Skip to content

[CI] Flaky test tracker: track per-test results via JUnit XML - #1602

Merged
vmoens merged 2 commits into
gh/vmoens/80/basefrom
gh/vmoens/80/head
Feb 19, 2026
Merged

[CI] Flaky test tracker: track per-test results via JUnit XML#1602
vmoens merged 2 commits 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):

Instead of only tracking which CI job failed (which you already
know from looking at CI), the tracker now identifies which specific
test is flaky.

Changes:

  • run_test.sh: add --junitxml to each pytest invocation, writing
    results to $RUNNER_ARTIFACT_DIR for upload.
  • test-linux.yml: add upload-artifact parameter to all four job
    types so JUnit XML is available as a downloadable artifact.
  • analyze_flaky_tests.py: rewritten to download test-results-*
    artifacts via gh run download, parse JUnit XML with
    xml.etree.ElementTree, and track per-test pass/fail across runs.
  • generate_dashboard.py / flaky-test-tracker.yml: minor wording
    updates for test-level (not job-level) reporting.

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

@vmoens
Update
96acc4c 
[ghstack-poisoned]
vmoens added a commit that referenced this pull request Feb 19, 2026
@vmoens @cursoragent
[CI] Flaky test tracker: track per-test results via JUnit XML
3de72af 
Instead of only tracking which CI *job* failed (which you already
know from looking at CI), the tracker now identifies which specific
*test* is flaky.

Changes:
- run_test.sh: add --junitxml to each pytest invocation, writing
  results to $RUNNER_ARTIFACT_DIR for upload.
- test-linux.yml: add upload-artifact parameter to all four job
  types so JUnit XML is available as a downloadable artifact.
- analyze_flaky_tests.py: rewritten to download test-results-*
  artifacts via `gh run download`, parse JUnit XML with
  xml.etree.ElementTree, and track per-test pass/fail across runs.
- generate_dashboard.py / flaky-test-tracker.yml: minor wording
  updates for test-level (not job-level) reporting.

Co-authored-by: Cursor <cursoragent@cursor.com>
ghstack-source-id: 9d82a8c
Pull-Request: #1602
@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 the CI label Feb 19, 2026
vmoens added a commit that referenced this pull request Feb 19, 2026
@vmoens @cursoragent
[CI] Flaky test tracker: track per-test results via JUnit XML
0c5f6f1 
Instead of only tracking which CI *job* failed (which you already
know from looking at CI), the tracker now identifies which specific
*test* is flaky.

Changes:
- run_test.sh: add --junitxml to each pytest invocation, writing
  results to $RUNNER_ARTIFACT_DIR for upload.
- test-linux.yml: add upload-artifact parameter to all four job
  types so JUnit XML is available as a downloadable artifact.
- analyze_flaky_tests.py: rewritten to download test-results-*
  artifacts via `gh run download`, parse JUnit XML with
  xml.etree.ElementTree, and track per-test pass/fail across runs.
- generate_dashboard.py / flaky-test-tracker.yml: minor wording
  updates for test-level (not job-level) reporting.

Co-authored-by: Cursor <cursoragent@cursor.com>
ghstack-source-id: 9d82a8c
Pull-Request: #1602
@vmoens
Update
4baa492 
[ghstack-poisoned]
vmoens added a commit that referenced this pull request Feb 19, 2026
@vmoens @cursoragent
[CI] Flaky test tracker: track per-test results via JUnit XML
447955c 
Instead of only tracking which CI *job* failed (which you already
know from looking at CI), the tracker now identifies which specific
*test* is flaky.

Changes:
- run_test.sh: add --junitxml to each pytest invocation, writing
  results to $RUNNER_ARTIFACT_DIR for upload.
- test-linux.yml: add upload-artifact parameter to all four job
  types so JUnit XML is available as a downloadable artifact.
- analyze_flaky_tests.py: rewritten to download test-results-*
  artifacts via `gh run download`, parse JUnit XML with
  xml.etree.ElementTree, and track per-test pass/fail across runs.
- generate_dashboard.py / flaky-test-tracker.yml: minor wording
  updates for test-level (not job-level) reporting.

Co-authored-by: Cursor <cursoragent@cursor.com>
ghstack-source-id: 330338a
Pull-Request: #1602
vmoens added a commit that referenced this pull request Feb 19, 2026
@vmoens @cursoragent
[CI] Flaky test tracker: track per-test results via JUnit XML
fc5a485 
Instead of only tracking which CI *job* failed (which you already
know from looking at CI), the tracker now identifies which specific
*test* is flaky.

Changes:
- run_test.sh: add --junitxml to each pytest invocation, writing
  results to $RUNNER_ARTIFACT_DIR for upload.
- test-linux.yml: add upload-artifact parameter to all four job
  types so JUnit XML is available as a downloadable artifact.
- analyze_flaky_tests.py: rewritten to download test-results-*
  artifacts via `gh run download`, parse JUnit XML with
  xml.etree.ElementTree, and track per-test pass/fail across runs.
- generate_dashboard.py / flaky-test-tracker.yml: minor wording
  updates for test-level (not job-level) reporting.

Co-authored-by: Cursor <cursoragent@cursor.com>
ghstack-source-id: 330338a
Pull-Request: #1602
@vmoens
vmoens merged commit 4baa492 into gh/vmoens/80/base Feb 19, 2026
41 of 53 checks passed
@vmoens
vmoens deleted the gh/vmoens/80/head branch February 19, 2026 14:10
@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: $\large\color{#35bf28}13$. Worsened: $\large\color{#d91a1a}11$.

