| Name | larq JSON |
| Version |
0.13.3
JSON |
| download |
| home_page | https://larq.dev/ |
| Summary | An Open Source Machine Learning Library for Training Binarized Neural Networks |
| upload_time | 2023-07-12 14:16:09 |
| maintainer | |
| docs_url | None |
| author | Plumerai |
| requires_python | >=3.7 |
| license | Apache 2.0 |
| keywords |
|
| VCS |
|
| bugtrack_url |
|
| requirements |
No requirements were recorded.
|
| Travis-CI |
No Travis.
|
| coveralls test coverage |
No coveralls.
|
<img src="https://user-images.githubusercontent.com/13285808/66865479-39c3b600-ef8f-11e9-9bd4-d47b8e432140.gif" alt="logo" height="100px" align="left"/>
<br/>
[](https://github.com/larq/larq/actions?workflow=Unittest) [](https://codecov.io/github/larq/larq?branch=main) [](https://pypi.org/project/larq/) [](https://pypi.org/project/larq/) [](https://github.com/larq/larq/blob/main/LICENSE) [](https://doi.org/10.21105/joss.01746) [](https://github.com/ambv/black)
Larq is an open-source deep learning library for training neural networks with extremely low precision weights and activations, such as Binarized Neural Networks (BNNs).
Existing deep neural networks use 32 bits, 16 bits or 8 bits to encode each weight and activation, making them large, slow and power-hungry.
This prohibits many applications in resource-constrained environments. Larq is the first step towards solving this. It is designed to provide an easy to use, composable way to train BNNs (1 bit) and other types of Quantized Neural Networks (QNNs) and is based on the `tf.keras` interface. Note that efficient inference using a trained BNN requires the use of an optimized inference engine; we provide these for several platforms in [Larq Compute Engine](https://github.com/larq/compute-engine).
_Larq is part of a family of libraries for BNN development; you can also check out [Larq Zoo](https://github.com/larq/zoo) for pretrained models and [Larq Compute Engine](https://github.com/larq/compute-engine) for deployment on mobile and edge devices._
## Getting Started
To build a QNN, Larq introduces the concept of [quantized layers](https://docs.larq.dev/larq/api/layers/) and [quantizers](https://docs.larq.dev/larq/api/quantizers/). A quantizer defines the way of transforming a full precision input to a quantized output and the pseudo-gradient method used for the backwards pass. Each quantized layer requires an `input_quantizer` and a `kernel_quantizer` that describe the way of quantizing the incoming activations and weights of the layer respectively. If both `input_quantizer` and `kernel_quantizer` are `None` the layer is equivalent to a full precision layer.
You can define a simple binarized fully-connected Keras model using the [Straight-Through Estimator](https://docs.larq.dev/larq/api/quantizers/#ste_sign) the following way:
```python
model = tf.keras.models.Sequential(
[
tf.keras.layers.Flatten(),
larq.layers.QuantDense(
512, kernel_quantizer="ste_sign", kernel_constraint="weight_clip"
),
larq.layers.QuantDense(
10,
input_quantizer="ste_sign",
kernel_quantizer="ste_sign",
kernel_constraint="weight_clip",
activation="softmax",
),
]
)
```
This layer can be used inside a [Keras model](https://www.tensorflow.org/guide/keras/overview#sequential_model) or with a [custom training loop](https://www.tensorflow.org/guide/keras/train_and_evaluate#part_ii_writing_your_own_training_evaluation_loops_from_scratch).
## Examples
Check out our examples on how to train a Binarized Neural Network in just a few lines of code:
- [Introduction to BNNs with Larq](https://docs.larq.dev/larq/tutorials/mnist/)
- [BinaryNet on CIFAR10](https://docs.larq.dev/larq/tutorials/binarynet_cifar10/)
## Installation
Before installing Larq, please install:
- [Python](https://www.python.org/) version `3.7`, `3.8`, `3.9`, or `3.10`
- [Tensorflow](https://www.tensorflow.org/install) version `1.14`, `1.15`, `2.0`, `2.1`, `2.2`, `2.3`, `2.4`, `2.5`, `2.6`, `2.7`, `2.8`, `2.9`, or `2.10`:
```shell
pip install tensorflow # or tensorflow-gpu
```
You can install Larq with Python's [pip](https://pip.pypa.io/en/stable/) package manager:
```shell
pip install larq
```
## About
Larq is being developed by a team of deep learning researchers and engineers at Plumerai to help accelerate both our own research and the general adoption of Binarized Neural Networks.
