Use shared kernels-test-utils and set metal3.1 compatibility 2b765de robtaylor-chipflow commited on 6 days ago
Improve fp32 precision in rotary embedding Metal kernel 3e008a3 robtaylor-chipflow commited on 11 days ago
Fix test reference: use scalars to avoid PyTorch view aliasing bug 0bee974 robtaylor-chipflow commited on 12 days ago
Fix test: capture CPU reference tensors before comparison cf2c51f robtaylor-chipflow commited on 13 days ago
Add Metal rotary embedding kernel matching vLLM interface 949658a robtaylor-chipflow commited on 13 days ago