feat: update according to openai doc
This commit is contained in:
parent
2ff8372439
commit
4ab359053a
|
@ -193,8 +193,8 @@ func handleAssistantConversation(c tele.Context, thread []*tele.Message) error {
|
||||||
logger.Warnw("failed to cache message", "error", err)
|
logger.Warnw("failed to cache message", "error", err)
|
||||||
}
|
}
|
||||||
|
|
||||||
nBytes := 0 // Used to estimated number of tokens. For now we treat 3 bytes as 1 token.
|
nBytes := 0 // Used to estimated number of tokens. For now we treat 3 bytes as 1 token.
|
||||||
nBytesMax := (4096 - 512) * 3 // Leave some space for the response
|
nBytesMax := 16384 * 3 // Leave some space for the response
|
||||||
|
|
||||||
sysMsg := prompts.Assistant()
|
sysMsg := prompts.Assistant()
|
||||||
chatReqMsgs := []openai.ChatMessage{
|
chatReqMsgs := []openai.ChatMessage{
|
||||||
|
@ -237,7 +237,6 @@ func handleAssistantConversation(c tele.Context, thread []*tele.Message) error {
|
||||||
Model: openai.ModelGpt4O,
|
Model: openai.ModelGpt4O,
|
||||||
Messages: chatReqMsgs,
|
Messages: chatReqMsgs,
|
||||||
Temperature: lo.ToPtr(0.42),
|
Temperature: lo.ToPtr(0.42),
|
||||||
MaxTokens: 2048,
|
|
||||||
User: assistantHashUserId(lastMsg.Sender.ID),
|
User: assistantHashUserId(lastMsg.Sender.ID),
|
||||||
}
|
}
|
||||||
|
|
||||||
|
|
|
@ -6,28 +6,40 @@ type ChatRole string
|
||||||
|
|
||||||
const (
|
const (
|
||||||
ChatRoleSystem ChatRole = "system"
|
ChatRoleSystem ChatRole = "system"
|
||||||
|
ChatRoleDeveloper ChatRole = "developer" // replaces `system` role for o1 and newer models
|
||||||
|
ChatRoleTool ChatRole = "tool"
|
||||||
ChatRoleAssistant ChatRole = "assistant"
|
ChatRoleAssistant ChatRole = "assistant"
|
||||||
ChatRoleUser ChatRole = "user"
|
ChatRoleUser ChatRole = "user"
|
||||||
)
|
)
|
||||||
|
|
||||||
|
type ReasoningEffort string
|
||||||
|
|
||||||
|
const (
|
||||||
|
ReasoningEffortLow ReasoningEffort = "low"
|
||||||
|
ReasoningEffortMedium ReasoningEffort = "medium"
|
||||||
|
ReasoningEffortHigh ReasoningEffort = "high"
|
||||||
|
)
|
||||||
|
|
||||||
type ChatMessage struct {
|
type ChatMessage struct {
|
||||||
Role ChatRole `json:"role"`
|
Role ChatRole `json:"role"`
|
||||||
Content string `json:"content"`
|
Content string `json:"content"`
|
||||||
}
|
}
|
||||||
|
|
||||||
type ChatRequest struct {
|
type ChatRequest struct {
|
||||||
Model string `json:"model"`
|
Model string `json:"model"`
|
||||||
Messages []ChatMessage `json:"messages"`
|
Messages []ChatMessage `json:"messages"`
|
||||||
Temperature *float64 `json:"temperature,omitempty"` // What sampling temperature to use, between 0 and 2.
|
Temperature *float64 `json:"temperature,omitempty"` // What sampling temperature to use, between 0 and 2.
|
||||||
TopP *float64 `json:"top_p,omitempty"` // Nucleus sampling. Specify this or temperature but not both.
|
TopP *float64 `json:"top_p,omitempty"` // Nucleus sampling. Specify this or temperature but not both.
