VALIDATION.md

Validation: flashrt-adaptive-norms

Required before publishing this package:

1. Source-extension correctness:

   python flashrt-adaptive-norms/tests/test_adaptive_norms.py \
     --backend source \
     --mode full
   

2. Source-extension benchmark:

   python flashrt-adaptive-norms/benchmarks/benchmark.py \
     --backend source \
     --shapes all
   

3. Kernel-builder artifact build:

   kernel-builder build-and-copy flashrt-adaptive-norms
   

4. Built-artifact correctness:

   PYTHONPATH=<artifact-path> \
   python flashrt-adaptive-norms/tests/test_adaptive_norms.py \
     --backend installed \
     --mode full
   

5. Built-artifact benchmark:

   python flashrt-adaptive-norms/benchmarks/benchmark.py \
     --backend installed \
     --artifact <artifact-path> \
     --shapes all
   

6. Multi-hardware matrix:

   Add hardware claims only after the same correctness and benchmark commands pass on that machine.

Style Broadcast Gate

Both APIs accept style as either (rows, 3*dim) or (1, 3*dim). The full source and installed-artifact runs must execute both layouts at rows 64/2520/4096. The broadcast path additionally requires bit-identical gate output, exact residual update, FP8 p99 absolute error zero, CUDA Graph replay, and installed-wrapper torch.compile(fullgraph=True) parity.