Update README.md

README.md

@@ -236,7 +236,7 @@ The BF16 parent and INT8 artifacts were evaluated on the same 14,543-row v2 vali
 | **ko masking coverage** | **0.998695** | **0.998662** | **-0.000033** |
 | **en masking coverage** | **0.996540** | **0.996739** | **+0.000199** |
 
-The INT8 artifact remained effectively equivalent to the BF16 parent: strict micro F1 changed by -0.000035 and masking coverage by +0.000049. The small positive and negative deltas reflect a few changed boundary decisions and do not show that either precision format is intrinsically more accurate. The intermediate FP32 ONNX parity check is retained in internal conversion reports because it isolates quantization effects after graph transfer.
+The INT8 artifact remained effectively equivalent to the BF16 parent: strict micro F1 changed by -0.000035 and masking coverage by +0.000049. The small positive and negative deltas reflect a few changed boundary decisions and do not show that either precision format is intrinsically more accurate.
 
 ## 2.3. Limitations and Deployment Guidance
 * The graph uses Microsoft contrib operators. ONNX Runtime 1.28.0 with `CPUExecutionProvider` was validated. Confirm operator placement and fallback behavior before selecting another provider.