Update README.md

README.md

@@ -62,9 +62,6 @@ As a result, where the original model's performance on the Needle in a Haystack
 | GLM-4-32B-Base-0414 | 100.0% | 99.9% | 66.1% | 3.1% | 0.4% |
 | GLM-4-32B-Base-32k | 100.0% | 99.9% | **98.3%** | **88.7%** | **76.5%** |
 
-## Summary
-The glm-4-32b-true32k model outperforms GLM-4-32B-Base-0414 across all benchmarks, with improvements ranging from 0.65% (hellaswag) to 5.54% (arc_challenge).
-
 ## Use Cases
 
 This model serves as a new base for continued training at 32K context