Skip to content

[CI] Fix flaky test tracker: use job-level API data, always notify, add badge - #1600

Merged
vmoens merged 1 commit into
gh/vmoens/80/basefrom
gh/vmoens/80/head
Feb 19, 2026
Merged

[CI] Fix flaky test tracker: use job-level API data, always notify, add badge#1600
vmoens merged 1 commit into
gh/vmoens/80/basefrom
gh/vmoens/80/head

Conversation

@vmoens

@vmoens vmoens commented Feb 19, 2026 •
edited
Loading

Copy link
Copy Markdown
Collaborator

Stack from ghstack (oldest at bottom):

The flaky test tracker was collecting 0 test results because the CI
(pytorch/test-infra) doesn't upload test result artifacts. Rewrote
analyze_flaky_tests.py to use the GitHub Actions API directly, tracking
job-level pass/fail outcomes (e.g. "test-cpu (3.13) failed 3/30 runs").

Also fixed the notify step to always run (so it can close the tracking
issue when all flaky tests are resolved), and added a shields.io flaky
test badge to README.md.

Co-authored-by: Cursor cursoragent@cursor.com

@vmoens
Update
3c906b5 
[ghstack-poisoned]
vmoens added a commit that referenced this pull request Feb 19, 2026
@vmoens @cursoragent
[CI] Fix flaky test tracker: use job-level API data, always notify, a…
73d01b0 
…dd badge

The flaky test tracker was collecting 0 test results because the CI
(pytorch/test-infra) doesn't upload test result artifacts. Rewrote
analyze_flaky_tests.py to use the GitHub Actions API directly, tracking
job-level pass/fail outcomes (e.g. "test-cpu (3.13) failed 3/30 runs").

Also fixed the notify step to always run (so it can close the tracking
issue when all flaky tests are resolved), and added a shields.io flaky
test badge to README.md.

Co-authored-by: Cursor <cursoragent@cursor.com>
ghstack-source-id: 54a62d9
Pull-Request: #1600
@meta-cla meta-cla Bot added the CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. label Feb 19, 2026
@github-actions github-actions Bot added CI documentation Improvements or additions to documentation labels Feb 19, 2026
vmoens added a commit that referenced this pull request Feb 19, 2026
@vmoens @cursoragent
[CI] Fix flaky test tracker: use job-level API data, always notify, a…
3327f31 
…dd badge

The flaky test tracker was collecting 0 test results because the CI
(pytorch/test-infra) doesn't upload test result artifacts. Rewrote
analyze_flaky_tests.py to use the GitHub Actions API directly, tracking
job-level pass/fail outcomes (e.g. "test-cpu (3.13) failed 3/30 runs").

Also fixed the notify step to always run (so it can close the tracking
issue when all flaky tests are resolved), and added a shields.io flaky
test badge to README.md.

Co-authored-by: Cursor <cursoragent@cursor.com>
ghstack-source-id: 54a62d9
Pull-Request: #1600
@vmoens
vmoens merged commit 3c906b5 into gh/vmoens/80/base Feb 19, 2026
38 of 54 checks passed
@vmoens
vmoens deleted the gh/vmoens/80/head branch February 19, 2026 11:40
@github-actions

github-actions Bot commented Feb 19, 2026

Copy link
Copy Markdown
Contributor

$\color{#D29922}\textsf{\Large&amp;#x26A0;\kern{0.2cm}\normalsize Warning}$ Result of CPU Benchmark Tests

Total Benchmarks: 243. Improved: $\large\color{#35bf28}19$. Worsened: $\large\color{#d91a1a}6$.

