Completion request should support seed

[AjinkyaZ]

ChatCompletionRequest supports a seed value as implemented in #525

go-openai/chat.go

Lines 186 to 221 in 03851d2

	type ChatCompletionRequest struct {
	Model string `json:"model"`
	Messages []ChatCompletionMessage `json:"messages"`
	MaxTokens int `json:"max_tokens,omitempty"`
	Temperature float32 `json:"temperature,omitempty"`
	TopP float32 `json:"top_p,omitempty"`
	N int `json:"n,omitempty"`
	Stream bool `json:"stream,omitempty"`
	Stop []string `json:"stop,omitempty"`
	PresencePenalty float32 `json:"presence_penalty,omitempty"`
	ResponseFormat *ChatCompletionResponseFormat `json:"response_format,omitempty"`
	Seed *int `json:"seed,omitempty"`
	FrequencyPenalty float32 `json:"frequency_penalty,omitempty"`
	// LogitBias is must be a token id string (specified by their token ID in the tokenizer), not a word string.
	// incorrect: `"logit_bias":{"You": 6}`, correct: `"logit_bias":{"1639": 6}`
	// refs: https://platform.openai.com/docs/api-reference/chat/create#chat/create-logit_bias
	LogitBias map[string]int `json:"logit_bias,omitempty"`
	// LogProbs indicates whether to return log probabilities of the output tokens or not.
	// If true, returns the log probabilities of each output token returned in the content of message.
	// This option is currently not available on the gpt-4-vision-preview model.
	LogProbs bool `json:"logprobs,omitempty"`
	// TopLogProbs is an integer between 0 and 5 specifying the number of most likely tokens to return at each
	// token position, each with an associated log probability.
	// logprobs must be set to true if this parameter is used.
	TopLogProbs int `json:"top_logprobs,omitempty"`
	User string `json:"user,omitempty"`
	// Deprecated: use Tools instead.
	Functions []FunctionDefinition `json:"functions,omitempty"`
	// Deprecated: use ToolChoice instead.
	FunctionCall any `json:"function_call,omitempty"`
	Tools []Tool `json:"tools,omitempty"`
	// This can be either a string or an ToolChoice object.
	ToolChoice any `json:"tool_choice,omitempty"`
	// Options for streaming response. Only set this when you set stream: true.
	StreamOptions *StreamOptions `json:"stream_options,omitempty"`
	}

However, the CompletionRequest does not:

go-openai/completion.go

Lines 134 to 154 in 03851d2

	type CompletionRequest struct {
	Model string `json:"model"`
	Prompt any `json:"prompt,omitempty"`
	Suffix string `json:"suffix,omitempty"`
	MaxTokens int `json:"max_tokens,omitempty"`
	Temperature float32 `json:"temperature,omitempty"`
	TopP float32 `json:"top_p,omitempty"`
	N int `json:"n,omitempty"`
	Stream bool `json:"stream,omitempty"`
	LogProbs int `json:"logprobs,omitempty"`
	Echo bool `json:"echo,omitempty"`
	Stop []string `json:"stop,omitempty"`
	PresencePenalty float32 `json:"presence_penalty,omitempty"`
	FrequencyPenalty float32 `json:"frequency_penalty,omitempty"`
	BestOf int `json:"best_of,omitempty"`
	// LogitBias is must be a token id string (specified by their token ID in the tokenizer), not a word string.
	// incorrect: `"logit_bias":{"You": 6}`, correct: `"logit_bias":{"1639": 6}`
	// refs: https://platform.openai.com/docs/api-reference/completions/create#completions/create-logit_bias
	LogitBias map[string]int `json:"logit_bias,omitempty"`
	User string `json:"user,omitempty"`
	}

while OpenAI docs do indicate support for it:
OpenAI docs for /completions https://platform.openai.com/docs/api-reference/completions/create#completions-create-seed