MCPcopy Create free account

hub / github.com/OpenNMT/CTranslate2 / functions

Functions2,508 in github.com/OpenNMT/CTranslate2

Functiondequantize_gemm_output_row
src/cpu/kernels.cc:654
Methoddevice
Device management.
include/ctranslate2/storage_view.h:83
Methoddevice
include/ctranslate2/layers/decoder.h:75
Methoddevice_index
include/ctranslate2/storage_view.h:87
Methoddiagnostics
include/nlohmann/json.hpp:4321
Methoddispatch
src/ops/multinomial.cc:22
Methoddiv
src/cpu/vec.h:127
Methoddiv
src/cpu/vec_avx512.h:130
Methoddiv
src/cpu/vec_neon.h:146
Methoddiv
src/cpu/vec_avx.h:172
Methoddiyfp
include/nlohmann/json.hpp:16847
Methoddtype
(self)
python/ctranslate2/specs/model_spec.py:683
Methoddtype
(self)
python/ctranslate2/specs/model_spec.py:737
Methodempty
include/ctranslate2/batch_reader.h:40
Methodencode
python/cpp/wav2vec2.cc:14
Methodencode
python/cpp/wav2vec2bert.cc:14
Methodencode
src/models/wav2vec2.cc:68
Methodencode
src/models/sequence_to_sequence.cc:216
Methodencode
src/models/wav2vec2bert.cc:68
Functionend_array
include/nlohmann/json.hpp:6841
Methodend_array
include/nlohmann/json.hpp:7069
Functionend_object
include/nlohmann/json.hpp:6819
Methodend_object
include/nlohmann/json.hpp:7016
Methodensure_dtype
src/models/model.cc:304
Methoderf
src/cpu/vec.h:103
Methoderf
src/cpu/vec_avx512.h:106
Methoderf
src/cpu/vec_neon.h:122
Methoderf
src/cpu/vec_avx.h:148
Functionerfc
Complementary error function. This function may be 1 ULP off the correctly rounded exact result for any rounding mode in <0.5% of inputs. See also:**
include/half_float/half.hpp:3988
Methodexception_message
! @param[in] format the current format @param[in] detail a detailed error message @param[in] context further context information
include/nlohmann/json.hpp:11988
Methodexception_message
include/nlohmann/json.hpp:12564
Methodexp
src/cpu/vec.h:83
Methodexp
src/cpu/vec_avx512.h:86
Methodexp
src/cpu/primitives.cc:358
Methodexp
src/cpu/vec_neon.h:102
Methodexp
src/cpu/vec_avx.h:128
Functionexp2_post
include/half_float/half.hpp:1688
Functionexp<TARGET_ISA>
src/cpu/kernels.cc:215
Functionexpm1
Exponential minus one. This function may be 1 ULP off the correctly rounded exact result in <0.05% of inputs for `std::round_to_nearest` and in <1% of
include/half_float/half.hpp:2997
Methodf31
include/half_float/half.hpp:1847
Functionfdim
Positive difference. This function is exact to rounding for all rounding modes. See also:** Documentation for [std::fdim](https://en.cppreference.com
include/half_float/half.hpp:2904
Functionfeclearexcept
Clear exception flags. This function works even if [automatic exception flag handling](\ref HALF_ERRHANDLING_FLAGS) is disabled, but in that case manu
include/half_float/half.hpp:4518
Functionfegetexceptflag
Save exception flags. This function works even if [automatic exception flag handling](\ref HALF_ERRHANDLING_FLAGS) is disabled, but in that case manua
include/half_float/half.hpp:4548
Functionfesetexceptflag
Restore exception flags. This only copies the specified exception state (including unset flags) without incurring any additional exception handling. T
include/half_float/half.hpp:4559
Functionfetestexcept
Test exception flags. This function works even if [automatic exception flag handling](\ref HALF_ERRHANDLING_FLAGS) is disabled, but in that case manua
include/half_float/half.hpp:4527
Functionfethrowexcept
Throw C++ exceptions based on set exception flags. This function manually throws a corresponding C++ exception if one of the specified flags is set, n
include/half_float/half.hpp:4572
Methodfill_batch_with_fixed_increment
src/batch_reader.cc:38
Methodfill_batch_with_variable_increment
src/batch_reader.cc:60
Methodfill_buffer
UTF-32
include/nlohmann/json.hpp:6250
Methodfill_buffer
UTF-16
include/nlohmann/json.hpp:6308
Methodfill_buffer
include/nlohmann/json.hpp:6397
Methodfinalize
src/devices.cc:206
Methodfinalize
Called after the work loop.
