Expand description
Raw, unsafe FFI bindings to LiteRT LM’s C API 1.0.0 (every public header:
engine.h, conversation.h, embedding_engine.h, model_info.h,
experimental.h, error_reporter.h), generated by bindgen at build
time. See the litertlm-rs crate for a safe wrapper.
Structs§
- Lite
RtLm Benchmark Info - Lite
RtLm Conversation - Lite
RtLm Conversation Config - Lite
RtLm Conversation Optional Args - Lite
RtLm Detokenize Result - Lite
RtLm Embedding Engine - Lite
RtLm Embedding Engine Settings - Lite
RtLm Embedding Options - Lite
RtLm Embedding Response - Lite
RtLm Embedding Responses - Lite
RtLm Engine - Lite
RtLm Engine Settings - Lite
RtLm Input Data - Lite
RtLm Json Response - Lite
RtLm Loaded File - Lite
RtLm NoRepeat Ngram Config - Lite
RtLm Repetition Penalty Config - Lite
RtLm Responses - Lite
RtLm Sampler Params - Lite
RtLm Session - Lite
RtLm Session Config - Lite
RtLm Session Debug Info - Lite
RtLm Speculative Decoding Config - Lite
RtLm Stream Chunk - Lite
RtLm Suppress Tokens Config - Lite
RtLm Thinking Config - Lite
RtLm Token Union - Lite
RtLm Token Unions - Lite
RtLm Tokenize Result
Constants§
- LITERT_
LM_ C_ API_ VERSION_ MAJOR - LITERT_
LM_ C_ API_ VERSION_ MINOR - LITERT_
LM_ C_ API_ VERSION_ PATCH - Lite
RtLm Activation Data Type_ kLite RtLm Activation Data Type Float16 - Lite
RtLm Activation Data Type_ kLite RtLm Activation Data Type Float32 - Lite
RtLm Activation Data Type_ kLite RtLm Activation Data Type Int8 - Lite
RtLm Activation Data Type_ kLite RtLm Activation Data Type Int16 - Lite
RtLm Backend Type_ kLite RtLm Backend Type Cpu - Lite
RtLm Backend Type_ kLite RtLm Backend Type Gpu - Lite
RtLm Backend Type_ kLite RtLm Backend Type Npu - Lite
RtLm Constraint Provider Type_ kLite RtLm Constraint Provider Type LlGuidance - Lite
RtLm Constraint Provider Type_ kLite RtLm Constraint Provider Type None - Lite
RtLm Constraint Type_ kLite RtLm Constraint Type Json Schema - Lite
RtLm Constraint Type_ kLite RtLm Constraint Type None - Lite
RtLm Constraint Type_ kLite RtLm Constraint Type Regex - Lite
RtLm Input Data Type_ kLite RtLm Input Data Type Audio - Lite
RtLm Input Data Type_ kLite RtLm Input Data Type Audio End - Lite
RtLm Input Data Type_ kLite RtLm Input Data Type Image - Lite
RtLm Input Data Type_ kLite RtLm Input Data Type Image End - Lite
RtLm Input Data Type_ kLite RtLm Input Data Type Text - Lite
RtLm Input Overflow Strategy_ kLite RtLm Input Overflow Strategy Chunk AndAverage - Lite
RtLm Input Overflow Strategy_ kLite RtLm Input Overflow Strategy Error - Lite
RtLm Input Overflow Strategy_ kLite RtLm Input Overflow Strategy Truncate - Lite
RtLm LogSeverity_ kLite RtLm LogSeverity Debug - Lite
RtLm LogSeverity_ kLite RtLm LogSeverity Error - Lite
RtLm LogSeverity_ kLite RtLm LogSeverity Fatal - Lite
RtLm LogSeverity_ kLite RtLm LogSeverity Info - Lite
RtLm LogSeverity_ kLite RtLm LogSeverity Silent - Lite
