ankk98/dspark-gemma4-12b-block7-Q4_0-GGUF

Quantization vs deepseek-ai/dspark_gemma4_12b_block7. ankk98/dspark-gemma4-12b-block7-Q4_0-GGUF is a GGUF quantization (Q4PURE) of the same weights: 32 tensors match exactly and 33 within quantization error.

Tensors changed: other 0/2, Embeddings 0/1, Norms 0/31, MLP 0/15, Attention 0/15, lm_head 0/1

On the Hub · Downloads on Model Pulse