szha commented on a change in pull request #11566: [MXNET-626] Add while_loop
URL: https://github.com/apache/incubator-mxnet/pull/11566#discussion_r203210492
##########
File path: python/mxnet/ndarray/contrib.py
##########
@@ -191,3 +191,170 @@ def check_input(inputs, in_type, msg):
if not_data_list and len(outputs) == 1:
outputs = outputs[0]
return (outputs, states)
+
+
+def while_loop(cond, func, loop_vars, max_iterations=None):
+ """Run a while loop with user-defined computation and loop condition.
+
+ This operator simulates a while loop which iterately does customized
computation
+ as long as the condition is satisfied.
+
+ `loop_vars` is a list of NDArrays on which the computation uses.
+
+ `cond` is a user-defined function, used as the loop condition.
+ It consumes `loop_vars`, and produces a scalar MXNet NDArray,
+ indicating the termination of the loop.
+ The loop ends when `cond` returns false (zero).
+ The `cond` is variadic, and its signature should be
+ `cond(*loop_vars) => NDArray`.
+
+ `func` is a user-defined function, used as the loop body.
+ It also consumes `loop_vars`, and produces `step_output` and
`new_loop_vars` at each step.
+ In each step, `step_output` should contain the same number elements.
+ Through all steps, the i-th element of `step_output` should have the same
shape and dtype.
+ Also, `new_loop_vars` should contain the same number of elements as
`loop_vars`,
+ and the corresponding element should have the same shape and dtype.
+ The `func` is variadic, and its signature should be
+ `func(*loop_vars) => (List[NDArray] step_output, List[NDArray]
new_loop_vars)`.
+
+ `max_iterations` is a scalar that defines the maximum number of iterations
allowed.
+
+ This function returns two lists.
+ The first list has the length of `|step_output|`,
+ in which the i-th element are all i-th elements of
+ `step_output` from all steps, stacked along axis 0.
+ The second list has the length of `|loop_vars|`,
+ which represents final states of loop variables.
+
+ Warning 1: for now, the axis 0 of all NDArrays in the first list are
`max_iterations`,
+ due to lack of dynamic shape inference.
+
+ Warning 2: when `cond` is never satisfied, we assume `step_output` is
empty,
+ because it cannot be inferred. This is different from the symbolic version.
+
+ Parameters
+ ----------
+ cond: a Python function.
+ The loop condition.
+ func: a Python function.
+ The loop body.
+ loop_vars: list of NDArrays.
+ The initial values of the loop variables.
+ max_iterations: a python int.
+ Maximum number of iterations.
+
+ Returns
+ -------
+ outputs: two lists, which both contains 0, 1 or more NDArrays.
Review comment:
```
Returns
------
outputs: list of NDArray
stacked output from each step
states: list of NDArray
final state
```
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
