diff --git a/Resources/mgmt-plane/L3N1YnNjcmlwdGlvbnMve30vcHJvdmlkZXJzL21pY3Jvc29mdC5haWluZnJhb3B0aW1pemF0aW9uL2xvY2F0aW9ucy97fS9yZWNvbW1lbmR2bXNrdQ==/2026-06-01-preview.json b/Resources/mgmt-plane/L3N1YnNjcmlwdGlvbnMve30vcHJvdmlkZXJzL21pY3Jvc29mdC5haWluZnJhb3B0aW1pemF0aW9uL2xvY2F0aW9ucy97fS9yZWNvbW1lbmR2bXNrdQ==/2026-06-01-preview.json index 1141a5291..7c86e2208 100644 --- a/Resources/mgmt-plane/L3N1YnNjcmlwdGlvbnMve30vcHJvdmlkZXJzL21pY3Jvc29mdC5haWluZnJhb3B0aW1pemF0aW9uL2xvY2F0aW9ucy97fS9yZWNvbW1lbmR2bXNrdQ==/2026-06-01-preview.json +++ b/Resources/mgmt-plane/L3N1YnNjcmlwdGlvbnMve30vcHJvdmlkZXJzL21pY3Jvc29mdC5haWluZnJhb3B0aW1pemF0aW9uL2xvY2F0aW9ucy97fS9yZWNvbW1lbmR2bXNrdQ==/2026-06-01-preview.json @@ -1 +1 @@ -{"plane": "mgmt-plane", "resources": [{"id": "/subscriptions/{}/providers/microsoft.aiinfraoptimization/locations/{}/recommendvmsku", "version": "2026-06-01-preview", "swagger": "mgmt-plane/aiinfraoptimization/ResourceProviders/Microsoft.AIInfraOptimization/Paths/L3N1YnNjcmlwdGlvbnMve3N1YnNjcmlwdGlvbklkfS9wcm92aWRlcnMvTWljcm9zb2Z0LkFJSW5mcmFPcHRpbWl6YXRpb24vbG9jYXRpb25zL3tsb2NhdGlvbn0vcmVjb21tZW5kVm1Ta3U=/V/MjAyNi0wNi0wMS1wcmV2aWV3"}], "commandGroups": [{"name": "aiinfraoptimization", "commands": [{"name": "recommendvmsku", "version": "2026-06-01-preview", "resources": [{"id": "/subscriptions/{}/providers/microsoft.aiinfraoptimization/locations/{}/recommendvmsku", "version": "2026-06-01-preview", "swagger": "mgmt-plane/aiinfraoptimization/ResourceProviders/Microsoft.AIInfraOptimization/Paths/L3N1YnNjcmlwdGlvbnMve3N1YnNjcmlwdGlvbklkfS9wcm92aWRlcnMvTWljcm9zb2Z0LkFJSW5mcmFPcHRpbWl6YXRpb24vbG9jYXRpb25zL3tsb2NhdGlvbn0vcmVjb21tZW5kVm1Ta3U=/V/MjAyNi0wNi0wMS1wcmV2aWV3"}], "argGroups": [{"name": "", "args": [{"type": "ResourceLocation", "var": "$Path.location", "options": ["l", "location"], "required": true, "idPart": "name"}, {"type": "SubscriptionId", "var": "$Path.subscriptionId", "options": ["subscription"], "required": true, "idPart": "subscription"}]}, {"name": "Body", "args": [{"type": "object", "var": "$body.deployment", "options": ["deployment"], "group": "Body", "help": {"short": "Deployment scale and Azure virtual machine procurement option."}, "args": [{"type": "string", "var": "$body.deployment.procurementOption", "options": ["procurement-option"], "help": {"short": "The Azure virtual machine procurement option."}, "enum": {"items": [{"name": "OnDemand", "value": "OnDemand"}, {"name": "Spot", "value": "Spot"}]}}, {"type": "integer32", "var": "$body.deployment.replicaCount", "options": ["replica-count"], "help": {"short": "The number of deployment replicas used for cost, quota, and placement calculations."}, "format": {"minimum": 1}}]}, {"type": "object", "var": "$body.model", "options": ["model"], "group": "Body", "help": {"short": "Model identity, sizing inputs, and optional maximum model context length."}, "args": [{"type": "integer32", "var": "$body.model.headDim", "options": ["head-dim"], "help": {"short": "The attention head dimension, used for architecture-aware key-value cache sizing. Supply all architecture fields together with a parameter count."}, "format": {"minimum": 1}}, {"type": "string", "var": "$body.model.huggingFaceModelId", "options": ["hugging-face-model-id"], "help": {"short": "The Hugging Face model identifier, for example `meta-llama/Llama-3.1-70B-Instruct`."}}, {"type": "integer32", "var": "$body.model.maxContextLength", "options": ["max-context-length"], "help": {"short": "The maximum context length the model supports, in tokens. Used only as an upper bound for the sizing context length when Hugging Face metadata provides no usable maximum."}, "format": {"minimum": 1}}, {"type": "integer32", "var": "$body.model.numKeyValueHeads", "options": ["num-key-value-heads"], "help": {"short": "The model key-value head count, used for architecture-aware key-value cache sizing. Supply all architecture fields together with a parameter count."}, "format": {"minimum": 1}}, {"type": "integer32", "var": "$body.model.numLayers", "options": ["num-layers"], "help": {"short": "The model layer count, used for architecture-aware key-value cache sizing. Supply all architecture fields together with a parameter count."}, "format": {"minimum": 1}}, {"type": "integer64", "var": "$body.model.parameterCount", "options": ["parameter-count"], "help": {"short": "The logical total model parameter count, including all mixture-of-experts experts. May be supplied with or without a model identifier."}, "format": {"minimum": 1}}, {"type": "string", "var": "$body.model.weightPrecision", "options": ["weight-precision"], "help": {"short": "The precision the model weights are served at. Matched case-insensitively."