Add cost-performance index (CPI)

results.py

@@ -16,78 +16,91 @@ instance_type_mappings = {
         "gpu": "1xNVIDIA A10G",
         "gpuRAM": "24 GB",
         "url": "https://instances.vantage.sh/aws/ec2/g5.xlarge",
+        "price": 1.006,
     },
     "g5.2xlarge": {
         "cloud": "AWS",
         "gpu": "1xNVIDIA A10G",
         "gpuRAM": "24 GB",
         "url": "https://instances.vantage.sh/aws/ec2/g5.2xlarge",
+        "price": 1.212,
     },
     "g5.12xlarge": {
         "cloud": "AWS",
         "gpu": "4xNVIDIA A10G",
         "gpuRAM": "96 GB",
         "url": "https://instances.vantage.sh/aws/ec2/g5.12xlarge",
+        "price": 5.672,
     },
     "g5.48xlarge": {
         "cloud": "AWS",
         "gpu": "8xNVIDIA A10G",
         "gpuRAM": "192 GB",
         "url": "https://instances.vantage.sh/aws/ec2/g5.48xlarge",
+        "price": 16.288,
     },
     "g6.2xlarge": {
         "cloud": "AWS",
         "gpu": "1xNVIDIA L4",
         "gpuRAM": "24 GB",
         "url": "https://instances.vantage.sh/aws/ec2/g6.2xlarge",
+        "price": 0.9776,
     },
     "g6.12xlarge": {
         "cloud": "AWS",
         "gpu": "4xNVIDIA L4",
         "gpuRAM": "96 GB",
         "url": "https://instances.vantage.sh/aws/ec2/g6.12xlarge",
+        "price": 4.602,
     },
     "g6.48xlarge": {
         "cloud": "AWS",
         "gpu": "8xNVIDIA L4",
         "gpuRAM": "192 GB",
         "url": "https://instances.vantage.sh/aws/ec2/g6.48xlarge",
+        "price": 13.35,
     },
     "g6e.2xlarge": {
         "cloud": "AWS",
         "gpu": "1xNVIDIA L40S",
         "gpuRAM": "48 GB",
         "url": "https://instances.vantage.sh/aws/ec2/g6e.2xlarge",
+        "price": 2.242,
     },
     "g6e.12xlarge": {
         "cloud": "AWS",
         "gpu": "4xNVIDIA L40S",
         "gpuRAM": "192 GB",
         "url": "https://instances.vantage.sh/aws/ec2/g6e.12xlarge",
+        "price": 10.493,
     },
     "g4dn.12xlarge": {
         "cloud": "AWS",
         "gpu": "4xNVIDIA T4",
         "gpuRAM": "64 GB",
         "url": "https://instances.vantage.sh/aws/ec2/g4dn.12xlarge",
+        "price": 3.912,
     },
     "p4d.24xlarge": {
         "cloud": "AWS",
         "gpu": "4xNVIDIA A100",
         "gpuRAM": "320 GB",
         "url": "https://instances.vantage.sh/aws/ec2/p4d.24xlarge",
+        "price": 32.773,
     },
     "p4de.24xlarge": {
         "cloud": "AWS",
         "gpu": "8xNVIDIA A100",
         "gpuRAM": "320 GB",
         "url": "https://instances.vantage.sh/aws/ec2/p4de.24xlarge",
+        "price": 40.966,
     },
     "p5.48xlarge": {
         "cloud": "AWS",
         "gpu": "8xNVIDIA H100",
         "gpuRAM": "640GB",
         "url": "https://instances.vantage.sh/aws/ec2/p5.48xlarge",
+        "price": 98.32,
     },
     "c6i.xlarge": {
         "cloud": "AWS",
@@ -100,36 +113,42 @@ instance_type_mappings = {
         "gpu": "-",
         "gpuRAM": "-",
         "url": "https://instances.vantage.sh/aws/ec2/c6i.2xlarge",
+        "price": 0.34,
     },
     "c6i.4xlarge": {
         "cloud": "AWS",
         "gpu": "-",
         "gpuRAM": "-",
         "url": "https://instances.vantage.sh/aws/ec2/c6i.4xlarge",
+        "price": 0.68,
     },
     "c7i.4xlarge": {
         "cloud": "AWS",
         "gpu": "-",
         "gpuRAM": "-",
         "url": "https://instances.vantage.sh/aws/ec2/c7i.4xlarge",
+        "price": 0.714,
     },
-    "inf2.2xlarge": {
+    "inf2.xlarge": {
         "cloud": "AWS",
         "gpu": "2xNeuronCore v2",
         "gpuRAM": "32 GB",
         "url": "https://instances.vantage.sh/aws/ec2/inf2.2xlarge",
+        "price": 0.7582,
     },
     "inf2.8xlarge": {
         "cloud": "AWS",
         "gpu": "1xNeuronCore v2",
         "gpuRAM": "32 GB",
         "url": "https://instances.vantage.sh/aws/ec2/inf2.8xlarge",
+        "price": 1.968,
     },
     "inf2.24xlarge": {
         "cloud": "AWS",
         "gpu": "12xNeuronCore v2",
         "gpuRAM": "192 GB",
         "url": "https://instances.vantage.sh/aws/ec2/inf2.24xlarge",
+        "price": 6.491,
     },
     "inf2.48xlarge": {
         "cloud": "AWS",
@@ -137,7 +156,6 @@ instance_type_mappings = {
         "gpuRAM": "384 GB",
         "url": "https://instances.vantage.sh/aws/ec2/inf2.48xlarge",
     },
-    "inf2.*": {"cloud": "AWS", "gpu": "-", "gpuRAM": "-", "url": ""},
 }
 
 results = {

results_arcee_agent.py

@@ -41,13 +41,6 @@ results_arcee_agent = {
             "status": "OK",
             "tokensPerSecond": "54.2",
         },
-        {
-            "instanceType": "inf2.*",
-            "container": "TGI 2.2.0",
-            "status": "not supported",
-            "tokensPerSecond": "-",
-            "notes": "Qwen2: TGI OK, Neuron SDK KO, optimum-neuron KO",
-        },
         {
             "instanceType": "g6e.2xlarge",
             "configurations": [

results_llama_spark.py

@@ -99,7 +99,7 @@ results_llama_spark = {
             "notes": '"MAX_INPUT_TOKENS": "40960", "MAX_TOTAL_TOKENS": "81920"\n\n64K/128K fails (even with 4-bit)',
         },
         {
-            "instanceType": "inf2.2xlarge",
+            "instanceType": "inf2.xlarge",
             "container": "LMI 0.29+transformers-neuronx 0.11.351",
             "quantization": "none",
             "status": "OK",