Expand to view detailed results
Name Max Mean Ops Ops on Repo HEAD Change
test_plain_set_nested 34.8010μs 14.7863μs 67.6303 KOps/s 68.5509 KOps/s $\color{#d91a1a}-1.34\%$
test_plain_set_stack_nested 32.0620μs 14.9649μs 66.8232 KOps/s 67.0937 KOps/s $\color{#d91a1a}-0.40\%$
test_plain_set_nested_inplace 51.5730μs 16.4336μs 60.8511 KOps/s 61.5965 KOps/s $\color{#d91a1a}-1.21\%$
test_plain_set_stack_nested_inplace 44.9520μs 16.3699μs 61.0879 KOps/s 61.7783 KOps/s $\color{#d91a1a}-1.12\%$
test_items 43.8220μs 5.8461μs 171.0547 KOps/s 174.1670 KOps/s $\color{#d91a1a}-1.79\%$
test_items_nested 0.5114ms 0.4520ms 2.2123 KOps/s 2.2154 KOps/s $\color{#d91a1a}-0.14\%$
test_items_nested_locked 0.5060ms 0.4572ms 2.1871 KOps/s 2.2038 KOps/s $\color{#d91a1a}-0.76\%$
test_items_nested_leaf 0.1368ms 95.2808μs 10.4953 KOps/s 10.3088 KOps/s $\color{#35bf28}+1.81\%$
test_items_stack_nested 0.5010ms 0.4541ms 2.2024 KOps/s 2.1979 KOps/s $\color{#35bf28}+0.20\%$
test_items_stack_nested_leaf 0.1285ms 95.2444μs 10.4993 KOps/s 10.4562 KOps/s $\color{#35bf28}+0.41\%$
test_items_stack_nested_locked 0.5325ms 0.4558ms 2.1939 KOps/s 2.1706 KOps/s $\color{#35bf28}+1.07\%$
test_keys 27.4010μs 4.2215μs 236.8833 KOps/s 237.6308 KOps/s $\color{#d91a1a}-0.31\%$
test_keys_nested 0.1625ms 0.1192ms 8.3874 KOps/s 8.3718 KOps/s $\color{#35bf28}+0.19\%$
test_keys_nested_locked 88.0454ms 0.1414ms 7.0708 KOps/s 7.7962 KOps/s $\textbf{\color{#d91a1a}-9.31\%}$
test_keys_nested_leaf 0.1457ms 0.1105ms 9.0521 KOps/s 8.9997 KOps/s $\color{#35bf28}+0.58\%$
test_keys_stack_nested 0.1592ms 0.1197ms 8.3534 KOps/s 8.3740 KOps/s $\color{#d91a1a}-0.25\%$
test_keys_stack_nested_leaf 0.1447ms 0.1100ms 9.0918 KOps/s 9.0876 KOps/s $\color{#35bf28}+0.05\%$
test_keys_stack_nested_locked 0.1819ms 0.1280ms 7.8101 KOps/s 7.7038 KOps/s $\color{#35bf28}+1.38\%$
test_values 6.1624μs 1.0225μs 977.9639 KOps/s 976.6124 KOps/s $\color{#35bf28}+0.14\%$
test_values_nested 73.2830μs 47.7202μs 20.9555 KOps/s 20.7887 KOps/s $\color{#35bf28}+0.80\%$
test_values_nested_locked 75.5130μs 50.6428μs 19.7461 KOps/s 19.5954 KOps/s $\color{#35bf28}+0.77\%$
test_values_nested_leaf 84.7840μs 53.9635μs 18.5310 KOps/s 18.5399 KOps/s $\color{#d91a1a}-0.05\%$
test_values_stack_nested 76.9530μs 47.7954μs 20.9225 KOps/s 20.9026 KOps/s $\color{#35bf28}+0.10\%$
test_values_stack_nested_leaf 84.1830μs 54.1062μs 18.4822 KOps/s 18.5897 KOps/s $\color{#d91a1a}-0.58\%$
test_values_stack_nested_locked 83.0830μs 51.1157μs 19.5635 KOps/s 19.6747 KOps/s $\color{#d91a1a}-0.57\%$
test_membership 6.0835μs 0.8530μs 1.1723 MOps/s 1.1946 MOps/s $\color{#d91a1a}-1.87\%$
test_membership_nested 40.9020μs 2.9021μs 344.5761 KOps/s 343.2626 KOps/s $\color{#35bf28}+0.38\%$
test_membership_nested_leaf 74.0330μs 2.9297μs 341.3364 KOps/s 344.4681 KOps/s $\color{#d91a1a}-0.91\%$
test_membership_stacked_nested 29.2510μs 2.8958μs 345.3226 KOps/s 343.8773 KOps/s $\color{#35bf28}+0.42\%$
test_membership_stacked_nested_leaf 23.9010μs 2.9122μs 343.3861 KOps/s 345.2518 KOps/s $\color{#d91a1a}-0.54\%$
test_membership_nested_last 33.3320μs 4.3274μs 231.0836 KOps/s 230.3285 KOps/s $\color{#35bf28}+0.33\%$
test_membership_nested_leaf_last 31.4620μs 4.3224μs 231.3520 KOps/s 230.1013 KOps/s $\color{#35bf28}+0.54\%$
test_membership_stacked_nested_last 23.0510μs 4.3141μs 231.7976 KOps/s 230.0626 KOps/s $\color{#35bf28}+0.75\%$
test_membership_stacked_nested_leaf_last 36.4720μs 4.3283μs 231.0354 KOps/s 229.6378 KOps/s $\color{#35bf28}+0.61\%$
test_nested_getleaf 49.8520μs 21.4907μs 46.5318 KOps/s 47.0328 KOps/s $\color{#d91a1a}-1.07\%$
test_nested_get 48.2020μs 20.4946μs 48.7933 KOps/s 49.8680 KOps/s $\color{#d91a1a}-2.16\%$
test_stacked_getleaf 45.3720μs 21.4040μs 46.7203 KOps/s 48.1807 KOps/s $\color{#d91a1a}-3.03\%$
test_stacked_get 45.3320μs 20.5111μs 48.7540 KOps/s 50.6724 KOps/s $\color{#d91a1a}-3.79\%$
test_nested_getitemleaf 54.5720μs 21.9919μs 45.4713 KOps/s 45.9361 KOps/s $\color{#d91a1a}-1.01\%$
test_nested_getitem 46.3610μs 20.6968μs 48.3166 KOps/s 49.3090 KOps/s $\color{#d91a1a}-2.01\%$
test_stacked_getitemleaf 95.0340μs 21.7568μs 45.9627 KOps/s 46.6408 KOps/s $\color{#d91a1a}-1.45\%$
test_stacked_getitem 55.6920μs 20.8913μs 47.8668 KOps/s 48.7611 KOps/s $\color{#d91a1a}-1.83\%$
test_lock_nested 8.0149ms 0.4805ms 2.0813 KOps/s 2.0962 KOps/s $\color{#d91a1a}-0.71\%$
test_lock_stack_nested 0.5617ms 0.4775ms 2.0944 KOps/s 2.0565 KOps/s $\color{#35bf28}+1.84\%$
test_unlock_nested 0.4505ms 0.3790ms 2.6388 KOps/s 2.5883 KOps/s $\color{#35bf28}+1.95\%$
test_unlock_stack_nested 0.4331ms 0.3823ms 2.6156 KOps/s 2.5532 KOps/s $\color{#35bf28}+2.45\%$
test_flatten_speed 0.1657ms 0.1238ms 8.0771 KOps/s 8.1636 KOps/s $\color{#d91a1a}-1.06\%$
test_unflatten_speed 0.6261ms 0.5691ms 1.7570 KOps/s 1.7776 KOps/s $\color{#d91a1a}-1.16\%$
test_common_ops 0.8385ms 0.6945ms 1.4399 KOps/s 1.4112 KOps/s $\color{#35bf28}+2.04\%$
test_creation 0.1076ms 2.9032μs 344.4426 KOps/s 341.8253 KOps/s $\color{#35bf28}+0.77\%$
test_creation_empty 34.6310μs 6.1395μs 162.8796 KOps/s 162.7461 KOps/s $\color{#35bf28}+0.08\%$
test_creation_nested_1 45.3020μs 10.4571μs 95.6288 KOps/s 95.4863 KOps/s $\color{#35bf28}+0.15\%$
test_creation_nested_2 38.0820μs 11.9683μs 83.5538 KOps/s 83.6340 KOps/s $\color{#d91a1a}-0.10\%$
test_creation_many_keys[10] 53.0620μs 18.4170μs 54.2978 KOps/s 55.5308 KOps/s $\color{#d91a1a}-2.22\%$
test_creation_many_keys[50] 0.1153ms 78.0668μs 12.8096 KOps/s 12.8640 KOps/s $\color{#d91a1a}-0.42\%$
test_creation_many_keys[100] 0.1926ms 0.1522ms 6.5692 KOps/s 6.5699 KOps/s $-0.01\%$
test_creation_nested_many_keys[10] 76.2540μs 39.2599μs 25.4713 KOps/s 25.4039 KOps/s $\color{#35bf28}+0.27\%$
test_creation_nested_many_keys[50] 0.2046ms 0.1597ms 6.2618 KOps/s 6.2663 KOps/s $\color{#d91a1a}-0.07\%$
test_clone 53.0520μs 13.5317μs 73.9005 KOps/s 71.6957 KOps/s $\color{#35bf28}+3.08\%$
test_getitem[int] 1.6945ms 14.4569μs 69.1711 KOps/s 55.3958 KOps/s $\textbf{\color{#35bf28}+24.87\%}$
test_getitem[slice_int] 0.1491ms 24.6510μs 40.5663 KOps/s 40.0851 KOps/s $\color{#35bf28}+1.20\%$
test_getitem[range] 0.1728ms 64.3025μs 15.5515 KOps/s 15.6821 KOps/s $\color{#d91a1a}-0.83\%$
test_getitem[tuple] 0.1494ms 23.9566μs 41.7422 KOps/s 41.1162 KOps/s $\color{#35bf28}+1.52\%$
test_getitem[list] 0.1901ms 60.1827μs 16.6161 KOps/s 16.9508 KOps/s $\color{#d91a1a}-1.97\%$
test_setitem_dim[int] 47.3220μs 26.5573μs 37.6545 KOps/s 37.3760 KOps/s $\color{#35bf28}+0.75\%$
test_setitem_dim[slice_int] 71.0930μs 45.0703μs 22.1875 KOps/s 22.5018 KOps/s $\color{#d91a1a}-1.40\%$
test_setitem_dim[range] 0.1410ms 93.6713μs 10.6756 KOps/s 10.5097 KOps/s $\color{#35bf28}+1.58\%$
test_setitem_dim[tuple] 61.5130μs 40.9414μs 24.4251 KOps/s 24.1622 KOps/s $\color{#35bf28}+1.09\%$
test_setitem 75.8230μs 18.2516μs 54.7896 KOps/s 54.1628 KOps/s $\color{#35bf28}+1.16\%$
test_set 51.7020μs 18.2713μs 54.7306 KOps/s 57.1189 KOps/s $\color{#d91a1a}-4.18\%$
test_set_shared 0.4817ms 0.2099ms 4.7632 KOps/s 4.7027 KOps/s $\color{#35bf28}+1.29\%$
test_update 0.2028ms 21.9805μs 45.4948 KOps/s 45.3340 KOps/s $\color{#35bf28}+0.35\%$
test_update_nested 93.9030μs 35.1123μs 28.4800 KOps/s 29.7188 KOps/s $\color{#d91a1a}-4.17\%$
test_update__nested 0.5337ms 36.4197μs 27.4577 KOps/s 28.5320 KOps/s $\color{#d91a1a}-3.77\%$
test_set_nested 53.5520μs 19.5007μs 51.2802 KOps/s 51.3562 KOps/s $\color{#d91a1a}-0.15\%$
test_set_nested_new 61.9830μs 24.8323μs 40.2701 KOps/s 40.8151 KOps/s $\color{#d91a1a}-1.34\%$
test_select 91.2340μs 42.6725μs 23.4343 KOps/s 23.9582 KOps/s $\color{#d91a1a}-2.19\%$
test_select_nested 0.1046ms 73.5683μs 13.5928 KOps/s 13.3435 KOps/s $\color{#35bf28}+1.87\%$
test_exclude_nested 0.1296ms 91.0660μs 10.9810 KOps/s 10.8913 KOps/s $\color{#35bf28}+0.82\%$
test_empty[True] 0.4395ms 0.3852ms 2.5962 KOps/s 2.5901 KOps/s $\color{#35bf28}+0.23\%$
test_empty[False] 8.0527μs 1.3154μs 760.2506 KOps/s 760.4694 KOps/s $\color{#d91a1a}-0.03\%$
test_to 0.1087ms 71.7755μs 13.9323 KOps/s 13.3363 KOps/s $\color{#35bf28}+4.47\%$
test_to_nonblocking 0.1133ms 68.7327μs 14.5491 KOps/s 14.7266 KOps/s $\color{#d91a1a}-1.21\%$
test_unbind_speed 0.3775ms 0.3296ms 3.0337 KOps/s 3.0390 KOps/s $\color{#d91a1a}-0.17\%$
test_unbind_speed_stack0 0.3732ms 0.3242ms 3.0843 KOps/s 3.0753 KOps/s $\color{#35bf28}+0.29\%$
test_unbind_speed_stack1 0.1022s 0.8281ms 1.2076 KOps/s 1.1946 KOps/s $\color{#35bf28}+1.09\%$
