Add Graviton4

results.py

@@ -131,32 +131,39 @@ instance_type_mappings = {
     },
     "c7g.2xlarge": {
         "cloud": "AWS",
-        "gpu": "-",
+        "gpu": "None - Graviton3, 16 vCPUs",
         "gpuRAM": "-",
         "url": "https://instances.vantage.sh/aws/ec2/c7g.2xlarge",
         "price": 0.29,
     },
     "c7g.4xlarge": {
         "cloud": "AWS",
-        "gpu": "-",
+        "gpu": "None - Graviton3, 16 vCPUs",
         "gpuRAM": "-",
         "url": "https://instances.vantage.sh/aws/ec2/c7g.4xlarge",
         "price": 0.58,
     },
     "c7g.8xlarge": {
         "cloud": "AWS",
-        "gpu": "-",
+        "gpu": "None - Graviton3, 32 vCPUs",
         "gpuRAM": "-",
         "url": "https://instances.vantage.sh/aws/ec2/c7g.8xlarge",
         "price": 1.16,
     },
     "c7g.16xlarge": {
         "cloud": "AWS",
-        "gpu": "-",
+        "gpu": "None - Graviton3, 64 vCPUs",
         "gpuRAM": "-",
         "url": "https://instances.vantage.sh/aws/ec2/c7g.16xlarge",
         "price": 2.32,
     },
+    "r8g.16xlarge": {
+        "cloud": "AWS",
+        "gpu": "None - Graviton4, 64 vCPUs",
+        "gpuRAM": "-",
+        "url": "https://instances.vantage.sh/aws/ec2/r8g.16xlarge",
+        "price": 3.77,
+    },
     "inf2.xlarge": {
         "cloud": "AWS",
         "gpu": "2xNeuronCore v2",

results_arcee_supernova.py

@@ -12,6 +12,14 @@ results_arcee_supernova = {
             "tokensPerSecond": "6.5",
             "notes": "",
         },
+        {
+            "instanceType": "r8g.16xlarge",
+            "quantization": "Q4_0_4_8",
+            "container": "llama.cpp 9/19/24",
+            "status": "OK",
+            "tokensPerSecond": "25",
+            "notes": "",
+        },
         {
             "instanceType": "g5.12xlarge",
             "quantization": "awq",

results_llama_supernova_lite.py

@@ -12,5 +12,29 @@ results_llama_supernova_lite = {
             "tokensPerSecond": "39.7",
             "notes": "requantized from Q4_K_S",
         },
+        {
+            "instanceType": "c7g.16xlarge",
+            "quantization": "Q4_0_8_8",
+            "container": "llama.cpp 9/18/24",
+            "status": "OK",
+            "tokensPerSecond": "45.5",
+            "notes": "",
+        },
+        {
+            "instanceType": "r8g.16xlarge",
+            "quantization": "Q4_0_4_8",
+            "container": "llama.cpp 9/119/24",
+            "status": "OK",
+            "tokensPerSecond": "65",
+            "notes": "",
+        },
+        {
+            "instanceType": "inf2.xlarge",
+            "quantization": "Q4_0_8_8",
+            "container": "llama.cpp 9/18/24",
+            "status": "OK",
+            "tokensPerSecond": "39.7",
+            "notes": "requantized from Q4_K_S",
+        },
     ],
 }