nmn 0.1.0__py3-none-any.whl

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
@@ -0,0 +1,76 @@
1
+ Metadata-Version: 2.4
2
+ Name: nmn
3
+ Version: 0.1.0
4
+ Summary: a neuron that matter
5
+ Project-URL: Homepage, https://github.com/mlnomadpy/nmn
6
+ Project-URL: Bug Tracker, https://github.com/mlnomadpy/my_package/issues
7
+ Author-email: Taha Bouhsine <yat@mlnomads.com>
8
+ License-File: LICENSE
9
+ Classifier: License :: OSI Approved :: GNU Affero General Public License v3
10
+ Classifier: Operating System :: OS Independent
11
+ Classifier: Programming Language :: Python :: 3
12
+ Requires-Python: >=3.8
13
+ Description-Content-Type: text/markdown
14
+
15
+ # nmn
16
+ Not the neurons we want, but the neurons we need
17
+
18
+ ## Overview
19
+
20
+ **nmn** provides neural network layers for multiple frameworks (Flax, NNX, Keras, PyTorch, TensorFlow) that do not require activation functions to learn non-linearity. The main goal is to enable deep learning architectures where the layer itself is inherently non-linear, inspired by the paper:
21
+
22
+ > Deep Learning 2.0: Artificial Neurons that Matter: Reject Correlation - Embrace Orthogonality
23
+
24
+ ## Supported Frameworks & Tasks
25
+
26
+ ### Flax (JAX)
27
+ - `YatNMN` layer implemented in `src/nmn/linen/nmn.py`
28
+ - **Tasks:**
29
+ - [x] Core layer implementation
30
+ - [ ] Recurrent layer (to be implemented)
31
+
32
+ ### NNX (Flax NNX)
33
+ - `YatNMN` layer implemented in `src/nmn/nnx/nmn.py`
34
+ - **Tasks:**
35
+ - [x] Core layer implementation
36
+ - [ ] Recurrent layer (to be implemented)
37
+
38
+ ### Keras
39
+ - `YatNMN` layer implemented in `src/nmn/keras/nmn.py`
40
+ - **Tasks:**
41
+ - [x] Core layer implementation
42
+ - [ ] Recurrent layer (to be implemented)
43
+
44
+ ### PyTorch
45
+ - `YatNMN` layer implemented in `src/nmn/torch/nmn.py`
46
+ - **Tasks:**
47
+ - [x] Core layer implementation
48
+ - [ ] Recurrent layer (to be implemented)
49
+
50
+ ### TensorFlow
51
+ - `YatNMN` layer implemented in `src/nmn/tf/nmn.py`
52
+ - **Tasks:**
53
+ - [x] Core layer implementation
54
+ - [ ] Recurrent layer (to be implemented)
55
+
56
+ ## Installation
57
+
58
+ ```bash
59
+ pip install nmn
60
+ ```
61
+
62
+ ## Usage Example (Flax)
63
+
64
+ ```python
65
+ from nmn.nnx.nmn import YatNMN
66
+ from nmn.nnx.yatconv import YatConv
67
+ # ... use as a Flax module ...
68
+ ```
69
+
70
+ ## Roadmap
71
+ - [ ] Implement recurrent layers for all frameworks
72
+ - [ ] Add more examples and benchmarks
73
+ - [ ] Improve documentation and API consistency
74
+
75
+ ## License
76
+ GNU Affero General Public License v3
@@ -0,0 +1,11 @@
1
+ nmn/__init__.py,sha256=F_5o-lCggdEdWfR1l1YC_jfR01mJmveugwUndoRx8n8,83
2
+ nmn/keras/nmn.py,sha256=E7V7kyFB09PfMG1Da_TA2FirOiTCeAXYp3JWACV8h_c,5908
3
+ nmn/linen/nmn.py,sha256=j4v6Z793wliE0xEAITde7jXu9Qras9u75NqdOSPSM4Q,3722
4
+ nmn/nnx/nmn.py,sha256=hZDgMnGnSnBSqMbk-z7qUt8QsHEM-2o6CVWacXZfz3E,4870
5
+ nmn/nnx/yatconv.py,sha256=EZx6g-KcuwrPNEVPl8YdQ16ZXkly_m0XvYCIoWVwFc0,11742
6
+ nmn/tf/nmn.py,sha256=A-K65z9_aN62tAy12b0553nXxrzOofK1umGMRGJYjqw,6036
7
+ nmn/torch/nmn.py,sha256=qOFOlH4_pCOQr_4ctGpEbnW3DAGQotijDTKu5aIEXaE,4609
8
+ nmn-0.1.0.dist-info/METADATA,sha256=jn_ZGPLThl5Smnq1_eAwfYPjvPUkiUl5mJqUXtwm840,2189
9
+ nmn-0.1.0.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
10
+ nmn-0.1.0.dist-info/licenses/LICENSE,sha256=kbZSd5WewnN2PSjvAC6DprP7pXx6NUNsnltmU2Mz1yA,34519
11
+ nmn-0.1.0.dist-info/RECORD,,
@@ -0,0 +1,4 @@
1
+ Wheel-Version: 1.0
2
+ Generator: hatchling 1.27.0
3
+ Root-Is-Purelib: true
4
+ Tag: py3-none-any