# Lava DL
__`lava-dl`__ is a library of deep learning tools within Lava that support offline training, online training and inference methods for various Deep Event-Based Networks.
There are two main strategies for training Deep Event-Based Networks: _direct training_ and _ANN to SNN converison_.
Directly training the network utilizes the information of precise timing of events. Direct training is very accurate and results in efficient networks. However, directly training networks take a lot of time and resources.
On the other hand, ANN to SNN conversion is especially suitable for rate coded SNNs where we can leverage fast training of ANNs. These converted SNNs, however, typically require increased latency compared to directly trained SNNs.
Lava-DL provides an improved version of [SLAYER](https://github.com/bamsumit/slayerPytorch) for direct training of deep event based networks and a new ANN-SNN accelerated training approach called [Bootstrap](https://github.com/lava-nc/lava-dl/blob/main/src/lava/lib/dl/bootstrap/README.md) to mitigate high latency issue of conventional ANN-SNN methods for training Deep Event-Based Networks.
The lava-dl training libraries are independent of the core lava library since Lava Processes cannot be trained directly at this point. Instead, lava-dl is first used to train the model which can then be converted to a network of Lava processes using the netx library using platform independent hdf5 network description.
The library presently consists of
1. `lava.lib.dl.slayer` for natively training Deep Event-Based Networks.
2. `lava.lib.dl.bootstrap` for training rate coded SNNs.
3. `lava.lib.dl.netx` for training and deployment of event-based deep neural networks on traditional as well as neuromorphic backends.
Lava-dl also has the following external, fully compatible, plugin.
1. [`lava.lib.dl.decolle`](https://github.com/kclip/lava-decolle) for training Deep SNNs with local learning and surrogate gradients. This extension is an implementation of [DECOLLE](https://github.com/nmi-lab/decolle-public) learning repo to be fully compatible to lava-dl training tools. Refer [here](https://github.com/kclip/lava-decolle) for the detailed description of the extension, examples and tutorials.
> J. Kaiser, H. Mostafa, and E. Neftci, _Synaptic Plasticity Dynamics for Deep Continuous Local Learning (DECOLLE)._ pp 424, Frontiers in Neuroscience 2020.
More tools will be added in the future.
## Lava-DL Workflow
<p align="center">
<img src="https://user-images.githubusercontent.com/29907126/204448416-1c222fa9-822a-4fd7-a2b8-eebcabf7316d.png" alt="Drawing" style="max-height: 400px;"/>
</p>
Typical Lava-DL workflow:
* **Training:** using `lava.lib.dl.{slayer/bootstrap/decolle}` which results in a _hdf5 network description_. Training usually follows an iterative cycle of architecture design, hyperparameter tuning, and backpropagation training.
* **Inference:** using `lava.lib.dl.netx` which generates lava proces from the hdf5 network description of the trained network and enables inference on different backends.
## Installation
### Cloning Lava-DL and Running from Source
**Note:** The instructions to follow will setup a virtual environement and install lava-dl and all dependencies in that virtual environment. Please [`setup git-lfs`](https://docs.github.com/en/repositories/working-with-files/managing-large-files/installing-git-large-file-storage) to ensure large files are pulled during the `git clone`.
#### [Linux/MacOS]
```bash
cd $HOME
git clone git@github.com:lava-nc/lava-dl.git
cd lava-dl
curl -sSL https://install.python-poetry.org | python3 -
poetry config virtualenvs.in-project true
poetry install
source .venv/bin/activate
pytest
```
#### [Windows]
```powershell
# Commands using PowerShell
cd $HOME
git clone git@github.com:lava-nc/lava-dl.git
cd lava-dl
python3 -m venv .venv
.venv\Scripts\activate
curl -sSL https://install.python-poetry.org | python3 -
pip install -U pip
poetry config virtualenvs.in-project true
poetry install
pytest
```
You should expect the following output after running the unit tests:
```
$ pytest
============================= test session starts ==============================
platform linux -- Python 3.9.10, pytest-7.0.1, pluggy-1.0.0
rootdir: /home/user/lava-dl, configfile: pyproject.toml, testpaths: tests
plugins: cov-3.0.0
collected 86 items
tests/lava/lib/dl/netx/test_blocks.py ... [ 3%]
tests/lava/lib/dl/netx/test_hdf5.py ... [ 6%]
tests/lava/lib/dl/slayer/neuron/test_adrf.py ....... [ 15%]
...... pytest output ...
