chore: import upstream snapshot with attribution

pkg/util/paramtable/function_param.go

@@ -0,0 +1,261 @@
+// Licensed to the LF AI & Data foundation under one
+// or more contributor license agreements. See the NOTICE file
+// distributed with this work for additional information
+// regarding copyright ownership. The ASF licenses this file
+// to you under the Apache License, Version 2.0 (the
+// "License"); you may not use this file except in compliance
+// with the License. You may obtain a copy of the License at
+//
+//     http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+package paramtable
+
+import (
+	"strings"
+)
+
+type functionConfig struct {
+	BatchFactor                   ParamItem  `refreshable:"true"`
+	ModelRequestTimeout           ParamItem  `refreshable:"true"`
+	TextEmbeddingProviders        ParamGroup `refreshable:"true"`
+	RerankModelProviders          ParamGroup `refreshable:"true"`
+	LocalResourcePath             ParamItem  `refreshable:"true"`
+	LinderaDownloadUrls           ParamGroup `refreshable:"true"`
+	ZillizProviders               ParamGroup `refreshable:"true"`
+	AnalyzerConcurrencyPerCPUCore ParamItem  `refreshable:"true"`
+	AnalyzerRunnerConcurrency     ParamItem  `refreshable:"true"`
+}
+
+func (p *functionConfig) init(base *BaseTable) {
+	p.BatchFactor = ParamItem{
+		Key:          "function.batch_factor.",
+		Version:      "2.6.7",
+		DefaultValue: "5",
+	}
+	p.BatchFactor.Init(base.mgr)
+
+	p.ModelRequestTimeout = ParamItem{
+		Key:          "function.model.requestTimeout",
+		Version:      "2.6.12",
+		DefaultValue: "30s",
+		Export:       true,
+		Doc:          "Global timeout for external model requests, e.g. 30s. Function param timeout_ms overrides it.",
+	}
+	p.ModelRequestTimeout.Init(base.mgr)
+
+	p.TextEmbeddingProviders = ParamGroup{
+		KeyPrefix: "function.textEmbedding.providers.",
+		Version:   "2.6.0",
+		Export:    true,
+		DocFunc: func(key string) string {
+			switch key {
+			case "tei.enable":
+				return "Whether to enable TEI model service"
+			case "tei.credential":
+				return "The name in the crendential configuration item"
+			case "azure_openai.credential":
+				return "The name in the crendential configuration item"
+			case "azure_openai.url":
+				return "Your azure openai embedding url, Default is the official embedding url"
+			case "azure_openai.resource_name":
+				return "Your azure openai resource name"
+			case "azure_openai.enable":
+				return "Whether to enable azure openai model service"
+			case "openai.credential":
+				return "The name in the crendential configuration item"
+			case "openai.url":
+				return "Your openai embedding url, Default is the official embedding url"
+			case "openai.enable":
+				return "Whether to enable openai model service"
+			case "dashscope.credential":
+				return "The name in the crendential configuration item"
+			case "dashscope.url":
+				return "Your dashscope embedding url, Default is the official embedding url"
+			case "dashscope.enable":
+				return "Whether to enable dashscope model service"
+			case "cohere.credential":
+				return "The name in the crendential configuration item"
+			case "cohere.url":
+				return "Your cohere embedding url, Default is the official embedding url"
+			case "cohere.enable":
+				return "Whether to enable cohere model service"
+			case "voyageai.credential":
+				return "The name in the crendential configuration item"
+			case "voyageai.url":
+				return "Your voyageai embedding url, Default is the official embedding url"
+			case "voyageai.enable":
+				return "Whether to enable voyageai model service"
+			case "siliconflow.url":
+				return "Your siliconflow embedding url, Default is the official embedding url"
+			case "siliconflow.credential":
+				return "The name in the crendential configuration item"
+			case "siliconflow.enable":
+				return "Whether to enable siliconflow model service"
+			case "bedrock.credential":
+				return "The name in the crendential configuration item"
+			case "bedrock.enable":
+				return "Whether to enable bedrock model service"
+			case "vertexai.url":
+				return "Your VertexAI embedding url"
+			case "vertexai.credential":
+				return "The name in the crendential configuration item"
+			case "vertexai.enable":
+				return "Whether to enable vertexai model service"
+			case "yc.credential":
+				return "The name in the credential configuration item"
+			case "yc.url":
+				return "Your Yandex Cloud text embedding url, Default is the official text embedding url"
+			case "yc.enable":
+				return "Whether to enable Yandex Cloud model service"
+			case "gemini.credential":
