lib: Add unit test for multistep planner graph #1204
Triggered via pull request
December 10, 2024 18:41
nfcampos
Status
Success
Total duration
54m 20s
Artifacts
–
Annotations
1 warning and 2 notices
|
benchmark
ubuntu-latest pipelines will use ubuntu-24.04 soon. For more details, see https://github.com/actions/runner-images/issues/10636
|
|
Benchmark results:
libs/langgraph/tests/test_pregel.py#L1
.........................................
fanout_to_subgraph_10x: Mean +- std dev: 62.1 ms +- 1.4 ms
.........................................
fanout_to_subgraph_10x_sync: Mean +- std dev: 52.4 ms +- 1.1 ms
.........................................
fanout_to_subgraph_10x_checkpoint: Mean +- std dev: 93.5 ms +- 7.3 ms
.........................................
fanout_to_subgraph_10x_checkpoint_sync: Mean +- std dev: 95.0 ms +- 1.9 ms
.........................................
fanout_to_subgraph_100x: Mean +- std dev: 616 ms +- 10 ms
.........................................
fanout_to_subgraph_100x_sync: Mean +- std dev: 512 ms +- 6 ms
.........................................
fanout_to_subgraph_100x_checkpoint: Mean +- std dev: 936 ms +- 39 ms
.........................................
fanout_to_subgraph_100x_checkpoint_sync: Mean +- std dev: 947 ms +- 16 ms
.........................................
react_agent_10x: Mean +- std dev: 31.1 ms +- 0.6 ms
.........................................
react_agent_10x_sync: Mean +- std dev: 22.8 ms +- 0.2 ms
.........................................
react_agent_10x_checkpoint: Mean +- std dev: 47.0 ms +- 0.8 ms
.........................................
react_agent_10x_checkpoint_sync: Mean +- std dev: 36.9 ms +- 0.4 ms
.........................................
react_agent_100x: Mean +- std dev: 347 ms +- 6 ms
.........................................
react_agent_100x_sync: Mean +- std dev: 274 ms +- 4 ms
.........................................
react_agent_100x_checkpoint: Mean +- std dev: 938 ms +- 8 ms
.........................................
react_agent_100x_checkpoint_sync: Mean +- std dev: 837 ms +- 7 ms
.........................................
wide_state_25x300: Mean +- std dev: 23.8 ms +- 0.5 ms
.........................................
wide_state_25x300_sync: Mean +- std dev: 14.9 ms +- 0.3 ms
.........................................
wide_state_25x300_checkpoint: Mean +- std dev: 287 ms +- 13 ms
.........................................
wide_state_25x300_checkpoint_sync: Mean +- std dev: 273 ms +- 12 ms
.........................................
wide_state_15x600: Mean +- std dev: 27.6 ms +- 0.4 ms
.........................................
wide_state_15x600_sync: Mean +- std dev: 17.3 ms +- 0.1 ms
.........................................
wide_state_15x600_checkpoint: Mean +- std dev: 486 ms +- 13 ms
.........................................
wide_state_15x600_checkpoint_sync: Mean +- std dev: 474 ms +- 14 ms
.........................................
wide_state_9x1200: Mean +- std dev: 27.6 ms +- 0.6 ms
.........................................
wide_state_9x1200_sync: Mean +- std dev: 17.3 ms +- 0.1 ms
.........................................
wide_state_9x1200_checkpoint: Mean +- std dev: 321 ms +- 13 ms
.........................................
wide_state_9x1200_checkpoint_sync: Mean +- std dev: 306 ms +- 13 ms
|
|
Comparison against main:
libs/langgraph/tests/test_pregel.py#L1
+----------------------------------------+---------+-----------------------+
| Benchmark | main | changes |
+========================================+=========+=======================+
| fanout_to_subgraph_100x | 628 ms | 616 ms: 1.02x faster |
+----------------------------------------+---------+-----------------------+
| fanout_to_subgraph_100x_checkpoint | 952 ms | 936 ms: 1.02x faster |
+----------------------------------------+---------+-----------------------+
| wide_state_15x600_sync | 17.3 ms | 17.3 ms: 1.00x slower |
+----------------------------------------+---------+-----------------------+
| fanout_to_subgraph_100x_sync | 510 ms | 512 ms: 1.00x slower |
+----------------------------------------+---------+-----------------------+
| wide_state_15x600 | 27.5 ms | 27.6 ms: 1.00x slower |
+----------------------------------------+---------+-----------------------+
| react_agent_100x_checkpoint_sync | 833 ms | 837 ms: 1.01x slower |
+----------------------------------------+---------+-----------------------+
| react_agent_10x_checkpoint_sync | 36.6 ms | 36.9 ms: 1.01x slower |
+----------------------------------------+---------+-----------------------+
| fanout_to_subgraph_10x_checkpoint_sync | 94.4 ms | 95.0 ms: 1.01x slower |
+----------------------------------------+---------+-----------------------+
| react_agent_10x_checkpoint | 46.6 ms | 47.0 ms: 1.01x slower |
+----------------------------------------+---------+-----------------------+
| react_agent_100x_checkpoint | 930 ms | 938 ms: 1.01x slower |
+----------------------------------------+---------+-----------------------+
| Geometric mean | (ref) | 1.00x slower |
+----------------------------------------+---------+-----------------------+
Benchmark hidden because not significant (18): fanout_to_subgraph_10x_checkpoint, react_agent_100x_sync, fanout_to_subgraph_10x, wide_state_9x1200_checkpoint_sync, wide_state_9x1200_sync, react_agent_10x_sync, react_agent_100x, wide_state_25x300, wide_state_25x300_sync, wide_state_25x300_checkpoint_sync, wide_state_15x600_checkpoint, wide_state_25x300_checkpoint, wide_state_9x1200_checkpoint, fanout_to_subgraph_100x_checkpoint_sync, react_agent_10x, fanout_to_subgraph_10x_sync, wide_state_9x1200, wide_state_15x600_checkpoint_sync
|