mradermacher/G3-4B-APE-2-GGUF

Same weights, converted (fewer layers) vs leftyfeep/G3-4B-APE-2. mradermacher/G3-4B-APE-2-GGUF matches part of leftyfeep/G3-4B-APE-2 (444 of its 883 tensors; the rest of leftyfeep/G3-4B-APE-2 has no counterpart). On the shared part: Same weights stored in another dtype (BF16 → F16/F32).

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

On the Hub · Downloads on Model Pulse