Release
-
v1.5.2: Fused Inference, FP16, and a Tokenizer That OOM'd a 64GB Machine
Inference +28%. Training +54% at dim≥1024. The tokenizer was O(n²) and nobody noticed until we fed it 48MB.
Inference +28%. Training +54% at dim≥1024. The tokenizer was O(n²) and nobody noticed until we fed it 48MB.