RtLm LogSeverity_ kLite RtLm LogSeverity Verbose - Lite
RtLm LogSeverity_ kLite RtLm LogSeverity Warning - Lite
RtLm Modality_ kLite RtLm Modality Audio - Lite
RtLm Modality_ kLite RtLm Modality Text - Lite
RtLm Modality_ kLite RtLm Modality Video - Lite
RtLm Modality_ kLite RtLm Modality Vision - Lite
RtLm Model Type_ kLite RtLm Model Type Embedding - Lite
RtLm Model Type_ kLite RtLm Model Type Llm - Lite
RtLm Model Type_ kLite RtLm Model Type Unknown - Lite
RtLm NpuBrand_ kLite RtLm NpuBrand Google Tensor - Lite
RtLm NpuBrand_ kLite RtLm NpuBrand Intel - Lite
RtLm NpuBrand_ kLite RtLm NpuBrand Media Tek - Lite
RtLm NpuBrand_ kLite RtLm NpuBrand Qualcomm - Lite
RtLm NpuBrand_ kLite RtLm NpuBrand Samsung - Lite
RtLm NpuBrand_ kLite RtLm NpuBrand Unknown - Lite
RtLm Sampler Type_ kLite RtLm Sampler Type Greedy - Lite
RtLm Sampler Type_ kLite RtLm Sampler Type TopK - Lite
RtLm Sampler Type_ kLite RtLm Sampler Type TopP - Lite
RtLm Sampler Type_ kLite RtLm Sampler Type Unspecified - Lite
RtLm Speculative Decoding Mode_ kLite RtLm Speculative Decoding Mode Adaptive - Lite
RtLm Speculative Decoding Mode_ kLite RtLm Speculative Decoding Mode Off - Lite
RtLm Speculative Decoding Mode_ kLite RtLm Speculative Decoding Mode On - Lite
RtLm Status Code_ kLite RtLm Status Aborted - Lite
RtLm Status Code_ kLite RtLm Status Already Exists - Lite
RtLm Status Code_ kLite RtLm Status Cancelled - Lite
RtLm Status Code_ kLite RtLm Status Data Loss - Lite
RtLm Status Code_ kLite RtLm Status Deadline Exceeded - Lite
RtLm Status Code_ kLite RtLm Status Failed Precondition - Lite
RtLm Status Code_ kLite RtLm Status Internal - Lite
RtLm Status Code_ kLite RtLm Status Invalid Argument - Lite
RtLm Status Code_ kLite RtLm Status NotFound - Lite
RtLm Status Code_ kLite RtLm Status Ok - Lite
RtLm Status Code_ kLite RtLm Status OutOf Range - Lite
RtLm Status Code_ kLite RtLm Status Permission Denied - Lite
RtLm Status Code_ kLite RtLm Status Resource Exhausted - Lite
RtLm Status Code_ kLite RtLm Status Unauthenticated - Lite
RtLm Status Code_ kLite RtLm Status Unavailable - Lite
RtLm Status Code_ kLite RtLm Status Unimplemented - Lite
RtLm Status Code_ kLite RtLm Status Unknown - Lite
RtLm Token Union Type_ kLite RtLm Token Union Type Ids - Lite
RtLm Token Union Type_ kLite RtLm Token Union Type String
Functions§
- litert_
lm_ ⚠benchmark_ info_ delete - litert_
lm_ ⚠benchmark_ info_ get_ decode_ token_ count_ at - litert_
lm_ ⚠benchmark_ info_ get_ decode_ tokens_ per_ sec_ at - litert_
lm_ ⚠benchmark_ info_ get_ max_ num_ tokens - litert_
lm_ ⚠benchmark_ info_ get_ num_ decode_ turns - litert_
lm_ ⚠benchmark_ info_ get_ num_ prefill_ turns - litert_
lm_ ⚠benchmark_ info_ get_ prefill_ token_ count_ at - litert_
lm_ ⚠benchmark_ info_ get_ prefill_ tokens_ per_ sec_ at - litert_
lm_ ⚠benchmark_ info_ get_ time_ to_ first_ token - litert_
