344 lines · 9.9 KB · Rust
RawDownload
use anyhow::{anyhow, Result};use futures::{io::BufReader, stream::BoxStream, AsyncBufReadExt, AsyncReadExt, StreamExt};use http_client::{AsyncBody, HttpClient, Method, Request as HttpRequest};use serde::{Deserialize, Serialize};use serde_json::Value;use std::convert::TryFrom;use strum::EnumIter; pub const MISTRAL_API_URL: &str = "https://api.mistral.ai/v1"; #[derive(Clone, Copy, Serialize, Deserialize, Debug, Eq, PartialEq)]#[serde(rename_all = "lowercase")]pub enum Role {    User,    Assistant,    System,    Tool,} impl TryFrom<String> for Role {    type Error = anyhow::Error;     fn try_from(value: String) -> Result<Self> {        match value.as_str() {            "user" => Ok(Self::User),            "assistant" => Ok(Self::Assistant),            "system" => Ok(Self::System),            "tool" => Ok(Self::Tool),            _ => Err(anyhow!("invalid role '{value}'")),        }    }} impl From<Role> for String {    fn from(val: Role) -> Self {        match val {            Role::User => "user".to_owned(),            Role::Assistant => "assistant".to_owned(),            Role::System => "system".to_owned(),            Role::Tool => "tool".to_owned(),        }    }} #[cfg_attr(feature = "schemars", derive(schemars::JsonSchema))]#[derive(Clone, Debug, Default, Serialize, Deserialize, PartialEq, EnumIter)]pub enum Model {    #[serde(rename = "codestral-latest", alias = "codestral-latest")]    CodestralLatest,    #[serde(rename = "mistral-large-latest", alias = "mistral-large-latest")]    MistralLargeLatest,    #[serde(rename = "mistral-small-latest", alias = "mistral-small-latest")]    MistralSmallLatest,    #[serde(rename = "open-mistral-nemo", alias = "open-mistral-nemo")]    OpenMistralNemo,    #[serde(rename = "open-codestral-mamba", alias = "open-codestral-mamba")]    #[default]    OpenCodestralMamba,     #[serde(rename = "custom")]    Custom {        name: String,        /// The name displayed in the UI, such as in the assistant panel model dropdown menu.        display_name: Option<String>,        max_tokens: usize,        max_output_tokens: Option<u32>,        max_completion_tokens: Option<u32>,    },} impl Model {    pub fn from_id(id: &str) -> Result<Self> {        match id {            "codestral-latest" => Ok(Self::CodestralLatest),            "mistral-large-latest" => Ok(Self::MistralLargeLatest),            "mistral-small-latest" => Ok(Self::MistralSmallLatest),            "open-mistral-nemo" => Ok(Self::OpenMistralNemo),            "open-codestral-mamba" => Ok(Self::OpenCodestralMamba),            _ => Err(anyhow!("invalid model id")),        }    }     pub fn id(&self) -> &str {        match self {            Self::CodestralLatest => "codestral-latest",            Self::MistralLargeLatest => "mistral-large-latest",            Self::MistralSmallLatest => "mistral-small-latest",            Self::OpenMistralNemo => "open-mistral-nemo",            Self::OpenCodestralMamba => "open-codestral-mamba",            Self::Custom { name, .. } => name,        }    }     pub fn display_name(&self) -> &str {        match self {            Self::CodestralLatest => "codestral-latest",            Self::MistralLargeLatest => "mistral-large-latest",            Self::MistralSmallLatest => "mistral-small-latest",            Self::OpenMistralNemo => "open-mistral-nemo",            Self::OpenCodestralMamba => "open-codestral-mamba",            Self::Custom {                name, display_name, ..            } => display_name.as_ref().unwrap_or(name),        }    }     pub fn max_token_count(&self) -> usize {        match self {            Self::CodestralLatest => 256000,            Self::MistralLargeLatest => 131000,            Self::MistralSmallLatest => 32000,            Self::OpenMistralNemo => 131000,            Self::OpenCodestralMamba => 256000,            Self::Custom { max_tokens, .. } => *max_tokens,        }    }     pub fn max_output_tokens(&self) -> Option<u32> {        match self {            Self::Custom {                max_output_tokens, ..            } => *max_output_tokens,            _ => None,        }    }} #[derive(Debug, Serialize, Deserialize)]pub struct Request {    pub model: String,    pub messages: Vec<RequestMessage>,    pub stream: bool,    #[serde(default, skip_serializing_if = "Option::is_none")]    pub max_tokens: Option<u32>,    #[serde(default, skip_serializing_if = "Option::is_none")]    pub temperature: Option<f32>,    #[serde(default, skip_serializing_if = "Option::is_none")]    pub response_format: Option<ResponseFormat>,    #[serde(default, skip_serializing_if = "Vec::is_empty")]    pub tools: Vec<ToolDefinition>,} #[derive(Debug, Serialize, Deserialize)]#[serde(rename_all = "snake_case")]pub enum ResponseFormat {    Text,    #[serde(rename = "json_object")]    JsonObject,} #[derive(Debug, Serialize, Deserialize)]#[serde(tag = "type", rename_all = "snake_case")]pub enum ToolDefinition {    Function { function: FunctionDefinition },} #[derive(Debug, Serialize, Deserialize)]pub struct FunctionDefinition {    pub name: String,    pub description: Option<String>,    pub parameters: Option<Value>,} #[derive(Debug, Serialize, Deserialize)]pub struct CompletionRequest {    pub model: String,    pub prompt: String,    pub max_tokens: u32,    pub temperature: f32,    #[serde(default, skip_serializing_if = "Option::is_none")]    pub prediction: Option<Prediction>,    #[serde(default, skip_serializing_if = "Option::is_none")]    pub rewrite_speculation: Option<bool>,} #[derive(Clone, Deserialize, Serialize, Debug)]#[serde(tag = "type", rename_all = "snake_case")]pub enum Prediction {    Content { content: String },} #[derive(Debug, Serialize, Deserialize)]#[serde(untagged)]pub enum ToolChoice {    Auto,    Required,    None,    Other(ToolDefinition),} #[derive(Serialize, Deserialize, Debug, Eq, PartialEq)]#[serde(tag = "role", rename_all = "lowercase")]pub enum RequestMessage {    Assistant {        content: Option<String>,        #[serde(default, skip_serializing_if = "Vec::is_empty")]        tool_calls: Vec<ToolCall>,    },    User {        content: String,    },    System {        content: String,    },    Tool {        content: String,        tool_call_id: String,    },} #[derive(Serialize, Deserialize, Debug, Eq, PartialEq)]pub struct ToolCall {    pub id: String,    #[serde(flatten)]    pub content: ToolCallContent,} #[derive(Serialize, Deserialize, Debug, Eq, PartialEq)]#[serde(tag = "type", rename_all = "lowercase")]pub enum ToolCallContent {    Function { function: FunctionContent },} #[derive(Serialize, Deserialize, Debug, Eq, PartialEq)]pub struct FunctionContent {    pub name: String,    pub arguments: String,} #[derive(Serialize, Deserialize, Debug)]pub struct CompletionChoice {    pub text: String,} #[derive(Serialize, Deserialize, Debug)]pub struct Response {    pub id: String,    pub object: String,    pub created: u64,    pub model: String,    pub choices: Vec<Choice>,    pub usage: Usage,} #[derive(Serialize, Deserialize, Debug)]pub struct Usage {    pub prompt_tokens: u32,    pub completion_tokens: u32,    pub total_tokens: u32,} #[derive(Serialize, Deserialize, Debug)]pub struct Choice {    pub index: u32,    pub message: RequestMessage,    pub finish_reason: Option<String>,} #[derive(Serialize, Deserialize, Debug)]pub struct StreamResponse {    pub id: String,    pub object: String,    pub created: u64,    pub model: String,    pub choices: Vec<StreamChoice>,} #[derive(Serialize, Deserialize, Debug)]pub struct StreamChoice {    pub index: u32,    pub delta: StreamDelta,    pub finish_reason: Option<String>,} #[derive(Serialize, Deserialize, Debug)]pub struct StreamDelta {    pub role: Option<Role>,    pub content: Option<String>,    #[serde(default, skip_serializing_if = "Option::is_none")]    pub tool_calls: Option<Vec<ToolCallChunk>>,    #[serde(default, skip_serializing_if = "Option::is_none")]    pub reasoning_content: Option<String>,} #[derive(Serialize, Deserialize, Debug, Eq, PartialEq)]pub struct ToolCallChunk {    pub index: usize,    pub id: Option<String>,    pub function: Option<FunctionChunk>,} #[derive(Serialize, Deserialize, Debug, Eq, PartialEq)]pub struct FunctionChunk {    pub name: Option<String>,    pub arguments: Option<String>,} pub async fn stream_completion(    client: &dyn HttpClient,    api_url: &str,    api_key: &str,    request: Request,) -> Result<BoxStream<'static, Result<StreamResponse>>> {    let uri = format!("{api_url}/chat/completions");    let request_builder = HttpRequest::builder()        .method(Method::POST)        .uri(uri)        .header("Content-Type", "application/json")        .header("Authorization", format!("Bearer {}", api_key));     let request = request_builder.body(AsyncBody::from(serde_json::to_string(&request)?))?;    let mut response = client.send(request).await?;     if response.status().is_success() {        let reader = BufReader::new(response.into_body());        Ok(reader            .lines()            .filter_map(|line| async move {                match line {                    Ok(line) => {                        let line = line.strip_prefix("data: ")?;                        if line == "[DONE]" {                            None                        } else {                            match serde_json::from_str(line) {                                Ok(response) => Some(Ok(response)),                                Err(error) => Some(Err(anyhow!(error))),                            }                        }                    }                    Err(error) => Some(Err(anyhow!(error))),                }            })            .boxed())    } else {        let mut body = String::new();        response.body_mut().read_to_string(&mut body).await?;        Err(anyhow!(            "Failed to connect to Mistral API: {} {}",            response.status(),            body,        ))    }}