This is xgboost_1.5.0.2. It seems iteration_range is not used in the predict function. See an example below. Furthermore, the docs on iteration_range is confusing, not for R. https://cran.r-project.org/web/packages/xgboost/xgboost.pdf
set.seed(1947)
x <- matrix(rnorm(100*2),100,2)
y <- sample(c(0,1),100,replace=TRUE)
fit0 <- xgboost(x, y, objective=“binary:logitraw”,
verbose=0, max.depth=5, nrounds=50)
pred1 <- predict(fit0, newdata=x, iteration_range=c(1, 1))
pred2 <- predict(fit0, newdata=x, iteration_range=c(1, 50))
summary(pred1 - pred2)
Min. 1st Qu. Median Mean 3rd Qu. Max.
0 0 0 0 0 0