Embeddings 1/2, Norms 44/44, Attention 44/44, MLP 44/44
2 other repos hold byte-for-byte identical weight files.