PoSTMEDIA/Rosetta-7B-Instruct

Full fine-tune vs PoSTMEDIA/Rosetta-7B-Base. Fine-tune: 100% of the weights changed, across 32 of 32 layers.

Tensors changed: MLP 96/96, Norms 129/129, Attention 128/128

1 derivatives checked

• PoSTMEDIA/Rosetta-7B-Instruct-NVFP4: Quantization

On the Hub · Downloads on Model Pulse