dphn

/

dolphin-2.9.4-gemma2-2b

like 38

Follow

Dolphin 162

It’s very impressive model in IQ3_M gguf. It keeps coherency in polish language, it fits in iPhone 14 with 32k context (Q4 context quantizantion) and keeps to be very coherent with long context (I haven’t yet tried full 32k). Just thank you! Perfect setup for low end edge.