model-quantization
Kev-style structured decisions on BitNet with a native I2_S CPU path (source preview)
离线可用的本地类型化决策：4 核 CPU 单题 15.6ms。Local & offline Jev / System One inference on CPU — ONNX + INT8, no torch at runtime. 支持 laya / kev / PlayJev
Measure quantization quality loss on Apple Silicon MLX — KL divergence, top-token flip rate and perplexity delta for KV-cache and weight quantization
RWKV-7 FP8 quantized inference - 6.4x decode speedup on Blackwell GPUs with <0.3% accuracy loss. Full FP8 E4M3 weight quantization with fused Triton kernels.
