Remix.run Logo
Alifatisk a day ago

You should mention that it is 4bit quant. Still very impressive!

geerlingguy a day ago | parent | next [-]

Kiki K2 was made to be optimized at 4-bit, though.

natrys a day ago | parent [-]

That's the Kimi K2 Thinking, this post seems to be talking about original Kimi K2 Instruct though, I don't think INT4 QAT (quantization aware training) version was released for this.

elif a day ago | parent | prev [-]

I think when you say trillion parameters, it's implied that it's quantized