lm_ ⚠benchmark_ info_ get_ total_ init_ time_ in_ second - litert_
lm_ ⚠clear_ last_ error - litert_
lm_ ⚠conversation_ cancel_ process - litert_
lm_ ⚠conversation_ clone - litert_
lm_ ⚠conversation_ config_ create - litert_
lm_ ⚠conversation_ config_ delete - litert_
lm_ ⚠conversation_ config_ set_ constraint_ provider - litert_
lm_ ⚠conversation_ config_ set_ enable_ constrained_ decoding - litert_
lm_ ⚠conversation_ config_ set_ extra_ context - litert_
lm_ ⚠conversation_ config_ set_ filter_ channel_ content_ from_ kv_ cache - litert_
lm_ ⚠conversation_ config_ set_ messages - litert_
lm_ ⚠conversation_ config_ set_ prompt_ template - litert_
lm_ ⚠conversation_ config_ set_ session_ config - litert_
lm_ ⚠conversation_ config_ set_ stream_ tool_ calls - litert_
lm_ ⚠conversation_ config_ set_ system_ message - litert_
lm_ ⚠conversation_ config_ set_ thinking_ config - litert_
lm_ ⚠conversation_ config_ set_ tools - litert_
lm_ ⚠conversation_ create - litert_
lm_ ⚠conversation_ delete - litert_
lm_ ⚠conversation_ get_ benchmark_ info - litert_
lm_ ⚠conversation_ get_ token_ count - litert_
lm_ ⚠conversation_ optional_ args_ create - litert_
lm_ ⚠conversation_ optional_ args_ delete - litert_
lm_ ⚠conversation_ optional_ args_ set_ constraint - litert_
lm_ ⚠conversation_ optional_ args_ set_ max_ output_ tokens - litert_
lm_ ⚠conversation_ optional_ args_ set_ no_ repeat_ ngram_ config - litert_
lm_ ⚠conversation_ optional_ args_ set_ repetition_ penalty_ config - litert_
lm_ ⚠conversation_ optional_ args_ set_ suppress_ tokens_ config - litert_
lm_ ⚠conversation_ optional_ args_ set_ thinking_ config - litert_
lm_ ⚠conversation_ optional_ args_ set_ visual_ token_ budget - litert_
lm_ ⚠conversation_ render_ message_ to_ string - litert_
lm_ ⚠conversation_ render_ preface_ to_ string - litert_
lm_ ⚠conversation_ send_ message - litert_
lm_ ⚠conversation_ send_ message_ stream - litert_
lm_ ⚠conversation_ wait_ until_ done - litert_
lm_ ⚠detokenize_ result_ delete - litert_
lm_ ⚠detokenize_ result_ get_ string - litert_
lm_ ⚠embedding_ engine_ compute_ embedding - litert_
lm_ ⚠embedding_ engine_ compute_ embedding_ batch - litert_
lm_ ⚠embedding_ engine_ create - litert_
lm_ ⚠embedding_ engine_ delete - litert_
lm_ ⚠embedding_ engine_ settings_ create - litert_
lm_ ⚠embedding_ engine_ settings_ delete - litert_
lm_ ⚠embedding_ engine_ settings_ set_ activation_ data_ type - litert_
lm_ ⚠embedding_ engine_ settings_ set_ audio_ litert_ dispatch_ lib_ dir - litert_
lm_ ⚠embedding_ engine_ settings_ set_ audio_ num_ threads - litert_
lm_ ⚠embedding_ engine_ settings_ set_ cache_ dir - litert_
lm_ ⚠embedding_ engine_ settings_ set_ litert_ dispatch_ lib_ dir - litert_
lm_ ⚠embedding_ engine_ settings_ set_ max_ input_ length - litert_
lm_ ⚠embedding_ engine_ settings_ set_ min_ input_ length - litert_