test_split 0.1023s 1.2631ms 791.6902 Ops/s 789.9048 Ops/s $\color{#35bf28}+0.23\%$
test_chunk 0.1024s 1.2056ms 829.4615 Ops/s 915.1872 Ops/s $\textbf{\color{#d91a1a}-9.37\%}$
test_to_cpu_blocking 19.8321ms 19.2087ms 52.0597 Ops/s 46.1222 Ops/s $\textbf{\color{#35bf28}+12.87\%}$
test_to_cpu_global_sync 11.4365ms 11.0983ms 90.1043 Ops/s 88.1775 Ops/s $\color{#35bf28}+2.19\%$
test_to_cpu_event_sync 12.4135ms 12.1039ms 82.6177 Ops/s 81.1133 Ops/s $\color{#35bf28}+1.85\%$
test_to_cpu_default 0.1145s 13.3243ms 75.0508 Ops/s 81.0927 Ops/s $\textbf{\color{#d91a1a}-7.45\%}$
test_consolidate[False-None] 4.2295ms 4.1235ms 242.5122 Ops/s 217.4216 Ops/s $\textbf{\color{#35bf28}+11.54\%}$
test_consolidate[default-None] 2.1162ms 2.0177ms 495.6026 Ops/s 482.2662 Ops/s $\color{#35bf28}+2.77\%$
test_consolidate[reduce-overhead-None] 2.1131ms 1.9382ms 515.9351 Ops/s 503.5815 Ops/s $\color{#35bf28}+2.45\%$
test_consolidate_njt[False-None] 8.7182ms 8.3809ms 119.3194 Ops/s 116.5894 Ops/s $\color{#35bf28}+2.34\%$
test_to[False-False-None] 2.1532ms 2.0598ms 485.4732 Ops/s 471.6915 Ops/s $\color{#35bf28}+2.92\%$
test_to[True-False-None] 2.1846ms 1.9379ms 516.0333 Ops/s 512.4286 Ops/s $\color{#35bf28}+0.70\%$
test_to[within-False-None] 6.2524ms 6.1396ms 162.8771 Ops/s 161.6566 Ops/s $\color{#35bf28}+0.75\%$
test_to[True-default-None] 8.5281ms 7.6389ms 130.9094 Ops/s 126.5318 Ops/s $\color{#35bf28}+3.46\%$
test_to_njt[False-False-None] 8.8104ms 8.4958ms 117.7053 Ops/s 115.3566 Ops/s $\color{#35bf28}+2.04\%$
test_to_njt[True-False-None] 7.3495ms 6.9298ms 144.3036 Ops/s 142.2568 Ops/s $\color{#35bf28}+1.44\%$
test_to_njt[within-False-None] 16.0480ms 15.3986ms 64.9409 Ops/s 63.6898 Ops/s $\color{#35bf28}+1.96\%$
test_creation[device0] 0.5182ms 0.1177ms 8.4945 KOps/s 8.4926 KOps/s $\color{#35bf28}+0.02\%$
test_creation_from_tensor 0.3966ms 0.1171ms 8.5374 KOps/s 8.6707 KOps/s $\color{#d91a1a}-1.54\%$
test_add_one[memmap_tensor0] 0.2835ms 6.5006μs 153.8308 KOps/s 143.9498 KOps/s $\textbf{\color{#35bf28}+6.86\%}$
test_contiguous[memmap_tensor0] 28.3410μs 0.7206μs 1.3876 MOps/s 2.1901 MOps/s $\textbf{\color{#d91a1a}-36.64\%}$
test_stack[memmap_tensor0] 29.2910μs 4.5687μs 218.8783 KOps/s 208.8901 KOps/s $\color{#35bf28}+4.78\%$
test_memmaptd_index 1.0123ms 0.2606ms 3.8368 KOps/s 3.7529 KOps/s $\color{#35bf28}+2.24\%$
test_memmaptd_index_astensor 0.5080ms 0.3561ms 2.8078 KOps/s 2.7493 KOps/s $\color{#35bf28}+2.13\%$
test_memmaptd_index_op 0.9479ms 0.5947ms 1.6814 KOps/s 1.6143 KOps/s $\color{#35bf28}+4.16\%$
test_serialize_model 0.1387s 0.1364s 7.3335 Ops/s 7.3122 Ops/s $\color{#35bf28}+0.29\%$
test_serialize_model_pickle 1.3488s 1.1944s 0.8373 Ops/s 0.8384 Ops/s $\color{#d91a1a}-0.14\%$
test_serialize_weights 0.1387s 0.1362s 7.3407 Ops/s 7.4017 Ops/s $\color{#d91a1a}-0.83\%$
test_serialize_weights_returnearly 0.4302s 96.9814ms 10.3113 Ops/s 5.5473 Ops/s $\textbf{\color{#35bf28}+85.88\%}$
test_serialize_weights_pickle 1.3573s 1.1966s 0.8357 Ops/s 0.8245 Ops/s $\color{#35bf28}+1.36\%$
test_reshape_pytree 0.2215ms 33.7552μs 29.6250 KOps/s 29.6130 KOps/s $\color{#35bf28}+0.04\%$
test_reshape_td 85.0530μs 44.6330μs 22.4049 KOps/s 22.1919 KOps/s $\color{#35bf28}+0.96\%$
test_view_pytree 0.2218ms 33.4002μs 29.9399 KOps/s 30.4159 KOps/s $\color{#d91a1a}-1.56\%$
test_view_td 91.7040μs 52.9345μs 18.8913 KOps/s 19.5012 KOps/s $\color{#d91a1a}-3.13\%$
test_unbind_pytree 0.2321ms 36.8571μs 27.1318 KOps/s 26.7276 KOps/s $\color{#35bf28}+1.51\%$
test_unbind_td 0.1020ms 48.5425μs 20.6005 KOps/s 20.4710 KOps/s $\color{#35bf28}+0.63\%$
test_split_pytree 0.1906ms 43.0699μs 23.2180 KOps/s 23.4192 KOps/s $\color{#d91a1a}-0.86\%$
test_split_td 0.1936ms 65.1516μs 15.3488 KOps/s 15.4393 KOps/s $\color{#d91a1a}-0.59\%$
test_add_pytree 0.2374ms 42.3545μs 23.6103 KOps/s 23.6453 KOps/s $\color{#d91a1a}-0.15\%$
test_add_td 0.2134ms 53.8319μs 18.5763 KOps/s 18.6502 KOps/s $\color{#d91a1a}-0.40\%$
test_compile_add_one_nested[tensordict-compile] 0.2129ms 0.1403ms 7.1264 KOps/s 6.6401 KOps/s $\textbf{\color{#35bf28}+7.32\%}$
test_compile_add_one_nested[tensordict-eager] 0.2718ms 0.1963ms 5.0941 KOps/s 5.2311 KOps/s $\color{#d91a1a}-2.62\%$
test_compile_add_one_nested[pytree-compile] 0.1597ms 0.1104ms 9.0542 KOps/s 8.9373 KOps/s $\color{#35bf28}+1.31\%$
test_compile_add_one_nested[pytree-eager] 0.4456ms 0.1796ms 5.5679 KOps/s 5.4512 KOps/s $\color{#35bf28}+2.14\%$
test_compile_copy_nested[tensordict-compile] 0.3555ms 29.8620μs 33.4873 KOps/s 28.4236 KOps/s $\textbf{\color{#35bf28}+17.82\%}$
test_compile_copy_nested[tensordict-eager] 89.3340μs 51.9984μs 19.2314 KOps/s 19.2598 KOps/s $\color{#d91a1a}-0.15\%$
test_compile_copy_nested[pytree-compile] 0.1057ms 9.7255μs 102.8224 KOps/s 102.1451 KOps/s $\color{#35bf28}+0.66\%$
test_compile_copy_nested[pytree-eager] 0.4703ms 70.4441μs 14.1957 KOps/s 14.6081 KOps/s $\color{#d91a1a}-2.82\%$
test_compile_add_one_flat[tensordict-compile] 0.6160ms 0.1793ms 5.5759 KOps/s 5.4054 KOps/s $\color{#35bf28}+3.16\%$
test_compile_add_one_flat[tensordict-eager] 0.3326ms 0.2537ms 3.9424 KOps/s 3.9055 KOps/s $\color{#35bf28}+0.94\%$
test_compile_add_one_flat[tensorclass-compile] 0.2666ms 0.1192ms 8.3921 KOps/s 8.3069 KOps/s $\color{#35bf28}+1.03\%$
test_compile_add_one_flat[tensorclass-eager] 0.1116ms 69.6970μs 14.3478 KOps/s 14.2693 KOps/s $\color{#35bf28}+0.55\%$
test_compile_add_one_flat[pytree-compile] 0.2271ms 0.1602ms 6.2441 KOps/s 5.9465 KOps/s $\textbf{\color{#35bf28}+5.00\%}$
test_compile_add_one_flat[pytree-eager] 0.8426ms 0.5265ms 1.8994 KOps/s 1.8345 KOps/s $\color{#35bf28}+3.54\%$
test_compile_add_self_flat[tensordict-eager] 0.4615ms 0.3076ms 3.2507 KOps/s 3.2324 KOps/s $\color{#35bf28}+0.57\%$
test_compile_add_self_flat[tensordict-compile] 0.2276ms 0.1820ms 5.4955 KOps/s 5.1616 KOps/s $\textbf{\color{#35bf28}+6.47\%}$
test_compile_add_self_flat[tensorclass-eager] 0.1371ms 85.6135μs 11.6804 KOps/s 11.1108 KOps/s $\textbf{\color{#35bf28}+5.13\%}$
test_compile_add_self_flat[tensorclass-compile] 0.2966ms 0.1209ms 8.2739 KOps/s 7.7618 KOps/s $\textbf{\color{#35bf28}+6.60\%}$
test_compile_add_self_flat[pytree-eager] 0.6474ms 0.4399ms 2.2734 KOps/s 2.2125 KOps/s $\color{#35bf28}+2.75\%$
test_compile_add_self_flat[pytree-compile] 0.3681ms 0.2022ms 4.9464 KOps/s 6.0695 KOps/s $\textbf{\color{#d91a1a}-18.50\%}$
test_compile_copy_flat[tensordict-compile] 88.7240μs 23.9819μs 41.6982 KOps/s 39.5151 KOps/s $\textbf{\color{#35bf28}+5.52\%}$
test_compile_copy_flat[tensordict-eager] 71.8030μs 40.8852μs 24.4587 KOps/s 24.5233 KOps/s $\color{#d91a1a}-0.26\%$
test_compile_copy_flat[pytree-compile] 61.3830μs 10.7887μs 92.6897 KOps/s 89.6513 KOps/s $\color{#35bf28}+3.39\%$
test_compile_copy_flat[pytree-eager] 0.4000ms 52.1933μs 19.1595 KOps/s 19.2858 KOps/s $\color{#d91a1a}-0.65\%$
test_compile_assign_and_add[tensordict-compile] 2.0075ms 0.1754ms 5.7011 KOps/s 5.4215 KOps/s $\textbf{\color{#35bf28}+5.16\%}$
test_compile_assign_and_add[tensordict-eager] 3.3956ms 3.2870ms 304.2250 Ops/s 298.5702 Ops/s $\color{#35bf28}+1.89\%$
test_compile_assign_and_add[pytree-compile] 1.9746ms 0.1613ms 6.1996 KOps/s 6.0536 KOps/s $\color{#35bf28}+2.41\%$
test_compile_assign_and_add[pytree-eager] 2.9280ms 2.7986ms 357.3216 Ops/s 349.9847 Ops/s $\color{#35bf28}+2.10\%$
test_compile_indexing[tensor-tensordict-compile] 0.1615ms 0.1105ms 9.0504 KOps/s 8.7644 KOps/s $\color{#35bf28}+3.26\%$
test_compile_indexing[tensor-tensordict-eager] 0.3143ms 73.5253μs 13.6008 KOps/s 13.6185 KOps/s $\color{#d91a1a}-0.13\%$
test_compile_indexing[tensor-tensorclass-compile] 0.2100ms 97.9658μs 10.2076 KOps/s 10.1622 KOps/s $\color{#35bf28}+0.45\%$
test_compile_indexing[tensor-tensorclass-eager] 0.2601ms 45.3055μs 22.0724 KOps/s 21.9002 KOps/s $\color{#35bf28}+0.79\%$
test_compile_indexing[tensor-pytree-compile] 0.1402ms 98.9369μs 10.1075 KOps/s 10.0320 KOps/s $\color{#35bf28}+0.75\%$
test_compile_indexing[tensor-pytree-eager] 0.2490ms 46.3158μs 21.5909 KOps/s 21.7667 KOps/s $\color{#d91a1a}-0.81\%$
test_compile_indexing[slice-tensordict-compile] 0.2053ms 55.9144μs 17.8845 KOps/s 17.3660 KOps/s $\color{#35bf28}+2.99\%$
test_compile_indexing[slice-tensordict-eager] 0.2274ms 27.6416μs 36.1773 KOps/s 35.3919 KOps/s $\color{#35bf28}+2.22\%$
test_compile_indexing[slice-tensorclass-compile] 0.1453ms 44.6438μs 22.3995 KOps/s 21.5559 KOps/s $\color{#35bf28}+3.91\%$
