fr0stbit3/laya-typed-decisions-gguf

Same weights, converted (fewer layers) vs convaiinnovations/laya-typed-decisions. fr0stbit3/laya-typed-decisions-gguf matches part of convaiinnovations/laya-typed-decisions (36 of its 206 tensors; the rest of convaiinnovations/laya-typed-decisions has no counterpart). On the shared part: Same weights stored in another dtype (F16 → F32).

Tensors changed: other 0/20, Norms 0/8, Attention 0/8

On the Hub · Downloads on Model Pulse