async_openai/types/completions/
completion.rs

1use std::collections::HashMap;
2
3use derive_builder::Builder;
4use serde::{Deserialize, Serialize};
5
6use crate::error::OpenAIError;
7
8use crate::types::chat::{
9    ChatCompletionStreamOptions, Choice, CompletionUsage, Prompt, StopConfiguration,
10};
11
12#[derive(Clone, Serialize, Deserialize, Default, Debug, Builder, PartialEq)]
13#[builder(name = "CreateCompletionRequestArgs")]
14#[builder(pattern = "mutable")]
15#[builder(setter(into, strip_option), default)]
16#[builder(derive(Debug))]
17#[builder(build_fn(error = "OpenAIError"))]
18#[cfg_attr(feature = "openapi", derive(utoipa::ToSchema))]
19#[cfg_attr(feature = "openapi", schema(as = async_openai::CreateCompletionRequest))]
20pub struct CreateCompletionRequest {
21    /// ID of the model to use. You can use the [List models](https://platform.openai.com/docs/api-reference/models/list) API to see all of your available models, or see our [Model overview](https://platform.openai.com/docs/models/overview) for descriptions of them.
22    pub model: String,
23
24    /// The prompt(s) to generate completions for, encoded as a string, array of strings, array of tokens, or array of token arrays.
25    ///
26    /// Note that <|endoftext|> is the document separator that the model sees during training, so if a prompt is not specified the model will generate as if from the beginning of a new document.
27    pub prompt: Prompt,
28
29    /// The suffix that comes after a completion of inserted text.
30    ///
31    /// This parameter is only supported for `gpt-3.5-turbo-instruct`.
32    #[serde(skip_serializing_if = "Option::is_none")]
33    pub suffix: Option<String>, // default: null
34
35    /// The maximum number of [tokens](https://platform.openai.com/tokenizer) that can be generated in the completion.
36    ///
37    /// The token count of your prompt plus `max_tokens` cannot exceed the model's context length. [Example Python code](https://cookbook.openai.com/examples/how_to_count_tokens_with_tiktoken) for counting tokens.
38    #[serde(skip_serializing_if = "Option::is_none")]
39    pub max_tokens: Option<u32>,
40
41    /// What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.
42    ///
43    /// We generally recommend altering this or `top_p` but not both.
44    #[serde(skip_serializing_if = "Option::is_none")]
45    #[cfg_attr(feature = "openapi", schema(minimum = 0, maximum = 2))]
46    pub temperature: Option<f32>, // min: 0, max: 2, default: 1,
47
48    /// An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.
49    ///
50    ///  We generally recommend altering this or `temperature` but not both.
51    #[serde(skip_serializing_if = "Option::is_none")]
52    #[cfg_attr(feature = "openapi", schema(minimum = 0, maximum = 1))]
53    pub top_p: Option<f32>, // min: 0, max: 1, default: 1
54
55    /// How many completions to generate for each prompt.
56
57    /// **Note:** Because this parameter generates many completions, it can quickly consume your token quota. Use carefully and ensure that you have reasonable settings for `max_tokens` and `stop`.
58    ///
59    #[serde(skip_serializing_if = "Option::is_none")]
60    #[cfg_attr(feature = "openapi", schema(minimum = 1, maximum = 128))]
61    pub n: Option<u8>, // min:1 max: 128, default: 1
62
63    /// Whether to stream back partial progress. If set, tokens will be sent as data-only [server-sent events](https://developer.mozilla.org/en-US/docs/Web/API/Server-sent_events/Using_server-sent_events#Event_stream_format)
64    /// as they become available, with the stream terminated by a `data: [DONE]` message.
65    #[serde(skip_serializing_if = "Option::is_none")]
66    pub stream: Option<bool>, // nullable: true
67
68    #[serde(skip_serializing_if = "Option::is_none")]
69    pub stream_options: Option<ChatCompletionStreamOptions>,
70
71    /// Include the log probabilities on the `logprobs` most likely output tokens, as well the chosen tokens. For example, if `logprobs` is 5, the API will return a list of the 5 most likely tokens. The API will always return the `logprob` of the sampled token, so there may be up to `logprobs+1` elements in the response.
72    ///
73    /// The maximum value for `logprobs` is 5.
74    #[serde(skip_serializing_if = "Option::is_none")]
75    #[cfg_attr(feature = "openapi", schema(minimum = 0, maximum = 5))]
76    pub logprobs: Option<u8>, // min:0 , max: 5, default: null, nullable: true
77
78    /// Echo back the prompt in addition to the completion
79    #[serde(skip_serializing_if = "Option::is_none")]
80    pub echo: Option<bool>,
81
82    ///  Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.
