Liquid AI ships quantized LFM2.5 checkpoints via quantization-aware distillation

Liquid AI released Q4_0-quantized LFM2.5 model checkpoints produced through quantization-aware distillation, aimed at efficient on-device inference.