tests/lava/lib/dl/slayer/neuron/dynamics/test_adaptive_threshold.py .... [ 80%]
. [ 81%]
tests/lava/lib/dl/slayer/neuron/dynamics/test_leaky_integrator.py ..... [ 87%]
tests/lava/lib/dl/slayer/neuron/dynamics/test_resonator.py ..... [ 93%]
tests/lava/lib/dl/slayer/utils/filter/test_conv_filter.py .. [ 95%]
tests/lava/lib/dl/slayer/utils/time/test_replicate.py . [ 96%]
tests/lava/lib/dl/slayer/utils/time/test_shift.py ... [100%]
=============================== warnings summary ===============================
...... pytest output ...
src/lava/lib/dl/slayer/utils/time/__init__.py 4 0 100%
src/lava/lib/dl/slayer/utils/time/replicate.py 6 0 100%
src/lava/lib/dl/slayer/utils/time/shift.py 59 16 73% 22-43, 50, 55, 75, 121, 128, 135, 139
src/lava/lib/dl/slayer/utils/utils.py 13 8 38% 14, 35-45
--------------------------------------------------------------------------------------------
TOTAL 4782 2535 47%
Required test coverage of 45.0% reached. Total coverage: 46.99%
======================= 86 passed, 3 warnings in 46.56s ========================
```
**Note:** If you see errors regarding `*.np` files or errors similar to "ValueError: Cannot load file containing pickled data when allow_pickle=False" please ensure git-lfs is installed. If you installed git-lfs after cloning the repository please fetch and pull, `git lfs fetch --all; git lfs pull` and try the tests again.
### [Alternative] Installing Lava via Conda
If you use the Conda package manager, you can simply install the Lava package
via:
```bash
conda install lava-dl -c conda-forge
```
Alternatively with intel numpy and scipy:
```bash
conda create -n lava-dl python=3.9 -c intel
conda activate lava-dl
conda install -n lava-dl -c intel numpy scipy
conda install -n lava-dl -c conda-forge lava-dl --freeze-installed
```
### [Alternative] Installing Lava from Binaries
If you only need the lava-dl package in your python environment, we will publish
Lava releases via
[GitHub Releases](https://github.com/lava-nc/lava-dl/releases). Please download
the package and install it.
Open a python terminal and run:
#### [Windows/MacOS/Linux]
```bash
$ python3 -m venv python3_venv
$ pip install -U pip
$ pip install lava-dl-0.2.0.tar.gz
```
## Getting Started
**End to end training tutorials**
* [Oxford spike train regression](https://github.com/lava-nc/lava-dl/blob/main/tutorials/lava/lib/dl/slayer/oxford/train.ipynb)
* [XOR regression](https://github.com/lava-nc/lava-dl/blob/main/tutorials/lava/lib/dl/slayer/xor_regression/xor_regression.ipynb)
* [MNIST digit classification](https://github.com/lava-nc/lava-dl/blob/main/tutorials/lava/lib/dl/bootstrap/mnist/train.ipynb)
* [NMNIST digit classification](https://github.com/lava-nc/lava-dl/blob/main/tutorials/lava/lib/dl/slayer/nmnist/train.ipynb)
* [PilotNet steering angle prediction](https://github.com/lava-nc/lava-dl/blob/main/tutorials/lava/lib/dl/slayer/pilotnet/train.ipynb)
* [Video object detection with sigma-detla YOLO](https://github.com/lava-nc/lava-dl/tree/main/tutorials/lava/lib/dl/slayer/tiny_yolo_sdnn)
* [Inference on GPU](https://github.com/lava-nc/lava-dl/blob/main/tutorials/lava/lib/dl/slayer/tiny_yolo_sdnn/inference.ipynb)
**Deep dive training tutorials**
* [Dynamics and Neurons](https://github.com/lava-nc/lava-dl/blob/main/tutorials/lava/lib/dl/slayer/neuron_dynamics/dynamics.ipynb)
**Inference tutorials**
* [Oxford Inference](https://github.com/lava-nc/lava-dl/blob/main/tutorials/lava/lib/dl/netx/oxford/run.ipynb)
* [PilotNet LIF Inference](https://github.com/lava-nc/lava-dl/blob/main/tutorials/lava/lib/dl/netx/pilotnet_snn/run.ipynb)
* [PilotNet LIF Benchmarking](https://github.com/lava-nc/lava-dl/blob/main/tutorials/lava/lib/dl/netx/pilotnet_snn/benchmark.ipynb)
* [PilotNet SDNN Inference](https://github.com/lava-nc/lava-dl/blob/main/tutorials/lava/lib/dl/netx/pilotnet_sdnn/run.ipynb)
* [PilotNet SDNN Benchmarking](https://github.com/lava-nc/lava-dl/blob/main/tutorials/lava/lib/dl/netx/pilotnet_sdnn/benchmark.ipynb)
* [YOLO-KP SDNN Inference](https://github.com/lava-nc/lava-dl/blob/main/tutorials/lava/lib/dl/netx/yolo_kp/run.ipynb)
## __`lava.lib.dl.slayer`__
`lava.lib.dl.slayer` is an enhanced version of [SLAYER](https://github.com/bamsumit/slayerPytorch). Most noteworthy enhancements are: support for _recurrent network structures_, a wider variety of _neuron models_ and _synaptic connections_ (a complete list of features is [here](https://github.com/lava-nc/lava-dl/blob/main/src/lava/lib/dl/slayer/README.md)). This version of SLAYER is built on top of the [PyTorch](https://pytorch.org/) deep learning framework, similar to its predecessor. For smooth integration with Lava, `lava.lib.dl.slayer` supports exporting trained models using the platform independent __hdf5 network exchange__ format.
