If you reverse the axis sense the bars will still start at zero but now the zero is on the other side of the plot.
Other changes I have made are:

• summarise has an argument .by, group_by is no longer needed;
• whenever you have both x and y coordinates, geom_col does what geom_bar does but with no need for stat = "identity";
• I have added a breaks argument since counts are integers, therefore, discrete.

suppressPackageStartupMessages(library(tidyverse))

iris %>%
  summarise(count = n(), .by = Sepal.Length) -> histBar

ggplot(data = histBar,
       mapping = aes(x = Sepal.Length, y = count)) +
  geom_col() +
  coord_flip() +
  scale_y_reverse(breaks = 0:12)

^{Created on 2024-09-07 with reprex v2.1.0}

Off-topic?

A plot going down just for fun.

iris %>%
  summarise(count = n(), .by = Sepal.Length) %>%
  ggplot(aes(x = Sepal.Length, y = count)) +
  geom_col() +
  scale_y_reverse(breaks = 0:12)

^{Created on 2024-09-07 with reprex v2.1.0}

And before the OP edit, the plot was a histogram.

ggplot(iris, mapping = aes(y = Sepal.Length)) +
  geom_histogram(bins = 30) +
  scale_x_reverse(breaks = 0:12)

^{Created on 2024-09-07 with reprex v2.1.0}