Llamacpp imatrix Quantizations of aya expanse 32b Using llama.cpp release b3930 for quantization. Original model: https://huggingface.co/CohereForAI/aya expanse 32b All quants made using imatrix option with dataset from here Run them in LM Studio Prompt format Download a file (not the whole branch) from below: Filename Quant type File Size Split Description aya expanse 32b f16.gguf f16 64.60GB true Full F16 weights. aya expanse 32b Q8 0.gguf Q8 0 34.33GB false Extremely high quality, generally unneeded but max available quant. aya expanse 32b Q6 K L.gguf Q6 K L 27.01GB false Uses Q8 0 for embed and output weights. Very high quality, near perfect, recommended . aya expanse 32b Q6 K.gguf Q6 K 26.51GB false Very high quality, near perfect, recommended . aya expanse 32b Q5 K L.gguf Q5 K L 23.56GB false Uses Q8 0 for embed and output weights. High quality, recommended . aya expanse 32b Q5 K M.gguf Q5 K M 23.05GB false High quality, recommended . aya expanse 32b Q5 K S.gguf Q5 K S 22.49GB false High quality, recommended . aya expanse 32b Q4 K L.gguf Q4 K L 20.31GB false Uses Q8 0 for embed and output weights. Good quality, recommended . aya expanse 32b Q4 K M.gguf Q4 K M 19.80GB false Go…
We use cookies for essential functionality and analytics. You can accept or reject analytics cookies.Cookie policy