12 lines
423 B
Plaintext
12 lines
423 B
Plaintext
|
Generate imatrix:
|
||
|
|
llama-imatrix -m input.gguf -f training_data.txt -o imatrix.dat
|
||
|
|
|
||
|
|
Use imatrix:
|
||
|
|
llama-quantize --imatrix imatrix.dat input.gguf output_iq4.gguf IQ4_XS
|
||
|
|
|
||
|
|
|
||
|
|
Generate imatrix:
|
||
|
|
llama-imatrix -m input.gguf -f illuminati_imatrix_v1.txt -o illuminati_imatrix_v1.dat
|
||
|
|
|
||
|
|
Use imatrix:
|
||
|
|
llama-quantize --imatrix illuminati_imatrix_v1.dat input.gguf B:\24B\DarkArtsForge__Asmodeus-24B-v2\output_iq4_xs.gguf IQ4_XS
|