zhangyiss/mlx
1
0
Fork 0
mirror of https://github.com/ml-explore/mlx.git synced 2025-10-31 16:21:27 +08:00
Code Issues Actions 1 Packages Projects Releases Wiki Activity
152 Commits 42 Branches 73 Tags
d7ac050f4b0af23d386e214371c4c7eda2c658f6
Commit Graph

98 Commits

Author SHA1 Message Date
Gabrijel Boduljak
c7edafb729 implemented InstanceNorm (#244) 
* implemented instancenorm

* implemented vector_norm in cpp

added linalg to mlx

* implemented vector_norm python binding

* renamed vector_norm to norm, implemented norm without provided ord

* completed the implementation of the norm

* added tests

* removed unused import in linalg.cpp

* updated python bindings

* added some tests for python bindings

* handling inf, -inf as numpy does, more extensive tests of compatibility with numpy

* added better docs and examples

* refactored mlx.linalg.norm bindings

* reused existing util for implementation of linalg.norm

* more tests

* fixed a bug with no ord and axis provided

* removed unused imports

* some style and API consistency updates to linalg norm

* remove unused includes

* fix python tests

* fixed a bug with frobenius norm of a complex-valued matrix

* complex for vector too

* addressed PR review comments

* fixed import order in __init__

* expected values in instancenorm tests are simple lists

* minor return expression style change

* added InstanceNorm to docs

* doc string nits

* added myself to individual contributors

---------

Co-authored-by: Awni Hannun <awni@apple.com>
 2024-01-03 12:21:15 -08:00
Awni Hannun
dff4a3833f Module checks the weight on load_weights (#337) 
* update module to check weights on load, also fix docs and reorganize tests

* nits + rebase

* a few more docs updates for Module

* use manual module file

* comment
 2024-01-02 18:55:42 -08:00
Diogo
0782a4573a Add Tensordot op (#344) 2024-01-02 17:15:00 -08:00
Angelos Katharopoulos
436bec9fd9 Fix the implementation of the Bilinear layer (#347) 2024-01-02 16:46:18 -08:00
Asaf Zorea
295ce9db09 Feature expand nn linear (#315) 
* Added an identity and bilinear layers
Added a reset_parameters option
Added normal init for bias

* pre-commit run

* add type hints for parameters and the return type
change Bilinear math to x_1 and x_2
change __call__ arguments to x and y instead of input and output
add explanation to the Initialization

* Remove unnecessary reshape

* Added 'i' to bilinear formula

* Changed bilinear computation to two matrix multiplications

* avoid saving intermediate results, kept y in bilinear for better clarity (can be replaced with x1)

* Changed math formula in Linear
Added more explanation to math formulas
Changed x1, x2 reshape to support all inputs sizes
 2024-01-02 06:08:53 -08:00
Josh Soref
44c1ce5e6a Spelling (#342) 
* spelling: accumulates

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: across

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: additional

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: against

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: among

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: array

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: at least

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: available

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: axes

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: basically

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: bfloat

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: bounds

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: broadcast

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: buffer

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: class

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: coefficients

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: collision

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: combinations

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: committing

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: computation

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: consider

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: constructing

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: conversions

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: correctly

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: corresponding

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: declaration

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: default

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: dependency

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: destination

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: destructor

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: dimensions

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: divided

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: element-wise

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: elements

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: endianness

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: equivalent

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: explicitly

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: github

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: indices

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: irregularly

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: memory

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: metallib

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: negative

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: notable

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: optional

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: otherwise

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: overridden

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: partially

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: partition

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: perform

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: perturbations

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: positively

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: primitive

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: repeat

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: repeats

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: respect

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: respectively

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: result

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: rounding

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: separate

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: skipping

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: structure

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: the

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: transpose

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: unnecessary

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: unneeded

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

* spelling: unsupported

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>

---------

Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>
 2024-01-01 21:08:17 -08:00
mutexuan
350095ce6e fix type cast error in item() for bfloat16 (#339) 
Co-authored-by: xuan <xuan@apple.com>
 2024-01-01 19:02:04 -08:00
Nripesh Niketan
e09bf35b28 feat: Add Dropout3d layer to nn.layers (#313) 
* feat: Add Dropout3d layer to nn.layers

* acknowledgement

* Add dropout tests to test_nn.py

* run pre-commit

* Add activation functions and dropout3d ops

* Add dropout tests for bfloat16 and float16
 2023-12-31 14:01:21 -08:00
Hazem Essam
e3b8da2a49 Added implementation for Scaled RoPE. (#261) 
* Added scale for RoPE

* Ran pre-commit

* Added RoPE scaling test

* Added docstring for scale parameter

* Modified docstrings
 2023-12-31 06:06:01 -08:00
Nripesh Niketan
5ad8fb7268 feat: add softsign, softmax, hardswish, logsoftmax activation function (#309) 
* feat: add softsign activation function

