Crate burn_lm_inference

Crate burn_lm_inference 

Source

Re-exports§

pub use crate::channels::mutex::MutexChannel;
pub use crate::channels::passthrough::SingleThreadedChannel;
pub use crate::client::InferenceClient;
pub use crate::message::Message;
pub use crate::message::MessageRole;
pub use crate::plugin::InferencePlugin;
pub use crate::server::InferenceServer;
pub use crate::server::InferenceServerConfig;
pub use crate::server::ServerConfigParsing;
pub use crate::stats::StatEntry;
pub use crate::stats::Stats;
pub use crate::stats::STATS_MARKER;
pub use backends::DTYPE_NAME;
pub use clap;
pub use serde_json;
pub use crate::errors::*;
pub use backends::burn_backend_types::*;

Modules§

backends
channels
client
errors
message
plugin
server
stats
utils

Structs§

GeneratedItemEmitter
An emitter is responsible to send generated items to the inference job channel.
InferenceJob
Defines a job to be run during inference.
JobHandle
The handle returned by [InferenceJob::start].
StdOutListener
The stdout listener directly writes the intermediary generated item to std::io::stdout.
TextGenerationListener
The text generation listener accumulate the generated text in a string that can be obtained at the end of the job with the handle finished method.

Enums§

GeneratedItem
Defines all the potential items that can be generated by an inference job.
InferenceTask
The potential tasks that can be executed by an inference job using the inference server.

Traits§

Any
A trait to emulate dynamic typing.
Backend
This trait defines all types and functions needed for a backend to be used with burn.
CommandFactory
Create a Command relevant for a user-defined container.
Deserialize
A data structure that can be deserialized from any data format supported by Serde.
FromArgMatches
Converts an instance of ArgMatches to a user-defined container.
InferenceJobListener
An inference job listener receive events while the inference job is running.
Parser
Parse command-line arguments into Self.

Type Aliases§

Prompt

Attribute Macros§

inference_server_config
This macro consumes the struct, extracts any #[config(default = ...)] attributes and regenerates a brand-new struct with:

Derive Macros§

Deserialize
InferenceServer
Parser
Generates the Parser implementation.