In future versions, SLAYER will get completely integrated into Lava to train Lava Processes directly. This will eliminate the need for explicitly exporting and importing the trained networks.
### Example Code
__Import modules__
```python
import lava.lib.dl.slayer as slayer
```
__Network Description__
```python
# like any standard pyTorch network
class Network(torch.nn.Module):
def __init__(self):
...
self.blocks = torch.nn.ModuleList([# sequential network blocks
slayer.block.sigma_delta.Input(sdnn_params),
slayer.block.sigma_delta.Conv(sdnn_params, 3, 24, 3),
slayer.block.sigma_delta.Conv(sdnn_params, 24, 36, 3),
slayer.block.rf_iz.Conv(rf_params, 36, 64, 3, delay=True),
slayer.block.rf_iz.Conv(sdnn_cnn_params, 64, 64, 3, delay=True),
slayer.block.rf_iz.Flatten(),
slayer.block.alif.Dense(alif_params, 64*40, 100, delay=True),
slayer.block.cuba.Recurrent(cuba_params, 100, 50),
slayer.block.cuba.KWTA(cuba_params, 50, 50, num_winners=5)
])
def forward(self, x):
for block in self.blocks:
# forward computation is as simple as calling the blocks in a loop
x = block(x)
return x
def export_hdf5(self, filename):
# network export to hdf5 format
h = h5py.File(filename, 'w')
layer = h.create_group('layer')
for i, b in enumerate(self.blocks):
b.export_hdf5(layer.create_group(f'{i}'))
```
__Training__
```python
net = Network()
assistant = slayer.utils.Assistant(net, error, optimizer, stats)
...
for epoch in range(epochs):
for i, (input, ground_truth) in enumerate(train_loader):
output = assistant.train(input, ground_truth)
...
for i, (input, ground_truth) in enumerate(test_loader):
output = assistant.test(input, ground_truth)
...
```
__Export the network__
```python
net.export_hdf5('network.net')
```
## __`lava.lib.dl.bootstrap`__
In general ANN-SNN conversion methods for rate based SNN result in high latency of the network during inference. This is because the rate interpretation of a spiking neuron using ReLU acitvation unit breaks down for short inference times. As a result, the network requires many time steps per sample to achieve adequate inference results.
`lava.lib.dl.bootstrap` enables rapid training of rate based SNNs by translating them to an equivalent dynamic ANN representation which leads to SNN performance close to the equivalent ANN and low latency inference. More details [here](https://github.com/lava-nc/lava-dl/blob/main/src/lava/lib/dl/bootstrap/README.md). It also supports _hybrid training_ with a mixed ANN-SNN network to minimize the ANN to SNN performance gap. This method is independent of the SNN model being used.
It has similar API as `lava.lib.dl.slayer` and supports exporting trained models using the platform independent __hdf5 network exchange__ format.
