feat: support ollama ai model (#1001)

Co-authored-by: Kent Dong <ch3cho@qq.com>
This commit is contained in:
Yifan Gao
2024-05-28 10:55:59 +08:00
committed by GitHub
parent 93966bf14b
commit 50f79c9099
4 changed files with 135 additions and 2 deletions

View File

@@ -93,6 +93,15 @@ Anthropic Claude 所对应的 `type` 为 `claude`。它特有的配置字段如
|-----------|--------|-----|-----|-------------------|
| `version` | string | 必填 | - | Claude 服务的 API 版本 |
#### Ollama
Ollama 所对应的 `type``ollama`。它特有的配置字段如下:
| 名称 | 数据类型 | 填写要求 | 默认值 | 描述 |
|-------------------|--------|------|-----|----------------------------------------------|
| `ollamaServerHost` | string | 必填 | - | Ollama 服务器的主机地址 |
| `ollamaServerPort` | number | 必填 | - | Ollama 服务器的端口号默认为11434 |
## 用法示例
### 使用 OpenAI 协议代理 Azure OpenAI 服务

View File

@@ -36,7 +36,7 @@ func main() {
}
func parseConfig(json gjson.Result, pluginConfig *config.PluginConfig, log wrapper.Log) error {
//log.Debugf("loading config: %s", json.String())
// log.Debugf("loading config: %s", json.String())
pluginConfig.FromJson(json)
if err := pluginConfig.Validate(); err != nil {

View File

@@ -0,0 +1,114 @@
package provider
import (
"errors"
"fmt"
"github.com/alibaba/higress/plugins/wasm-go/extensions/ai-proxy/util"
"github.com/alibaba/higress/plugins/wasm-go/pkg/wrapper"
"github.com/higress-group/proxy-wasm-go-sdk/proxywasm"
"github.com/higress-group/proxy-wasm-go-sdk/proxywasm/types"
)
// ollamaProvider is the provider for Ollama service.
const (
ollamaChatCompletionPath = "/v1/chat/completions"
)
type ollamaProviderInitializer struct {
}
func (m *ollamaProviderInitializer) ValidateConfig(config ProviderConfig) error {
if config.ollamaServerHost == "" {
return errors.New("missing ollamaServerHost in provider config")
}
if config.ollamaServerPort == 0 {
return errors.New("missing ollamaServerPort in provider config")
}
return nil
}
func (m *ollamaProviderInitializer) CreateProvider(config ProviderConfig) (Provider, error) {
serverPortStr := fmt.Sprintf("%d", config.ollamaServerPort)
serviceDomain := config.ollamaServerHost + ":" + serverPortStr
return &ollamaProvider{
config: config,
serviceDomain: serviceDomain,
contextCache: createContextCache(&config),
}, nil
}
type ollamaProvider struct {
config ProviderConfig
serviceDomain string
contextCache *contextCache
}
func (m *ollamaProvider) GetProviderType() string {
return providerTypeOllama
}
func (m *ollamaProvider) OnRequestHeaders(ctx wrapper.HttpContext, apiName ApiName, log wrapper.Log) (types.Action, error) {
if apiName != ApiNameChatCompletion {
return types.ActionContinue, errUnsupportedApiName
}
_ = util.OverwriteRequestPath(ollamaChatCompletionPath)
_ = util.OverwriteRequestHost(m.serviceDomain)
_ = proxywasm.RemoveHttpRequestHeader("Content-Length")
return types.ActionContinue, nil
}
func (m *ollamaProvider) OnRequestBody(ctx wrapper.HttpContext, apiName ApiName, body []byte, log wrapper.Log) (types.Action, error) {
if apiName != ApiNameChatCompletion {
return types.ActionContinue, errUnsupportedApiName
}
if m.config.modelMapping == nil && m.contextCache == nil {
return types.ActionContinue, nil
}
request := &chatCompletionRequest{}
if err := decodeChatCompletionRequest(body, request); err != nil {
return types.ActionContinue, err
}
model := request.Model
if model == "" {
return types.ActionContinue, errors.New("missing model in chat completion request")
}
mappedModel := getMappedModel(model, m.config.modelMapping, log)
if mappedModel == "" {
return types.ActionContinue, errors.New("model becomes empty after applying the configured mapping")
}
request.Model = mappedModel
if m.contextCache != nil {
err := m.contextCache.GetContent(func(content string, err error) {
defer func() {
_ = proxywasm.ResumeHttpRequest()
}()
if err != nil {
log.Errorf("failed to load context file: %v", err)
_ = util.SendResponse(500, util.MimeTypeTextPlain, fmt.Sprintf("failed to load context file: %v", err))
}
insertContextMessage(request, content)
if err := replaceJsonRequestBody(request, log); err != nil {
_ = util.SendResponse(500, util.MimeTypeTextPlain, fmt.Sprintf("failed to replace request body: %v", err))
}
}, log)
if err == nil {
return types.ActionPause, nil
} else {
return types.ActionContinue, err
}
} else {
if err := replaceJsonRequestBody(request, log); err != nil {
_ = util.SendResponse(500, util.MimeTypeTextPlain, fmt.Sprintf("failed to replace request body: %v", err))
return types.ActionContinue, err
}
_ = proxywasm.ResumeHttpRequest()
return types.ActionPause, nil
}
}

