Update README.md

README.md

@@ -35,11 +35,36 @@ Quick testing shows optimized can:
   - Answer/solve at lower quant.
   - Come up with a better answer/stronger reasoning
   - Use less tokens to "reason" ... up to 50% less.
-  - Faster and smaller quant size.
+  - Faster and smaller quant size (VS "MAX" with output tensor and embed at BF16).
+
+Quants - "EDGE of REASON":
+
+IQ1_M - Works, but limited reasoning (reasoning operates, but has a tough time (if at all) coming up with the right answer for some problems).
+
+...
+
+IQ2_S - Moderate reasoning ; impressive performance for both reasoning AND this quant.
+
+...
+
+For best performance IQ3_M or IQ4_XS/NL or Q4s.
+
+...
+
+For TOP performance, Q6/Q8.
+
+...
+
+All quants have be optimized with:
+
+- NEO Imatrix Dataset
+- BF16 Output tensor (full precision)
+
+I found this config worked best with this specific model and "reasoning" in general.
 
 ---
 
-Reka's excellent reasoning model with MAXed quants, and NEO Imatrix dataset.
+Reka's excellent reasoning model with MAX (level 1) quants, and NEO Imatrix dataset.
 
 128k context.