|
||||||
N int `json:"n,omitempty"` // How many chat completion choices to generate for each input message.
|
N int `json:"n,omitempty"` // How many chat completion choices to generate for each input message.
|
||||||
Stream bool `json:"stream,omitempty"` // If set, partial message deltas will be sent as data-only server-sent events as they become available.
|
Stream bool `json:"stream,omitempty"` // If set, partial message deltas will be sent as data-only server-sent events as they become available.
|
||||||
Stop []string `json:"stop,omitempty"` // Up to 4 sequences where the API will stop generating further tokens.
|
Stop []string `json:"stop,omitempty"` // Up to 4 sequences where the API will stop generating further tokens.
|
||||||
MaxTokens int `json:"max_tokens,omitempty"`
|
MaxTokens int `json:"max_tokens,omitempty"` // Deprecated: in favor of `max_completion_tokens`
|
||||||
PresencePenalty *float64 `json:"presence_penalty,omitempty"` // Number between -2.0 and 2.0.
|
MaxCompletionTokens int `json:"max_completion_tokens,omitempty"` // Including visible output tokens and reasoning tokens.
|
||||||
FrequencyPenalty *float64 `json:"frequency_penalty,omitempty"` // Number between -2.0 and 2.0.
|
PresencePenalty *float64 `json:"presence_penalty,omitempty"` // Number between -2.0 and 2.0.
|
||||||
LogitBias map[string]float64 `json:"logit_bias,omitempty"` // Modify the likelihood of specified tokens appearing in the completion.
|
FrequencyPenalty *float64 `json:"frequency_penalty,omitempty"` // Number between -2.0 and 2.0.
|
||||||
User string `json:"user,omitempty"` // A unique identifier representing your end-user, which can help OpenAI to monitor and detect abuse.
|
LogitBias map[string]float64 `json:"logit_bias,omitempty"` // Modify the likelihood of specified tokens appearing in the completion.
|
||||||
|
User string `json:"user,omitempty"` // A unique identifier representing your end-user, which can help OpenAI to monitor and detect abuse.
|
||||||
|
ReasoningEffort ReasoningEffort `json:"reasoning_effort,omitempty"` // Constrains effort on reasoning for reasoning models.
|
||||||
}
|
}
|
||||||
|
|
||||||
type ChatResponseChoice struct {
|
type ChatResponseChoice struct {
|
||||||
|
|
|
@ -20,7 +20,7 @@ func NewClient(apiKey string) *Client {
|
||||||
cli := resty.New().
|
cli := resty.New().
|
||||||
SetTransport(&http.Transport{
|
SetTransport(&http.Transport{
|
||||||
Proxy: http.ProxyFromEnvironment,
|
Proxy: http.ProxyFromEnvironment,
|
||||||
ResponseHeaderTimeout: 10 * time.Second,
|
ResponseHeaderTimeout: 90 * time.Second,
|
||||||
}).
|
}).
|
||||||
SetBaseURL("https://api.openai.com").
|
SetBaseURL("https://api.openai.com").
|
||||||
SetHeader("Authorization", "Bearer "+apiKey).
|
SetHeader("Authorization", "Bearer "+apiKey).
|
||||||
|
|
|
@ -2,6 +2,7 @@ package openai
|
||||||
|
|
||||||
const (
|
const (
|
||||||
ModelGpt4O = "gpt-4o" // Safe default
|
ModelGpt4O = "gpt-4o" // Safe default
|
||||||
ModelO1Preview = "o1-preview" // Expensive
|
ModelO1Preview = "o1-preview" // Expensive reasoning model
|
||||||
ModelO1Mini = "o1-mini"
|
ModelO1Mini = "o1-mini" // Cheaper reasoning model
|
||||||
|
ModelO3Mini = "o3-mini" // Cheaper yet powerful reasoning model
|
||||||
)
|
)
|
||||||
|
|
Loading…
Reference in New Issue