BidirLM/BidirLM-0.6B-Embedding
Full fine-tune vs BidirLM/BidirLM-0.6B-Base. Fine-tune: 100% of the weights changed, across 28 of 28 layers.
Tensors changed: Embeddings 1/1, Norms 79/113, MLP 84/84, Attention 112/112
Full fine-tune vs BidirLM/BidirLM-0.6B-Base. Fine-tune: 100% of the weights changed, across 28 of 28 layers.
Tensors changed: Embeddings 1/1, Norms 79/113, MLP 84/84, Attention 112/112