83    #[serde(skip_serializing_if = "Option::is_none")]
84    pub stop: Option<StopConfiguration>,
85
86    /// Number between -2.0 and 2.0. Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.
87    ///
88    /// [See more information about frequency and presence penalties.](https://platform.openai.com/docs/guides/text-generation/parameter-details)
89    #[serde(skip_serializing_if = "Option::is_none")]
90    #[cfg_attr(feature = "openapi", schema(minimum = -2, maximum = 2))]
91    pub presence_penalty: Option<f32>, // min: -2.0, max: 2.0, default 0
92
93    /// Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.
94    ///
95    /// [See more information about frequency and presence penalties.](https://platform.openai.com/docs/guides/text-generation/parameter-details)
96    #[serde(skip_serializing_if = "Option::is_none")]
97    #[cfg_attr(feature = "openapi", schema(minimum = -2, maximum = 2))]
98    pub frequency_penalty: Option<f32>, // min: -2.0, max: 2.0, default: 0
99
100    /// Generates `best_of` completions server-side and returns the "best" (the one with the highest log probability per token). Results cannot be streamed.
101    ///
102    /// When used with `n`, `best_of` controls the number of candidate completions and `n` specifies how many to return – `best_of` must be greater than `n`.
103    ///
104    /// **Note:** Because this parameter generates many completions, it can quickly consume your token quota. Use carefully and ensure that you have reasonable settings for `max_tokens` and `stop`.
105    #[serde(skip_serializing_if = "Option::is_none")]
106    #[cfg_attr(feature = "openapi", schema(minimum = 0, maximum = 20))]
107    pub best_of: Option<u8>, //min: 0, max: 20, default: 1
108
109    /// Modify the likelihood of specified tokens appearing in the completion.
110    ///
111    /// Accepts a json object that maps tokens (specified by their token ID in the GPT tokenizer) to an associated bias value from -100 to 100. You can use this [tokenizer tool](/tokenizer?view=bpe) (which works for both GPT-2 and GPT-3) to convert text to token IDs. Mathematically, the bias is added to the logits generated by the model prior to sampling. The exact effect will vary per model, but values between -1 and 1 should decrease or increase likelihood of selection; values like -100 or 100 should result in a ban or exclusive selection of the relevant token.
112    ///
113    /// As an example, you can pass `{"50256": -100}` to prevent the <|endoftext|> token from being generated.
114    #[serde(skip_serializing_if = "Option::is_none")]
115    #[cfg_attr(feature = "openapi", schema(schema_with = crate::types::shared::schema::logit_bias))]
116    pub logit_bias: Option<HashMap<String, serde_json::Value>>, // default: null
117
118    /// A unique identifier representing your end-user, which will help OpenAI to monitor and detect abuse. [Learn more](https://platform.openai.com/docs/usage-policies/end-user-ids).
119    #[serde(skip_serializing_if = "Option::is_none")]
120    pub user: Option<String>,
121
122    /// If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same `seed` and parameters should return the same result.
123    ///
124    /// Determinism is not guaranteed, and you should refer to the `system_fingerprint` response parameter to monitor changes in the backend.
125    #[serde(skip_serializing_if = "Option::is_none")]
126    pub seed: Option<i64>,
127}
128
129#[derive(Debug, Deserialize, Clone, PartialEq, Serialize)]
130#[cfg_attr(feature = "openapi", derive(utoipa::ToSchema))]
131#[cfg_attr(feature = "openapi", schema(as = async_openai::CreateCompletionResponse))]
132pub struct CreateCompletionResponse {
133    /// A unique identifier for the completion.
134    pub id: String,
135    pub choices: Vec<Choice>,
136    /// The Unix timestamp (in seconds) of when the completion was created.
137    pub created: u32,
138
139    /// The model used for completion.
140    pub model: String,
141    /// This fingerprint represents the backend configuration that the model runs with.
142    ///
143    /// Can be used in conjunction with the `seed` request parameter to understand when backend changes have been
144    /// made that might impact determinism.
145    pub system_fingerprint: Option<String>,
146
147    /// The object type, which is always "text_completion"
148    pub object: String,
149    pub usage: Option<CompletionUsage>,
150}
151
152/// Parsed server side events stream until an \[DONE\] is received from server.
153#[cfg(feature = "_api")]
154pub type CompletionResponseStream = crate::types::stream::StreamResponse<CreateCompletionResponse>;
async_openai/types/completions/completion.rs

async_openai/types/completions/
completion.rs
