fix bf16 inference speed/VRAM claim to match measured benchmark

README.md

@@ -63,9 +63,10 @@ headaches.
 - **Real-world usability over metric chasing.** OTF models are trained on the
   full Real-ESRGAN degradation chain and restore real photos, not just clean
   bicubic downsamples.
-- **Competitive quality at lower cost.** In our tests it matches HAT on PSNR
-  while using roughly half the peak VRAM and running ~1.6x faster in bf16
-  inference.
+- **Competitive quality at lower cost.** In our tests it matches HAT_M on PSNR
+  (25.02 vs 25.05 dB, within noise) while using ~half the training VRAM
+  (1.44 vs 2.84 GB), and at bf16 inference it is ~1.2x faster with ~2.7x less
+  VRAM.
 - **One file, exports cleanly.** Single-file pure PyTorch; dynamic ONNX on the
   first try; no fusion step or fused/unfused checkpoint pairs.
 - **Stable, simple training.** i-LN keeps bf16 AMP stable without fp32