* run pre-commit

* Add Softsign activation function

* Add Softsign activation function

* Add documentation for ReLU6, Softplus, and Softsign activations

* Update activation functions in neural network layers

* Add LogSoftmax and Hardswish activations

* run pre-commit

* Update activations.py

* Added acknowledgements

* Fix activation function comments

* Fix activation functions in neural network layers
 2023-12-29 11:49:36 -08:00
Chunyang Wen
2aedf3e791 Minor refactor for tree_map and tree_unflatten (#311) 
* Minor refact for tree_map and tree_unflatten

* Remove the if statement

---------

Co-authored-by: Chunyang Wen <chunyang_wen@apple.com>
 2023-12-28 20:55:10 -08:00
Chunyang Wen
473b6b43b4 Use defaultdict (#307) 
Co-authored-by: Chunyang Wen <chunyang_wen@apple.com>
 2023-12-28 14:46:13 -08:00
Angelos Katharopoulos
d29770eeaa Update batchnorm to have the running stats in parameters (#305) 2023-12-28 14:31:10 -08:00
Chunyang Wen
040c3bafab Add missing f str (#306) 
Co-authored-by: Chunyang Wen <chunyang_wen@apple.com>
 2023-12-28 06:09:34 -08:00
Chunyang Wen
05767b026f Add information for dropout probability (#304) 
Co-authored-by: Chunyang Wen <chunyang_wen@apple.com>
 2023-12-27 21:51:30 -08:00
Bahaa
ff2b58e299 Add support for repeat (#278) 
* add repeat function

* fix styling

* optimizing repeat

* fixed minor issues

* not sure why that folder is there xD

* fixed now for sure

* test repeat not repeat test

* Fixed

---------

Co-authored-by: Bahaa Eddin tabbakha <bahaa@Bahaas-MacBook-Pro.local>
 2023-12-27 13:11:38 -08:00
YUN, Junwoo
4417e37ede Transformer fix (#167) 
* add transformer with dropout, fix transformer ffm, layernorm order

* precommit changes

* precommit changes

* add docstring, activation, norm_first

* run precommit

* run precommit

* add doctstring

* precommit

* style nits in docs

---------

Co-authored-by: junwoo-yun <junwoo.yun@bagelcode.com>
Co-authored-by: Awni Hannun <awni@apple.com>
 2023-12-27 08:48:36 -08:00
Angelos Katharopoulos
79c95b6919 Fix load compilation (#298) 2023-12-27 06:20:45 -08:00
Diogo
1f6ab6a556 Safetensor support (#215) 
Co-authored-by: Awni Hannun <awni@apple.com>
 2023-12-27 02:06:55 -08:00
Gabrijel Boduljak
6b0d30bb85 linalg.norm (#187) 
* implemented vector_norm in cpp

added linalg to mlx

* implemented vector_norm python binding

* renamed vector_norm to norm, implemented norm without provided ord

* completed the implementation of the norm

* added tests

* removed unused import in linalg.cpp

* updated python bindings

* added some tests for python bindings

* handling inf, -inf as numpy does, more extensive tests of compatibility with numpy

* added better docs and examples

* refactored mlx.linalg.norm bindings

* reused existing util for implementation of linalg.norm

* more tests

* fixed a bug with no ord and axis provided

* removed unused imports

* some style and API consistency updates to linalg norm

* remove unused includes

* fix python tests

* fixed a bug with frobenius norm of a complex-valued matrix

* complex for vector too

---------

Co-authored-by: Awni Hannun <awni@apple.com>
 2023-12-26 19:42:04 -08:00
Angelos Katharopoulos
447bc089b9 Fix tolerance in de-/quantization test (#295) 2023-12-26 19:21:05 -08:00
Daniel Strobusch
d58ac083f3 expose itemsize and nbytes as for numpy arrays (#284) 
see:
  * https://numpy.org/doc/stable/reference/generated/numpy.ndarray.nbytes.html
  * https://numpy.org/doc/stable/reference/generated/numpy.ndarray.itemsize.html

relates to https://github.com/ml-explore/mlx-examples/pull/174
 2023-12-25 10:34:28 -08:00
__mo_san__
a123c3c7d2 implement-batch-norm-layer (#217) 
- Add batch normalization layer

---------

Co-authored-by: Robert McCraith <mccraithrobert@gmail.com>
Co-authored-by: Awni Hannun <awni@apple.com>
 2023-12-25 07:32:53 -08:00
Zach Schillaci
22fee5a383 Remove redundant assert in losses.py (#281) 2023-12-24 08:39:08 -08:00
Daniel Strobusch
7365d142a3 random.uniform must respect dtype, even if lower precision than "low" (#280) 
Fix an edge case where random uniform returns a float32 array, even if a lower precision dtype is wanted due to adding the float32 "low" array.
 2023-12-24 07:04:43 -08:00
Vidit Agarwal
8c3da54c7d Fix failing test for log cosh loss (#275) 
* fix assert statement in log_cosh_loss

