mradermacher/QiMing-Gemma-3-Socratic-4b-GGUF

Same weights, converted (fewer layers) vs aifeifei798/QiMing-Gemma-3-Socratic-4b. mradermacher/QiMing-Gemma-3-Socratic-4b-GGUF matches part of aifeifei798/QiMing-Gemma-3-Socratic-4b (444 of its 883 tensors; the rest of aifeifei798/QiMing-Gemma-3-Socratic-4b has no counterpart). On the shared part: Same weights stored in another dtype (BF16 → F16/F32).

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

On the Hub · Downloads on Model Pulse