CMSManhattan/JiRackDeltaNet_180b

Partial fine-tune vs Qwen/Qwen3.8-Flash-Next. Partial change: 0% of the weights changed (attention, mlp, norms, other), in 1 of 48 layers.

Tensors changed: Norms 3/184, other 2/5, Attention 3/492, MLP 8/384, Embeddings 0/129, lm_head 0/1

On the Hub · Downloads on Model Pulse