}, "enum": {"items": [{"name": "BF16", "value": "BF16"}, {"name": "FP16", "value": "FP16"}, {"name": "FP32", "value": "FP32"}, {"name": "FP4", "value": "FP4"}, {"name": "FP8", "value": "FP8"}, {"name": "INT4", "value": "INT4"}, {"name": "INT8", "value": "INT8"}]}}]}, {"type": "object", "var": "$body.optimization", "options": ["optimization"], "group": "Body", "help": {"short": "Ranking preference and optional deployment-wide hourly cost limit."}, "args": [{"type": "object", "var": "$body.optimization.maxDeploymentCostPerHour", "options": ["max-deployment-cost-per-hour"], "help": {"short": "The deployment-wide hourly cost ceiling."}, "args": [{"type": "float64", "var": "$body.optimization.maxDeploymentCostPerHour.amount", "options": ["amount"], "help": {"short": "The maximum cost per hour for the whole deployment. Must be greater than zero when supplied."}}, {"type": "string", "var": "$body.optimization.maxDeploymentCostPerHour.currencyCode", "options": ["currency-code"], "help": {"short": "The ISO 4217 currency code the amount is expressed in. The recommendation service currently supports `USD`."}, "format": {"pattern": "^[A-Za-z]{3}$"}}]}, {"type": "string", "var": "$body.optimization.rankingPreference", "options": ["ranking-preference"], "help": {"short": "The preference used to rank the recommended options."}, "enum": {"items": [{"name": "Cost", "value": "Cost"}, {"name": "Latency", "value": "Latency"}]}}]}, {"type": "object", "var": "$body.subscriptionOptions", "options": ["subscription-options"], "group": "Body", "help": {"short": "The subscription-scoped lookups to include in the recommendation."}, "args": [{"type": "boolean", "var": "$body.subscriptionOptions.includePlacement", "options": ["include-placement"], "help": {"short": "Whether to query placement scores for candidate virtual machine sizes with available quota."}, "default": {"value": false}}, {"type": "boolean", "var": "$body.subscriptionOptions.includePricing", "options": ["include-pricing"], "help": {"short": "Whether to fetch subscription-scoped virtual machine pricing."}, "default": {"value": false}}, {"type": "boolean", "var": "$body.subscriptionOptions.includeQuota", "options": ["include-quota"], "help": {"short": "Whether to evaluate subscription quota for candidate virtual machine sizes."}, "default": {"value": false}}]}, {"type": "array", "var": "$body.targetVmSizes", "options": ["target-vm-sizes"], "group": "Body", "help": {"short": "The Azure virtual machine candidate set. When supplied, only these virtual machine sizes are ranked. Entries must be non-blank and unique, compared case-insensitively."}, "format": {"unique": true, "maxLength": 50, "minLength": 1}, "item": {"type": "string", "format": {"maxLength": 128}}}, {"type": "object", "var": "$body.workload", "options": ["workload"], "group": "Body", "help": {"short": "Inference request concurrency and token shape assumptions."}, "args": [{"type": "integer32", "var": "$body.workload.maxConcurrentRequestsPerReplica", "options": ["max-concurrent-requests-per-replica"], "help": {"short": "The maximum number of simultaneous inference requests handled by each deployment replica."}, "format": {"minimum": 1}}, {"type": "object", "var": "$body.workload.tokenBudget", "options": ["token-budget"], "help": {"short": "The prompt and output token budget used for key-value cache sizing."}, "args": [{"type": "integer32", "var": "$body.workload.tokenBudget.maxOutputTokens", "options": ["max-output-tokens"], "help": {"short": "The maximum number of generated output tokens per inference request. Must be supplied together with `promptTokens`, and the two must total from 1 through 131072."}, "format": {"maximum": 131072, "minimum": 0}}, {"type": "integer32", "var": "$body.workload.tokenBudget.promptTokens", "options": ["prompt-tokens"], "help": {"short": "The number of prompt, or input, tokens per inference request. Must be supplied together with `maxOutputTokens`, and the two must total from 1 through 131072."}, "format": {"maximum": 131072, "minimum": 0}}]}]}]}], "operations": [{"operationId": "Recommendations_RecommendVmSku", "http": {"path": "/subscriptions/{subscriptionId}/providers/Microsoft.AIInfraOptimization/locations/{location}/recommendVmSku", "request": {"method": "post", "path": {"params": [{"type": "string", "name": "location", "arg": "$Path.location", "required": true, "format": {"minLength": 1}}, {"type": "uuid", "name": "subscriptionId", "arg": "$Path.subscriptionId", "required": true}]}, "query": {"consts": [{"readOnly": true, "const": true, "default": {"value": "2026-06-01-preview"}, "type": "string", "name": "api-version", "required": true, "format": {"minLength": 