MCPcopy Create free account

hub / github.com/UbiquitousLearning/mllm / functions

Functions5,971 in github.com/UbiquitousLearning/mllm

⨍ Functions 5,971◇ Types & classes 1,855↳ Endpoints 4
↓ 7 callersMethoddecode
mllm/models/vocos/modeling_vocos.hpp:322
↓ 7 callersFunctiondefine_ir
(name: str, op: Cls, val: Cls, attr: Cls)
mllm/compile/ir/rtti_kind_gen.py:182
↓ 7 callersMethodexecute
Executes a command in the persistent shell. :param command: The shell command to execute. :return: The output (stdout and std
pymllm/utils/adb.py:79
↓ 7 callersFunctionfloat_to_fp16s
mllm/backends/qnn/custom-op-package/LLaMAPackage/src/ops/LLaMASuperSiLU.cpp:102
↓ 7 callersMethodgetKind
mllm/compile/symbolic_expr/AST.hpp:43
↓ 7 callersMethodgetOpType
mllm/core/BaseOp.cpp:24
↓ 7 callersFunctioninitLocal
some OS has no en_US.UTF-8 but has C.UTF-8.
mllm/preprocessor/tokenizers/Unicode.hpp:22
↓ 7 callersMethodmean
mllm/core/Tensor.cpp:428
↓ 7 callersMethodmeta
mllm/utils/Argparse.hpp:111
↓ 7 callersMethodmin
mllm/core/Tensor.cpp:413
↓ 7 callersMethodregisterDispatcher
mllm/engine/DispatcherManager.cpp:29
↓ 7 callersFunctionscatter2Shards
mllm/nn/Functional.cpp:146
↓ 7 callersMethodsetBelongsTo
mllm/compile/ir/Node.cpp:46
↓ 7 callersFunctionsetQuantScale
mllm/backends/qnn/QNNUtils.hpp:143
↓ 7 callersMethodtensor
mllm/preprocessor/visual/Image.cpp:86
↓ 7 callersMethodwait
! \brief Blocks the calling thread until this event completes. * * Wraps clWaitForEvents(). */
mllm/backends/opencl/vendors/OpenCL-Headers/CL/opencl.hpp:3759
↓ 6 callersMethodClose
()
mllm-cli/mllm/c.go:93
↓ 6 callersMethodT
mllm/core/Tensor.cpp:440
↓ 6 callersMethod__init__
(self, config)
pymllm/backends/qualcomm/transformers/qwen2/modeling_qwen2.py:47
↓ 6 callersMethod__init__
(self, config)
pymllm/backends/qualcomm/transformers/llama/modeling_llama.py:164
↓ 6 callersMethodaddConstant
mllm/core/Tensor.cpp:301
↓ 6 callersMethodblank
mllm/compile/ir/IRPrinter.hpp:41
↓ 6 callersMethodcomma
mllm/compile/ir/IRPrinter.hpp:34
↓ 6 callersFunctioncompute_strides
mllm/backends/cpu/kernels/arm/permute.cpp:13
↓ 6 callersMethodconvert_to_deploy
In-place replacement of self.weight: Float Parameter -> Int Buffer
pymllm/backends/qualcomm/transformers/core/rms_norm.py:51
↓ 6 callersMethoddimensions
\brief Queries the number of dimensions in the range.
mllm/backends/opencl/vendors/OpenCL-Headers/CL/opencl.hpp:6155
↓ 6 callersMethoddisable_fake_quant
(self)
pymllm/backends/qualcomm/transformers/qwen3/runner.py:227
↓ 6 callersMethoddisable_observer
Freeze scale and zero_point calculation.