lm_ ⚠embedding_ engine_ settings_ set_ num_ threads - litert_
lm_ ⚠embedding_ engine_ settings_ set_ vision_ litert_ dispatch_ lib_ dir - litert_
lm_ ⚠embedding_ engine_ settings_ set_ vision_ tokens_ per_ image - litert_
lm_ ⚠embedding_ options_ create - litert_
lm_ ⚠embedding_ options_ delete - litert_
lm_ ⚠embedding_ options_ get_ input_ overflow_ strategy - litert_
lm_ ⚠embedding_ options_ get_ insert_ special_ tokens - litert_
lm_ ⚠embedding_ options_ get_ normalize - litert_
lm_ ⚠embedding_ options_ get_ output_ size - litert_
lm_ ⚠embedding_ options_ get_ vision_ tokens_ per_ image - litert_
lm_ ⚠embedding_ options_ set_ input_ overflow_ strategy - litert_
lm_ ⚠embedding_ options_ set_ insert_ special_ tokens - litert_
lm_ ⚠embedding_ options_ set_ normalize - litert_
lm_ ⚠embedding_ options_ set_ output_ size - litert_
lm_ ⚠embedding_ options_ set_ vision_ tokens_ per_ image - litert_
lm_ ⚠embedding_ response_ delete - litert_
lm_ ⚠embedding_ response_ get_ size - litert_
lm_ ⚠embedding_ response_ get_ values - litert_
lm_ ⚠embedding_ responses_ delete - litert_
lm_ ⚠embedding_ responses_ get_ at - litert_
lm_ ⚠embedding_ responses_ get_ size - litert_
lm_ ⚠engine_ create - litert_
lm_ ⚠engine_ create_ session - litert_
lm_ ⚠engine_ delete - litert_
lm_ ⚠engine_ detokenize - litert_
lm_ ⚠engine_ get_ start_ token - litert_
lm_ ⚠engine_ get_ stop_ tokens - litert_
lm_ ⚠engine_ settings_ create - litert_
lm_ ⚠engine_ settings_ create_ from_ raw_ file_ descriptor - litert_
lm_ ⚠engine_ settings_ delete - litert_
lm_ ⚠engine_ settings_ enable_ benchmark - litert_
lm_ ⚠engine_ settings_ set_ activation_ data_ type - litert_
lm_ ⚠engine_ settings_ set_ audio_ lora_ rank - litert_
lm_ ⚠engine_ settings_ set_ audio_ num_ threads - litert_
lm_ ⚠engine_ settings_ set_ cache_ dir - litert_
lm_ ⚠engine_ settings_ set_ enable_ speculative_ decoding - litert_
lm_ ⚠engine_ settings_ set_ enable_ ynnpack - litert_
lm_ ⚠engine_ settings_ set_ gpu_ decode_ steps_ per_ sync - litert_
lm_ ⚠engine_ settings_ set_ gpu_ enable_ metal_ residency_ set - litert_
lm_ ⚠engine_ settings_ set_ gpu_ wait_ for_ weight_ uploads - litert_
lm_ ⚠engine_ settings_ set_ litert_ dispatch_ lib_ dir - litert_
lm_ ⚠engine_ settings_ set_ lora_ rank - litert_
lm_ ⚠engine_ settings_ set_ max_ num_ images - litert_
lm_ ⚠engine_ settings_ set_ max_ num_ tokens - litert_
lm_ ⚠engine_ settings_ set_ max_ vision_ tokens_ per_ image - litert_
lm_ ⚠engine_ settings_ set_ num_ decode_ tokens - litert_
lm_ ⚠engine_ settings_ set_ num_ prefill_ tokens - litert_
lm_ ⚠engine_ settings_ set_ num_ threads - litert_
lm_ ⚠engine_ settings_ set_ parallel_ file_ section_ loading - litert_
lm_ ⚠engine_ settings_ set_ prefill_ chunk_ size - litert_
lm_ ⚠engine_ settings_ set_ single_ threaded_ execution - litert_
lm_ ⚠engine_ settings_ set_ speculative_ decoding_ config - litert_
