Struct spider::features::openai_common::GPTConfigs

pub struct GPTConfigs {
    pub prompt: Prompt,
    pub model: String,
    pub max_tokens: u16,
    pub temperature: Option<f32>,
    pub user: Option<String>,
    pub top_p: Option<f32>,
    pub prompt_url_map: Option<HashMap<CaseInsensitiveString, Self>>,
    pub extra_ai_data: bool,
    pub paths_map: bool,
    pub screenshot: bool,
    pub api_key: Option<String>,
    pub cache: Option<AICache>,
}

The GPT configs to use for dynamic Javascript execution and other functionality.

Fields

prompt: Prompt

The prompt to use for the Chat. Example: Search for movies. This will attempt to get the code required to perform the action on the page.

model: String

The model to use. Example: gpt-4-1106-preview or gpt-3.5-turbo-16k

max_tokens: u16

The max tokens to use for the request.

temperature: Option<f32>

The temperature between 0 - 2.

user: Option<String>

The user for the request.

top_p: Option<f32>

The top priority for the request.

prompt_url_map: Option<HashMap<CaseInsensitiveString, Self>>

Prompts to use for certain urls. If this is set only the urls that match exactly are ran.

extra_ai_data: bool

Extra data, this will merge the prompts and try to get the content for you. Example: extracting data from the page.

paths_map: bool

Map to paths. If the prompt_url_map has a key called /blog and all blog pages are found like /blog/something the same prompt is perform unless an exact match is found.

screenshot: bool

Take a screenshot of the page after each JS script execution. The screenshot is stored as a base64.

api_key: Option<String>

The API key to use for the request.

cache: Option<AICache>

Use caching to cache the prompt. This does nothing without the ‘cache_openai’ flag enabled.

Implementations

impl GPTConfigs

pub fn new(model: &str, prompt: &str, max_tokens: u16) -> GPTConfigs

GPTConfigs for OpenAI chrome dynamic scripting.

pub fn new_cache( model: &str, prompt: &str, max_tokens: u16, cache: Option<AICache> ) -> GPTConfigs

GPTConfigs for OpenAI chrome dynamic scripting and caching.

pub fn new_multi<I, S>(model: &str, prompt: I, max_tokens: u16) -> GPTConfigs

where I: IntoIterator<Item = S>, S: AsRef<str>,

GPTConfigs for OpenAI chrome dynamic scripting multi chain prompts.

pub fn new_multi_cache<I, S>( model: &str, prompt: I, max_tokens: u16, cache: Option<AICache> ) -> GPTConfigs

where I: IntoIterator<Item = S>, S: AsRef<str>,

GPTConfigs for OpenAI chrome dynamic scripting multi chain prompts with prompt caching. The feature flag ‘cache_openai’ is required.

pub fn set_extra(&mut self, extra_ai_data: bool) -> &mut Self

Set extra AI data to return results.

Trait Implementations

impl Clone for GPTConfigs

fn clone(&self) -> GPTConfigs

Returns a copy of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for GPTConfigs

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for GPTConfigs

fn default() -> GPTConfigs

Returns the “default value” for a type.