Code
Hub
Workspaces
Following
Trending
Connect
MCP
copy
Create free account
hub
/
github.com/ml-explore/mlx
/ functions
Functions
4,946 in github.com/ml-explore/mlx
⨍
Functions
4,946
◇
Types & classes
770
↳
Endpoints
5
↓ 47 callers
Function
Op
mlx/backend/cpu/simd/neon_fp16_simd.h:99
↓ 47 callers
Function
broadcast_shapes
mlx/utils.cpp:136
↓ 47 callers
Method
get_library
mlx/backend/metal/device.cpp:527
↓ 47 callers
Method
ptr
mlx/backend/cuda/event.h:71
↓ 47 callers
Method
rand
(shape)
python/tests/test_blas.py:1372
↓ 47 callers
Function
where
mlx/ops.cpp:1874
↓ 46 callers
Method
next
Iteration helper */
mlx/backend/metal/kernels/steel/attn/loader.h:258
↓ 45 callers
Method
dequantize
(self)
python/mlx/nn/layers/quantized.py:379
↓ 45 callers
Function
moveaxis
mlx/ops.cpp:1506
↓ 45 callers
Function
split
mlx/ops.cpp:1058
↓ 44 callers
Method
dispatch_threadgroups
mlx/backend/metal/device.cpp:356
↓ 41 callers
Method
length
mlx/3rdparty/pocketfft.h:2566
↓ 41 callers
Function
malloc_async
mlx/backend/cuda/allocator.cpp:397
↓ 40 callers
Method
load_safe
mlx/backend/metal/kernels/steel/gemm/nax.h:743
↓ 40 callers
Method
pop_back
mlx/small_vector.h:336
↓ 40 callers
Function
size
The number of elements in the array. */
mlx/array.h:100
↓ 40 callers
Function
square
mlx/ops.cpp:2983
↓ 39 callers
Function
is_comparable_with_array
python/src/utils.h:44
↓ 39 callers
Function
promote_types
mlx/dtype.cpp:86
↓ 39 callers
Method
step
mlx/backend/common/utils.h:104
↓ 37 callers
Function
broadcast_arrays
Broadcast the input arrays against one another while ignoring the * axes specified in `ignore_axes`. Note, this API is internal only. * The `ignore_
mlx/ops.cpp:1630
↓ 37 callers
Method
jvp
mlx/fast.cpp:29
↓ 37 callers
Function
key
mlx/random.cpp:30
↓ 37 callers
Function
min
mlx/ops.h:726
↓ 37 callers
Function
throw_invalid_operation
python/src/utils.h:61
↓ 35 callers
Function
at_least_float
mlx/ops.cpp:57
↓ 35 callers
Method
load
mlx/export.cpp:494
↓ 35 callers
Method
size
mlx/3rdparty/pocketfft.h:229
↓ 34 callers
Function
bench
(f, *args)
benchmarks/python/comparative/bench_mlx.py:37
↓ 34 callers
Function
copy_cpu
mlx/backend/cpu/copy.cpp:312
↓ 34 callers
Function
equal
mlx/ops.cpp:1758
↓ 34 callers
Function
gather
mlx/ops.h:1060
↓ 34 callers
Method
raw_ptr
mlx/backend/cuda/allocator.cpp:410
↓ 33 callers
Function
bench
(f, *args)
benchmarks/python/comparative/bench_torch.py:36
↓ 33 callers
Function
exp
mlx/backend/cpu/simd/math.h:28
↓ 33 callers
Function
grad
* Returns a function which computes the gradient of the input function with * respect to a vector of input arrays. * * The function being differ
mlx/transforms.h:128
↓ 33 callers
Function
prod
mlx/ops.h:692
↓ 33 callers
Function
take_along_axis
mlx/ops.cpp:3464
↓ 32 callers
Function
any
mlx/ops.h:568
↓ 32 callers
Function
copy
mlx/backend/cpu/copy.cpp:179
↓ 32 callers
Function
dtype_to_cuda_type
mlx/backend/cuda/utils.cpp:28
↓ 32 callers
Method
export_function
mlx/export.cpp:758
↓ 32 callers
Function
negative
mlx/ops.cpp:2779
↓ 31 callers
Function
bits
mlx/random.h:50
↓ 31 callers
Function
contiguous_copy_gpu
mlx/backend/gpu/copy.cpp:37
↓ 31 callers
Method
erase
mlx/compile.cpp:370
↓ 31 callers
Function
fn
(x)
examples/python/qqmm.py:96
↓ 31 callers
Function
normal
mlx/random.h:104
↓ 31 callers
Method
sort
mlx/backend/metal/kernels/sort.h:67
↓ 31 callers
Function
swapaxes
mlx/ops.cpp:1533
↓ 30 callers
Method
assertCmpNumpy
( self, args: List[Union[Tuple[int], Any]], mx_fn: Callable[..., mx.array], np
python/tests/mlx_tests.py:79
↓ 30 callers
Method
maximum
python/src/array.cpp:59
↓ 30 callers
Function
select
