Low/Med/High
#1
by
Hamora
- opened
How can we set the low/medium/high in the model wen inferencing the test time in llamacpp?
I must admit I don't have any clue what you are asking. Is this a model specific thing? If yes, you would need to ask on the original model page, as we only provide the quants.
Thank you for the quant XBai-4o-Q8_0.gguf. Works a treat.
Someone already asked in the original model a few days ago: https://huggingface.co/MetaStoneTec/XBai-o4/discussions/2 - nobody answered so now way for us to know. I would assume low based on how long the thinking in the responses is but it could as well be medium if with low they mean super short thinking.