[DRAFT] update evaluate to be concurrent #697
isahers1Annotations
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:
python/langsmith/evaluation/_arunner.py#L1
...........
WARNING: the benchmark result may be unstable
* the standard deviation (89.6 ms) is 13% of the mean (685 ms)
Try to rerun the benchmark with more runs, values and/or loops.
Run 'python -m pyperf system tune' command to reduce the system jitter.
Use pyperf stats, pyperf dump and pyperf hist to analyze results.
Use --quiet option to hide these warnings.
create_5_000_run_trees: Mean +- std dev: 685 ms +- 90 ms
...........
WARNING: the benchmark result may be unstable
* the standard deviation (151 ms) is 11% of the mean (1.42 sec)
Try to rerun the benchmark with more runs, values and/or loops.
Run 'python -m pyperf system tune' command to reduce the system jitter.
Use pyperf stats, pyperf dump and pyperf hist to analyze results.
Use --quiet option to hide these warnings.
create_10_000_run_trees: Mean +- std dev: 1.42 sec +- 0.15 sec
...........
WARNING: the benchmark result may be unstable
* the standard deviation (190 ms) is 13% of the mean (1.45 sec)
Try to rerun the benchmark with more runs, values and/or loops.
Run 'python -m pyperf system tune' command to reduce the system jitter.
Use pyperf stats, pyperf dump and pyperf hist to analyze results.
Use --quiet option to hide these warnings.
create_20_000_run_trees: Mean +- std dev: 1.45 sec +- 0.19 sec
...........
dumps_class_nested_py_branch_and_leaf_200x400: Mean +- std dev: 695 us +- 11 us
...........
dumps_class_nested_py_leaf_50x100: Mean +- std dev: 25.1 ms +- 0.4 ms
...........
dumps_class_nested_py_leaf_100x200: Mean +- std dev: 105 ms +- 3 ms
...........
dumps_dataclass_nested_50x100: Mean +- std dev: 25.4 ms +- 0.3 ms
...........
WARNING: the benchmark result may be unstable
* the standard deviation (16.7 ms) is 23% of the mean (71.9 ms)
Try to rerun the benchmark with more runs, values and/or loops.
Run 'python -m pyperf system tune' command to reduce the system jitter.
Use pyperf stats, pyperf dump and pyperf hist to analyze results.
Use --quiet option to hide these warnings.
dumps_pydantic_nested_50x100: Mean +- std dev: 71.9 ms +- 16.7 ms
...........
dumps_pydanticv1_nested_50x100: Mean +- std dev: 200 ms +- 3 ms
|
|
Comparison against main:
python/langsmith/evaluation/_arunner.py#L1
+-----------------------------------------------+----------+------------------------+
| Benchmark | main | changes |
+===============================================+==========+========================+
| dumps_pydanticv1_nested_50x100 | 220 ms | 200 ms: 1.10x faster |
+-----------------------------------------------+----------+------------------------+
| create_5_000_run_trees | 731 ms | 685 ms: 1.07x faster |
+-----------------------------------------------+----------+------------------------+
| dumps_dataclass_nested_50x100 | 25.5 ms | 25.4 ms: 1.01x faster |
+-----------------------------------------------+----------+------------------------+
| dumps_class_nested_py_leaf_50x100 | 25.1 ms | 25.1 ms: 1.00x slower |
+-----------------------------------------------+----------+------------------------+
| create_10_000_run_trees | 1.41 sec | 1.42 sec: 1.00x slower |
+-----------------------------------------------+----------+------------------------+
| dumps_class_nested_py_leaf_100x200 | 104 ms | 105 ms: 1.01x slower |
+-----------------------------------------------+----------+------------------------+
| dumps_class_nested_py_branch_and_leaf_200x400 | 690 us | 695 us: 1.01x slower |
+-----------------------------------------------+----------+------------------------+
| create_20_000_run_trees | 1.39 sec | 1.45 sec: 1.04x slower |
+-----------------------------------------------+----------+------------------------+
| dumps_pydantic_nested_50x100 | 65.9 ms | 71.9 ms: 1.09x slower |
+-----------------------------------------------+----------+------------------------+
| Geometric mean | (ref) | 1.00x faster |
+-----------------------------------------------+----------+------------------------+
|