Apogeum12 / benchmark_ytLip

Simulating Neuron from model Hodgkin–Huxley with help ODE Solver.

Benchmark simulating biological Neuron Model:

Multicore cpu and gpu Benchmark. But in fact it is simulation hodgkin huxley model neuron. Where for computing Ordinary Differential Equations i used Differentiable ODE solver method.

Script

Benchmark in script version for Linux:

Installation

• In Directly benchmark_ytLip/

sudo apt install python3-venv

python3 -m venv benchmark_script

source benchmark_script/bin/activate 

Install package:

pip3 -r package.txt

 sudo apt install python3-tk 

Install torchdiffeq

pip3 install git+https://github.com/rtqichen/torchdiffeq

Benchmark in script version for Windows:

Installation

(IN PROGRESS)

App verssion

Maybe in future ^ ^

Run:

In directly folder benchmark_script

• [--h] Num Hiden Neuron; [--i] Num input Size; [--d] Device choice; [--t] Test Size
• If Dont put Neuron Hidden num then Hidden_size is equal => 2 * sqrt(Input_size)

python3 bench.py --i 2048 --d cpu --t 10
## Plan
- Notebook cpu version (DONE)
- Support gpu (DONE)
- Shell Script (Done)
- App Linux (In progress)
- App Windows
- ...