test_compile_indexing[slice-tensorclass-eager] 0.2524ms 22.5551μs 44.3359 KOps/s 44.9405 KOps/s $\color{#d91a1a}-1.35\%$
test_compile_indexing[slice-pytree-compile] 87.1540μs 46.0052μs 21.7367 KOps/s 21.4508 KOps/s $\color{#35bf28}+1.33\%$
test_compile_indexing[slice-pytree-eager] 0.2940ms 22.4969μs 44.4506 KOps/s 43.5880 KOps/s $\color{#35bf28}+1.98\%$
test_compile_indexing[int-tensordict-compile] 95.9840μs 56.8918μs 17.5772 KOps/s 16.8052 KOps/s $\color{#35bf28}+4.59\%$
test_compile_indexing[int-tensordict-eager] 0.2344ms 26.9996μs 37.0376 KOps/s 35.9132 KOps/s $\color{#35bf28}+3.13\%$
test_compile_indexing[int-tensorclass-compile] 81.3130μs 45.2014μs 22.1232 KOps/s 21.8865 KOps/s $\color{#35bf28}+1.08\%$
test_compile_indexing[int-tensorclass-eager] 0.2810ms 22.1404μs 45.1663 KOps/s 43.8828 KOps/s $\color{#35bf28}+2.92\%$
test_compile_indexing[int-pytree-compile] 94.2130μs 45.8146μs 21.8271 KOps/s 21.3627 KOps/s $\color{#35bf28}+2.17\%$
test_compile_indexing[int-pytree-eager] 0.2556ms 22.4322μs 44.5788 KOps/s 43.7712 KOps/s $\color{#35bf28}+1.85\%$
test_mod_add[eager] 0.1059ms 51.5836μs 19.3860 KOps/s 19.5037 KOps/s $\color{#d91a1a}-0.60\%$
test_mod_add[compile] 0.1525ms 0.1046ms 9.5617 KOps/s 9.3364 KOps/s $\color{#35bf28}+2.41\%$
test_mod_add[compile-overhead] 0.2357ms 0.1492ms 6.7011 KOps/s 6.6629 KOps/s $\color{#35bf28}+0.57\%$
test_mod_wrap[eager] 0.4789ms 0.2890ms 3.4599 KOps/s 3.4137 KOps/s $\color{#35bf28}+1.35\%$
test_mod_wrap[compile] 0.4018ms 0.3498ms 2.8588 KOps/s 2.8284 KOps/s $\color{#35bf28}+1.07\%$
test_mod_wrap[compile-overhead] 7.2653ms 4.0200ms 248.7574 Ops/s 247.9600 Ops/s $\color{#35bf28}+0.32\%$
test_mod_wrap_and_backward[eager] 1.6050ms 1.4969ms 668.0256 Ops/s 617.4412 Ops/s $\textbf{\color{#35bf28}+8.19\%}$
test_mod_wrap_and_backward[compile] 1.5861ms 1.4480ms 690.5850 Ops/s 633.9266 Ops/s $\textbf{\color{#35bf28}+8.94\%}$
test_mod_wrap_and_backward[compile-overhead] 1.2664ms 0.8915ms 1.1217 KOps/s 990.9671 Ops/s $\textbf{\color{#35bf28}+13.19\%}$
test_seq_add[eager] 0.2474ms 0.1539ms 6.4979 KOps/s 6.3050 KOps/s $\color{#35bf28}+3.06\%$
test_seq_add[compile] 0.1915ms 0.1145ms 8.7323 KOps/s 8.4077 KOps/s $\color{#35bf28}+3.86\%$
test_seq_add[compile-overhead] 0.4342ms 0.1552ms 6.4439 KOps/s 6.3191 KOps/s $\color{#35bf28}+1.98\%$
test_seq_wrap[eager] 0.8039ms 0.5241ms 1.9081 KOps/s 1.8932 KOps/s $\color{#35bf28}+0.79\%$
test_seq_wrap[compile] 0.4395ms 0.3691ms 2.7095 KOps/s 2.6805 KOps/s $\color{#35bf28}+1.08\%$
test_seq_wrap[compile-overhead] 0.3315ms 0.2681ms 3.7305 KOps/s 3.6897 KOps/s $\color{#35bf28}+1.11\%$
test_func_call_runtime[False-eager] 0.9153ms 0.8345ms 1.1983 KOps/s 1.1812 KOps/s $\color{#35bf28}+1.45\%$
test_func_call_runtime[False-compile] 1.0129ms 0.9049ms 1.1051 KOps/s 1.0795 KOps/s $\color{#35bf28}+2.37\%$
test_func_call_runtime[False-compile-overhead] 0.5154ms 0.4629ms 2.1603 KOps/s 2.1382 KOps/s $\color{#35bf28}+1.03\%$
test_func_call_runtime[True-eager] 1.1503ms 1.0814ms 924.7372 Ops/s 914.9211 Ops/s $\color{#35bf28}+1.07\%$
test_func_call_runtime[True-compile] 0.9926ms 0.9208ms 1.0860 KOps/s 1.0709 KOps/s $\color{#35bf28}+1.41\%$
test_func_call_runtime[True-compile-overhead] 0.5322ms 0.4736ms 2.1113 KOps/s 2.0764 KOps/s $\color{#35bf28}+1.68\%$
test_func_call_cm_runtime[False-eager] 0.9101ms 0.8366ms 1.1954 KOps/s 1.1422 KOps/s $\color{#35bf28}+4.66\%$
test_func_call_cm_runtime[False-compile] 0.9724ms 0.9148ms 1.0932 KOps/s 1.0788 KOps/s $\color{#35bf28}+1.34\%$
test_func_call_cm_runtime[False-compile-overhead] 0.5046ms 0.4645ms 2.1527 KOps/s 2.1524 KOps/s $\color{#35bf28}+0.01\%$
test_func_call_cm_runtime[True-eager] 1.3233ms 1.2271ms 814.9373 Ops/s 795.9900 Ops/s $\color{#35bf28}+2.38\%$
test_func_call_cm_runtime[True-compile] 1.0046ms 0.9562ms 1.0458 KOps/s 1.0338 KOps/s $\color{#35bf28}+1.16\%$
test_func_call_cm_runtime[True-compile-overhead] 0.5526ms 0.5070ms 1.9722 KOps/s 1.9374 KOps/s $\color{#35bf28}+1.79\%$
test_vmap_func_call_cm_runtime[eager] 2.8961ms 2.3584ms 424.0238 Ops/s 416.8813 Ops/s $\color{#35bf28}+1.71\%$
test_vmap_func_call_cm_runtime[compile] 1.0526ms 0.9696ms 1.0313 KOps/s 1.0094 KOps/s $\color{#35bf28}+2.18\%$
test_vmap_func_call_cm_runtime[compile-overhead] 0.6094ms 0.5125ms 1.9511 KOps/s 1.9110 KOps/s $\color{#35bf28}+2.10\%$
test_distributed 0.7574ms 0.1525ms 6.5593 KOps/s 5.3151 KOps/s $\textbf{\color{#35bf28}+23.41\%}$
test_tdmodule 70.6030μs 27.8621μs 35.8911 KOps/s 34.1749 KOps/s $\textbf{\color{#35bf28}+5.02\%}$
test_tdmodule_dispatch 75.9730μs 44.8616μs 22.2908 KOps/s 21.8685 KOps/s $\color{#35bf28}+1.93\%$
test_tdseq 48.7320μs 26.7128μs 37.4353 KOps/s 36.3320 KOps/s $\color{#35bf28}+3.04\%$
test_tdseq_dispatch 70.9630μs 46.7330μs 21.3981 KOps/s 21.1304 KOps/s $\color{#35bf28}+1.27\%$
test_instantiation_functorch 2.1501ms 2.0590ms 485.6834 Ops/s 481.0666 Ops/s $\color{#35bf28}+0.96\%$
test_exec_functorch 0.2226ms 0.1797ms 5.5644 KOps/s 5.4898 KOps/s $\color{#35bf28}+1.36\%$
test_exec_functional_call 0.2100ms 0.1601ms 6.2453 KOps/s 5.9868 KOps/s $\color{#35bf28}+4.32\%$
test_exec_td_decorator 0.4512ms 0.2366ms 4.2262 KOps/s 4.1509 KOps/s $\color{#35bf28}+1.81\%$
test_vmap_mlp_speed_decorator[True-True] 1.0230ms 0.8216ms 1.2171 KOps/s 1.2043 KOps/s $\color{#35bf28}+1.07\%$
test_vmap_mlp_speed_decorator[True-False] 1.0210ms 0.8193ms 1.2206 KOps/s 1.2058 KOps/s $\color{#35bf28}+1.23\%$
test_vmap_mlp_speed_decorator[False-True] 0.9368ms 0.7107ms 1.4071 KOps/s 1.3931 KOps/s $\color{#35bf28}+1.00\%$
test_vmap_mlp_speed_decorator[False-False] 0.8725ms 0.7107ms 1.4070 KOps/s 1.3896 KOps/s $\color{#35bf28}+1.25\%$
test_vmap_transformer_speed_decorator[True-True] 20.7133ms 20.5733ms 48.6068 Ops/s 48.0981 Ops/s $\color{#35bf28}+1.06\%$
test_vmap_transformer_speed_decorator[True-False] 21.1048ms 20.5667ms 48.6222 Ops/s 48.1154 Ops/s $\color{#35bf28}+1.05\%$
test_vmap_transformer_speed_decorator[False-True] 21.1090ms 20.3921ms 49.0386 Ops/s 48.6548 Ops/s $\color{#35bf28}+0.79\%$
test_vmap_transformer_speed_decorator[False-False] 21.1123ms 20.4244ms 48.9610 Ops/s 48.5828 Ops/s $\color{#35bf28}+0.78\%$
test_to_module_speed[True] 2.0698ms 1.4793ms 676.0154 Ops/s 681.8290 Ops/s $\color{#d91a1a}-0.85\%$
test_to_module_speed[False] 2.0180ms 1.4567ms 686.4857 Ops/s 691.4065 Ops/s $\color{#d91a1a}-0.71\%$
test_tc_init 81.0240μs 45.6470μs 21.9073 KOps/s 21.6690 KOps/s $\color{#35bf28}+1.10\%$
test_tc_init_tensor_only 43.9410μs 9.8686μs 101.3312 KOps/s 101.2610 KOps/s $\color{#35bf28}+0.07\%$
test_tc_init_nested 0.1707ms 91.2324μs 10.9610 KOps/s 10.9628 KOps/s $\color{#d91a1a}-0.02\%$
test_tc_init_many_fields 48.3120μs 16.3633μs 61.1124 KOps/s 60.1048 KOps/s $\color{#35bf28}+1.68\%$
test_tc_first_layer_tensor 32.2210μs 1.8536μs 539.4980 KOps/s 539.8701 KOps/s $\color{#d91a1a}-0.07\%$
test_tc_first_layer_tensor_only 5.2446μs 0.7598μs 1.3161 MOps/s 1.3260 MOps/s $\color{#d91a1a}-0.75\%$
test_tc_first_layer_tensor_set 29.6110μs 3.9457μs 253.4429 KOps/s 252.7718 KOps/s $\color{#35bf28}+0.27\%$
test_tc_first_layer_tensor_only_set 20.9700μs 3.1539μs 317.0651 KOps/s 312.5725 KOps/s $\color{#35bf28}+1.44\%$
test_tc_first_layer_nontensor 33.3720μs 6.1805μs 161.8005 KOps/s 161.6718 KOps/s $\color{#35bf28}+0.08\%$
test_tc_second_layer_tensor 32.4310μs 4.3612μs 229.2933 KOps/s 225.1664 KOps/s $\color{#35bf28}+1.83\%$
test_tc_second_layer_nontensor 43.8610μs 8.7936μs 113.7186 KOps/s 113.2073 KOps/s $\color{#35bf28}+0.45\%$
test_unbind 0.2582s 15.9080ms 62.8615 Ops/s 66.9343 Ops/s $\textbf{\color{#d91a1a}-6.08\%}$
test_full_like 4.5070ms 4.3848ms 228.0597 Ops/s 229.5585 Ops/s $\color{#d91a1a}-0.65\%$
test_zeros_like 4.4870ms 4.3696ms 228.8514 Ops/s 228.4059 Ops/s $\color{#35bf28}+0.20\%$
test_ones_like 5.0486ms 4.3900ms 227.7895 Ops/s 227.9577 Ops/s $\color{#d91a1a}-0.07\%$
test_clone 6.7480ms 6.5238ms 153.2838 Ops/s 153.1493 Ops/s $\color{#35bf28}+0.09\%$
test_squeeze 0.1952ms 14.5364μs 68.7927 KOps/s 70.4406 KOps/s $\color{#d91a1a}-2.34\%$
test_unsqueeze 0.1644ms 0.1138ms 8.7887 KOps/s 9.0613 KOps/s $\color{#d91a1a}-3.01\%$
test_split 0.2404ms 0.1826ms 5.4759 KOps/s 5.4191 KOps/s $\color{#35bf28}+1.05\%$
test_permute 0.2693ms 0.2123ms 4.7094 KOps/s 4.8811 KOps/s $\color{#d91a1a}-3.52\%$
test_stack 52.0843ms 51.4912ms 19.4208 Ops/s 19.4674 Ops/s $\color{#d91a1a}-0.24\%$
test_cat 43.5106ms 43.1358ms 23.1826 Ops/s 19.5025 Ops/s $\textbf{\color{#35bf28}+18.87\%}$

