TaQuants/Tema_Q-X4-8B-Thinking-TaQuants-GGUF
Quantization (fewer layers) vs temaq-org/Tema_Q-X4-8B-Thinking. TaQuants/Tema_Q-X4-8B-Thinking-TaQuants-GGUF matches part of temaq-org/Tema_Q-X4-8B-Thinking (677 of its 2130 tensors; the rest of temaq-org/Tema_Q-X4-8B-Thinking has no counterpart). On the shared part: TaQuants/Tema_Q-X4-8B-Thinking-TaQuants-GGUF is a GGUF quantization (Q6_K/IQ3_S) of the same weights: 297 tensors match exactly and 380 within quantization error.
Tensors changed: Embeddings 0/2, Norms 0/296, MLP 0/126, other 0/85, Attention 0/168