News
Liquid AI releases four LFM2.5 QAD Q4_0 checkpoints
Liquid AI has released four LFM2.5 QAD Q4_0 checkpoints, covering 230M, 350M, 1.2B-Instruct, and 2.6B scales, with the goal of improving quality and efficiency in low-bit deployments.
Liquid AI has released four LFM2.5 QAD Q4_0 quantization checkpoints, covering parameter scales of 230M, 350M, 1.2B-Instruct, and 2.6B.
The QAD scheme trains the model to adapt to low bit weights during quantization-aware distillation to improve the quality of Q4_0 deployments. Liquid AI has published self-test results relative to conventional quantization methods, but the actual benefits on different devices and tasks still need to be independently verified.
source:Liquid AI technical articles。