pymllm/backends/qualcomm/transformers/core/qdq.py:76
↓ 6 callersMethodenable_fake_quant
(self)
pymllm/backends/qualcomm/transformers/qwen3/runner.py:224
↓ 6 callersMethodencode
mllm/models/minicpm_o2_6/tokenization_chattts.cpp:34
↓ 6 callersMethodequalConstant
mllm/core/Tensor.cpp:466
↓ 6 callersMethodeval
mllm/compile/symbolic_expr/Eval.cpp:66
↓ 6 callersMethodflush
mllm/backends/opencl/vendors/OpenCL-Headers/CL/opencl.hpp:9573
↓ 6 callersFunctionfreeMultiPtr
mllm/backends/qnn/QNNModel.cpp:13
↓ 6 callersMethodfunc
mllm/backends/qnn/aot/QnnWrappersAPI.hpp:160
↓ 6 callersMethodgetCurrentSeqCnt
mllm/nn/layers/KVCache.cpp:30
↓ 6 callersFunctiongetLinearIndex
Helper function to convert multi-dimensional index to linear index
mllm/backends/cpu/ops/MaskedScatterOp.cpp:54
↓ 6 callersMethodgetWrapper
mllm/backends/qnn/aot/QnnWrappersAPI.hpp:53
↓ 6 callersFunctionhsum
mllm/backends/cpu/kernels/common/llamafile/llamafile_sgemm.cpp:151
↓ 6 callersMethodlookupCustomizedOpId
mllm/engine/Context.cpp:157
↓ 6 callersMethodmanualCacheLengthUpdate
algorithms/lazy_vlm/HKVCache.cpp:190
↓ 6 callersMethodmulConstant
mllm/core/Tensor.cpp:315
↓ 6 callersFunctionnoSharingSingleInAndSingleOutQuantAnnoAttr
mllm/backends/qnn/aot/passes/LLMQuantRecipePass.cpp:147
↓ 6 callersFunctionparseQcomTargetMachineFromJSONFile
mllm/backends/qnn/aot/QnnTargetMachineParser.cpp:89
↓ 6 callersMethodpush_file
Push file to device
pymllm/utils/adb.py:257
↓ 6 callersMethodsaveContext
mllm/backends/qnn/QNNBackend.cpp:543
↓ 6 callersFunctionsave_dot
tests/engine/RadixTreeTest.cpp:40
↓ 6 callersMethodsetOpType
mllm/core/BaseOp.cpp:26
↓ 6 callersFunctionsigmoid
mllm/nn/Functional.cpp:227
↓ 6 callersMethodsqueeze
(self, dim: int)
pymllm/ffi/__init__.py:218
↓ 6 callersFunctionsum
mllm/nn/Functional.cpp:123
↓ 6 callersMethodsum
mllm/core/Tensor.cpp:423
↓ 6 callersFunctiontestCase
mllm-ext-opset/cpu/radix_swa_sink/tests/main.cpp:66
↓ 6 callersFunctiontestCase
mllm-ext-opset/cpu/fa2_swa_sink/tests/main.cpp:64
↓ 6 callersMethodupdateHiddenStateCache
algorithms/lazy_vlm/HKVCache.cpp:169
↓ 6 callersFunctionvec_pack_64_uint2_values
mllm/backends/cpu/kernels/arm/quantize/bitspack/u2.hpp:76
↓ 6 callersFunctionvec_pack_64_uint6_values
mllm/backends/cpu/kernels/arm/quantize/bitspack/u6.hpp:126
↓ 6 callersFunctionvec_unpack_64_uint2_values
mllm/backends/cpu/kernels/arm/quantize/bitspack/u2.hpp:94
↓ 6 callersFunctionvec_unpack_64_uint6_values
mllm/backends/cpu/kernels/arm/quantize/bitspack/u6.hpp:157
↓ 6 callersMethodverify
mllm-kernel/include/mllm_kernel/tensor.hpp:187
↓ 6 callersFunctionview
mllm/nn/Functional.cpp:48
↓ 6 callersMethodvisitHiddenStateCache
algorithms/lazy_vlm/HKVCache.cpp:183
↓ 6 callersFunctionwhere
mllm/nn/Functional.cpp:222
↓ 6 callersMethodworkspace_size
mllm/backends/cpu/kernels/arm/linear/kai.cpp:52
↓ 5 callersMethod__init__
(self, config)
pymllm/backends/qualcomm/transformers/qwen3/modeling_qwen3.py:62
↓ 5 callersFunction__mllm_blas_matmul_fp32_gemv_nt_t_decode_small_d_qk
Optimized for decoding. Q: [1, D] K: [S, D] D is small in mllm's case(small language model). D=64, 96, 128 ...