1}}]}, "body": {"json": {"schema": {"type": "object", "name": "body", "required": true, "props": [{"type": "object", "name": "deployment", "arg": "$body.deployment", "props": [{"type": "string", "name": "procurementOption", "arg": "$body.deployment.procurementOption", "enum": {"items": [{"value": "OnDemand"}, {"value": "Spot"}]}}, {"type": "integer32", "name": "replicaCount", "arg": "$body.deployment.replicaCount", "format": {"minimum": 1}}]}, {"type": "object", "name": "model", "arg": "$body.model", "props": [{"type": "integer32", "name": "headDim", "arg": "$body.model.headDim", "format": {"minimum": 1}}, {"type": "string", "name": "huggingFaceModelId", "arg": "$body.model.huggingFaceModelId"}, {"type": "integer32", "name": "maxContextLength", "arg": "$body.model.maxContextLength", "format": {"minimum": 1}}, {"type": "integer32", "name": "numKeyValueHeads", "arg": "$body.model.numKeyValueHeads", "format": {"minimum": 1}}, {"type": "integer32", "name": "numLayers", "arg": "$body.model.numLayers", "format": {"minimum": 1}}, {"type": "integer64", "name": "parameterCount", "arg": "$body.model.parameterCount", "format": {"minimum": 1}}, {"type": "string", "name": "weightPrecision", "arg": "$body.model.weightPrecision", "enum": {"items": [{"value": "BF16"}, {"value": "FP16"}, {"value": "FP32"}, {"value": "FP4"}, {"value": "FP8"}, {"value": "INT4"}, {"value": "INT8"}]}}]}, {"type": "object", "name": "optimization", "arg": "$body.optimization", "props": [{"type": "object", "name": "maxDeploymentCostPerHour", "arg": "$body.optimization.maxDeploymentCostPerHour", "props": [{"type": "float64", "name": "amount", "arg": "$body.optimization.maxDeploymentCostPerHour.amount"}, {"type": "string", "name": "currencyCode", "arg": "$body.optimization.maxDeploymentCostPerHour.currencyCode", "format": {"pattern": "^[A-Za-z]{3}$"}}]}, {"type": "string", "name": "rankingPreference", "arg": "$body.optimization.rankingPreference", "enum": {"items": [{"value": "Cost"}, {"value": "Latency"}]}}]}, {"type": "object", "name": "subscriptionOptions", "arg": "$body.subscriptionOptions", "props": [{"default": {"value": false}, "type": "boolean", "name": "includePlacement", "arg": "$body.subscriptionOptions.includePlacement"}, {"default": {"value": false}, "type": "boolean", "name": "includePricing", "arg": "$body.subscriptionOptions.includePricing"}, {"default": {"value": false}, "type": "boolean", "name": "includeQuota", "arg": "$body.subscriptionOptions.includeQuota"}]}, {"type": "array", "name": "targetVmSizes", "arg": "$body.targetVmSizes", "format": {"unique": true, "maxLength": 50, "minLength": 1}, "item": {"type": "string", "format": {"maxLength": 128}}}, {"type": "object", "name": "workload", "arg": "$body.workload", "props": [{"type": "integer32", "name": "maxConcurrentRequestsPerReplica", "arg": "$body.workload.maxConcurrentRequestsPerReplica", "format": {"minimum": 1}}, {"type": "object", "name": "tokenBudget", "arg": "$body.workload.tokenBudget", "props": [{"type": "integer32", "name": "maxOutputTokens", "arg": "$body.workload.tokenBudget.maxOutputTokens", "format": {"maximum": 131072, "minimum": 0}}, {"type": "integer32", "name": "promptTokens", "arg": "$body.workload.tokenBudget.promptTokens", "format": {"maximum": 131072, "minimum": 0}}]}]}], "clientFlatten": true}}}}, "responses": [{"statusCode": [200], "body": {"json": {"var": "$Instance", "schema": {"type": "object", "props": [{"type": "array", "name": "options", "required": true, "item": {"type": "object", "props": [{"type": "boolean", "name": "azureMlRecommended"}, {"type": "object", "name": "benchmark", "props": [{"type": "string", "name": "benchmarkModel"}, {"type": "string", "name": "level", "required": true, "enum": {"items": [{"value": "Benchmarked"}, {"value": "TopologyMatched"}]}}, {"type": "string", "name": "modelClass", "required": true}, {"type": "string", "name": "operatingSystem"}, {"type": "float64", "name": "p99TpotMs"}, {"type": "float64", "name": "p99TtftMs"}, {"type": "string", "name": "scenario", "required": true}, {"type": "string", "name": "software"}, {"type": "string", "name": "submissionId"}, {"type": "string", "name": "submitter"}, {"type": "integer32", "name": "submitterCount", "required": true}, {"type": "string", "name": "system"}, {"type": "float64", "name": "throughput"}, {"type": "string", "name": "throughputUnit"}, {"type": "string", "name": "weightDataTypes"}]}, {"type": "object", "name": "cost", "required": true, "props": [{"type": "string", "name": "currencyCode", "required": true, "format": {"pattern": "^[A-Za-z]{3}$"}}, {"type": "float64", "name": "deploymentHourly", "required": true}, {"type": "object", "name": "error", "props": [{"type": "string", "name": "code", "required": true, "enum": {"items": [{"value": "DataUnavailable"}, {"value": "DependencyError"}, {"value": "EvaluationBudgetExceeded"}, {"value": "PartitionCapacityExceeded"}, {"value": "Throttled"}, {"value": "TimedOut"}]}}, {"type": "string", "name": "message", "required": true}], "cls": "RecommendVmSkuStageError_read"}, {"type": "dateTime", "name": "priceAsOf", "format": {"protocol": "iso"}}, {"type": "string", "name": "source", "required": true, "enum": {"items": [{"value": "AzureRetail"}, {"value": "AzureSubscription"}, {"value": "StaticPreview"}]}}, {"type": "float64", "name": "vmHourly", "required": true}]}, {"type": "string", "name": "documentationUrl"}, {"type": "object", "name": "performanceEstimate", "props": [{"type": "float64", "name": "deploymentOutputTokensPerSecond", "required": true}, {"type": "float64", "name": "ttftMs", "required": true}]}, {"type": "object", "name": "placement", "props": [{"type": "@RecommendVmSkuStageError_read", "name": "error"}, {"type": "string", "name": "reason"}, {"type": "array", "name": "recommendedAvailabilityZones", "item": {"type": "string"}}, {"type": "string", "name": "status", "required": true, "enum": {"items": [{"value": "Evaluated"}, {"value": "Failed"}, {"value": "NotEvaluated"}, {"value": "Unsupported"}]}}]}, {"type": "object", "name": "quota", "props": [{"type": "@RecommendVmSkuStageError_read", "name": "error"}, {"type": "object", "name": "increase", "props": [{"type": "array", "name": "limits", "required": true, "item": {"type": "object", "props": [{"type": "integer64", "name": "currentLimit", "required": true}, {"type": "integer64", "name": "currentUsage", "required": true}, {"type": "string", "name": "name", "required": true}, {"type": "integer64", "name": "requiredLimit", "required": true}]}, "identifiers": ["name"]}, {"type": "string", "name": "portalUrl", "required": true}]}, {"type": "integer32", "name": "maxAvailableReplicaCount"}, {"type": "string", "name": "reason"}, {"type": "string", "name": "status", "required": true, "enum": {"items": [{"value": "Available"}, {"value": "Unavailable"}, {"value": "Unknown"}]}}]}, {"type": "object", "name": "readiness", "required": true, "props": [{"type": "string", "name": "reason", "required": true}, {"type": "string", "name": "status", "required": true, "enum": {"items": [{"value": "QuotaIncreaseRequired"}, {"value": "ReadyToAttempt"}, {"value": "Unavailable"}, {"value": "VerificationRequired"}]}}]}, {"type": "integer32", "name": "recommendationRank", "required": true}, {"type": "object", "name": "topology", "required": true, "props": [{"type": "object", "name": "accelerator", "required": true, "props": [{"type": "integer32", "name": "countPerVm", "required": true}, {"type": "integer32", "name": "memoryGiBPerAccelerator", "required": true}, {"type": "string", "name": "model", "required": true}]}, {"type": "integer32", "name": "estimatedNodeCountPerReplica", "required": true}, {"type": "integer64", "name": "estimatedTotalNodeCount", "required": true}]}, {"type": "string", "name": "vmSize", "required": true}]}, "identifiers": ["vmSize"]}, {"type": "object", "name": "sizing", "required": true, "props": [{"type": "object", "name": "basis", "required": true, "props": [{"type": "string", "name": "contextLengthSource", "required": true, "enum": {"items": [{"value": "Default"}, {"value": "Request"}]}}, {"type": "string", "name": "modelWeightsCalculation", "required": true, "enum": {"items": [{"value": "ArtifactSize"}, {"value": "EstimatedFromParameterCount"}]}}, {"type": "string", "name": "parameterCountSource", "required": true, "enum": {"items": [{"value": "HuggingFace"}, {"value": "Request"}]}}]}, {"type": "object", "name": "memory", "required": true, "props": [{"type": "float64", "name": "kvCacheGiBPerRequest", "required": true}, {"type": "float64", "name": "kvCacheGiBTotal", "required": true}, {"type": "float64", "name": "modelAndKvCacheGiB", "required": true}, {"type": "float64", "name": "modelWeightsGiB", "required": true}]}, {"type": "object", "name": "model", "required": true, "props": [{"type": "string", "name": "huggingFaceModelId"}, {"type": "integer64", "name": "parameterCount", "required": true}, {"type": "string", "name": "weightPrecision", "required": true, "enum": {"items": [{"value": "BF16"}, {"value": "FP16"}, {"value": "FP32"}, {"value": "FP4"}, {"value": "FP8"}, {"value": "INT4"}, {"value": "INT8"}]}}]}, {"type": "object", "name": "workload", "required": true, "props": [{"type": "integer32", "name": "contextLength", "required": true}, {"type": "integer32", "name": "maxConcurrentRequestsPerReplica", "required": true}]}]}]}}}}, {"isError": true, "body": {"json": {"schema": {"type": "@MgmtErrorFormat"}}}}]}}], "outputs": [{"type": "object", "ref": "$Instance", "clientFlatten": true}], "confirmation": ""}]}]} \ No newline at end of file +{"plane": "mgmt-plane", "resources": [{"id": "/subscriptions/{}/providers/microsoft.aiinfraoptimization/locations/{}/recommendvmsku", "version": "2026-06-01-preview", "swagger": "mgmt-plane/aiinfraoptimization/ResourceProviders/Microsoft.AIInfraOptimization/Paths/L3N1YnNjcmlwdGlvbnMve3N1YnNjcmlwdGlvbklkfS9wcm92aWRlcnMvTWljcm9zb2Z0LkFJSW5mcmFPcHRpbWl6YXRpb24vbG9jYXRpb25zL3tsb2NhdGlvbn0vcmVjb21tZW5kVm1Ta3U=/V/MjAyNi0wNi0wMS1wcmV2aWV3"}], "commandGroups": [{"name": "aiinfraoptimization", "commands": [{"name": "recommendvmsku", "version": "2026-06-01-preview", "resources": [{"id": "/subscriptions/{}/providers/microsoft.aiinfraoptimization/locations/{}/recommendvmsku", "version": "2026-06-01-preview", "swagger": "mgmt-plane/aiinfraoptimization/ResourceProviders/Microsoft.AIInfraOptimization/Paths/L3N1YnNjcmlwdGlvbnMve3N1YnNjcmlwdGlvbklkfS9wcm92aWRlcnMvTWljcm9zb2Z0LkFJSW5mcmFPcHRpbWl6YXRpb24vbG9jYXRpb25zL3tsb2NhdGlvbn0vcmVjb21tZW5kVm1Ta3U=/V/MjAyNi0wNi0wMS1wcmV2aWV3"}], "argGroups": [{"name": "", "args": [{"type": "ResourceLocation", "var": "$Path.location", "options": ["l", "location"], "required": true, "idPart": "name"}, {"type": "SubscriptionId", "var": "$Path.subscriptionId", "options": ["subscription"], "required": true, "idPart": "subscription"}]}, {"name": "Body", "args": [{"type": "object", "var": "$body.deployment", "options": ["deployment"], "group": "Body", "help": {"short": "Deployment scale and Azure virtual machine procurement option."}, "args": [{"type": "string", "var": "$body.deployment.procurementOption", "options": ["procurement-option"], "help": {"short": "The Azure virtual machine procurement option."}, "enum": {"items": [{"name": "OnDemand", "value": "OnDemand"}, {"name": "Spot", "value": "Spot"}]}}, {"type": "integer32", "var": "$body.deployment.replicaCount", "options": ["replica-count"], "help": {"short": "The number of deployment replicas used for cost, quota, and placement calculations."}, "format": {"minimum": 1}}]}, {"type": "object", "var": "$body.model", "options": ["model"], "group": "Body", "help": {"short": "Model identity, sizing inputs, and optional maximum model context length."}, "args": [{"type": "integer32", "var": "$body.model.headDim", "options": ["head-dim"], "help": {"short": "The attention head dimension, used for architecture-aware key-value cache sizing. Supply all architecture fields together with a parameter count."}, "format": {"minimum": 1}}, {"type": "string", "var": "$body.model.huggingFaceModelId", "options": ["hugging-face-model-id"], "help": {"short": "The Hugging Face model identifier, for example `meta-llama/Llama-3.1-70B-Instruct`."}}, {"type": "integer32", "var": "$body.model.maxContextLength", "options": ["max-context-length"], "help": {"short": "The maximum context length the model supports, in tokens. Used only as an upper bound for the sizing context length when Hugging Face metadata provides no usable maximum."}, "format": {"minimum": 1}}, {"type": "integer32", "var": "$body.model.numKeyValueHeads", "options": ["num-key-value-heads"], "help": {"short": "The model key-value head count, used for architecture-aware key-value cache sizing. Supply all architecture fields together with a parameter count."}, "format": {"minimum": 1}}, {"type": "integer32", "var": "$body.model.numLayers", "options": ["num-layers"], "help": {"short": "The model layer count, used for architecture-aware key-value cache sizing. Supply all architecture fields together with a parameter count."}, "format": {"minimum": 1}}, {"type": "integer64", "var": "$body.model.parameterCount", "options": ["parameter-count"], "help": {"short": "The logical total model parameter count, including all mixture-of-experts experts. May be supplied with or without a model identifier."}, "format": {"minimum": 1}}, {"type": "string", "var": "$body.model.weightPrecision", "options": ["weight-precision"], "help": {"short": "The precision the model weights are served at. Matched case-insensitively."}, "enum": {"items": [{"name": "BF16", "value": "BF16"}, {"name": "FP16", "value": "FP16"}, {"name": "FP32", "value": "FP32"}, {"name": "FP4", "value": "FP4"}, {"name": "FP8", "value": "FP8"}, {"name": "INT4", "value": "INT4"}, {"name": "INT8", "value": "INT8"}]}}]}, {"type": "object", "var": "$body.optimization", "options": ["optimization"], "group": "Body", "help": {"short": "Ranking preference and optional deployment-wide hourly cost limit."}, "args": [{"type": "object", "var": "$body.optimization.maxDeploymentCostPerHour", "options": ["max-deployment-cost-per-hour"], "help": {"short": "The deployment-wide hourly cost ceiling."}, "args": [{"type": "float64", "var": "$body.optimization.maxDeploymentCostPerHour.amount", "options": ["amount"], "help": {"short": "The maximum cost per hour for the whole deployment. Must be greater than zero when supplied."