diff --git a/inference_benchmarks/inference_metrics.csv b/inference_benchmarks/inference_metrics.csv
index 192235b..7c59d96 100644
--- a/inference_benchmarks/inference_metrics.csv
+++ b/inference_benchmarks/inference_metrics.csv
@@ -599,3 +599,5 @@ run_1763526867_non_optimized,1763526867.6077204,False,105,0,0.000243425369262695
 run_1763526867_optimized,1763526867.7184992,True,105,0,0.10547971725463867,0,0,86.33642578125,,cuda
 run_1763526870_non_optimized,1763526870.5837529,False,110,0,0.00023603439331054688,0,0,73.24609375,,cuda
 run_1763526870_optimized,1763526870.6944253,True,110,0,0.10529303550720215,0,0,86.57470703125,,cuda
+run_1763936322_non_optimized,1763936322.5285127,False,53,47,0.34784841537475586,135.11632631519785,7.401030114356508,83.77001953125,,cuda
+run_1763936322_optimized,1763936322.6566308,True,53,47,0.12265253067016602,383.1963331143258,2.609628312131192,86.13916015625,,cuda
diff --git a/inference_benchmarks/inference_metrics.json b/inference_benchmarks/inference_metrics.json
index 45fec15..ebcc141 100644
--- a/inference_benchmarks/inference_metrics.json
+++ b/inference_benchmarks/inference_metrics.json
@@ -7799,6 +7799,45 @@
       "memory_used_mb": 86.57470703125,
       "gpu_utilization": null,
       "device": "cuda"
+    },
+    {
+      "run_name": "run_1763936322_non_optimized",
+      "timestamp": 1763936322.5285127,
+      "optimized": false,
+      "prompt_length": 53,
+      "generated_length": 47,
+      "total_time": 0.34784841537475586,
+      "tokens_per_second": 135.11632631519785,
+      "time_per_token": 7.401030114356508,
+      "memory_used_mb": 83.77001953125,
+      "gpu_utilization": null,
+      "device": "cuda"
+    },
+    {
+      "run_name": "run_1763936322_optimized",
+      "timestamp": 1763936322.6566308,
+      "optimized": true,
+      "prompt_length": 53,
+      "generated_length": 47,
+      "total_time": 0.12265253067016602,
+      "tokens_per_second": 383.1963331143258,
+      "time_per_token": 2.609628312131192,
+      "memory_used_mb": 86.13916015625,
+      "gpu_utilization": null,
+      "device": "cuda"
+    },
+    {
+      "run_name": "run_1763936325_non_optimized",
+      "timestamp": 1763936325.8276925,
+      "optimized": false,
+      "prompt_length": 66,
+      "generated_length": 34,
+      "total_time": 0.3239288330078125,
+      "tokens_per_second": 104.96132648735221,
+      "time_per_token": 9.527318617876839,
+      "memory_used_mb": 83.7705078125,
+      "gpu_utilization": null,
+      "device": "cuda"
     }
   ]
 }
\ No newline at end of file
diff --git a/inference_benchmarks/optimization_comparison.png b/inference_benchmarks/optimization_comparison.png
index fc86ed4..bdf6079 100644
Binary files a/inference_benchmarks/optimization_comparison.png and b/inference_benchmarks/optimization_comparison.png differ
diff --git a/inference_benchmarks/performance_over_time.png b/inference_benchmarks/performance_over_time.png
index 45a653b..81dfe49 100644
Binary files a/inference_benchmarks/performance_over_time.png and b/inference_benchmarks/performance_over_time.png differ