mllm/backends/cpu/kernels/arm/mllm_blas/mllm_blas_sgemm.cpp:40
↓ 5 callersMethodas_uint
mllm/preprocessor/tokenizers/llama_cpp_unicode/unicode.h:67
↓ 5 callersFunctioncalculateBroadcastInfo
mllm/backends/cpu/ops/ElewiseOps.cpp:11
↓ 5 callersMethoderr
! \brief Get error code associated with exception * * \return The error code. */
mllm/backends/opencl/vendors/OpenCL-Headers/CL/opencl.hpp:780
↓ 5 callersFunctionfloat_to_bits
mllm/backends/qnn/custom-op-package/LLaMAPackage/src/ops/RMSNorm.cpp:191
↓ 5 callersMethodforward
examples/qwen3_qnn_aot/modeling_qwen_qnn_aot.hpp:160
↓ 5 callersMethodgetContext
mllm/compile/ir/Node.cpp:420
↓ 5 callersMethodgetGraphInputTensorWrappers
Get input/output tensor wrappers
mllm/backends/qnn/QNNModel.hpp:74
↓ 5 callersMethodget_weak
mllm/utils/WeakOwner.hpp:55
↓ 5 callersMethodhas_value
mllm-kernel/include/mllm_kernel/tensor.hpp:176
↓ 5 callersMethodisRedirect
mllm/core/BaseOp.hpp:72
↓ 5 callersFunctionmakeBytes2UnicodeMap
mllm/preprocessor/tokenizers/Unicode.cpp:78
↓ 5 callersFunctionquantize_q4_0_nr_bl
mllm/backends/cpu/kernels/arm/linear/ggml/gemm_aarch64.cpp:322
↓ 5 callersMethodregisterAllocator
mllm/engine/MemoryManager.cpp:16
↓ 5 callersMethodregisterBackend
mllm/engine/Context.cpp:35
↓ 5 callersMethodsampleGreedy
mllm/models/ARGeneration.cpp:334
↓ 5 callersMethodsetAbsoluteName
mllm/nn/AbstractNnNode.cpp:23
↓ 5 callersMethodsetCurrentSeqCnt
mllm/nn/layers/KVCache.cpp:26
↓ 5 callersMethodsetDevice
mllm/compile/ir/Node.cpp:256
↓ 5 callersFunctionsetLogLevel
mllm/mllm.cpp:68
↓ 5 callersMethodstr
mllm/compile/symbolic_expr/AST.hpp:215
↓ 5 callersMethodtobytes
(self)
pymllm/ffi/__init__.py:138
↓ 5 callersMethodtrace
examples/qwen2_qnn_aot/modeling_qwen2_qnn_aot.hpp:422
↓ 5 callersMethodtrace
examples/qwen3_qnn_aot/modeling_qwen_qnn_aot.hpp:427
↓ 5 callersFunctionunicode_cpt_flags
mllm/preprocessor/tokenizers/llama_cpp_unicode/unicode.cpp:624
↓ 5 callersFunctionunicode_cpts_from_utf8
mllm/preprocessor/tokenizers/llama_cpp_unicode/unicode.cpp:616
↓ 4 callersMethodSessionID
()
mllm-cli/mllm/c.go:143
↓ 4 callersFunction__mllm_blas_matmul_fp32_gemv_nt_nt_decode_small_d_wv
Optimized for decoding. W: [B, H, 1, S] V: [B, H, S, D] D is small in mllm's case(small language model). D=64, 96, 128 ...
mllm/backends/cpu/kernels/arm/mllm_blas/mllm_blas_sgemm.cpp:147
↓ 4 callersMethod__setattr__
(self, name: str, value)
pymllm/nn/_module.py:16
↓ 4 callersMethodaddStaticTensor
mllm/backends/qnn/QNNModel.cpp:194
↓ 4 callersMethodallocSize
mllm/backends/cpu/CPUAllocator.cpp:67
↓ 4 callersMethodassign
mllm/compile/ir/IRPrinter.hpp:31
↓ 4 callersMethodasyncSubmit
mllm/engine/DispatcherManager.cpp:23
↓ 4 callersMethodbegin
mllm/engine/prefix_cache/RadixTree.hpp:95
↓ 4 callersFunctioncalculateBroadcastInfo
mllm/backends/opencl/ops/ElewiseOp.cpp:20
← previousnext →301–400 of 5,971, ranked by callers