Bifrost-AI/Mirage-OpenReasoning-Nemotron-7B

Full fine-tune vs nvidia/OpenReasoning-Nemotron-7B. Fine-tune: 93% of the weights changed, across 28 of 28 layers.

Tensors changed: Embeddings 0/1, Norms 43/57, MLP 84/84, Attention 195/196, lm_head 1/1

1 derivatives checked

• mradermacher/Mirage-OpenReasoning-Nemotron-7B-GGUF: Same weights, converted

On the Hub · Downloads on Model Pulse