@github-actions

github-actions Bot commented Feb 19, 2026

Copy link
Copy Markdown
Contributor

$\color{#D29922}\textsf{\Large&amp;#x26A0;\kern{0.2cm}\normalsize Warning}$ Result of GPU Benchmark Tests

Total Benchmarks: 243. Improved: $\large\color{#35bf28}34$. Worsened: $\large\color{#d91a1a}7$.

Expand to view detailed results
Name Max Mean Ops Ops on Repo HEAD Change
test_plain_set_nested 34.7610μs 14.8999μs 67.1146 KOps/s 67.1795 KOps/s $\color{#d91a1a}-0.10\%$
test_plain_set_stack_nested 34.6710μs 15.1696μs 65.9212 KOps/s 66.9848 KOps/s $\color{#d91a1a}-1.59\%$
test_plain_set_nested_inplace 55.5300μs 16.5605μs 60.3846 KOps/s 61.5538 KOps/s $\color{#d91a1a}-1.90\%$
test_plain_set_stack_nested_inplace 40.7800μs 16.3852μs 61.0307 KOps/s 61.2231 KOps/s $\color{#d91a1a}-0.31\%$
test_items 33.1710μs 5.8521μs 170.8787 KOps/s 166.8869 KOps/s $\color{#35bf28}+2.39\%$
test_items_nested 0.5257ms 0.4664ms 2.1442 KOps/s 2.2100 KOps/s $\color{#d91a1a}-2.98\%$
test_items_nested_locked 0.5536ms 0.4720ms 2.1185 KOps/s 2.1797 KOps/s $\color{#d91a1a}-2.81\%$
test_items_nested_leaf 0.1541ms 95.5178μs 10.4693 KOps/s 10.1539 KOps/s $\color{#35bf28}+3.11\%$
test_items_stack_nested 0.5368ms 0.4676ms 2.1387 KOps/s 2.1739 KOps/s $\color{#d91a1a}-1.62\%$
test_items_stack_nested_leaf 0.1337ms 97.3353μs 10.2738 KOps/s 10.0670 KOps/s $\color{#35bf28}+2.05\%$
test_items_stack_nested_locked 0.5370ms 0.4710ms 2.1230 KOps/s 2.1463 KOps/s $\color{#d91a1a}-1.08\%$
test_keys 27.1210μs 4.4406μs 225.1956 KOps/s 235.6038 KOps/s $\color{#d91a1a}-4.42\%$
test_keys_nested 0.1756ms 0.1208ms 8.2811 KOps/s 8.2113 KOps/s $\color{#35bf28}+0.85\%$
test_keys_nested_locked 89.3855ms 0.1423ms 7.0261 KOps/s 7.6905 KOps/s $\textbf{\color{#d91a1a}-8.64\%}$
test_keys_nested_leaf 0.1589ms 0.1109ms 9.0138 KOps/s 8.9734 KOps/s $\color{#35bf28}+0.45\%$
test_keys_stack_nested 0.1904ms 0.1211ms 8.2559 KOps/s 8.3648 KOps/s $\color{#d91a1a}-1.30\%$
test_keys_stack_nested_leaf 0.1519ms 0.1117ms 8.9547 KOps/s 8.9943 KOps/s $\color{#d91a1a}-0.44\%$
test_keys_stack_nested_locked 0.2315ms 0.1298ms 7.7046 KOps/s 7.7711 KOps/s $\color{#d91a1a}-0.86\%$
test_values 7.0240μs 1.0249μs 975.7283 KOps/s 972.0673 KOps/s $\color{#35bf28}+0.38\%$
test_values_nested 79.6110μs 47.8616μs 20.8936 KOps/s 20.8334 KOps/s $\color{#35bf28}+0.29\%$
test_values_nested_locked 0.1054ms 51.1591μs 19.5469 KOps/s 19.4535 KOps/s $\color{#35bf28}+0.48\%$
test_values_nested_leaf 93.1810μs 54.8024μs 18.2474 KOps/s 18.2733 KOps/s $\color{#d91a1a}-0.14\%$
test_values_stack_nested 81.2010μs 47.9621μs 20.8498 KOps/s 20.7363 KOps/s $\color{#35bf28}+0.55\%$
test_values_stack_nested_leaf 91.4110μs 54.6136μs 18.3105 KOps/s 18.2365 KOps/s $\color{#35bf28}+0.41\%$
test_values_stack_nested_locked 0.7617ms 50.7269μs 19.7134 KOps/s 19.7640 KOps/s $\color{#d91a1a}-0.26\%$
test_membership 5.7135μs 0.8559μs 1.1684 MOps/s 1.1778 MOps/s $\color{#d91a1a}-0.80\%$
test_membership_nested 37.5300μs 2.9472μs 339.3038 KOps/s 343.5003 KOps/s $\color{#d91a1a}-1.22\%$
test_membership_nested_leaf 27.3400μs 2.9037μs 344.3875 KOps/s 339.5415 KOps/s $\color{#35bf28}+1.43\%$
test_membership_stacked_nested 40.0400μs 2.9062μs 344.0949 KOps/s 341.3478 KOps/s $\color{#35bf28}+0.80\%$
test_membership_stacked_nested_leaf 50.4400μs 2.8730μs 348.0679 KOps/s 341.0795 KOps/s $\color{#35bf28}+2.05\%$
test_membership_nested_last 32.2300μs 4.3218μs 231.3851 KOps/s 230.3992 KOps/s $\color{#35bf28}+0.43\%$
test_membership_nested_leaf_last 29.8900μs 4.3655μs 229.0703 KOps/s 230.9715 KOps/s $\color{#d91a1a}-0.82\%$
test_membership_stacked_nested_last 33.7400μs 4.3660μs 229.0411 KOps/s 229.7004 KOps/s $\color{#d91a1a}-0.29\%$
test_membership_stacked_nested_leaf_last 44.3300μs 4.3619μs 229.2587 KOps/s 227.7654 KOps/s $\color{#35bf28}+0.66\%$
test_nested_getleaf 50.3800μs 21.3124μs 46.9211 KOps/s 46.1213 KOps/s $\color{#35bf28}+1.73\%$
test_nested_get 49.5400μs 20.3080μs 49.2416 KOps/s 48.1395 KOps/s $\color{#35bf28}+2.29\%$
test_stacked_getleaf 65.6110μs 21.0346μs 47.5407 KOps/s 46.2798 KOps/s $\color{#35bf28}+2.72\%$
test_stacked_get 46.5510μs 20.1005μs 49.7500 KOps/s 48.2966 KOps/s $\color{#35bf28}+3.01\%$
test_nested_getitemleaf 48.9100μs 22.1163μs 45.2156 KOps/s 44.5843 KOps/s $\color{#35bf28}+1.42\%$
test_nested_getitem 75.8010μs 20.3919μs 49.0392 KOps/s 47.3259 KOps/s $\color{#35bf28}+3.62\%$
test_stacked_getitemleaf 48.8410μs 22.3203μs 44.8023 KOps/s 44.9786 KOps/s $\color{#d91a1a}-0.39\%$
test_stacked_getitem 57.5910μs 20.9414μs 47.7522 KOps/s 46.2569 KOps/s $\color{#35bf28}+3.23\%$
test_lock_nested 7.8134ms 0.4844ms 2.0644 KOps/s 2.0771 KOps/s $\color{#d91a1a}-0.61\%$
test_lock_stack_nested 0.5614ms 0.4837ms 2.0673 KOps/s 2.0451 KOps/s $\color{#35bf28}+1.09\%$
test_unlock_nested 0.5428ms 0.3896ms 2.5664 KOps/s 2.5797 KOps/s $\color{#d91a1a}-0.52\%$
test_unlock_stack_nested 0.4553ms 0.3900ms 2.5638 KOps/s 2.5315 KOps/s $\color{#35bf28}+1.28\%$
test_flatten_speed 0.1602ms 0.1223ms 8.1776 KOps/s 8.1146 KOps/s $\color{#35bf28}+0.78\%$
test_unflatten_speed 0.6520ms 0.5828ms 1.7160 KOps/s 1.7235 KOps/s $\color{#d91a1a}-0.43\%$
test_common_ops 0.8045ms 0.6883ms 1.4529 KOps/s 1.4273 KOps/s $\color{#35bf28}+1.79\%$
test_creation 0.1139ms 2.9289μs 341.4240 KOps/s 338.7648 KOps/s $\color{#35bf28}+0.78\%$
test_creation_empty 83.5510μs 6.1857μs 161.6635 KOps/s 159.6473 KOps/s $\color{#35bf28}+1.26\%$
test_creation_nested_1 35.0700μs 10.5267μs 94.9963 KOps/s 94.3057 KOps/s $\color{#35bf28}+0.73\%$
test_creation_nested_2 44.5300μs 11.9199μs 83.8932 KOps/s 84.2547 KOps/s $\color{#d91a1a}-0.43\%$
test_creation_many_keys[10] 41.0700μs 18.2131μs 54.9056 KOps/s 54.9536 KOps/s $\color{#d91a1a}-0.09\%$
test_creation_many_keys[50] 0.1236ms 78.5643μs 12.7284 KOps/s 12.8380 KOps/s $\color{#d91a1a}-0.85\%$
test_creation_many_keys[100] 0.1931ms 0.1526ms 6.5514 KOps/s 6.5062 KOps/s $\color{#35bf28}+0.69\%$
test_creation_nested_many_keys[10] 71.1800μs 38.9000μs 25.7069 KOps/s 25.4010 KOps/s $\color{#35bf28}+1.20\%$
test_creation_nested_many_keys[50] 0.2477ms 0.1585ms 6.3075 KOps/s 6.2623 KOps/s $\color{#35bf28}+0.72\%$
test_clone 49.6110μs 13.7440μs 72.7591 KOps/s 72.4834 KOps/s $\color{#35bf28}+0.38\%$
test_getitem[int] 1.5584ms 14.6787μs 68.1261 KOps/s 55.7043 KOps/s $\textbf{\color{#35bf28}+22.30\%}$
test_getitem[slice_int] 0.1527ms 24.9206μs 40.1274 KOps/s 40.4040 KOps/s $\color{#d91a1a}-0.68\%$
test_getitem[range] 0.1874ms 64.2411μs 15.5664 KOps/s 16.0253 KOps/s $\color{#d91a1a}-2.86\%$
