89  Downloads Updated  1 year ago

q4_k_m quantization only. Now using 8k context size