whisper4dart_bindings_generated library

Classes

ggml_backend

ggml_backend_buffer

ggml_backend_buffer_type

ggml_backend_buffer_usage

Backend buffer

ggml_backend_dev_caps

functionality supported by the device

ggml_backend_dev_props

all the device properties

ggml_backend_dev_type

Backend device

ggml_backend_device

ggml_backend_event

ggml_backend_feature

Get a list of feature flags supported by the backend (returns a NULL-terminated array)

ggml_backend_graph_copy

Utils

ggml_backend_reg

ggml_backend_sched

ggml_bf16_t

google brain half-precision bfloat16

ggml_cgraph

ggml_context

ggml_cplan

the compute plan that needs to be prepared for ggml_graph_compute() since https://github.com/ggerganov/ggml/issues/287

ggml_ftype

model file types

ggml_init_params

ggml_log_level

ggml_numa_strategy

numa strategies

ggml_object

ggml_object_type

ggml_op

available tensor operations:

ggml_op_pool

ggml_prec

precision

ggml_sched_priority

scheduling priorities

ggml_sort_order

sort rows

ggml_status

ggml_tensor

n-dimensional tensor

ggml_tensor_flag

this tensor...

ggml_threadpool

ggml_threadpool_params

threadpool params Use ggml_threadpool_params_default() or ggml_threadpool_params_init() to populate the defaults

ggml_type

NOTE: always add types at the end of the enum to keep backward compatibility

ggml_type_traits

ggml_type_traits_cpu

ggml_unary_op

UnnamedStruct1

UnnamedStruct2

whisper_ahead

whisper_aheads

whisper_alignment_heads_preset

whisper_context

C interface

whisper_context_params

whisper_full_params

Parameters for the whisper_full() function If you change the order or add new parameters, make sure to update the default values in whisper.cpp: whisper_full_default_params()

whisper_grammar_element

whisper_gretype

grammar element type

whisper_model_loader

whisper_sampling_strategy

Available sampling strategies

whisper_state

whisper_timings

Performance information from the default state.

whisper_token_data

WhisperDartBindings

Bindings for src/whisper4dart.h.

Constants

GGML_DEFAULT_GRAPH_SIZE → const int

GGML_DEFAULT_N_THREADS → const int

GGML_EXIT_ABORTED → const int

GGML_EXIT_SUCCESS → const int

GGML_FILE_MAGIC → const int

GGML_FILE_VERSION → const int

GGML_KQ_MASK_PAD → const int

GGML_MAX_DIMS → const int

GGML_MAX_N_THREADS → const int

GGML_MAX_NAME → const int

GGML_MAX_OP_PARAMS → const int

GGML_MAX_PARAMS → const int

GGML_MAX_SRC → const int

GGML_MEM_ALIGN → const int

GGML_N_TASKS_MAX → const int

GGML_QNT_VERSION → const int

GGML_QNT_VERSION_FACTOR → const int

GGML_ROPE_TYPE_MROPE → const int

GGML_ROPE_TYPE_NEOX → const int

GGML_ROPE_TYPE_VISION → const int

WHISPER_CHUNK_SIZE → const int

WHISPER_HOP_LENGTH → const int

WHISPER_N_FFT → const int

WHISPER_SAMPLE_RATE → const int

Typedefs

FILE = _IO_FILE

ggml_abort_callback = Pointer<NativeFunction<Bool Function(Pointer<Void> data)>>

Abort callback If not NULL, called before ggml computation If it returns true, the computation is aborted

ggml_backend_buffer_t = Pointer<ggml_backend_buffer>

ggml_backend_buffer_type_t = Pointer<ggml_backend_buffer_type>

ggml_backend_dev_t = Pointer<ggml_backend_device>

ggml_backend_eval_callback = Pointer<NativeFunction<Bool Function(Int node_index, Pointer<ggml_tensor> t1, Pointer<ggml_tensor> t2, Pointer<Void> user_data)>>

ggml_backend_event_t = Pointer<ggml_backend_event>

ggml_backend_graph_plan_t = Pointer<Void>

ggml_backend_reg_t = Pointer<ggml_backend_reg>

ggml_backend_sched_eval_callback = Pointer<NativeFunction<Bool Function(Pointer<ggml_tensor> t, Bool ask, Pointer<Void> user_data)>>

