Lynxpda/DeepSeek-V4-Flash-0731-DSpark-Drafter-Q2_K_S-GGUF
Quantization vs unsloth/DeepSeek-V4-Flash-0731-GGUF. Lynxpda/DeepSeek-V4-Flash-0731-DSpark-Drafter-Q2_K_S-GGUF is a GGUF quantization (Q2_K/Q8_0) of the same weights: 46 tensors match exactly and 34 within quantization error.
Tensors changed: MLP 0/30, Attention 0/27, Norms 0/14, other 0/9