### Example Code
__Import modules__
```python
import lava.lib.dl.bootstrap as bootstrap
```
__Network Description__
```python
# like any standard pyTorch network
class Network(torch.nn.Module):
def __init__(self):
...
self.blocks = torch.nn.ModuleList([# sequential network blocks
bootstrap.block.cuba.Input(sdnn_params),
bootstrap.block.cuba.Conv(sdnn_params, 3, 24, 3),
bootstrap.block.cuba.Conv(sdnn_params, 24, 36, 3),
bootstrap.block.cuba.Conv(rf_params, 36, 64, 3),
bootstrap.block.cuba.Conv(sdnn_cnn_params, 64, 64, 3),
bootstrap.block.cuba.Flatten(),
bootstrap.block.cuba.Dense(alif_params, 64*40, 100),
bootstrap.block.cuba.Dense(cuba_params, 100, 10),
])
def forward(self, x, mode):
...
for block, m in zip(self.blocks, mode):
x = block(x, mode=m)
return x
def export_hdf5(self, filename):
# network export to hdf5 format
h = h5py.File(filename, 'w')
layer = h.create_group('layer')
for i, b in enumerate(self.blocks):
b.export_hdf5(layer.create_group(f'{i}'))
```
__Training__
```python
net = Network()
scheduler = bootstrap.routine.Scheduler()
...
for epoch in range(epochs):
for i, (input, ground_truth) in enumerate(train_loader):
mode = scheduler.mode(epoch, i, net.training)
output = net.forward(input, mode)
...
loss.backward()
for i, (input, ground_truth) in enumerate(test_loader):
mode = scheduler.mode(epoch, i, net.training)
output = net.forward(input, mode)
...
```
__Export the network__
```python
net.export_hdf5('network.net')
```
## __`lava.lib.dl.netx`__
For inference using Lava, `lava.lib.dl.netx` provides an automated API for loading SLAYER-trained models as Lava Processes, which can be directly run on a desired backend. `lava.lib.dl.netx` imports models saved via SLAYER using the hdf5 network exchange format. The details of hdf5 network description specification can be found [here](https://github.com/lava-nc/lava-dl/blob/main/src/lava/lib/dl/netx/README.md).
### Example Code
__Import modules__
```python
from lava.lib.dl.netx import hdf5
```
__Load the trained network__
```python
# Import the model as a Lava Process
net = hdf5.Network(net_config='network.net')
```
__Attach Processes for Input-Output interaction__
```python
from lava.proc import io
# Instantiate the processes
dataloader = io.dataloader.SpikeDataloader(dataset=test_set)
output_logger = io.sink.RingBuffer(shape=net.out_layer.shape, buffer=num_steps)
gt_logger = io.sink.RingBuffer(shape=(1,), buffer=num_steps)
# Connect the input to the network:
dataloader.ground_truth.connect(gt_logger.a_in)
dataloader.s_out.connect(net.in_layer.neuron.a_in)
# Connect network-output to the output process
net.out_layer.out.connect(output_logger.a_in)
```
__Run the network__
```python
from lava.magma import run_configs as rcfg
from lava.magma import run_conditions as rcnd
net.run(condition=rcnd.RunSteps(total_run_time), run_cfg=rcfg.Loihi1SimCfg())
output = output_logger.data.get()
gts = gt_logger.data.get()
net.stop()
```
Raw data
{
"_id": null,
"home_page": "https://lava-nc.org/",
"name": "lava-dl",
"maintainer": "",
"docs_url": null,
"requires_python": ">=3.8,<3.11",
"maintainer_email": "",
"keywords": "neuromorphic,ai,artificial intelligence,neural models,spiking neural networks,deep learning",
"author": "Intel's Neuromorphic Computing Lab and the open source community",
"author_email": "lava@intel.com",
"download_url": "",
"platform": null,
"description": "# Lava DL\n\n__`lava-dl`__ is a library of deep learning tools within Lava that support offline training, online training and inference methods for various Deep Event-Based Networks.\n\nThere are two main strategies for training Deep Event-Based Networks: _direct training_ and _ANN to SNN converison_. \n\nDirectly training the network utilizes the information of precise timing of events. Direct training is very accurate and results in efficient networks. However, directly training networks take a lot of time and resources.\n\nOn the other hand, ANN to SNN conversion is especially suitable for rate coded SNNs where we can leverage fast training of ANNs. These converted SNNs, however, typically require increased latency compared to directly trained SNNs.