model card: long-mix numbers

README.md

@@ -33,9 +33,11 @@ inverse and the chunked inference with linear fades run in the host —
 
 - The exported network is asserted equal to `model(chunk)` before export.
 - Spectrogram layout and masked inverse: 151–155 dB PSNR against torch.
-- Each chunk through Core AI on the GPU: 96–110 dB PSNR against upstream's output; whole clips
-  end to end 97–107 dB. Lower than a convolutional model's 140 dB because eight transformer
-  layers of fp32 attention accumulate GPU-versus-CPU rounding, still far below anything audible.
+- Each of 72 chunks (a 10-second clip and a 6-minute mix) through Core AI on the GPU: 74–110 dB
+  PSNR against upstream's output; whole clips end to end 97–114 dB on every stem. Lower than a
+  convolutional model's 140 dB because eight transformer layers of fp32 attention accumulate
+  GPU-versus-CPU rounding — 1e-4 relative on the worst chunk, 1e-5 typical, repeatable run to
+  run, far below anything audible.
 - The host runs every chunk twice and settles a mismatch with a third run, because Core AI's GPU
   was measured to return a slightly wrong result now and then on other models.