test_getitem[tuple] 0.1436ms 24.3651μs 41.0423 KOps/s 41.3800 KOps/s $\color{#d91a1a}-0.82\%$
test_getitem[list] 0.1869ms 61.2321μs 16.3313 KOps/s 17.2177 KOps/s $\textbf{\color{#d91a1a}-5.15\%}$
test_setitem_dim[int] 46.8000μs 26.4616μs 37.7907 KOps/s 38.1144 KOps/s $\color{#d91a1a}-0.85\%$
test_setitem_dim[slice_int] 64.5410μs 43.6957μs 22.8856 KOps/s 22.6883 KOps/s $\color{#35bf28}+0.87\%$
test_setitem_dim[range] 0.1535ms 95.2013μs 10.5041 KOps/s 10.5888 KOps/s $\color{#d91a1a}-0.80\%$
test_setitem_dim[tuple] 65.5100μs 40.8158μs 24.5003 KOps/s 23.2731 KOps/s $\textbf{\color{#35bf28}+5.27\%}$
test_setitem 47.9300μs 17.9864μs 55.5976 KOps/s 54.7872 KOps/s $\color{#35bf28}+1.48\%$
test_set 55.6010μs 17.2260μs 58.0516 KOps/s 57.4615 KOps/s $\color{#35bf28}+1.03\%$
test_set_shared 0.4877ms 0.2120ms 4.7175 KOps/s 4.8207 KOps/s $\color{#d91a1a}-2.14\%$
test_update 0.3139ms 22.0117μs 45.4304 KOps/s 45.6800 KOps/s $\color{#d91a1a}-0.55\%$
test_update_nested 70.8400μs 34.4056μs 29.0651 KOps/s 29.2628 KOps/s $\color{#d91a1a}-0.68\%$
test_update__nested 0.4793ms 35.0967μs 28.4927 KOps/s 28.8732 KOps/s $\color{#d91a1a}-1.32\%$
test_set_nested 61.8800μs 19.0087μs 52.6074 KOps/s 51.9517 KOps/s $\color{#35bf28}+1.26\%$
test_set_nested_new 60.9610μs 24.4265μs 40.9391 KOps/s 40.4587 KOps/s $\color{#35bf28}+1.19\%$
test_select 77.3600μs 42.2783μs 23.6528 KOps/s 23.4647 KOps/s $\color{#35bf28}+0.80\%$
test_select_nested 0.1248ms 73.5514μs 13.5959 KOps/s 13.4008 KOps/s $\color{#35bf28}+1.46\%$
test_exclude_nested 0.1322ms 91.3709μs 10.9444 KOps/s 10.7003 KOps/s $\color{#35bf28}+2.28\%$
test_empty[True] 0.5559ms 0.3899ms 2.5646 KOps/s 2.5269 KOps/s $\color{#35bf28}+1.49\%$
test_empty[False] 8.2950μs 1.3304μs 751.6785 KOps/s 743.0199 KOps/s $\color{#35bf28}+1.17\%$
test_to 0.1041ms 73.3037μs 13.6419 KOps/s 13.6583 KOps/s $\color{#d91a1a}-0.12\%$
test_to_nonblocking 0.1200ms 68.8183μs 14.5310 KOps/s 15.4137 KOps/s $\textbf{\color{#d91a1a}-5.73\%}$
test_unbind_speed 0.3952ms 0.3291ms 3.0388 KOps/s 2.9975 KOps/s $\color{#35bf28}+1.38\%$
test_unbind_speed_stack0 0.4010ms 0.3285ms 3.0442 KOps/s 3.0513 KOps/s $\color{#d91a1a}-0.23\%$
test_unbind_speed_stack1 0.1035s 0.8445ms 1.1841 KOps/s 1.1832 KOps/s $\color{#35bf28}+0.08\%$
test_split 0.1040s 1.2751ms 784.2295 Ops/s 785.3581 Ops/s $\color{#d91a1a}-0.14\%$
test_chunk 0.1037s 1.2220ms 818.3286 Ops/s 914.1377 Ops/s $\textbf{\color{#d91a1a}-10.48\%}$
test_to_cpu_blocking 19.8716ms 19.4248ms 51.4805 Ops/s 42.8767 Ops/s $\textbf{\color{#35bf28}+20.07\%}$
test_to_cpu_global_sync 11.7240ms 11.3437ms 88.1545 Ops/s 88.7982 Ops/s $\color{#d91a1a}-0.72\%$
test_to_cpu_event_sync 0.1153s 13.5581ms 73.7564 Ops/s 81.5792 Ops/s $\textbf{\color{#d91a1a}-9.59\%}$
test_to_cpu_default 12.8276ms 12.3399ms 81.0377 Ops/s 81.2674 Ops/s $\color{#d91a1a}-0.28\%$
test_consolidate[False-None] 4.3242ms 4.1389ms 241.6092 Ops/s 216.0457 Ops/s $\textbf{\color{#35bf28}+11.83\%}$
test_consolidate[default-None] 2.4550ms 2.0291ms 492.8295 Ops/s 480.4112 Ops/s $\color{#35bf28}+2.58\%$
test_consolidate[reduce-overhead-None] 2.1048ms 1.9741ms 506.5715 Ops/s 498.3775 Ops/s $\color{#35bf28}+1.64\%$
test_consolidate_njt[False-None] 9.3543ms 8.4436ms 118.4322 Ops/s 118.7653 Ops/s $\color{#d91a1a}-0.28\%$
test_to[False-False-None] 2.4971ms 2.0712ms 482.8219 Ops/s 482.1534 Ops/s $\color{#35bf28}+0.14\%$
test_to[True-False-None] 2.3678ms 1.9502ms 512.7657 Ops/s 516.5660 Ops/s $\color{#d91a1a}-0.74\%$
test_to[within-False-None] 6.5094ms 6.1840ms 161.7067 Ops/s 160.7050 Ops/s $\color{#35bf28}+0.62\%$
test_to[True-default-None] 8.4440ms 7.8419ms 127.5204 Ops/s 127.3403 Ops/s $\color{#35bf28}+0.14\%$
test_to_njt[False-False-None] 8.9948ms 8.6040ms 116.2245 Ops/s 116.1562 Ops/s $\color{#35bf28}+0.06\%$
test_to_njt[True-False-None] 7.2670ms 6.9413ms 144.0646 Ops/s 142.3139 Ops/s $\color{#35bf28}+1.23\%$
test_to_njt[within-False-None] 16.4888ms 15.7098ms 63.6547 Ops/s 63.8790 Ops/s $\color{#d91a1a}-0.35\%$
test_creation[device0] 0.3463ms 0.1220ms 8.1954 KOps/s 8.6363 KOps/s $\textbf{\color{#d91a1a}-5.10\%}$
test_creation_from_tensor 0.5536ms 0.1185ms 8.4404 KOps/s 8.5320 KOps/s $\color{#d91a1a}-1.07\%$
test_add_one[memmap_tensor0] 0.1622ms 6.4665μs 154.6443 KOps/s 149.9910 KOps/s $\color{#35bf28}+3.10\%$
test_contiguous[memmap_tensor0] 28.2900μs 0.6708μs 1.4907 MOps/s 2.1466 MOps/s $\textbf{\color{#d91a1a}-30.55\%}$
test_stack[memmap_tensor0] 30.4000μs 4.5720μs 218.7244 KOps/s 216.3261 KOps/s $\color{#35bf28}+1.11\%$
test_memmaptd_index 0.9726ms 0.2646ms 3.7798 KOps/s 3.7573 KOps/s $\color{#35bf28}+0.60\%$
test_memmaptd_index_astensor 0.5216ms 0.3593ms 2.7829 KOps/s 2.7875 KOps/s $\color{#d91a1a}-0.17\%$
test_memmaptd_index_op 0.7584ms 0.6011ms 1.6636 KOps/s 1.6533 KOps/s $\color{#35bf28}+0.62\%$
test_serialize_model 0.1387s 0.1364s 7.3310 Ops/s 7.2693 Ops/s $\color{#35bf28}+0.85\%$
test_serialize_model_pickle 1.3507s 1.2188s 0.8205 Ops/s 0.8267 Ops/s $\color{#d91a1a}-0.76\%$
test_serialize_weights 0.1384s 0.1361s 7.3497 Ops/s 7.2812 Ops/s $\color{#35bf28}+0.94\%$
test_serialize_weights_returnearly 0.4226s 90.8903ms 11.0023 Ops/s 5.5452 Ops/s $\textbf{\color{#35bf28}+98.41\%}$
test_serialize_weights_pickle 1.3659s 1.1923s 0.8387 Ops/s 0.8184 Ops/s $\color{#35bf28}+2.49\%$
test_reshape_pytree 0.2035ms 33.4064μs 29.9344 KOps/s 29.8496 KOps/s $\color{#35bf28}+0.28\%$
test_reshape_td 70.8200μs 43.9546μs 22.7507 KOps/s 22.5803 KOps/s $\color{#35bf28}+0.75\%$
test_view_pytree 0.2229ms 33.5705μs 29.7880 KOps/s 30.2053 KOps/s $\color{#d91a1a}-1.38\%$
test_view_td 92.4110μs 53.7764μs 18.5955 KOps/s 19.2428 KOps/s $\color{#d91a1a}-3.36\%$
test_unbind_pytree 0.2270ms 37.6260μs 26.5774 KOps/s 26.4499 KOps/s $\color{#35bf28}+0.48\%$
test_unbind_td 0.1752ms 49.9109μs 20.0357 KOps/s 19.9082 KOps/s $\color{#35bf28}+0.64\%$
test_split_pytree 0.2513ms 43.1123μs 23.1952 KOps/s 23.2042 KOps/s $\color{#d91a1a}-0.04\%$
test_split_td 0.1292ms 65.4365μs 15.2820 KOps/s 15.0772 KOps/s $\color{#35bf28}+1.36\%$
test_add_pytree 0.2014ms 42.6000μs 23.4742 KOps/s 23.2386 KOps/s $\color{#35bf28}+1.01\%$
test_add_td 86.2210μs 53.0562μs 18.8479 KOps/s 17.6507 KOps/s $\textbf{\color{#35bf28}+6.78\%}$
test_compile_add_one_nested[tensordict-compile] 0.2514ms 0.1430ms 6.9908 KOps/s 6.5449 KOps/s $\textbf{\color{#35bf28}+6.81\%}$
test_compile_add_one_nested[tensordict-eager] 0.3009ms 0.1967ms 5.0830 KOps/s 5.2267 KOps/s $\color{#d91a1a}-2.75\%$
