GoogleCloudPlatform / ai-infra-cluster-provisioning

Intro

Conditional complexity (also called cyclomatic complexity) is a term used to measure the complexity of software. The term refers to the number of possible paths through a program function. A higher value ofter means higher maintenance and testing costs (infosecinstitute.com).
Conditional complexity is calculated by counting all conditions in the program that can affect the execution path (e.g. if statement, loops, switches, and/or operators, try and catch blocks...).
Conditional complexity is measured at the unit level (methods, functions...).
Units are classified in four categories based on the measured McCabe index: 1-5 (simple units), 6-10 (medium complex units), 11-25 (complex units), 26+ (very complex units).

Learn more...

Conditional Complexity Overall

There are 82 units with 653 lines of code in units (6.8% of code).

0 very complex units (0 lines of code)
0 complex units (0 lines of code)
0 medium complex units (0 lines of code)
3 simple units (29 lines of code)
79 very simple units (624 lines of code)

Legend:

51+

26-50

11-25

6-10

1-5

Alternative Visuals

3D view of all units

Conditional Complexity per Extension

51+

26-50

11-25

6-10

1-5

Conditional Complexity per Logical Component

primary logical decomposition

51+

26-50

11-25

6-10

1-5

Most Complex Units

Top 50 most complex units

Unit	# lines	McCabe index	# params
def validate_args() in sample_workloads/lit-gpt-demo/openwebtext.py	11	7	3
def _should_rank_record_comm() in sample_workloads/lit-gpt-demo/utilities/monitor_collectives.py	9	6	4
def _should_rank_record_comm() in sample_workloads/megatron-gke/docker/monitor_collectives.py	9	6	4
def get_lr() in sample_workloads/lit-gpt-demo/openwebtext.py	9	5	5
def shunt_torch_communication() in sample_workloads/lit-gpt-demo/utilities/monitor_collectives.py	14	5	0
def traced_batch_isend_irecv() in sample_workloads/lit-gpt-demo/utilities/monitor_collectives.py	10	5	1
def __init__() in sample_workloads/lit-gpt-demo/utilities/monitor_collectives.py	18	5	7
def _should_rank_print() in sample_workloads/lit-gpt-demo/utilities/monitor_collectives.py	10	5	3
def _is_crossnode_comm() in sample_workloads/lit-gpt-demo/utilities/monitor_collectives.py	13	4	2
def traced_batch_isend_irecv() in sample_workloads/megatron-gke/docker/monitor_collectives.py	7	4	1
def _is_crossnode_comm() in sample_workloads/megatron-gke/docker/monitor_collectives.py	13	4	2
def __init__() in sample_workloads/megatron-gke/docker/monitor_collectives.py	13	4	6
def _should_rank_print() in sample_workloads/megatron-gke/docker/monitor_collectives.py	8	4	3
def on_train_batch_start() in sample_workloads/lit-gpt-demo/utilities/nsight_callbacks.py	8	3	3
def traced_broadcast_object_list() in sample_workloads/lit-gpt-demo/utilities/monitor_collectives.py	11	3	4
def traced_all_to_all() in sample_workloads/lit-gpt-demo/utilities/monitor_collectives.py	10	3	4
def traced_all_to_all_single() in sample_workloads/lit-gpt-demo/utilities/monitor_collectives.py	19	3	6
def _get_original_p2p_op() in sample_workloads/lit-gpt-demo/utilities/monitor_collectives.py	5	3	1
def traced_broadcast_object_list() in sample_workloads/megatron-gke/docker/monitor_collectives.py	11	3	4
def traced_all_to_all() in sample_workloads/megatron-gke/docker/monitor_collectives.py	10	3	4
def traced_all_to_all_single() in sample_workloads/megatron-gke/docker/monitor_collectives.py	19	3	6
def _get_original_p2p_op() in sample_workloads/megatron-gke/docker/monitor_collectives.py	5	3	1
def run() in sample_workloads/pingpong/pingpong.py	10	3	1
def validate() in sample_workloads/lit-gpt-demo/openwebtext.py	12	2	4
def __iter__() in sample_workloads/lit-gpt-demo/openwebtext.py	7	2	1
def on_after_backward() in sample_workloads/lit-gpt-demo/utilities/nsight_callbacks.py	3	2	1
def _shunt_torch_communication_calls() in sample_workloads/lit-gpt-demo/utilities/monitor_collectives.py	27	2	0
def traced_barrier() in sample_workloads/lit-gpt-demo/utilities/monitor_collectives.py	4	2	3
def traced_broadcast() in sample_workloads/lit-gpt-demo/utilities/monitor_collectives.py	6	2	4
def traced_gather() in sample_workloads/lit-gpt-demo/utilities/monitor_collectives.py	9	2	5
def traced_scatter() in sample_workloads/lit-gpt-demo/utilities/monitor_collectives.py	9	2	5
def traced_reduce() in sample_workloads/lit-gpt-demo/utilities/monitor_collectives.py	6	2	5
def traced_reduce_scatter() in sample_workloads/lit-gpt-demo/utilities/monitor_collectives.py	11	2	5
def traced_reduce_scatter_tensor() in sample_workloads/lit-gpt-demo/utilities/monitor_collectives.py	11	2	5
def traced_all_reduce() in sample_workloads/lit-gpt-demo/utilities/monitor_collectives.py	7	2	4
def traced_all_gather() in sample_workloads/lit-gpt-demo/utilities/monitor_collectives.py	7	2	4
def traced_all_gather_into_tensor() in sample_workloads/lit-gpt-demo/utilities/monitor_collectives.py	7	2	4
def traced_isend() in sample_workloads/lit-gpt-demo/utilities/monitor_collectives.py	5	2	4
def traced_irecv() in sample_workloads/lit-gpt-demo/utilities/monitor_collectives.py	5	2	4
def traced_send() in sample_workloads/lit-gpt-demo/utilities/monitor_collectives.py	5	2	4
def traced_recv() in sample_workloads/lit-gpt-demo/utilities/monitor_collectives.py	5	2	4
def _emit_call_description() in sample_workloads/lit-gpt-demo/utilities/monitor_collectives.py	7	2	6
def shunt_torch_communication() in sample_workloads/megatron-gke/docker/monitor_collectives.py	8	2	0
def _shunt_torch_communication_calls() in sample_workloads/megatron-gke/docker/monitor_collectives.py	27	2	0
def traced_barrier() in sample_workloads/megatron-gke/docker/monitor_collectives.py	4	2	3
def traced_broadcast() in sample_workloads/megatron-gke/docker/monitor_collectives.py	6	2	4
def traced_gather() in sample_workloads/megatron-gke/docker/monitor_collectives.py	9	2	5
def traced_scatter() in sample_workloads/megatron-gke/docker/monitor_collectives.py	9	2	5
def traced_reduce() in sample_workloads/megatron-gke/docker/monitor_collectives.py	6	2	5
def traced_reduce_scatter() in sample_workloads/megatron-gke/docker/monitor_collectives.py	11	2	5