mradermacher/Veiled-Calla-4B-GGUF

Same weights, converted (fewer layers) vs soob3123/Veiled-Calla-4B. mradermacher/Veiled-Calla-4B-GGUF matches part of soob3123/Veiled-Calla-4B (444 of its 883 tensors; the rest of soob3123/Veiled-Calla-4B has no counterpart). On the shared part: Same weights stored in another dtype (F32 → F16/F32).

Tensors changed: Embeddings 0/1, Norms 0/205, MLP 0/102, Attention 0/136

On the Hub · Downloads on Model Pulse