lm_ ⚠engine_ settings_ set_ supported_ audio_ lora_ ranks - litert_
lm_ ⚠engine_ settings_ set_ supported_ lora_ ranks - litert_
lm_ ⚠engine_ settings_ set_ use_ ringbuffers_ local_ attention - litert_
lm_ ⚠engine_ tokenize - litert_
lm_ ⚠experimental_ conversation_ get_ session_ debug_ info - litert_
lm_ ⚠experimental_ engine_ update_ gpu_ enable_ metal_ residency_ set - litert_
lm_ ⚠experimental_ is_ debugger_ enabled - litert_
lm_ ⚠experimental_ session_ debug_ info_ delete - litert_
lm_ ⚠experimental_ session_ debug_ info_ get_ capture_ dir - litert_
lm_ ⚠experimental_ session_ get_ debug_ info - litert_
lm_ ⚠get_ api_ version - litert_
lm_ ⚠get_ last_ error_ message - litert_
lm_ ⚠input_ data_ create - litert_
lm_ ⚠input_ data_ delete - litert_
lm_ ⚠json_ response_ delete - litert_
lm_ ⚠json_ response_ get_ string - litert_
lm_ ⚠loaded_ file_ create - litert_
lm_ ⚠loaded_ file_ delete - litert_
lm_ ⚠loaded_ file_ embedding_ dimension - litert_
lm_ ⚠loaded_ file_ embedding_ signature_ selection - litert_
lm_ ⚠loaded_ file_ has_ speculative_ decoding_ support - litert_
lm_ ⚠loaded_ file_ is_ dynamic_ context - litert_
lm_ ⚠loaded_ file_ max_ context_ tokens - litert_
lm_ ⚠loaded_ file_ max_ vision_ token_ budget - litert_
lm_ ⚠loaded_ file_ min_ runtime_ version - litert_
lm_ ⚠loaded_ file_ modality_ npu_ brand - litert_
lm_ ⚠loaded_ file_ modality_ soc_ name - litert_
lm_ ⚠loaded_ file_ modality_ supported_ backends - litert_
lm_ ⚠loaded_ file_ model_ type - litert_
lm_ ⚠loaded_ file_ sampler_ temperature - litert_
lm_ ⚠loaded_ file_ sampler_ top_ k - litert_
lm_ ⚠loaded_ file_ sampler_ top_ p - litert_
lm_ ⚠loaded_ file_ sampler_ type - litert_
lm_ ⚠loaded_ file_ supports_ function_ calling - litert_
lm_ ⚠loaded_ file_ supports_ input_ modality - litert_
lm_ ⚠loaded_ file_ supports_ thinking - litert_
lm_ ⚠loaded_ file_ vision_ signature_ selection - litert_
lm_ ⚠no_ repeat_ ngram_ config_ create - litert_
lm_ ⚠no_ repeat_ ngram_ config_ delete - litert_
lm_ ⚠no_ repeat_ ngram_ config_ set_ no_ repeat_ ngram_ size - litert_
lm_ ⚠no_ repeat_ ngram_ config_ set_ window_ size - litert_
lm_ ⚠repetition_ penalty_ config_ create - litert_
lm_ ⚠repetition_ penalty_ config_ delete - litert_
lm_ ⚠repetition_ penalty_ config_ set_ frequency_ penalty - litert_
lm_ ⚠repetition_ penalty_ config_ set_ presence_ penalty - litert_
lm_ ⚠repetition_ penalty_ config_ set_ repetition_ penalty - litert_
lm_ ⚠repetition_ penalty_ config_ set_ window_ size - litert_
lm_ ⚠responses_ delete - litert_
lm_ ⚠responses_ get_ num_ candidates - litert_
lm_ ⚠responses_ get_ num_ token_ scores_ at - litert_
lm_ ⚠responses_ get_ response_ text_ at - litert_
lm_ ⚠responses_ get_ score_ at - litert_
lm_ ⚠responses_ get_ token_ length_ at - litert_
lm_ ⚠responses_ get_ token_ scores_ at - litert_
lm_ ⚠responses_ has_ score_ at - litert_
