FreeFly19

void setup() {
  Serial.begin(230400,SERIAL_8N1);
  Serial1.begin(230400, SERIAL_8N1, 21, 22);
}

int n;
void loop() {
  n = Serial1.available();
  if (n != 0)
  {

FreeFly19

import json
from argparse import ArgumentParser


def convert(input_file, output_file):
    with open(input_file, 'r') as fin:
        data2 = json.load(fin)

    features = data2['features']
    with open(output_file, 'w') as fout:

FreeFly19

import numpy as np
import torch
import torchvision
from torch.utils.data import DataLoader
from torchvision.transforms import ToTensor

train_dataset = torchvision.datasets.MNIST('data', train=True, transform=ToTensor(), download=True)
test_dataset = torchvision.datasets.MNIST('data', train=False, transform=ToTensor(), download=True)

batch_size = 16

FreeFly19

import torch

w = torch.tensor([5.], requires_grad=True)
lr = 0.6

w_mom_grad = 0
momentum_coef = 0.9
weight_decay = 0.001

for i in range(10):

FreeFly19

import torch

x = torch.tensor([1.5,2.8,15.,19.5])
y = torch.tensor([0.,0.,1.,1.])

w = torch.tensor([-.3], requires_grad=True)
b = torch.tensor([0.1232154], requires_grad=True)


def model(x):

FreeFly19

import time

import av
import tqdm
from av.video import VideoStream


desired_fps = 2

timestamp = int(time.time())

FreeFly19

client
# put the line below to disable tunneling outcoming trafic
route-nopull
....

FreeFly19

https://unix.stackexchange.com/questions/528400/how-can-i-stream-my-desktop-screen-to-dev-video1-as-a-fake-webcam-on-linux

FreeFly19

import tensorflow as tf

train_ds = tf.keras.utils.image_dataset_from_directory(
  train_dir,
  validation_split=0.1,
  subset="training",
  label_mode="categorical",
  #shuffle=True,
  seed=123,
  image_size=(image_size, image_size),

FreeFly19

import keras.backend as K
def f1(y_true, y_pred):
    true_positives = K.sum(K.round(K.clip(y_true * y_pred, 0, 1)))
    possible_positives = K.sum(K.round(K.clip(y_true, 0, 1)))
    predicted_positives = K.sum(K.round(K.clip(y_pred, 0, 1)))
    precision = true_positives / (predicted_positives + K.epsilon())
    recall = true_positives / (possible_positives + K.epsilon())
    f1_val = 2*(precision*recall)/(precision+recall+K.epsilon())
    return f1_val