keisuke-miyako/granite-embedding-107m-multilingual-gguf-q8_0
Quantization vs ibm-granite/granite-embedding-107m-multilingual. keisuke-miyako/granite-embedding-107m-multilingual-gguf-q8_0 is a GGUF quantization (Q8_0) of the same weights: 63 tensors match exactly and 37 within quantization error.
Tensors changed: Embeddings 0/4, Norms 0/24, Attention 0/48, other 0/24