\n\nLava-DL provides an improved version of [SLAYER](https://github.com/bamsumit/slayerPytorch) for direct training of deep event based networks and a new ANN-SNN accelerated training approach called [Bootstrap](https://github.com/lava-nc/lava-dl/blob/main/src/lava/lib/dl/bootstrap/README.md) to mitigate high latency issue of conventional ANN-SNN methods for training Deep Event-Based Networks.\n\nThe lava-dl training libraries are independent of the core lava library since Lava Processes cannot be trained directly at this point. Instead, lava-dl is first used to train the model which can then be converted to a network of Lava processes using the netx library using platform independent hdf5 network description.\n\nThe library presently consists of\n\n1. `lava.lib.dl.slayer` for natively training Deep Event-Based Networks.\n2. `lava.lib.dl.bootstrap` for training rate coded SNNs.\n3. `lava.lib.dl.netx` for training and deployment of event-based deep neural networks on traditional as well as neuromorphic backends.\n\nLava-dl also has the following external, fully compatible, plugin.\n1. [`lava.lib.dl.decolle`](https://github.com/kclip/lava-decolle) for training Deep SNNs with local learning and surrogate gradients. This extension is an implementation of [DECOLLE](https://github.com/nmi-lab/decolle-public) learning repo to be fully compatible to lava-dl training tools. Refer [here](https://github.com/kclip/lava-decolle) for the detailed description of the extension, examples and tutorials. \n > J. Kaiser, H. Mostafa, and E. Neftci, _Synaptic Plasticity Dynamics for Deep Continuous Local Learning (DECOLLE)._ pp 424, Frontiers in Neuroscience 2020.\n\nMore tools will be added in the future.\n\n## Lava-DL Workflow\n\n<p align=\"center\">\n<img src=\"https://user-images.githubusercontent.com/29907126/204448416-1c222fa9-822a-4fd7-a2b8-eebcabf7316d.png\" alt=\"Drawing\" style=\"max-height: 400px;\"/>\n</p>\n\nTypical Lava-DL workflow:\n* **Training:** using `lava.lib.dl.{slayer/bootstrap/decolle}` which results in a _hdf5 network description_. Training usually follows an iterative cycle of architecture design, hyperparameter tuning, and backpropagation training.\n* **Inference:** using `lava.lib.dl.netx` which generates lava proces from the hdf5 network description of the trained network and enables inference on different backends.\n\n## Installation\n\n### Cloning Lava-DL and Running from Source\n\n**Note:** The instructions to follow will setup a virtual environement and install lava-dl and all dependencies in that virtual environment. Please [`setup git-lfs`](https://docs.github.com/en/repositories/working-with-files/managing-large-files/installing-git-large-file-storage) to ensure large files are pulled during the `git clone`.\n\n#### [Linux/MacOS]\n```bash\ncd $HOME\ngit clone git@github.com:lava-nc/lava-dl.git\ncd lava-dl\ncurl -sSL https://install.python-poetry.org | python3 -\npoetry config virtualenvs.in-project true\npoetry install\nsource .venv/bin/activate\npytest\n```\n#### [Windows]\n```powershell\n# Commands using PowerShell\ncd $HOME\ngit clone git@github.com:lava-nc/lava-dl.git\ncd lava-dl\npython3 -m venv .venv\n.venv\\Scripts\\activate\ncurl -sSL https://install.python-poetry.org | python3 -\npip install -U pip\npoetry config virtualenvs.in-project true\npoetry install\npytest\n```\n\nYou should expect the following output after running the unit tests:\n```\n$ pytest\n============================= test session starts ==============================\nplatform linux -- Python 3.9.10, pytest-7.0.1, pluggy-1.0.0\nrootdir: /home/user/lava-dl, configfile: pyproject.toml, testpaths: tests\nplugins: cov-3.0.0\ncollected 86 items\n\ntests/lava/lib/dl/netx/test_blocks.py ... [ 3%]\ntests/lava/lib/dl/netx/test_hdf5.py ... [ 6%]\ntests/lava/lib/dl/slayer/neuron/test_adrf.py ....... [ 15%]\n\n...... pytest output ...