FadedRedStar/LFM2.5-VL-1.6B-heretic-imatrix-GGUF
Quantization (fewer layers) vs coder3101/LFM2.5-VL-1.6B-heretic. FadedRedStar/LFM2.5-VL-1.6B-heretic-imatrix-GGUF matches part of coder3101/LFM2.5-VL-1.6B-heretic (138 of its 589 tensors; the rest of coder3101/LFM2.5-VL-1.6B-heretic has no counterpart). On the shared part: FadedRedStar/LFM2.5-VL-1.6B-heretic-imatrix-GGUF is a GGUF quantization (Q5_K/Q6_K) of the same weights: 45 tensors match exactly and 93 within quantization error.
Tensors changed: Embeddings 0/2, other 0/20, MLP 0/48, Norms 0/44, Attention 0/24