lm_ ⚠responses_ has_ token_ length_ at - litert_
lm_ ⚠responses_ has_ token_ scores_ at - litert_
lm_ ⚠sampler_ params_ create - litert_
lm_ ⚠sampler_ params_ delete - litert_
lm_ ⚠sampler_ params_ set_ seed - litert_
lm_ ⚠sampler_ params_ set_ temperature - litert_
lm_ ⚠sampler_ params_ set_ top_ k - litert_
lm_ ⚠sampler_ params_ set_ top_ p - litert_
lm_ ⚠session_ cancel_ process - litert_
lm_ ⚠session_ clone - litert_
lm_ ⚠session_ config_ create - litert_
lm_ ⚠session_ config_ delete - litert_
lm_ ⚠session_ config_ set_ apply_ prompt_ template - litert_
lm_ ⚠session_ config_ set_ audio_ lora_ path - litert_
lm_ ⚠session_ config_ set_ enable_ speculative_ decoding - litert_
lm_ ⚠session_ config_ set_ lora_ path - litert_
lm_ ⚠session_ config_ set_ max_ output_ tokens - litert_
lm_ ⚠session_ config_ set_ sampler_ params - litert_
lm_ ⚠session_ config_ set_ speculative_ decoding_ config - litert_
lm_ ⚠session_ delete - litert_
lm_ ⚠session_ generate_ content - litert_
lm_ ⚠session_ generate_ content_ stream - litert_
lm_ ⚠session_ get_ benchmark_ info - litert_
lm_ ⚠session_ rewind_ to_ checkpoint - litert_
lm_ ⚠session_ rewind_ to_ step - litert_
lm_ ⚠session_ run_ decode - litert_
lm_ ⚠session_ run_ decode_ async - litert_
lm_ ⚠session_ run_ prefill - litert_
lm_ ⚠session_ run_ text_ scoring - litert_
lm_ ⚠session_ save_ checkpoint - litert_
lm_ ⚠set_ min_ log_ level - litert_
lm_ ⚠speculative_ decoding_ config_ create - litert_
lm_ ⚠speculative_ decoding_ config_ delete - litert_
lm_ ⚠speculative_ decoding_ config_ set_ mode - litert_
lm_ ⚠stream_ chunk_ get_ error - litert_
lm_ ⚠stream_ chunk_ get_ text - litert_
lm_ ⚠stream_ chunk_ is_ final - litert_
lm_ ⚠suppress_ tokens_ config_ create - litert_
lm_ ⚠suppress_ tokens_ config_ delete - litert_
lm_ ⚠suppress_ tokens_ config_ set_ suppress_ tokens - litert_
lm_ ⚠thinking_ config_ create - litert_
lm_ ⚠thinking_ config_ delete - litert_
lm_ ⚠thinking_ config_ set_ enable_ thinking - litert_
lm_ ⚠thinking_ config_ set_ thinking_ token_ budget - litert_
lm_ ⚠token_ union_ delete - litert_
lm_ ⚠token_ union_ get_ ids - litert_
lm_ ⚠token_ union_ get_ string - litert_
lm_ ⚠token_ union_ get_ type - litert_
lm_ ⚠token_ unions_ delete - litert_
lm_ ⚠token_ unions_ get_ num_ tokens - litert_
lm_ ⚠token_ unions_ get_ token_ at - litert_
lm_ ⚠tokenize_ result_ delete - litert_
lm_ ⚠tokenize_ result_ get_ num_ tokens - litert_
lm_ ⚠tokenize_ result_ get_ tokens
Type Aliases§
- Lite
RtLm Activation Data Type - Lite
RtLm Backend Type - Lite
RtLm Constraint Provider Type - Lite
RtLm Constraint Type - Lite
RtLm Input Data Type - Lite
RtLm Input Overflow Strategy - Lite
RtLm LogSeverity - Lite
RtLm Modality - Lite
RtLm Model Type - Lite
RtLm NpuBrand - Lite
RtLm Sampler Type - Lite
RtLm Speculative Decoding Mode - Lite
RtLm Status Code - Lite
RtLm Stream Callback - Lite
RtLm Token Union Type