}}, {"type": "string", "var": "$body.optimization.maxDeploymentCostPerHour.currencyCode", "options": ["currency-code"], "help": {"short": "The ISO 4217 currency code the amount is expressed in. The recommendation service currently supports `USD`."}, "format": {"pattern": "^[A-Za-z]{3}$"}}]}, {"type": "string", "var": "$body.optimization.rankingPreference", "options": ["ranking-preference"], "help": {"short": "The preference used to rank the recommended options."}, "enum": {"items": [{"name": "Cost", "value": "Cost"}, {"name": "Latency", "value": "Latency"}]}}]}, {"type": "object", "var": "$body.subscriptionOptions", "options": ["subscription-options"], "group": "Body", "help": {"short": "The subscription-scoped lookups to include in the recommendation."}, "args": [{"type": "boolean", "var": "$body.subscriptionOptions.includePlacement", "options": ["include-placement"], "help": {"short": "Whether to query placement scores for candidate virtual machine sizes with available quota."}, "default": {"value": false}}, {"type": "boolean", "var": "$body.subscriptionOptions.includePricing", "options": ["include-pricing"], "help": {"short": "Whether to fetch subscription-scoped virtual machine pricing."}, "default": {"value": false}}, {"type": "boolean", "var": "$body.subscriptionOptions.includeQuota", "options": ["include-quota"], "help": {"short": "Whether to evaluate subscription quota for candidate virtual machine sizes."}, "default": {"value": false}}]}, {"type": "array", "var": "$body.targetVmSizes", "options": ["target-vm-sizes"], "group": "Body", "help": {"short": "The Azure virtual machine candidate set. When supplied, only these virtual machine sizes are ranked. Entries must be non-blank and unique, compared case-insensitively."}, "format": {"unique": true, "maxLength": 50, "minLength": 1}, "item": {"type": "string", "format": {"maxLength": 128}}}, {"type": "object", "var": "$body.workload", "options": ["workload"], "group": "Body", "help": {"short": "Inference request concurrency and token shape assumptions."}, "args": [{"type": "integer32", "var": "$body.workload.maxConcurrentRequestsPerReplica", "options": ["max-concurrent-requests-per-replica"], "help": {"short": "The maximum number of simultaneous inference requests handled by each deployment replica."}, "format": {"minimum": 1}}, {"type": "object", "var": "$body.workload.tokenBudget", "options": ["token-budget"], "help": {"short": "The prompt and output token budget used for key-value cache sizing."}, "args": [{"type": "integer32", "var": "$body.workload.tokenBudget.maxOutputTokens", "options": ["max-output-tokens"], "group": "", "help": {"short": "The maximum number of generated output tokens per inference request. Must be supplied together with `promptTokens`, and the two must total from 1 through 2147483647 without exceeding the model's maximum context length."}, "format": {"maximum": 2147483647, "minimum": 0}}, {"type": "integer32", "var": "$body.workload.tokenBudget.promptTokens", "options": ["prompt-tokens"], "group": "", "help": {"short": "The number of prompt, or input, tokens per inference request. Must be supplied together with `maxOutputTokens`, and the two must total from 1 through 2147483647 without exceeding the model's maximum context length."}, "format": {"maximum": 2147483647, "minimum": 0}}]}]}]}], "operations": [{"operationId": "Recommendations_RecommendVmSku", "http": {"path": "/subscriptions/{subscriptionId}/providers/Microsoft.AIInfraOptimization/locations/{location}/recommendVmSku", "request": {"method": "post", "path": {"params": [{"type": "string", "name": "location", "arg": "$Path.location", "required": true, "format": {"minLength": 1}}, {"type": "uuid", "name": "subscriptionId", "arg": "$Path.subscriptionId", "required": true}]}, "query": {"consts": [{"readOnly": true, "const": true, "default": {"value": "2026-06-01-preview"}, "type": "string", "name": "api-version", "required": true, "format": {"minLength": 1}}]}, "body": {"json": {"schema": {"type": "object", "name": "body", "required": true, "props": [{"type": "object", "name": "deployment", "arg": "$body.deployment", "props": [{"type": "string", "name": "procurementOption", "arg": "$body.deployment.procurementOption", "enum": {"items": [{"value": "OnDemand"}, {"value": "Spot"}]}}, {"type": "integer32", "name": "replicaCount", "arg": "$body.deployment.replicaCount", "format": {"minimum": 1}}]}, {"type": "object", "name": "model", "arg": "$body.model", "props": [{"type": "integer32", "name": "headDim", "arg": "$body.model.headDim", "format": {"minimum": 1}}, {"type": "string", "name": "huggingFaceModelId", "arg": "$body.model.huggingFaceModelId"}, {"type": "integer32", "name": "maxContextLength", "arg": "$body.model.maxContextLength", "format": {"minimum": 