Raw data
{
"_id": null,
"home_page": "https://larq.dev/",
"name": "larq",
"maintainer": "",
"docs_url": null,
"requires_python": ">=3.7",
"maintainer_email": "",
"keywords": "",
"author": "Plumerai",
"author_email": "opensource@plumerai.com",
"download_url": "https://files.pythonhosted.org/packages/82/2f/ca60a6d6c36bb1da390c4b2cd89bd9f05e6bd85a8fed0634d100d6220f99/larq-0.13.3.tar.gz",
"platform": null,
"description": "<img src=\"https://user-images.githubusercontent.com/13285808/66865479-39c3b600-ef8f-11e9-9bd4-d47b8e432140.gif\" alt=\"logo\" height=\"100px\" align=\"left\"/>\n<br/>\n\n[](https://github.com/larq/larq/actions?workflow=Unittest) [](https://codecov.io/github/larq/larq?branch=main) [](https://pypi.org/project/larq/) [](https://pypi.org/project/larq/) [](https://github.com/larq/larq/blob/main/LICENSE) [](https://doi.org/10.21105/joss.01746) [](https://github.com/ambv/black)\n\nLarq is an open-source deep learning library for training neural networks with extremely low precision weights and activations, such as Binarized Neural Networks (BNNs).\n\nExisting deep neural networks use 32 bits, 16 bits or 8 bits to encode each weight and activation, making them large, slow and power-hungry.\nThis prohibits many applications in resource-constrained environments. Larq is the first step towards solving this. It is designed to provide an easy to use, composable way to train BNNs (1 bit) and other types of Quantized Neural Networks (QNNs) and is based on the `tf.keras` interface. Note that efficient inference using a trained BNN requires the use of an optimized inference engine; we provide these for several platforms in [Larq Compute Engine](https://github.com/larq/compute-engine).\n\n_Larq is part of a family of libraries for BNN development; you can also check out [Larq Zoo](https://github.com/larq/zoo) for pretrained models and [Larq Compute Engine](https://github.com/larq/compute-engine) for deployment on mobile and edge devices._\n\n## Getting Started\n\nTo build a QNN, Larq introduces the concept of [quantized layers](https://docs.larq.dev/larq/api/layers/) and [quantizers](https://docs.larq.dev/larq/api/quantizers/). A quantizer defines the way of transforming a full precision input to a quantized output and the pseudo-gradient method used for the backwards pass. Each quantized layer requires an `input_quantizer` and a `kernel_quantizer` that describe the way of quantizing the incoming activations and weights of the layer respectively. If both `input_quantizer` and `kernel_quantizer` are `None` the layer is equivalent to a full precision layer.\n\nYou can define a simple binarized fully-connected Keras model using the [Straight-Through Estimator](https://docs.larq.dev/larq/api/quantizers/#ste_sign) the following way:\n\n```python\nmodel = tf.keras.models.Sequential(\n [\n tf.keras.layers.Flatten(),\n larq.layers.QuantDense(\n 512, kernel_quantizer=\"ste_sign\", kernel_constraint=\"weight_clip\"\n ),\n larq.layers.QuantDense(\n 10,\n input_quantizer=\"ste_sign\",\n kernel_quantizer=\"ste_sign\",\n kernel_constraint=\"weight_clip\",\n activation=\"softmax\",\n ),\n ]\n)\n```\n\nThis layer can be used inside a [Keras model](https://www.tensorflow.org/guide/keras/overview#sequential_model) or with a [custom training loop](https://www.tensorflow.org/guide/keras/train_and_evaluate#part_ii_writing_your_own_training_evaluation_loops_from_scratch).\n\n## Examples\n\nCheck out our examples on how to train a Binarized Neural Network in just a few lines of code:\n\n- [Introduction to BNNs with Larq](https://docs.larq.dev/larq/tutorials/mnist/)\n- [BinaryNet on CIFAR10](https://docs.larq.dev/larq/tutorials/binarynet_cifar10/)\n\n## Installation\n\nBefore installing Larq, please install:\n\n- [Python](https://www.python.org/) version `3.7`, `3.8`, `3.9`, or `3.10`\n- [Tensorflow](https://www.tensorflow.org/install) version `1.14`, `1.15`, `2.0`, `2.1`, `2.2`, `2.3`, `2.4`, `2.5`, `2.6`, `2.7`, `2.8`, `2.9`, or `2.10`:\n ```shell\n pip install tensorflow # or tensorflow-gpu\n ```\n\nYou can install Larq with Python's [pip](https://pip.pypa.io/en/stable/) package manager:\n\n```shell\npip install larq\n```\n\n## About\n\nLarq is being developed by a team of deep learning researchers and engineers at Plumerai to help accelerate both our own research and the general adoption of Binarized Neural Networks.\n",
"bugtrack_url": null,
"license": "Apache 2.0",
"summary": "An Open Source Machine Learning Library for Training Binarized Neural Networks",
"version": "0.13.3",
"project_urls": {
"Homepage": "https://larq.dev/"
},
"split_keywords": [],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "eea773f3fcfea3b618f755dcd621d43073a3367428e9d16b7b6e0c34d5307c73",
"md5": "f0ac0e38ca4a4f2d31531b617c99941a",
"sha256": "400b1381a98fe51db04678299cee1296a16afcf6c24c6445b0a5cf74d9850d03"
},
"downloads": -1,
"filename": "larq-0.13.3-py3-none-any.whl",
"has_sig": false,
"md5_digest": "f0ac0e38ca4a4f2d31531b617c99941a",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.7",
"size": 66233,
"upload_time": "2023-07-12T14:16:08",
"upload_time_iso_8601": "2023-07-12T14:16:08.210577Z",
"url": "https://files.pythonhosted.org/packages/ee/a7/73f3fcfea3b618f755dcd621d43073a3367428e9d16b7b6e0c34d5307c73/larq-0.13.3-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "822fca60a6d6c36bb1da390c4b2cd89bd9f05e6bd85a8fed0634d100d6220f99",
"md5": "5dc9548b7039c9bdc428d432122019df",
"sha256": "d0f594c4f75c0efc4e54e39650cf8ed559b1af66a48b4d4fcdca38a05fe44a35"
},
"downloads": -1,
"filename": "larq-0.13.3.tar.gz",
"has_sig": false,
"md5_digest": "5dc9548b7039c9bdc428d432122019df",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.7",
"size": 58370,
"upload_time": "2023-07-12T14:16:09",
"upload_time_iso_8601": "2023-07-12T14:16:09.900752Z",
"url": "https://files.pythonhosted.org/packages/82/2f/ca60a6d6c36bb1da390c4b2cd89bd9f05e6bd85a8fed0634d100d6220f99/larq-0.13.3.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2023-07-12 14:16:09",
"github": false,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"lcname": "larq"
}