\n\ntests/lava/lib/dl/slayer/neuron/dynamics/test_adaptive_threshold.py .... [ 80%]\n. [ 81%]\ntests/lava/lib/dl/slayer/neuron/dynamics/test_leaky_integrator.py ..... [ 87%]\ntests/lava/lib/dl/slayer/neuron/dynamics/test_resonator.py ..... [ 93%]\ntests/lava/lib/dl/slayer/utils/filter/test_conv_filter.py .. [ 95%]\ntests/lava/lib/dl/slayer/utils/time/test_replicate.py . [ 96%]\ntests/lava/lib/dl/slayer/utils/time/test_shift.py ... [100%]\n\n=============================== warnings summary ===============================\n\n...... pytest output ...\n\nsrc/lava/lib/dl/slayer/utils/time/__init__.py 4 0 100%\nsrc/lava/lib/dl/slayer/utils/time/replicate.py 6 0 100%\nsrc/lava/lib/dl/slayer/utils/time/shift.py 59 16 73% 22-43, 50, 55, 75, 121, 128, 135, 139\nsrc/lava/lib/dl/slayer/utils/utils.py 13 8 38% 14, 35-45\n--------------------------------------------------------------------------------------------\nTOTAL 4782 2535 47%\n\nRequired test coverage of 45.0% reached. Total coverage: 46.99%\n======================= 86 passed, 3 warnings in 46.56s ========================\n```\n\n**Note:** If you see errors regarding `*.np` files or errors similar to \"ValueError: Cannot load file containing pickled data when allow_pickle=False\" please ensure git-lfs is installed. If you installed git-lfs after cloning the repository please fetch and pull, `git lfs fetch --all; git lfs pull` and try the tests again.\n\n### [Alternative] Installing Lava via Conda\nIf you use the Conda package manager, you can simply install the Lava package\nvia:\n```bash\nconda install lava-dl -c conda-forge\n```\n\nAlternatively with intel numpy and scipy:\n\n```bash\nconda create -n lava-dl python=3.9 -c intel\nconda activate lava-dl\nconda install -n lava-dl -c intel numpy scipy\nconda install -n lava-dl -c conda-forge lava-dl --freeze-installed\n```\n\n### [Alternative] Installing Lava from Binaries\n\nIf you only need the lava-dl package in your python environment, we will publish\nLava releases via\n[GitHub Releases](https://github.com/lava-nc/lava-dl/releases). Please download\nthe package and install it.\n\nOpen a python terminal and run:\n\n#### [Windows/MacOS/Linux]\n```bash\n$ python3 -m venv python3_venv\n$ pip install -U pip\n$ pip install lava-dl-0.2.0.tar.gz\n```\n\n## Getting Started\n\n**End to end training tutorials**\n* [Oxford spike train regression](https://github.com/lava-nc/lava-dl/blob/main/tutorials/lava/lib/dl/slayer/oxford/train.ipynb)\n* [XOR regression](https://github.com/lava-nc/lava-dl/blob/main/tutorials/lava/lib/dl/slayer/xor_regression/xor_regression.ipynb)\n* [MNIST digit classification](https://github.com/lava-nc/lava-dl/blob/main/tutorials/lava/lib/dl/bootstrap/mnist/train.ipynb)\n* [NMNIST digit classification](https://github.com/lava-nc/lava-dl/blob/main/tutorials/lava/lib/dl/slayer/nmnist/train.ipynb)\n* [PilotNet steering angle prediction](https://github.com/lava-nc/lava-dl/blob/main/tutorials/lava/lib/dl/slayer/pilotnet/train.ipynb)\n* [Video object detection with sigma-detla YOLO](https://github.com/lava-nc/lava-dl/tree/main/tutorials/lava/lib/dl/slayer/tiny_yolo_sdnn)\n * [Inference on GPU](https://github.com/lava-nc/lava-dl/blob/main/tutorials/lava/lib/dl/slayer/tiny_yolo_sdnn/inference.ipynb)\n\n**Deep dive training tutorials**\n* [Dynamics and Neurons](https://github.com/lava-nc/lava-dl/blob/main/tutorials/lava/lib/dl/slayer/neuron_dynamics/dynamics.ipynb)\n\n**Inference tutorials**\n* [Oxford Inference](https://github.com/lava-nc/lava-dl/blob/main/tutorials/lava/lib/dl/netx/oxford/run.ipynb)\n* [PilotNet LIF Inference](https://github.com/lava-nc/lava-dl/blob/main/tutorials/lava/lib/dl/netx/pilotnet_snn/run.ipynb)\n* [PilotNet LIF Benchmarking](https://github.com/lava-nc/lava-dl/blob/main/tutorials/lava/lib/dl/netx/pilotnet_snn/benchmark.ipynb)\n* [PilotNet SDNN Inference](https://github.com/lava-nc/lava-dl/blob/main/tutorials/lava/lib/dl/netx/pilotnet_sdnn/run.ipynb)\n* [PilotNet SDNN Benchmarking](https://github.com/lava-nc/lava-dl/blob/main/tutorials/lava/lib/dl/netx/pilotnet_sdnn/benchmark.ipynb)\n* [YOLO-KP SDNN Inference](https://github.com/lava-nc/lava-dl/blob/main/tutorials/lava/lib/dl/netx/yolo_kp/run.ipynb)\n\n## __`lava.lib.dl.slayer`__ \n\n`lava.lib.dl.slayer` is an enhanced version of [SLAYER](https://github.com/bamsumit/slayerPytorch). Most noteworthy enhancements