Save the date! Google I/O returns May 18-20 Register now


Benchmarks a


ds = tfds.load('mnist', split='train').batch(32).prefetch()
tfds.core.benchmark(ds, batch_size=32)


  • Total execution time
  • Setup time (first warmup batch)
  • Number of examples/sec

ds Dataset to benchmark
num_iter Number of iteration to perform (iteration might be batched)
batch_size Batch size of the dataset, used to normalize iterations

statistics The recorded statistics, for eventual post-processing