1}}, {"type": "integer32", "name": "numKeyValueHeads", "arg": "$body.model.numKeyValueHeads", "format": {"minimum": 1}}, {"type": "integer32", "name": "numLayers", "arg": "$body.model.numLayers", "format": {"minimum": 1}}, {"type": "integer64", "name": "parameterCount", "arg": "$body.model.parameterCount", "format": {"minimum": 1}}, {"type": "string", "name": "weightPrecision", "arg": "$body.model.weightPrecision", "enum": {"items": [{"value": "BF16"}, {"value": "FP16"}, {"value": "FP32"}, {"value": "FP4"}, {"value": "FP8"}, {"value": "INT4"}, {"value": "INT8"}]}}]}, {"type": "object", "name": "optimization", "arg": "$body.optimization", "props": [{"type": "object", "name": "maxDeploymentCostPerHour", "arg": "$body.optimization.maxDeploymentCostPerHour", "props": [{"type": "float64", "name": "amount", "arg": "$body.optimization.maxDeploymentCostPerHour.amount"}, {"type": "string", "name": "currencyCode", "arg": "$body.optimization.maxDeploymentCostPerHour.currencyCode", "format": {"pattern": "^[A-Za-z]{3}$"}}]}, {"type": "string", "name": "rankingPreference", "arg": "$body.optimization.rankingPreference", "enum": {"items": [{"value": "Cost"}, {"value": "Latency"}]}}]}, {"type": "object", "name": "subscriptionOptions", "arg": "$body.subscriptionOptions", "props": [{"default": {"value": false}, "type": "boolean", "name": "includePlacement", "arg": "$body.subscriptionOptions.includePlacement"}, {"default": {"value": false}, "type": "boolean", "name": "includePricing", "arg": "$body.subscriptionOptions.includePricing"}, {"default": {"value": false}, "type": "boolean", "name": "includeQuota", "arg": "$body.subscriptionOptions.includeQuota"}]}, {"type": "array", "name": "targetVmSizes", "arg": "$body.targetVmSizes", "format": {"unique": true, "maxLength": 50, "minLength": 1}, "item": {"type": "string", "format": {"maxLength": 128}}}, {"type": "object", "name": "workload", "arg": "$body.workload", "props": [{"type": "integer32", "name": "maxConcurrentRequestsPerReplica", "arg": "$body.workload.maxConcurrentRequestsPerReplica", "format": {"minimum": 1}}, {"type": "object", "name": "tokenBudget", "arg": "$body.workload.tokenBudget", "props": [{"type": "integer32", "name": "maxOutputTokens", "arg": "$body.workload.tokenBudget.maxOutputTokens", "format": {"maximum": 2147483647, "minimum": 0}}, {"type": "integer32", "name": "promptTokens", "arg": "$body.workload.tokenBudget.promptTokens", "format": {"maximum": 2147483647, "minimum": 0}}]}]}], "clientFlatten": true}}}}, "responses": [{"statusCode": [200], "body": {"json": {"var": "$Instance", "schema": {"type": "object", "props": [{"type": "array", "name": "options", "required": true, "item": {"type": "object", "props": [{"type": "boolean", "name": "azureMlRecommended"}, {"type": "object", "name": "benchmark", "props": [{"type": "string", "name": "benchmarkModel"}, {"type": "string", "name": "level", "required": true, "enum": {"items": [{"value": "Benchmarked"}, {"value": "TopologyMatched"}]}}, {"type": "string", "name": "modelClass", "required": true}, {"type": "string", "name": "operatingSystem"}, {"type": "float64", "name": "p99TpotMs"}, {"type": "float64", "name": "p99TtftMs"}, {"type": "string", "name": "scenario", "required": true}, {"type": "string", "name": "software"}, {"type": "string", "name": "submissionId"}, {"type": "string", "name": "submitter"}, {"type": "integer32", "name": "submitterCount", "required": true}, {"type": "string", "name": "system"}, {"type": "float64", "name": "throughput"}, {"type": "string", "name": "throughputUnit"}, {"type": "string", "name": "weightDataTypes"}]}, {"type": "object", "name": "cost", "required": true, "props": [{"type": "string", "name": "currencyCode", "required": true, "format": {"pattern": "^[A-Za-z]{3}$"}}, {"type": "float64", "name": "deploymentHourly", "required": true}, {"type": "object", "name": "error", "props": [{"type": "string", "name": "code", "required": true, "enum": {"items": [{"value": "DataUnavailable"}, {"value": "DependencyError"}, {"value": "EvaluationBudgetExceeded"}, {"value": "PartitionCapacityExceeded"}, {"value": "Throttled"}, {"value": "TimedOut"}]}}, {"type": "string", "name": "message", "required": true}], "cls": "RecommendVmSkuStageError_read"}, {"type": "dateTime", "name": "priceAsOf", "format": {"protocol": "iso"}}, {"type": "string", "name": "source", "required": true, "enum": {"items": [{"value": "AzureRetail"}, {"value": "AzureSubscription"}, {"value": "StaticPreview"}]}}, {"type": "float64", "name": "vmHourly", "required": true}]}, {"type": "string", "name": "documentationUrl"}, {"type": "object", "name": "performanceEstimate", "props": [{"type": "float64", "name": "deploymentOutputTokensPerSecond", "required": true}, {"type": "float64", "name": "ttftMs", "required