test_compile_add_one_nested[pytree-compile] 0.1536ms 0.1085ms 9.2204 KOps/s 9.0235 KOps/s $\color{#35bf28}+2.18\%$
test_compile_add_one_nested[pytree-eager] 0.4352ms 0.1828ms 5.4716 KOps/s 5.4574 KOps/s $\color{#35bf28}+0.26\%$
test_compile_copy_nested[tensordict-compile] 0.4912ms 30.7758μs 32.4930 KOps/s 30.2733 KOps/s $\textbf{\color{#35bf28}+7.33\%}$
test_compile_copy_nested[tensordict-eager] 81.6110μs 52.6120μs 19.0071 KOps/s 19.2591 KOps/s $\color{#d91a1a}-1.31\%$
test_compile_copy_nested[pytree-compile] 0.1555ms 9.9216μs 100.7900 KOps/s 101.0982 KOps/s $\color{#d91a1a}-0.30\%$
test_compile_copy_nested[pytree-eager] 0.4682ms 70.0641μs 14.2726 KOps/s 14.1855 KOps/s $\color{#35bf28}+0.61\%$
test_compile_add_one_flat[tensordict-compile] 0.2318ms 0.1781ms 5.6141 KOps/s 5.3333 KOps/s $\textbf{\color{#35bf28}+5.27\%}$
test_compile_add_one_flat[tensordict-eager] 0.3160ms 0.2561ms 3.9046 KOps/s 3.8947 KOps/s $\color{#35bf28}+0.25\%$
test_compile_add_one_flat[tensorclass-compile] 0.2019ms 0.1180ms 8.4729 KOps/s 8.2607 KOps/s $\color{#35bf28}+2.57\%$
test_compile_add_one_flat[tensorclass-eager] 0.1180ms 69.3661μs 14.4163 KOps/s 14.1186 KOps/s $\color{#35bf28}+2.11\%$
test_compile_add_one_flat[pytree-compile] 0.2041ms 0.1591ms 6.2851 KOps/s 3.7877 KOps/s $\textbf{\color{#35bf28}+65.93\%}$
test_compile_add_one_flat[pytree-eager] 0.8047ms 0.5336ms 1.8741 KOps/s 1.8724 KOps/s $\color{#35bf28}+0.09\%$
test_compile_add_self_flat[tensordict-eager] 0.3932ms 0.3083ms 3.2438 KOps/s 3.2112 KOps/s $\color{#35bf28}+1.02\%$
test_compile_add_self_flat[tensordict-compile] 0.3184ms 0.1794ms 5.5728 KOps/s 5.1310 KOps/s $\textbf{\color{#35bf28}+8.61\%}$
test_compile_add_self_flat[tensorclass-eager] 0.1445ms 87.1446μs 11.4752 KOps/s 11.6126 KOps/s $\color{#d91a1a}-1.18\%$
test_compile_add_self_flat[tensorclass-compile] 0.2943ms 0.1212ms 8.2540 KOps/s 7.9857 KOps/s $\color{#35bf28}+3.36\%$
test_compile_add_self_flat[pytree-eager] 0.6524ms 0.4338ms 2.3053 KOps/s 2.2236 KOps/s $\color{#35bf28}+3.67\%$
test_compile_add_self_flat[pytree-compile] 0.8755ms 0.1597ms 6.2600 KOps/s 6.1170 KOps/s $\color{#35bf28}+2.34\%$
test_compile_copy_flat[tensordict-compile] 0.1682ms 24.0712μs 41.5435 KOps/s 38.8067 KOps/s $\textbf{\color{#35bf28}+7.05\%}$
test_compile_copy_flat[tensordict-eager] 88.5210μs 41.8059μs 23.9201 KOps/s 24.1031 KOps/s $\color{#d91a1a}-0.76\%$
test_compile_copy_flat[pytree-compile] 33.6900μs 10.8640μs 92.0473 KOps/s 92.1576 KOps/s $\color{#d91a1a}-0.12\%$
test_compile_copy_flat[pytree-eager] 0.3923ms 52.6081μs 19.0085 KOps/s 18.9303 KOps/s $\color{#35bf28}+0.41\%$
test_compile_assign_and_add[tensordict-compile] 1.9778ms 0.1745ms 5.7322 KOps/s 5.4663 KOps/s $\color{#35bf28}+4.86\%$
test_compile_assign_and_add[tensordict-eager] 3.4265ms 3.2714ms 305.6785 Ops/s 295.8108 Ops/s $\color{#35bf28}+3.34\%$
test_compile_assign_and_add[pytree-compile] 2.0517ms 0.1642ms 6.0900 KOps/s 5.9980 KOps/s $\color{#35bf28}+1.53\%$
test_compile_assign_and_add[pytree-eager] 2.9706ms 2.8187ms 354.7792 Ops/s 354.5315 Ops/s $\color{#35bf28}+0.07\%$
test_compile_indexing[tensor-tensordict-compile] 0.2618ms 0.1088ms 9.1875 KOps/s 8.4913 KOps/s $\textbf{\color{#35bf28}+8.20\%}$
test_compile_indexing[tensor-tensordict-eager] 0.3171ms 73.6188μs 13.5835 KOps/s 12.6739 KOps/s $\textbf{\color{#35bf28}+7.18\%}$
test_compile_indexing[tensor-tensorclass-compile] 0.1902ms 96.8017μs 10.3304 KOps/s 9.9945 KOps/s $\color{#35bf28}+3.36\%$
test_compile_indexing[tensor-tensorclass-eager] 0.2560ms 45.2498μs 22.0995 KOps/s 20.3040 KOps/s $\textbf{\color{#35bf28}+8.84\%}$
test_compile_indexing[tensor-pytree-compile] 0.1624ms 97.0761μs 10.3012 KOps/s 9.9652 KOps/s $\color{#35bf28}+3.37\%$
test_compile_indexing[tensor-pytree-eager] 0.2745ms 47.4081μs 21.0934 KOps/s 20.2891 KOps/s $\color{#35bf28}+3.96\%$
test_compile_indexing[slice-tensordict-compile] 0.2052ms 55.3212μs 18.0762 KOps/s 17.2427 KOps/s $\color{#35bf28}+4.83\%$
test_compile_indexing[slice-tensordict-eager] 0.2268ms 28.4818μs 35.1102 KOps/s 32.2752 KOps/s $\textbf{\color{#35bf28}+8.78\%}$
test_compile_indexing[slice-tensorclass-compile] 0.1406ms 45.5696μs 21.9444 KOps/s 21.7707 KOps/s $\color{#35bf28}+0.80\%$
test_compile_indexing[slice-tensorclass-eager] 0.2649ms 22.9835μs 43.5094 KOps/s 43.6154 KOps/s $\color{#d91a1a}-0.24\%$
test_compile_indexing[slice-pytree-compile] 89.9110μs 45.7463μs 21.8597 KOps/s 20.3310 KOps/s $\textbf{\color{#35bf28}+7.52\%}$
test_compile_indexing[slice-pytree-eager] 0.2740ms 23.0242μs 43.4326 KOps/s 41.9761 KOps/s $\color{#35bf28}+3.47\%$
test_compile_indexing[int-tensordict-compile] 93.0710μs 55.8020μs 17.9205 KOps/s 16.2578 KOps/s $\textbf{\color{#35bf28}+10.23\%}$
test_compile_indexing[int-tensordict-eager] 0.2391ms 28.0547μs 35.6446 KOps/s 32.9560 KOps/s $\textbf{\color{#35bf28}+8.16\%}$
test_compile_indexing[int-tensorclass-compile] 81.5410μs 46.3007μs 21.5980 KOps/s 20.5879 KOps/s $\color{#35bf28}+4.91\%$
test_compile_indexing[int-tensorclass-eager] 0.2772ms 22.7896μs 43.8796 KOps/s 42.8410 KOps/s $\color{#35bf28}+2.42\%$
test_compile_indexing[int-pytree-compile] 95.4610μs 46.9189μs 21.3134 KOps/s 20.7217 KOps/s $\color{#35bf28}+2.86\%$
test_compile_indexing[int-pytree-eager] 0.3071ms 23.2010μs 43.1016 KOps/s 43.5632 KOps/s $\color{#d91a1a}-1.06\%$
test_mod_add[eager] 94.1210μs 49.6780μs 20.1296 KOps/s 18.5648 KOps/s $\textbf{\color{#35bf28}+8.43\%}$
test_mod_add[compile] 0.5528ms 0.1065ms 9.3876 KOps/s 9.3203 KOps/s $\color{#35bf28}+0.72\%$
test_mod_add[compile-overhead] 0.3189ms 0.1488ms 6.7205 KOps/s 6.5385 KOps/s $\color{#35bf28}+2.78\%$
test_mod_wrap[eager] 0.4030ms 0.2935ms 3.4073 KOps/s 3.3775 KOps/s $\color{#35bf28}+0.88\%$
test_mod_wrap[compile] 0.4149ms 0.3486ms 2.8687 KOps/s 2.8278 KOps/s $\color{#35bf28}+1.45\%$
test_mod_wrap[compile-overhead] 6.4946ms 3.5416ms 282.3617 Ops/s 253.3321 Ops/s $\textbf{\color{#35bf28}+11.46\%}$
test_mod_wrap_and_backward[eager] 1.6553ms 1.5100ms 662.2599 Ops/s 616.0845 Ops/s $\textbf{\color{#35bf28}+7.49\%}$
test_mod_wrap_and_backward[compile] 1.5769ms 1.4593ms 685.2461 Ops/s 636.6183 Ops/s $\textbf{\color{#35bf28}+7.64\%}$
test_mod_wrap_and_backward[compile-overhead] 1.2508ms 0.8936ms 1.1191 KOps/s 992.5807 Ops/s $\textbf{\color{#35bf28}+12.74\%}$
test_seq_add[eager] 0.2120ms 0.1583ms 6.3153 KOps/s 6.3025 KOps/s $\color{#35bf28}+0.20\%$
test_seq_add[compile] 0.1742ms 0.1152ms 8.6797 KOps/s 8.4099 KOps/s $\color{#35bf28}+3.21\%$
test_seq_add[compile-overhead] 0.2088ms 0.1599ms 6.2550 KOps/s 6.2791 KOps/s $\color{#d91a1a}-0.38\%$