+				return "The name in the credential configuration item"
+			case "gemini.url":
+				return "Your Gemini embedding url, Default is the official embedding url"
+			case "gemini.enable":
+				return "Whether to enable Gemini model service"
+			case "huggingface.credential":
+				return "The name in the credential configuration item"
+			case "huggingface.url":
+				return "Your Hugging Face Inference Providers router URL, default is https://router.huggingface.co"
+			case "huggingface.enable":
+				return "Whether to enable Hugging Face text embedding service"
+			default:
+				return ""
+			}
+		},
+	}
+	p.TextEmbeddingProviders.Init(base.mgr)
+
+	p.RerankModelProviders = ParamGroup{
+		KeyPrefix: "function.rerank.model.providers.",
+		Version:   "2.6.0",
+		Export:    true,
+		DocFunc: func(key string) string {
+			switch key {
+			case "tei.credential":
+				return "The name in the crendential configuration item"
+			case "tei.enable":
+				return "Whether to enable TEI rerank service"
+			case "vllm.credential":
+				return "The name in the crendential configuration item"
+			case "vllm.enable":
+				return "Whether to enable vllm rerank service"
+			case "siliconflow.credential":
+				return "The name in the crendential configuration item"
+			case "siliconflow.url":
+				return "Your siliconflow rerank url, Default is the official rerank url"
+			case "siliconflow.enable":
+				return "Whether to enable siliconflow model service"
+			case "voyageai.credential":
+				return "The name in the crendential configuration item"
+			case "voyageai.url":
+				return "Your voyageai rerank url, Default is the official rerank url"
+			case "voyageai.enable":
+				return "Whether to enable voyageai model service"
+			case "cohere.credential":
+				return "The name in the crendential configuration item"
+			case "cohere.url":
+				return "Your cohere rerank url, Default is the official rerank url"
+			case "cohere.enable":
+				return "Whether to enable cohere model service"
+			case "huggingface.credential":
+				return "The name in the credential configuration item"
+			case "huggingface.url":
+				return "Your Hugging Face Inference Providers router URL, default is https://router.huggingface.co"
+			case "huggingface.enable":
+				return "Whether to enable Hugging Face rerank service"
+			default:
+				return ""
+			}
+		},
+	}
+	p.RerankModelProviders.Init(base.mgr)
+
+	p.LocalResourcePath = ParamItem{
+		Key:          "function.analyzer.local_resource_path",
+		Version:      "2.5.16",
+		Export:       true,
+		DefaultValue: "/var/lib/milvus/analyzer",
+	}
+	p.LocalResourcePath.Init(base.mgr)
+
+	p.LinderaDownloadUrls = ParamGroup{
+		KeyPrefix: "function.analyzer.lindera.download_urls.",
+		Version:   "2.5.16",
+	}
+	p.LinderaDownloadUrls.Init(base.mgr)
+
+	p.ZillizProviders = ParamGroup{
+		KeyPrefix: "function.models.zilliz.",
+		Version:   "2.6.5",
+	}
+	p.ZillizProviders.Init(base.mgr)
+
+	p.AnalyzerConcurrencyPerCPUCore = ParamItem{
+		Key:          "function.analyzer.concurrency_per_cpu_core",
+		Version:      "2.6.8",
+		Export:       true,
+		Doc:          "The concurrency per cpu core for analyzer, pipeline not included",
+		DefaultValue: "8",
+	}
+	p.AnalyzerConcurrencyPerCPUCore.Init(base.mgr)
+
+	p.AnalyzerRunnerConcurrency = ParamItem{
+		Key:          "function.analyzer.runner_concurrency",
+		Version:      "2.6.8",
+		Export:       true,
+		Doc:          "The concurrency for each function runner to tokenize text",
+		DefaultValue: "8",
+	}
+	p.AnalyzerRunnerConcurrency.Init(base.mgr)
+}
+
+func (p *functionConfig) GetTextEmbeddingProviderConfig(providerName string) map[string]string {
+	matchedParam := make(map[string]string)
+
+	params := p.TextEmbeddingProviders.GetValue()
+	prefix := providerName + "."
+
+	for k, v := range params {
+		if strings.HasPrefix(k, prefix) {
+			matchedParam[strings.TrimPrefix(k, prefix)] = v
+		}
+	}
+	return matchedParam
+}
+
+func (p *functionConfig) GetBatchFactor() int {
+	factor := p.BatchFactor.GetAsInt()
+	if factor <= 0 {
+		factor = 1
+	}
+	return factor
+}
+
+func (p *functionConfig) GetAnalyzerRunnerConcurrency() int {
+	concurrency := p.AnalyzerRunnerConcurrency.GetAsInt()
+	if concurrency <= 0 {
+		concurrency = 1
+	}
+	return concurrency
+}
+
+func (p *functionConfig) GetRerankModelProviders(providerName string) map[string]string {
+	matchedParam := make(map[string]string)
+
+	params := p.RerankModelProviders.GetValue()
+	prefix := providerName + "."
+
+	for k, v := range params {
+		if strings.HasPrefix(k, prefix) {
+			matchedParam[strings.TrimPrefix(k, prefix)] = v
+		}
+	}
+	return matchedParam
+}