avemio/German-RAG-UAE-LARGE-V1-TRIPLES-MERGED-HESSIAN-AI

Full fine-tune vs WhereIsAI/UAE-Large-V1. Fine-tune: 90% of the weights changed, across 24 of 24 layers.

Tensors changed: Embeddings 4/5, Norms 96/96, Attention 192/192, other 96/98

1 derivatives checked

• mradermacher/German-RAG-UAE-LARGE-V1-TRIPLES-MERGED-HESSIAN-AI-GGUF: Same weights, converted

On the Hub · Downloads on Model Pulse