": true}]}, {"type": "object", "name": "placement", "props": [{"type": "@RecommendVmSkuStageError_read", "name": "error"}, {"type": "string", "name": "reason"}, {"type": "array", "name": "recommendedAvailabilityZones", "item": {"type": "string"}}, {"type": "string", "name": "status", "required": true, "enum": {"items": [{"value": "Evaluated"}, {"value": "Failed"}, {"value": "NotEvaluated"}, {"value": "Unsupported"}]}}]}, {"type": "object", "name": "quota", "props": [{"type": "@RecommendVmSkuStageError_read", "name": "error"}, {"type": "object", "name": "increase", "props": [{"type": "array", "name": "limits", "required": true, "item": {"type": "object", "props": [{"type": "integer64", "name": "currentLimit", "required": true}, {"type": "integer64", "name": "currentUsage", "required": true}, {"type": "string", "name": "name", "required": true}, {"type": "integer64", "name": "requiredLimit", "required": true}]}, "identifiers": ["name"]}, {"type": "string", "name": "portalUrl", "required": true}]}, {"type": "integer32", "name": "maxAvailableReplicaCount"}, {"type": "string", "name": "reason"}, {"type": "string", "name": "status", "required": true, "enum": {"items": [{"value": "Available"}, {"value": "Unavailable"}, {"value": "Unknown"}]}}]}, {"type": "object", "name": "readiness", "required": true, "props": [{"type": "string", "name": "reason", "required": true}, {"type": "string", "name": "status", "required": true, "enum": {"items": [{"value": "QuotaIncreaseRequired"}, {"value": "ReadyToAttempt"}, {"value": "Unavailable"}, {"value": "VerificationRequired"}]}}]}, {"type": "integer32", "name": "recommendationRank", "required": true}, {"type": "object", "name": "topology", "required": true, "props": [{"type": "object", "name": "accelerator", "required": true, "props": [{"type": "integer32", "name": "countPerVm", "required": true}, {"type": "integer32", "name": "memoryGiBPerAccelerator", "required": true}, {"type": "string", "name": "model", "required": true}]}, {"type": "integer32", "name": "estimatedNodeCountPerReplica", "required": true}, {"type": "integer64", "name": "estimatedTotalNodeCount", "required": true}]}, {"type": "string", "name": "vmSize", "required": true}]}, "identifiers": ["vmSize"]}, {"type": "object", "name": "sizing", "required": true, "props": [{"type": "object", "name": "basis", "required": true, "props": [{"type": "string", "name": "contextLengthSource", "required": true, "enum": {"items": [{"value": "Default"}, {"value": "Request"}]}}, {"type": "string", "name": "modelWeightsCalculation", "required": true, "enum": {"items": [{"value": "ArtifactSize"}, {"value": "EstimatedFromParameterCount"}]}}, {"type": "string", "name": "parameterCountSource", "required": true, "enum": {"items": [{"value": "HuggingFace"}, {"value": "Request"}]}}]}, {"type": "object", "name": "memory", "required": true, "props": [{"type": "float64", "name": "kvCacheGiBPerRequest", "required": true}, {"type": "float64", "name": "kvCacheGiBTotal", "required": true}, {"type": "float64", "name": "modelAndKvCacheGiB", "required": true}, {"type": "float64", "name": "modelWeightsGiB", "required": true}]}, {"type": "object", "name": "model", "required": true, "props": [{"type": "string", "name": "huggingFaceModelId"}, {"type": "integer64", "name": "parameterCount", "required": true}, {"type": "string", "name": "weightPrecision", "required": true, "enum": {"items": [{"value": "BF16"}, {"value": "FP16"}, {"value": "FP32"}, {"value": "FP4"}, {"value": "FP8"}, {"value": "INT4"}, {"value": "INT8"}]}}]}, {"type": "object", "name": "workload", "required": true, "props": [{"type": "integer32", "name": "contextLength", "required": true}, {"type": "integer32", "name": "maxConcurrentRequestsPerReplica", "required": true}]}]}]}}}}, {"isError": true, "body": {"json": {"schema": {"type": "@MgmtErrorFormat"}}}}]}}], "outputs": [{"type": "object", "ref": "$Instance", "clientFlatten": true}], "confirmation": ""}]}]} \ No newline at end of file diff --git a/Resources/mgmt-plane/L3N1YnNjcmlwdGlvbnMve30vcHJvdmlkZXJzL21pY3Jvc29mdC5haWluZnJhb3B0aW1pemF0aW9uL2xvY2F0aW9ucy97fS9yZWNvbW1lbmR2bXNrdQ==/2026-06-01-preview.xml b/Resources/mgmt-plane/L3N1YnNjcmlwdGlvbnMve30vcHJvdmlkZXJzL21pY3Jvc29mdC5haWluZnJhb3B0aW1pemF0aW9uL2xvY2F0aW9ucy97fS9yZWNvbW1lbmR2bXNrdQ==/2026-06-01-preview.xml index e4c458038..16e5a3583 100644 --- a/Resources/mgmt-plane/L3N1YnNjcmlwdGlvbnMve30vcHJvdmlkZXJzL21pY3Jvc29mdC5haWluZnJhb3B0aW1pemF0aW9uL2xvY2F0aW9ucy97fS9yZWNvbW1lbmR2bXNrdQ==/2026-06-01-preview.xml +++ b/Resources/mgmt-plane/L3N1YnNjcmlwdGlvbnMve30vcHJvdmlkZXJzL21pY3Jvc29mdC5haWluZnJhb3B0aW1pemF0aW9uL2xvY2F0aW9ucy97fS9yZWNvbW1lbmR2bXNrdQ==/2026-06-01-preview.xml @@ -111,13 +111,13 @@ - - - + + + - - - + + + @@ -217,10 +217,10 @@ - + - +