are: support for _recurrent network structures_, a wider variety of _neuron models_ and _synaptic connections_ (a complete list of features is [here](https://github.com/lava-nc/lava-dl/blob/main/src/lava/lib/dl/slayer/README.md)). This version of SLAYER is built on top of the [PyTorch](https://pytorch.org/) deep learning framework, similar to its predecessor. For smooth integration with Lava, `lava.lib.dl.slayer` supports exporting trained models using the platform independent __hdf5 network exchange__ format. \n\nIn future versions, SLAYER will get completely integrated into Lava to train Lava Processes directly. This will eliminate the need for explicitly exporting and importing the trained networks. \n\n### Example Code\n\n__Import modules__\n```python\nimport lava.lib.dl.slayer as slayer\n```\n__Network Description__\n```python\n# like any standard pyTorch network\nclass Network(torch.nn.Module):\n def __init__(self):\n ...\n self.blocks = torch.nn.ModuleList([# sequential network blocks \n slayer.block.sigma_delta.Input(sdnn_params), \n slayer.block.sigma_delta.Conv(sdnn_params, 3, 24, 3),\n slayer.block.sigma_delta.Conv(sdnn_params, 24, 36, 3),\n slayer.block.rf_iz.Conv(rf_params, 36, 64, 3, delay=True),\n slayer.block.rf_iz.Conv(sdnn_cnn_params, 64, 64, 3, delay=True),\n slayer.block.rf_iz.Flatten(),\n slayer.block.alif.Dense(alif_params, 64*40, 100, delay=True),\n slayer.block.cuba.Recurrent(cuba_params, 100, 50),\n slayer.block.cuba.KWTA(cuba_params, 50, 50, num_winners=5)\n ])\n\n def forward(self, x):\n for block in self.blocks: \n # forward computation is as simple as calling the blocks in a loop\n x = block(x)\n return x\n\n def export_hdf5(self, filename):\n # network export to hdf5 format\n h = h5py.File(filename, 'w')\n layer = h.create_group('layer')\n for i, b in enumerate(self.blocks):\n b.export_hdf5(layer.create_group(f'{i}'))\n```\n__Training__\n```python\nnet = Network()\nassistant = slayer.utils.Assistant(net, error, optimizer, stats)\n...\nfor epoch in range(epochs):\n for i, (input, ground_truth) in enumerate(train_loader):\n output = assistant.train(input, ground_truth)\n ...\n for i, (input, ground_truth) in enumerate(test_loader):\n output = assistant.test(input, ground_truth)\n ...\n```\n__Export the network__\n```python\nnet.export_hdf5('network.net')\n```\n\n## __`lava.lib.dl.bootstrap`__\n\nIn general ANN-SNN conversion methods for rate based SNN result in high latency of the network during inference. This is because the rate interpretation of a spiking neuron using ReLU acitvation unit breaks down for short inference times. As a result, the network requires many time steps per sample to achieve adequate inference results.\n\n`lava.lib.dl.bootstrap` enables rapid training of rate based SNNs by translating them to an equivalent dynamic ANN representation which leads to SNN performance close to the equivalent ANN and low latency inference. More details [here](https://github.com/lava-nc/lava-dl/blob/main/src/lava/lib/dl/bootstrap/README.md). It also supports _hybrid training_ with a mixed ANN-SNN network to minimize the ANN to SNN performance gap. This method is independent of the SNN model being used.\n\nIt has similar API as `lava.lib.dl.slayer` and supports exporting trained models using the platform independent __hdf5 network exchange__ format.\n\n### Example Code\n\n__Import modules__\n```python\nimport lava.lib.dl.bootstrap as bootstrap\n```\n__Network Description__\n```python\n# like any standard pyTorch network\nclass Network(torch.nn.Module):\n def __init__(self):\n ...\n self.blocks = torch.nn.ModuleList([# sequential network blocks \n bootstrap.block.cuba.Input(sdnn_params), \n bootstrap.block.cuba.Conv(sdnn_params, 3, 24, 3),\n bootstrap.block.cuba.Conv(sdnn_params, 24, 36, 3),\n bootstrap.block.cuba.Conv(rf_params, 36, 64, 3),\n bootstrap.block.cuba.Conv(sdnn_cnn_params, 64, 64, 3),\n bootstrap.block.cuba.Flatten(),\n bootstrap.block.cuba.Dense(alif_params, 64*40, 100),\n bootstrap.block.cuba.Dense(cuba_params, 100, 10),\n ])\n\n def forward(self, x, mode):\n ...