mlx/backend/cpu/simd/neon_fp16_simd.h:176
↓ 30 callers
Function
slice_update
Update a slice from the source array */
mlx/ops.cpp:822
↓ 30 callers
Function
sync_if_needed
(x)
benchmarks/python/comparative/bench_torch.py:47
↓ 30 callers
Method
tensor
mlx/backend/cuda/cudnn_utils.h:111
↓ 29 callers
Method
add_temporary
mlx/backend/metal/device.cpp:334
↓ 29 callers
Function
default_stream
mlx/stream.cpp:40
↓ 29 callers
Function
power
mlx/ops.cpp:3863
↓ 29 callers
Function
serialize
mlx/export.cpp:114
↓ 28 callers
Function
complex_mul_conj
Complex mul followed by conjugate
mlx/backend/metal/kernels/fft/radix.h:24
↓ 28 callers
Function
log
mlx/ops.cpp:3097
↓ 27 callers
Method
erase
mlx/backend/cuda/lru_cache.h:108
↓ 27 callers
Function
new_scoped_memory_pool
mlx/backend/metal/device.cpp:824
↓ 27 callers
Function
scatter
mlx/ops.h:1208
↓ 26 callers
Function
eval
mlx/backend/cpu/eval.cpp:10
↓ 26 callers
Function
isnan
mlx/ops.cpp:1839
↓ 26 callers
Function
less
mlx/ops.cpp:1799
↓ 26 callers
Function
maximum
mlx/ops.cpp:2944
↓ 26 callers
Function
to_arrays
python/src/utils.cpp:50
↓ 25 callers
Function
complex_mul
mlx/backend/metal/kernels/fft/radix.h:19
↓ 25 callers
Method
size
mlx/distributed/jaccl/lib/jaccl/ring.h:33
↓ 25 callers
Function
tree_map
Applies ``fn`` to the leaves of the Python tree ``tree`` and returns a new collection with the results. If ``rest`` is provided, every item i
python/mlx/utils.py:8
↓ 24 callers
Function
copy
mlx/ops.cpp:287
↓ 24 callers
Function
dispatch_bool
mlx/backend/metal/kernels/steel/utils/integral_constant.h:94
↓ 24 callers
Function
elem_to_loc
mlx/backend/metal/kernels/utils.h:98
↓ 24 callers
Method
load_unsafe
Load from device memory into threadgroup memory - without bound checking */
mlx/backend/metal/kernels/steel/gemm/loader.h:74
↓ 24 callers
Function
unary_fp
mlx/backend/cpu/unary.h:141
↓ 23 callers
Function
cos
mlx/ops.cpp:3007
↓ 23 callers
Function
is_available
mlx/device.cpp:28
↓ 23 callers
Method
size
mlx/distributed/ring/ring.cpp:473
↓ 23 callers
Method
store_result
Store results from simdgroup_matrix results into device memory */
mlx/backend/metal/kernels/steel/gemm/mma.h:535
↓ 23 callers
Function
unflatten
mlx/ops.cpp:404
↓ 22 callers
Method
apply_gradients
Apply the gradients to the parameters and return the updated parameters. Can be used to update a model via ``model.update(opt.apply_g
python/mlx/optimizers/optimizers.py:85
↓ 22 callers
Function
cumsum
mlx/ops.cpp:3873
↓ 22 callers
Method
get_peak_memory
mlx/backend/cuda/allocator.cpp:351
↓ 22 callers
Function
get_reduce_axes
python/src/utils.h:31
↓ 21 callers
Method
add_temporaries
mlx/backend/metal/device.cpp:338
↓ 21 callers
Function
concatenate
mlx/ops.cpp:1198
↓ 21 callers
Function
elem_to_loc_broadcast
mlx/backend/metal/kernels/steel/utils.h:7
↓ 21 callers
Method
elems
mlx/backend/metal/kernels/steel/gemm/mma.h:263
↓ 21 callers
Method
emplace
mlx/export.cpp:638
↓ 21 callers
Function
fft
Compute the one-dimensional Fourier Transform. */
mlx/fft.h:52
↓ 21 callers
Function
fftn
mlx/fft.cpp:154
↓ 21 callers
Function
layer_norm
(x, weight, bias, eps)
python/tests/test_fast.py:57
↓ 21 callers
Method
minimum
python/src/array.cpp:63
↓ 21 callers
Function
sin
mlx/ops.cpp:3001
↓ 21 callers
Function
softmax
Softmax of an array. */
mlx/ops.h:1304
↓ 21 callers
Method
store_result_safe
mlx/backend/metal/kernels/steel/gemm/mma.h:568
↓ 21 callers
Function
tree_flatten
Flattens a Python tree to a list of key, value tuples. The keys are using the dot notation to define trees of arbitrary depth and complexity.
python/mlx/utils.py:118
← previous
next →
101–200 of 4,946, ranked by callers