test_seq_wrap[eager] 0.6310ms 0.5493ms 1.8205 KOps/s 1.8357 KOps/s $\color{#d91a1a}-0.83\%$
test_seq_wrap[compile] 0.4409ms 0.3689ms 2.7108 KOps/s 2.7001 KOps/s $\color{#35bf28}+0.40\%$
test_seq_wrap[compile-overhead] 0.3303ms 0.2670ms 3.7459 KOps/s 3.7011 KOps/s $\color{#35bf28}+1.21\%$
test_func_call_runtime[False-eager] 0.9291ms 0.8457ms 1.1825 KOps/s 1.1186 KOps/s $\textbf{\color{#35bf28}+5.71\%}$
test_func_call_runtime[False-compile] 1.0053ms 0.9123ms 1.0961 KOps/s 1.0893 KOps/s $\color{#35bf28}+0.62\%$
test_func_call_runtime[False-compile-overhead] 0.5130ms 0.4586ms 2.1806 KOps/s 2.1193 KOps/s $\color{#35bf28}+2.89\%$
test_func_call_runtime[True-eager] 1.1638ms 1.0802ms 925.7268 Ops/s 905.2301 Ops/s $\color{#35bf28}+2.26\%$
test_func_call_runtime[True-compile] 1.0107ms 0.9182ms 1.0891 KOps/s 1.0752 KOps/s $\color{#35bf28}+1.29\%$
test_func_call_runtime[True-compile-overhead] 0.5483ms 0.4732ms 2.1133 KOps/s 2.0641 KOps/s $\color{#35bf28}+2.39\%$
test_func_call_cm_runtime[False-eager] 0.9330ms 0.8439ms 1.1849 KOps/s 1.0978 KOps/s $\textbf{\color{#35bf28}+7.93\%}$
test_func_call_cm_runtime[False-compile] 0.9836ms 0.9160ms 1.0917 KOps/s 1.0729 KOps/s $\color{#35bf28}+1.75\%$
test_func_call_cm_runtime[False-compile-overhead] 0.5151ms 0.4598ms 2.1747 KOps/s 2.1229 KOps/s $\color{#35bf28}+2.44\%$
test_func_call_cm_runtime[True-eager] 1.3467ms 1.2400ms 806.4433 Ops/s 803.2506 Ops/s $\color{#35bf28}+0.40\%$
test_func_call_cm_runtime[True-compile] 1.0302ms 0.9553ms 1.0468 KOps/s 1.0351 KOps/s $\color{#35bf28}+1.13\%$
test_func_call_cm_runtime[True-compile-overhead] 0.5473ms 0.5048ms 1.9808 KOps/s 1.9324 KOps/s $\color{#35bf28}+2.50\%$
test_vmap_func_call_cm_runtime[eager] 2.8546ms 2.3708ms 421.7948 Ops/s 418.3430 Ops/s $\color{#35bf28}+0.83\%$
test_vmap_func_call_cm_runtime[compile] 1.0916ms 0.9813ms 1.0190 KOps/s 1.0146 KOps/s $\color{#35bf28}+0.44\%$
test_vmap_func_call_cm_runtime[compile-overhead] 0.5756ms 0.5097ms 1.9619 KOps/s 1.9019 KOps/s $\color{#35bf28}+3.16\%$
test_distributed 0.5349ms 0.1529ms 6.5398 KOps/s 6.4141 KOps/s $\color{#35bf28}+1.96\%$
test_tdmodule 81.2710μs 27.9051μs 35.8358 KOps/s 35.2168 KOps/s $\color{#35bf28}+1.76\%$
test_tdmodule_dispatch 75.8910μs 45.5119μs 21.9723 KOps/s 21.7771 KOps/s $\color{#35bf28}+0.90\%$
test_tdseq 47.0410μs 27.1902μs 36.7780 KOps/s 36.8758 KOps/s $\color{#d91a1a}-0.27\%$
test_tdseq_dispatch 69.6310μs 47.5179μs 21.0447 KOps/s 20.7341 KOps/s $\color{#35bf28}+1.50\%$
test_instantiation_functorch 2.2402ms 2.0713ms 482.7985 Ops/s 476.1336 Ops/s $\color{#35bf28}+1.40\%$
test_exec_functorch 0.2365ms 0.1796ms 5.5675 KOps/s 5.4848 KOps/s $\color{#35bf28}+1.51\%$
test_exec_functional_call 0.2131ms 0.1626ms 6.1499 KOps/s 6.1352 KOps/s $\color{#35bf28}+0.24\%$
test_exec_td_decorator 0.4511ms 0.2372ms 4.2164 KOps/s 4.1490 KOps/s $\color{#35bf28}+1.62\%$
test_vmap_mlp_speed_decorator[True-True] 1.0366ms 0.8287ms 1.2067 KOps/s 1.1889 KOps/s $\color{#35bf28}+1.49\%$
test_vmap_mlp_speed_decorator[True-False] 0.9981ms 0.8262ms 1.2104 KOps/s 1.1895 KOps/s $\color{#35bf28}+1.75\%$
test_vmap_mlp_speed_decorator[False-True] 0.9021ms 0.7164ms 1.3958 KOps/s 1.3800 KOps/s $\color{#35bf28}+1.14\%$
test_vmap_mlp_speed_decorator[False-False] 0.8908ms 0.7184ms 1.3920 KOps/s 1.3740 KOps/s $\color{#35bf28}+1.31\%$
test_vmap_transformer_speed_decorator[True-True] 21.0616ms 20.6956ms 48.3194 Ops/s 48.1532 Ops/s $\color{#35bf28}+0.35\%$
test_vmap_transformer_speed_decorator[True-False] 21.5430ms 20.7856ms 48.1102 Ops/s 48.0501 Ops/s $\color{#35bf28}+0.13\%$
test_vmap_transformer_speed_decorator[False-True] 21.1548ms 20.5268ms 48.7169 Ops/s 48.4008 Ops/s $\color{#35bf28}+0.65\%$
test_vmap_transformer_speed_decorator[False-False] 21.1479ms 20.5214ms 48.7295 Ops/s 48.3113 Ops/s $\color{#35bf28}+0.87\%$
test_to_module_speed[True] 2.1016ms 1.4878ms 672.1192 Ops/s 672.8974 Ops/s $\color{#d91a1a}-0.12\%$
test_to_module_speed[False] 1.9791ms 1.4412ms 693.8892 Ops/s 688.3902 Ops/s $\color{#35bf28}+0.80\%$
test_tc_init 80.7100μs 45.6618μs 21.9001 KOps/s 21.3460 KOps/s $\color{#35bf28}+2.60\%$
test_tc_init_tensor_only 40.2010μs 9.9768μs 100.2329 KOps/s 100.6511 KOps/s $\color{#d91a1a}-0.42\%$
test_tc_init_nested 0.1496ms 90.9225μs 10.9984 KOps/s 10.4531 KOps/s $\textbf{\color{#35bf28}+5.22\%}$
test_tc_init_many_fields 49.1300μs 16.6331μs 60.1211 KOps/s 59.9915 KOps/s $\color{#35bf28}+0.22\%$
test_tc_first_layer_tensor 17.9300μs 1.8421μs 542.8611 KOps/s 537.0655 KOps/s $\color{#35bf28}+1.08\%$
test_tc_first_layer_tensor_only 4.2814μs 0.7618μs 1.3127 MOps/s 1.3031 MOps/s $\color{#35bf28}+0.74\%$
test_tc_first_layer_tensor_set 29.2500μs 3.9474μs 253.3301 KOps/s 253.7911 KOps/s $\color{#d91a1a}-0.18\%$
test_tc_first_layer_tensor_only_set 24.3200μs 3.0991μs 322.6795 KOps/s 313.3253 KOps/s $\color{#35bf28}+2.99\%$
test_tc_first_layer_nontensor 29.1000μs 6.2797μs 159.2425 KOps/s 159.4074 KOps/s $\color{#d91a1a}-0.10\%$
test_tc_second_layer_tensor 2.6825ms 4.4997μs 222.2370 KOps/s 220.7640 KOps/s $\color{#35bf28}+0.67\%$
test_tc_second_layer_nontensor 39.7900μs 8.8477μs 113.0235 KOps/s 112.8300 KOps/s $\color{#35bf28}+0.17\%$
test_unbind 0.2587s 16.0420ms 62.3362 Ops/s 55.7678 Ops/s $\textbf{\color{#35bf28}+11.78\%}$
test_full_like 4.5525ms 4.3631ms 229.1931 Ops/s 228.4860 Ops/s $\color{#35bf28}+0.31\%$
test_zeros_like 4.6141ms 4.3780ms 228.4146 Ops/s 94.8134 Ops/s $\textbf{\color{#35bf28}+140.91\%}$
test_ones_like 5.0501ms 4.3719ms 228.7316 Ops/s 94.7100 Ops/s $\textbf{\color{#35bf28}+141.51\%}$
test_clone 6.5976ms 6.4424ms 155.2207 Ops/s 67.1623 Ops/s $\textbf{\color{#35bf28}+131.11\%}$
test_squeeze 69.1510μs 14.4727μs 69.0954 KOps/s 65.6598 KOps/s $\textbf{\color{#35bf28}+5.23\%}$
test_unsqueeze 0.1789ms 0.1140ms 8.7736 KOps/s 8.7672 KOps/s $\color{#35bf28}+0.07\%$
test_split 0.2588ms 0.1852ms 5.4008 KOps/s 5.1411 KOps/s $\textbf{\color{#35bf28}+5.05\%}$
test_permute 0.2911ms 0.2017ms 4.9588 KOps/s 4.5798 KOps/s $\textbf{\color{#35bf28}+8.28\%}$
test_stack 51.7453ms 51.0766ms 19.5785 Ops/s 19.4465 Ops/s $\color{#35bf28}+0.68\%$
test_cat 51.4762ms 51.0630ms 19.5836 Ops/s 19.5372 Ops/s $\color{#35bf28}+0.24\%$

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

No reviews

Assignees

No one assigned

Labels

CI CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. documentation Improvements or additions to documentation

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

1 participant

@vmoens