View File

@@ -23,6 +23,7 @@ const (
providerTypeBaichuan = "baichuan"
providerTypeYi = "yi"
providerTypeDeepSeek = "deepseek"
providerTypeOllama = "ollama"
protocolOpenAI = "openai"
protocolOriginal = "original"
@@ -60,6 +61,7 @@ var (
providerTypeBaichuan: &baichuanProviderInitializer{},
providerTypeYi: &yiProviderInitializer{},
providerTypeDeepSeek: &deepseekProviderInitializer{},
providerTypeOllama: &ollamaProviderInitializer{},
}
)
@@ -89,7 +91,7 @@ type ResponseBodyHandler interface {
type ProviderConfig struct {
// @Title zh-CN AI服务提供商
// @Description zh-CN AI服务提供商类型目前支持的取值为"moonshot"、"qwen"、"openai"、"azure"、"baichuan"、"yi"
// @Description zh-CN AI服务提供商类型目前支持的取值为"moonshot"、"qwen"、"openai"、"azure"、"baichuan"、"yi"、"ollama"
typ string `required:"true" yaml:"type" json:"type"`
// @Title zh-CN API Tokens
// @Description zh-CN 在请求AI服务时用于认证的API Token列表。不同的AI服务提供商可能有不同的名称。部分供应商只支持配置一个API Token如Azure OpenAI
@@ -109,6 +111,12 @@ type ProviderConfig struct {
// @Title zh-CN 启用通义千问搜索服务
// @Description zh-CN 仅适用于通义千问服务,表示是否启用通义千问的互联网搜索功能。
qwenEnableSearch bool `required:"false" yaml:"qwenEnableSearch" json:"qwenEnableSearch"`
// @Title zh-CN Ollama Server IP/Domain
// @Description zh-CN 仅适用于 Ollama 服务。Ollama 服务器的主机地址。
ollamaServerHost string `required:"false" yaml:"ollamaServerHost" json:"ollamaServerHost"`
// @Title zh-CN Ollama Server Port
// @Description zh-CN 仅适用于 Ollama 服务。Ollama 服务器的端口号。
ollamaServerPort uint32 `required:"false" yaml:"ollamaServerPort" json:"ollamaServerPort"`
// @Title zh-CN 模型名称映射表
// @Description zh-CN 用于将请求中的模型名称映射为目标AI服务商支持的模型名称。支持通过“*”来配置全局映射
modelMapping map[string]string `required:"false" yaml:"modelMapping" json:"modelMapping"`
@@ -137,6 +145,8 @@ func (c *ProviderConfig) FromJson(json gjson.Result) {
c.qwenFileIds = append(c.qwenFileIds, fileId.String())
}
c.qwenEnableSearch = json.Get("qwenEnableSearch").Bool()
c.ollamaServerHost = json.Get("ollamaServerHost").String()
c.ollamaServerPort = uint32(json.Get("ollamaServerPort").Uint())
c.modelMapping = make(map[string]string)
for k, v := range json.Get("modelMapping").Map() {
c.modelMapping[k] = v.String()