include/ctranslate2/thread_pool.h:69
Methodfinalize
include/ctranslate2/replica_pool.h:355
Functionfix_pybind11_signatures
( app, what, name, obj, options, signature, return_annotation )
docs/conf.py:44
Functionfixed2half
include/half_float/half.hpp:884
Functionflash_fwd_kernel
include/ctranslate2/ops/flash-attention/flash_fwd_launch_template.h:15
Functionflash_fwd_splitkv_combine_kernel
include/ctranslate2/ops/flash-attention/flash_fwd_launch_template.h:37
Functionflash_fwd_splitkv_kernel
include/ctranslate2/ops/flash-attention/flash_fwd_launch_template.h:27
Functionfloat2half
include/half_float/half.hpp:1064
Functionfloat2half_impl
include/half_float/half.hpp:907
Functionfma
Fused multiply add. This function is exact to rounding for all rounding modes. See also:** Documentation for [std::fma](https://en.cppreference.com/w
include/half_float/half.hpp:2809
Functionfmax
Maximum of half expressions. See also:** Documentation for [std::fmax](https://en.cppreference.com/w/cpp/numeric/math/fmax). \param x first operand \p
include/half_float/half.hpp:2878
Functionfmin
Minimum of half expressions. See also:** Documentation for [std::fmin](https://en.cppreference.com/w/cpp/numeric/math/fmin). \param x first operand \p
include/half_float/half.hpp:2890
Functionfmod
Remainder of division. See also:** Documentation for [std::fmod](https://en.cppreference.com/w/cpp/numeric/math/fmod). \param x first operand \param y
include/half_float/half.hpp:2743
Methodforward_batch
python/cpp/generator.cc:98
Methodforward_batch_async
src/generator.cc:43
Methodforward_impl
src/models/language_model.cc:352
Functionfp_test_name
tests/test_utils.h:79
Methodfree
src/cpu/allocator.cc:36
Methodfree
src/cuda/allocator.cc:67
Methodfree
src/cuda/allocator.cc:104
Methodfree
include/ctranslate2/allocator.h:19
Functionfrom_bson
include/nlohmann/json.hpp:23734
Functionfrom_cbor
include/nlohmann/json.hpp:23538
Methodfrom_config
Creates a Transformer decoder model specification. Args: num_layers: Number of decoder layers. num_heads: Number of atten
python/ctranslate2/specs/transformer_spec.py:629
Methodfrom_json
include/nlohmann/json.hpp:5766
Methodfrom_json_file
src/vocabulary.cc:32
Functionfrom_json_tuple_impl_base
include/nlohmann/json.hpp:4933
Functionfrom_msgpack
include/nlohmann/json.hpp:23595
Methodfrom_text_file
src/vocabulary.cc:9
Functionfrom_ubjson
include/nlohmann/json.hpp:23649
Functionfront
@brief access the first element @sa https://json.nlohmann.me/api/basic_json/front/
include/nlohmann/json.hpp:21560
Functionfuse_linear
(spec, layers)
python/ctranslate2/converters/utils.py:4
Functionfuse_linear_prequant
(spec, layers, axis)
python/ctranslate2/converters/utils.py:38
Functiongamma
include/half_float/half.hpp:1937
Functiongather_batch_inplace
src/ops/gather.cc:30
Methodgelu
src/cpu/primitives.cc:305
Functiongelu<TARGET_ISA>
src/cpu/kernels.cc:240
Methodgelu_sigmoid
src/cpu/primitives.cc:323
Functiongelu_sigmoid<TARGET_ISA>
src/cpu/kernels.cc:250
Methodgelu_tanh
src/cpu/primitives.cc:314
Functiongelu_tanh<TARGET_ISA>
src/cpu/kernels.cc:245
Methodgemm
src/cpu/primitives.cc:646
Methodgemm_batch_strided
src/cpu/primitives.cc:1077
Methodgemm_pack_b
src/cpu/primitives.cc:554
Methodgenerate
src/models/language_model.cc:55
Functiongenerator_async_generate_tokens
Yields tokens asynchronously as they are generated by the model. Arguments: prompt: Batch of start tokens. If the decoder starts from a
python/ctranslate2/extensions.py:351
Functiongenerator_generate_iterable
Generates from an iterable of tokenized prompts. This method is built on top of :meth:`ctranslate2.Generator.generate_batch` to efficiently r
python/ctranslate2/extensions.py:130
Functiongenerator_generate_tokens
Yields tokens as they are generated by the model. Arguments: prompt: Batch of start tokens. If the decoder starts from a special st
python/ctranslate2/extensions.py:270
← previousnext →1,401–1,500 of 2,508, ranked by callers