* reformatted by pre-commit black
 2023-12-23 16:26:46 -08:00
Vidit Agarwal
acf1721b98 Corrected the example of value_and_grad (#274) 
* Corrected the example for mx.value_and_grad

* Reformat through pre-commit/black
 2023-12-23 11:06:38 -08:00
Finn Voorhees
f91f450141 Fix argmax returns documentation (#263) 2023-12-22 20:33:17 -08:00
Nicholas Santavas
d35fa1db41 Add Hinge, Huber and LogCosh losses (#199) 2023-12-22 10:28:10 -08:00
Justin Deschenaux
e8deca84e0 Add dropout2d (#250) 2023-12-22 08:02:29 -08:00
Angelos Katharopoulos
1d053e0d1d Fix the alibi test that was left unchanged (#252) 2023-12-21 14:59:25 -08:00
Hazem Essam
0aa65c7a6b Added ALiBi implementation (#232) 2023-12-21 14:36:38 -08:00
Angelos Katharopoulos
2c7df6795e Make sure that arrays are freed when saving (#247) 2023-12-21 14:08:24 -08:00
Angelos Katharopoulos
b3916cbf2b Improve names of quantization arguments (#235) 
* Change the default quantization group_size to 64
* Rename groups to group_size and width to bits
 2023-12-20 16:53:53 -08:00
Angelos Katharopoulos
57fe918cf8 Adds C++ and nn quantization utilities (#230) 
* Add C++ de-/quantize ops
* Add quantize functions to the docs and tests
* Add a QuantizedLinear module
 2023-12-20 14:17:38 -08:00
Justin Deschenaux
4912ff3ec2 Add Lion optimizer (#209) 
* Add Lion optimizer
* Update acknowledgements also with past contributions
 2023-12-20 13:54:58 -08:00
Awni Hannun
f40d17047d Indexing bug (#233) 
* fix

* test
 2023-12-20 10:44:01 -08:00
Angelos Katharopoulos
2807c6aff0 Implements divide for integer types and adds floor_divide op (#228) 
* Add floor_divide
* Add floor_divide to the tests
* Add floor_divide to the docs
 2023-12-19 20:12:19 -08:00
Diogo
137f55bf28 fail early if readinto does not exist (#221) 2023-12-19 13:27:17 -08:00
Emircan Erol
e549f84532 Triplet Loss (#211) 
* Triplet Loss

* Requested Changes

* Margin to alpha
 2023-12-19 12:37:12 -08:00
Angelos Katharopoulos
dfa9f4bc58 An initial quantized matmul implementation (#205) 
* Add quantized matvec
* Add quantized matrix matrix with 2nd matrix transposed
* Add quantized matmul tests
* Add a slow cpu quantized matmul
* Add a slightly faster vectorized cpu version
 2023-12-18 23:18:57 -08:00
Abe Leininger
e6872a4149 Added linspace (#181) 
* linspace ops support

---------

Co-authored-by: Awni Hannun <awni@apple.com>
 2023-12-18 19:57:55 -08:00
Juarez Bochi
f4f6e17d45 Fix cross-attention (#210) 
* Fix cross-attention

With the current code, ln2 is a no-op. Its output should be passed to the cross-attention layer

* Add name to contributors
 2023-12-18 12:27:27 -08:00
Angelos Katharopoulos
4d4af12c6f Adds round op and primitive (#203) 2023-12-18 11:32:48 -08:00
jojopuppet
18cca64c81 Add smoothed L1 loss and enhancements to cross entropy loss (#166) 
* Add smooth_l1_loss
* Add labels moothing for cross entropy loss

---------

Co-authored-by: Awni Hannun <awni@apple.com>
 2023-12-18 07:26:21 -08:00
Cyril Zakka, MD
8eb56beb3a Added clip function (#159) 
* Added clip

* Added Python bindings

* Formatting

* Added cpp tests

* Added Python tests

* python bindings work

* rebase

---------

Co-authored-by: Awni Hannun <awni@apple.com>
 2023-12-17 20:00:29 -08:00
Awni Hannun
ee0c2835c5 Docs updates (#198) 
Reorganize NN docs + a few other tidbits.
 2023-12-17 13:20:55 -08:00
Awni Hannun
90d04072b7 fix build w/ flatten (#195) 2023-12-17 11:58:45 -08:00
__mo_san__
52e1589a52 implemented Flatten Module (#149) 
* implemented flatten op

---------

Co-authored-by: Awni Hannun <awni@apple.com>
 2023-12-16 21:54:37 -08:00
YUN, Junwoo
eebd7c275d Add optimizers (AdaMax, AdaDelta, RMSprop) and ordering optimizer classes (#142) 
* Add AdaMax, AdaDelta, RMSprop
 2023-12-16 21:43:15 -08:00