mradermacher/Llama-3.1-8B-UltraLong-4M-Instruct-GGUF

Same weights, converted vs nvidia/Llama-3.1-Nemotron-8B-UltraLong-4M-Instruct. Same weights stored in another dtype (F32 → F16/F32), plus 1 new tensors.

Tensors changed: Embeddings 0/1, Norms 0/65, MLP 0/96, Attention 0/128, lm_head 0/1

On the Hub · Downloads on Model Pulse