\n for block, m in zip(self.blocks, mode):\n x = block(x, mode=m)\n\n return x\n\n def export_hdf5(self, filename):\n # network export to hdf5 format\n h = h5py.File(filename, 'w')\n layer = h.create_group('layer')\n for i, b in enumerate(self.blocks):\n b.export_hdf5(layer.create_group(f'{i}'))\n```\n__Training__\n```python\nnet = Network()\nscheduler = bootstrap.routine.Scheduler()\n...\nfor epoch in range(epochs):\n for i, (input, ground_truth) in enumerate(train_loader):\n mode = scheduler.mode(epoch, i, net.training)\n output = net.forward(input, mode)\n ...\n loss.backward()\n for i, (input, ground_truth) in enumerate(test_loader):\n mode = scheduler.mode(epoch, i, net.training)\n output = net.forward(input, mode)\n ...\n```\n__Export the network__\n```python\nnet.export_hdf5('network.net')\n```\n\n## __`lava.lib.dl.netx`__ \n\nFor inference using Lava, `lava.lib.dl.netx` provides an automated API for loading SLAYER-trained models as Lava Processes, which can be directly run on a desired backend. `lava.lib.dl.netx` imports models saved via SLAYER using the hdf5 network exchange format. The details of hdf5 network description specification can be found [here](https://github.com/lava-nc/lava-dl/blob/main/src/lava/lib/dl/netx/README.md).\n\n### Example Code\n\n__Import modules__\n```python\nfrom lava.lib.dl.netx import hdf5\n```\n__Load the trained network__\n```python\n# Import the model as a Lava Process\nnet = hdf5.Network(net_config='network.net')\n```\n__Attach Processes for Input-Output interaction__\n```python\nfrom lava.proc import io\n\n# Instantiate the processes\ndataloader = io.dataloader.SpikeDataloader(dataset=test_set)\noutput_logger = io.sink.RingBuffer(shape=net.out_layer.shape, buffer=num_steps)\ngt_logger = io.sink.RingBuffer(shape=(1,), buffer=num_steps)\n\n# Connect the input to the network:\ndataloader.ground_truth.connect(gt_logger.a_in)\ndataloader.s_out.connect(net.in_layer.neuron.a_in)\n\n# Connect network-output to the output process\nnet.out_layer.out.connect(output_logger.a_in)\n```\n__Run the network__\n```python\nfrom lava.magma import run_configs as rcfg\nfrom lava.magma import run_conditions as rcnd\n\nnet.run(condition=rcnd.RunSteps(total_run_time), run_cfg=rcfg.Loihi1SimCfg())\noutput = output_logger.data.get()\ngts = gt_logger.data.get()\nnet.stop()\n```\n\n\n",
"bugtrack_url": null,
"license": "(BSD-3-Clause)",
"summary": "A library of deep learning tools, which consists of lava.lib.dl.slayer and lava.lib.dl.netx for training and deployment of event-based deep neural networks on traditional as well as neuromorphic backends. Lava-DL is part of Lava Framework",
"version": "0.5.0",
"project_urls": {
"Discussions": "https://github.com/lava-nc/lava-dl/discussions",
"Frequently Asked Questions": "https://github.com/lava-nc/lava-dl/wiki/Frequently-Asked-Questions-(FAQ)",
"Homepage": "https://lava-nc.org/",
"Issue and Bug Tracker": "https://github.com/lava-nc/lava-dl/issues",
"Questions and Answers": "https://github.com/lava-nc/lava-dl/discussions/categories/q-a",
"Repository": "https://github.com/lava-nc/lava-dl"
},
"split_keywords": [
"neuromorphic",
"ai",
"artificial intelligence",
"neural models",
"spiking neural networks",
"deep learning"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "cfba391e48532ba421a5d32cab254d4200cf23b8a9b29d14a2bef16f45c1f1af",
"md5": "1a918dca1368045eb602eb453708a7b9",
"sha256": "fd31875f082892008357ece29f95f0021611433e2295ebd60477417525f97362"
},
"downloads": -1,
"filename": "lava_dl-0.5.0-py3-none-any.whl",
"has_sig": false,
"md5_digest": "1a918dca1368045eb602eb453708a7b9",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.8,<3.11",
"size": 68447945,
"upload_time": "2023-11-15T06:37:30",
"upload_time_iso_8601": "2023-11-15T06:37:30.682643Z",
"url": "https://files.pythonhosted.org/packages/cf/ba/391e48532ba421a5d32cab254d4200cf23b8a9b29d14a2bef16f45c1f1af/lava_dl-0.5.0-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2023-11-15 06:37:30",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "lava-nc",
"github_project": "lava-dl",
"travis_ci": false,
"coveralls": false,
"github_actions": true,
"lcname": "lava-dl"
}
JSON
