Remote distribution of neural networks
Remote distribution of multiple neural network models to various client devices over a network can be implemented by identifying a native neural network and remotely converting the native neural network to a target neural network based on a given client device operating environment. The native neural network can be configured for execution using efficient parameters, and the target neural network can use less efficient but more precise parameters.
1. A method comprising:
detecting, via a device, a request to execute a task;
identifying a native neural network model associated with the task;
converting the native neural network model to a target neural network model based on a computational resource of the device;
using the target neural network model to execute the task; and
causing display of an output as a result of the execution of the task.
2. The method of claim 1 , the converting the native neural network model to the target neural network model comprises:
reassembling the native neural network model block-by-block into the target neural network model based on the computational resource of the device.
3. The method of claim 1 , wherein the native neural network model is integrated within an application installed on the device.
4. The method of claim 1 , wherein the request to execute the task is detected based on detection of a user interface (UI) based event that triggers the task.
5. The method of claim 4 , wherein the UI based event comprises one or more of a scrolling of a plurality of icons configured to initiate a plurality of image effects, navigation to a pre-specified user interface, and activation of a messaging client application associated with the device.
6. The method of claim 1 , comprising:
determining the target neural network model based on an operating system of the device having a library for parallel execution of neural networks on a plurality of hardware processors.
7. The method of claim 1 , comprising:
specifying a plurality of input parameters of the target neural network model; and
compiling the target neural network model using the plurality of input parameters.
8. The method of claim 1 , wherein the task comprises modification of an image using an image effect associated with the native neural network model.
9. The method of claim 8 , comprising:
identifying the computational resource of the device; and
applying, using the target neural network model, the image effect to the image to generate a modified image based on a plurality of input parameters specified for the target neural network model.
10. The method of claim 1 , wherein the computational resource of the device comprises one or more of a screen size, bandwidth quality, GPU availability, one or more hardware acceleration libraries, multi-core parallel processing capability on a CPU, and multi-core concurrent processing capability on the CPU.
11. A system comprising:
a memory storing instructions; and
one or more hardware processors communicatively coupled to the memory and configured by the instructions to perform operations comprising:
detecting, via a device, a request to execute a task;
identifying a native neural network model associated with the task;
converting the native neural network model to a target neural network model based on a computational resource of the device;
using the target neural network model to execute the task; and
causing display of an output as a result of the execution of the task.
12. The system of claim 11 , the converting the native neural network model to the target neural network model comprises:
reassembling the native neural network model block-by-block into the target neural network model based on the computational resource of the device.
13. The system of claim 11 , wherein the native neural network model is integrated within an application installed on the device.
14. The system of claim 11 , wherein the request to execute the task is detected based on detection of a user interface (UI) based event that triggers the task.
15. The system of claim 14 , wherein the UI based event comprises one or more of a scrolling of a plurality of icons configured to initiate a plurality of image effects, navigation to a pre-specified user interface, and activation of a messaging client application associated with the device.
16. The system of claim 11 , wherein the operations comprise:
determining the target neural network model based on an operating system of the device having a library for parallel execution of neural networks on a plurality of hardware processors.
17. The system of claim 11 , wherein the operations comprise:
specifying a plurality of input parameters of the target neural network model; and
compiling the target neural network model based on the plurality of input parameters.
18. The system of claim 11 , wherein the task comprises modification of an image using an image effect associated with the native neural network model.
19. The system of claim 18 , wherein the operations comprise:
identifying the computational resource of the device; and
applying, using the target neural network model, the image effect to the image to generate a modified image based on a plurality of input parameters specified for the target neural network model.
20. A non-transitory machine-readable storage medium comprising instructions that, when executed by a machine, cause the machine to perform operations comprising:
detecting, via a device, a request to execute a task;
identifying a native neural network model associated with the task;
converting the native neural network model to a target neural network model based on a computational resource of the device;
using the target neural network model to execute the task; and
causing display of an output as a result of the execution of the task.