Skip to main content

Github CI/CD

Toy Neural Network Generator

Installation

$ pip install tnng

Simple Model Generator

#!/usr/bin/env python
import torch
import torch.nn as nn
import torchex.nn as exnn
from tnng import Generator, MultiHeadLinkedListLayer

m = MultiHeadLinkedListLayer()
# all layers can be lazy evaluation.
m.append([exnn.Linear(64), exnn.Linear(128), exnn.Linear(256)])
m.append([nn.ReLU(), nn.ELU()])
m.append([exnn.Linear(16), exnn.Linear(32), exnn.Linear(64),])
m.append([nn.ReLU(), nn.ELU()])
m.append([exnn.Linear(10)])

g = Generator(m)

x = torch.randn(128, 256)

class Model(nn.Module):
    def __init__(self, idx=0):
        super(Model, self).__init__()
        self.model = nn.ModuleList([l[0] for l in g[idx]])

    def forward(self, x):
        for m in self.model:
            x = m(x)
        return x

m = Model(0)
o = m(x)

'''
ModuleList(
  (0): Linear(in_features=256, out_features=64, bias=True)
  (1): ReLU()
  (2): Linear(in_features=64, out_features=16, bias=True)
  (3): ReLU()
  (4): Linear(in_features=16, out_features=10, bias=True)
)
'''

Multimodal Model Generator

#!/usr/bin/env python
import torch
import torch.nn as nn
import torchex.nn as exnn
from tnng import Generator, MultiHeadLinkedListLayer

m = MultiHeadLinkedListLayer()
m1 = MultiHeadLinkedListLayer()
# all layers can be lazy evaluation.
m.append([exnn.Linear(64), exnn.Linear(128), exnn.Linear(256)])
m.append([nn.ReLU(), nn.ELU()])
m.append([exnn.Linear(16), exnn.Linear(32), exnn.Linear(64),])
m.append([nn.ReLU(), nn.ELU()])

m1.append([exnn.Conv2d(16, 1), exnn.Conv2d(32, 1), exnn.Conv2d(64, 1)])
m1.append([nn.MaxPool2d(2), nn.AvgPool2d(2)])
m1.append([nn.ReLU(), nn.ELU(), nn.Identity()])
m1.append([exnn.Conv2d(32, 1), exnn.Conv2d(64, 1), exnn.Conv2d(128, 1)])
m1.append([nn.MaxPool2d(2), nn.AvgPool2d(2)])
m1.append([exnn.Flatten(),])

m = m + m1
m.append([exnn.Linear(128)])
m.append([nn.ReLU(), nn.ELU(), nn.Identity()])
m.append([exnn.Linear(10)])


g = Generator(m)
class Model(nn.Module):
    def __init__(self, idx=0):
        super(Model, self).__init__()
        self.model = g[idx]
        for layers in self.model:
            for layer in layers:
                self.add_module(f'{layer}', layer)

    def forward(self, x, img):
        for m in self.model:
            if len(m) == 2:
                if m[0] is not None:
                    x = m[0](x)
                img = m[1](img)
            elif len(m) == 1 and m[0] is None:
                x = torch.cat((x, img), 1)
            else:
                x = m[0](x)
        return x

x = torch.randn(128, 256)
img = torch.randn(128, 3, 28, 28)
m = Model()
o = m(x, img)
print(o.shape)

Metadata

Release files for tnng 0.3.1

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for tnng 0.3.1
File Size Uploaded Provenance Actions
tnng-0.3.1.tar.gz 4.8 kB Without provenance Details for tnng-0.3.1.tar.gz

Built distribution (wheel)

Table of built distributions (wheels) for tnng 0.3.1
File Interpreter ABI Platform Actions
tnng-0.3.1-py2.py3-none-any.whl Python 3, Python 2 none any Details for tnng-0.3.1-py2.py3-none-any.whl

Total release size: 9.9 kB

Release files / tnng-0.3.1.tar.gz

Download URL tnng-0.3.1.tar.gz
Size 4.8 kB
Tags Source
SHA-256 checksum
How to use checksums
f7380b5df81d8ccf526a08c0e04df06cfb92364aa39f9ac887261df9e1cc3fc8
BLAKE2b-256 checksum
How to use checksums
4a1d21655d67eb98706a9643264834820484b745e3e3d901f04708df8fedcd7e
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/3.1.1 pkginfo/1.5.0.1 requests/2.22.0 setuptools/42.0.2.post20191203 requests-toolbelt/0.9.1 tqdm/4.41.1 CPython/3.7.5

Release files / tnng-0.3.1-py2.py3-none-any.whl

Download URL tnng-0.3.1-py2.py3-none-any.whl
Size 5.1 kB
Tags Python 2 Python 3
SHA-256 checksum
How to use checksums
d43980d7c61c0c534bcc973e1c3042200cfefcce03537ddfb0ba555d735abd1f
BLAKE2b-256 checksum
How to use checksums
0f6068e601918cab51e8b0daba4e29f515c6702746cad7f4a8e238562a2178a9
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/3.1.1 pkginfo/1.5.0.1 requests/2.22.0 setuptools/42.0.2.post20191203 requests-toolbelt/0.9.1 tqdm/4.41.1 CPython/3.7.5

Security for tnng 0.3.1

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

No provenance found

None of the files in this release were uploaded with provenance metadata.

A note on package security

PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.

To learn more about protecting yourself from malicious actors, visit our package security documentation.

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page