CoolFace
Modelpublic

candypunk/NanoJev-Web

sourceHugging Faceotherupdated 4d agoView on Hugging Face
1likes26downloads
runtime-1.1.json158 linesDownload Raw Back to evals
1{2  "runtimeVersion": "1.1.0",3  "weightsUnchanged": true,4  "checkpoint": "browser-head-v5",5  "hardware": "Apple M3 Max",6  "unifiedMemoryGB": 36,7  "macOS": "26.5.2",8  "node": "24.19.0",9  "browser": "agent-browser 0.37.1",10  "inference": "local PyTorch MPS FP32",11  "runs": [12    {13      "mode": "model",14      "status": "completed",15      "elapsedMs": 6997,16      "metrics": {17        "modelDecisions": 10,18        "executedActions": 9,19        "deterministicActions": 0,20        "blockedRetries": 0,21        "browserCalls": 55,22        "browserMs": 5114,23        "modelMs": 1878,24        "inputTokens": 7074,25        "outputTokens": 0,26        "localApiCostUsd": 027      },28      "visibleSuccess": true,29      "completedSteps": 630    },31    {32      "mode": "hybrid",33      "status": "completed",34      "elapsedMs": 3319,35      "metrics": {36        "modelDecisions": 5,37        "executedActions": 9,38        "deterministicActions": 5,39        "blockedRetries": 0,40        "browserCalls": 45,41        "browserMs": 2452,42        "modelMs": 864,43        "inputTokens": 3175,44        "outputTokens": 0,45        "localApiCostUsd": 046      },47      "visibleSuccess": true,48      "completedSteps": 649    },50    {51      "mode": "hybrid",52      "status": "completed",53      "elapsedMs": 3358,54      "metrics": {55        "modelDecisions": 5,56        "executedActions": 9,57        "deterministicActions": 5,58        "blockedRetries": 0,59        "browserCalls": 45,60        "browserMs": 2488,61        "modelMs": 867,62        "inputTokens": 3175,63        "outputTokens": 0,64        "localApiCostUsd": 065      },66      "visibleSuccess": true,67      "completedSteps": 668    },69    {70      "mode": "model",71      "status": "completed",72      "elapsedMs": 4594,73      "metrics": {74        "modelDecisions": 10,75        "executedActions": 9,76        "deterministicActions": 0,77        "blockedRetries": 0,78        "browserCalls": 55,79        "browserMs": 2942,80        "modelMs": 1650,81        "inputTokens": 7074,82        "outputTokens": 0,83        "localApiCostUsd": 084      },85      "visibleSuccess": true,86      "completedSteps": 687    },88    {89      "mode": "model",90      "status": "completed",91      "elapsedMs": 4640,92      "metrics": {93        "modelDecisions": 10,94        "executedActions": 9,95        "deterministicActions": 0,96        "blockedRetries": 0,97        "browserCalls": 55,98        "browserMs": 2970,99        "modelMs": 1667,100        "inputTokens": 7074,101        "outputTokens": 0,102        "localApiCostUsd": 0103      },104      "visibleSuccess": true,105      "completedSteps": 6106    },107    {108      "mode": "hybrid",109      "status": "completed",110      "elapsedMs": 3377,111      "metrics": {112        "modelDecisions": 5,113        "executedActions": 9,114        "deterministicActions": 5,115        "blockedRetries": 0,116        "browserCalls": 45,117        "browserMs": 2507,118        "modelMs": 867,119        "inputTokens": 3175,120        "outputTokens": 0,121        "localApiCostUsd": 0122      },123      "visibleSuccess": true,124      "completedSteps": 6125    }126  ],127  "medianElapsedMs": {128    "model": 4640,129    "hybrid": 3358130  },131  "relativeMedianReduction": 0.27629310344827585,132  "scope": "Three runs per mode on one synthetic five-field modal workflow; loop timing excludes setup and final evidence. Not a comparison with the original 15-field benchmark or a general reliability claim.",133  "checks": {134    "attachedSessionPreserved": true,135    "compactCliOutput": true,136    "independentReceipts": 7,137    "exactSavedData": true,138    "ambiguousLabelsRejected": true,139    "legacyRevealNormalized": true,140    "animationObservation": {141      "settled": true,142      "waitedMs": 267.19999998807907143    }144  },145  "requestContractTests": 6,146  "runtimeTests": 10,147  "externalModelCalls": 0,148  "defaultExecutionMode": "hybrid",149  "hybridDefaultValidation": {150    "runtimeUnitTestsPassed": 11,151    "cliWithoutModeFlagPassed": true,152    "assignedInputsExecutedDeterministically": 5,153    "exactServerReceipts": 1,154    "attachedSessionPreserved": true,155    "modelOnlyOverridePassed": true156  }157}158