mradermacher/Holo2-4B-GGUF

Same weights, converted (fewer layers) vs Hcompany/Holo2-4B. mradermacher/Holo2-4B-GGUF matches part of Hcompany/Holo2-4B (399 of its 714 tensors; the rest of Hcompany/Holo2-4B has no counterpart). On the shared part: Same weights stored in another dtype (BF16 → F16/F32).

Tensors changed: Embeddings 0/1, Norms 0/145, MLP 0/108, Attention 0/144, lm_head 0/1

On the Hub · Downloads on Model Pulse