Evaluation callback for each node in the graph (set with ggml_backend_sched_set_eval_callback) when ask == true, the scheduler wants to know if the user wants to observe this node this allows the scheduler to batch nodes together in order to evaluate them in a single call

ggml_backend_sched_t = Pointer<ggml_backend_sched>

The backend scheduler allows for multiple backend devices to be used together Handles compute buffer allocation, assignment of tensors to backends, and copying of tensors between backends The backends are selected based on:

ggml_backend_t = Pointer<ggml_backend>

ggml_binary_op_f32_t = Pointer<NativeFunction<Void Function(Int, Pointer<Float>, Pointer<Float>, Pointer<Float>)>>

ggml_custom1_op_f32_t = Pointer<NativeFunction<Void Function(Pointer<ggml_tensor>, Pointer<ggml_tensor>)>>

ggml_custom1_op_t = Pointer<NativeFunction<Void Function(Pointer<ggml_tensor> dst, Pointer<ggml_tensor> a, Int ith, Int nth, Pointer<Void> userdata)>>

custom operators v2

ggml_custom2_op_f32_t = Pointer<NativeFunction<Void Function(Pointer<ggml_tensor>, Pointer<ggml_tensor>, Pointer<ggml_tensor>)>>

ggml_custom2_op_t = Pointer<NativeFunction<Void Function(Pointer<ggml_tensor> dst, Pointer<ggml_tensor> a, Pointer<ggml_tensor> b, Int ith, Int nth, Pointer<Void> userdata)>>

ggml_custom3_op_f32_t = Pointer<NativeFunction<Void Function(Pointer<ggml_tensor>, Pointer<ggml_tensor>, Pointer<ggml_tensor>, Pointer<ggml_tensor>)>>

ggml_custom3_op_t = Pointer<NativeFunction<Void Function(Pointer<ggml_tensor> dst, Pointer<ggml_tensor> a, Pointer<ggml_tensor> b, Pointer<ggml_tensor> c, Int ith, Int nth, Pointer<Void> userdata)>>

ggml_fp16_t = Uint16

ieee 754-2008 half-precision float16 todo: make this not an integral type

ggml_from_float_t = Pointer<NativeFunction<Void Function(Pointer<Float> x, Pointer<Void> y, Int64 k)>>

ggml_guid_t = Pointer<Pointer<Uint8>>

ggml_log_callback = Pointer<NativeFunction<Void Function(Int32 level, Pointer<Char> text, Pointer<Void> user_data)>>

TODO these functions were sandwiched in the old optimization interface, is there a better place for them?

ggml_threadpool_t = Pointer<ggml_threadpool>

ggml_to_float_t = Pointer<NativeFunction<Void Function(Pointer<Void> x, Pointer<Float> y, Int64 k)>>

ggml_unary_op_f32_t = Pointer<NativeFunction<Void Function(Int, Pointer<Float>, Pointer<Float>)>>

custom operators

ggml_vec_dot_t = Pointer<NativeFunction<Void Function(Int n, Pointer<Float> s, Size bs, Pointer<Void> x, Size bx, Pointer<Void> y, Size by, Int nrc)>>

Internal types and functions exposed for tests and benchmarks

whisper_encoder_begin_callback = Pointer<NativeFunction<Bool Function(Pointer<whisper_context> ctx, Pointer<whisper_state> state, Pointer<Void> user_data)>>

Encoder begin callback If not NULL, called before the encoder starts If it returns false, the computation is aborted

whisper_logits_filter_callback = Pointer<NativeFunction<Void Function(Pointer<whisper_context> ctx, Pointer<whisper_state> state, Pointer<whisper_token_data> tokens, Int n_tokens, Pointer<Float> logits, Pointer<Void> user_data)>>

Logits filter callback Can be used to modify the logits before sampling If not NULL, called after applying temperature to logits

whisper_new_segment_callback = Pointer<NativeFunction<Void Function(Pointer<whisper_context> ctx, Pointer<whisper_state> state, Int n_new, Pointer<Void> user_data)>>

Text segment callback Called on every newly generated text segment Use the whisper_full_...() functions to obtain the text segments

whisper_progress_callback = Pointer<NativeFunction<Void Function(Pointer<whisper_context> ctx, Pointer<whisper_state> state, Int progress, Pointer<Void> user_data)>>

Progress callback

whisper_token = Int32