Expand to view detailed results
Name Max Mean Ops Ops on Repo HEAD Change
test_plain_set_nested 29.6410μs 14.0890μs 70.9772 KOps/s 71.8119 KOps/s $\color{#d91a1a}-1.16\%$
test_plain_set_stack_nested 41.4210μs 14.3388μs 69.7406 KOps/s 70.2612 KOps/s $\color{#d91a1a}-0.74\%$
test_plain_set_nested_inplace 39.6110μs 15.8286μs 63.1769 KOps/s 63.9306 KOps/s $\color{#d91a1a}-1.18\%$
test_plain_set_stack_nested_inplace 51.1610μs 15.6600μs 63.8572 KOps/s 64.1199 KOps/s $\color{#d91a1a}-0.41\%$
test_items 23.2710μs 5.5662μs 179.6565 KOps/s 180.6591 KOps/s $\color{#d91a1a}-0.55\%$
test_items_nested 0.5179ms 0.4350ms 2.2987 KOps/s 2.2838 KOps/s $\color{#35bf28}+0.65\%$
test_items_nested_locked 0.4921ms 0.4395ms 2.2754 KOps/s 2.2494 KOps/s $\color{#35bf28}+1.16\%$
test_items_nested_leaf 0.1278ms 92.3817μs 10.8247 KOps/s 11.0113 KOps/s $\color{#d91a1a}-1.69\%$
test_items_stack_nested 0.4980ms 0.4399ms 2.2732 KOps/s 2.2666 KOps/s $\color{#35bf28}+0.29\%$
test_items_stack_nested_leaf 0.1249ms 91.4355μs 10.9367 KOps/s 10.9649 KOps/s $\color{#d91a1a}-0.26\%$
test_items_stack_nested_locked 0.4979ms 0.4425ms 2.2599 KOps/s 2.2480 KOps/s $\color{#35bf28}+0.53\%$
test_keys 31.7910μs 4.1766μs 239.4281 KOps/s 242.9631 KOps/s $\color{#d91a1a}-1.45\%$
test_keys_nested 0.1531ms 0.1174ms 8.5187 KOps/s 8.5868 KOps/s $\color{#d91a1a}-0.79\%$
test_keys_nested_locked 88.9326ms 0.1378ms 7.2553 KOps/s 7.9867 KOps/s $\textbf{\color{#d91a1a}-9.16\%}$
test_keys_nested_leaf 0.1403ms 0.1079ms 9.2712 KOps/s 9.2076 KOps/s $\color{#35bf28}+0.69\%$
test_keys_stack_nested 0.1499ms 0.1178ms 8.4920 KOps/s 8.5673 KOps/s $\color{#d91a1a}-0.88\%$
test_keys_stack_nested_leaf 0.1671ms 0.1084ms 9.2216 KOps/s 9.2444 KOps/s $\color{#d91a1a}-0.25\%$
test_keys_stack_nested_locked 0.1864ms 0.1261ms 7.9281 KOps/s 8.0311 KOps/s $\color{#d91a1a}-1.28\%$
test_values 8.5542μs 0.9984μs 1.0016 MOps/s 1.0114 MOps/s $\color{#d91a1a}-0.97\%$
test_values_nested 76.9610μs 46.6490μs 21.4367 KOps/s 21.5465 KOps/s $\color{#d91a1a}-0.51\%$
test_values_nested_locked 74.5820μs 49.4243μs 20.2329 KOps/s 20.3784 KOps/s $\color{#d91a1a}-0.71\%$
test_values_nested_leaf 90.6820μs 52.8805μs 18.9106 KOps/s 19.1312 KOps/s $\color{#d91a1a}-1.15\%$
test_values_stack_nested 80.3410μs 46.9355μs 21.3059 KOps/s 21.5673 KOps/s $\color{#d91a1a}-1.21\%$
test_values_stack_nested_leaf 0.1234ms 52.5543μs 19.0279 KOps/s 19.1880 KOps/s $\color{#d91a1a}-0.83\%$
test_values_stack_nested_locked 97.7110μs 49.6865μs 20.1262 KOps/s 20.2932 KOps/s $\color{#d91a1a}-0.82\%$
test_membership 6.8383μs 0.8153μs 1.2265 MOps/s 1.2471 MOps/s $\color{#d91a1a}-1.65\%$
test_membership_nested 41.6910μs 2.7245μs 367.0425 KOps/s 365.2611 KOps/s $\color{#35bf28}+0.49\%$
test_membership_nested_leaf 24.0555μs 2.6611μs 375.7787 KOps/s 362.4304 KOps/s $\color{#35bf28}+3.68\%$
test_membership_stacked_nested 34.2310μs 2.7281μs 366.5521 KOps/s 361.9779 KOps/s $\color{#35bf28}+1.26\%$
test_membership_stacked_nested_leaf 59.9410μs 2.7384μs 365.1832 KOps/s 366.0329 KOps/s $\color{#d91a1a}-0.23\%$
test_membership_nested_last 73.3520μs 4.0756μs 245.3610 KOps/s 243.5666 KOps/s $\color{#35bf28}+0.74\%$
test_membership_nested_leaf_last 22.6610μs 4.1217μs 242.6206 KOps/s 242.4940 KOps/s $\color{#35bf28}+0.05\%$
test_membership_stacked_nested_last 48.4210μs 4.0782μs 245.2044 KOps/s 245.7649 KOps/s $\color{#d91a1a}-0.23\%$
test_membership_stacked_nested_leaf_last 32.9300μs 4.0794μs 245.1341 KOps/s 244.7189 KOps/s $\color{#35bf28}+0.17\%$
test_nested_getleaf 53.1710μs 20.1902μs 49.5291 KOps/s 48.5782 KOps/s $\color{#35bf28}+1.96\%$
test_nested_get 48.2710μs 19.1816μs 52.1333 KOps/s 50.8669 KOps/s $\color{#35bf28}+2.49\%$
test_stacked_getleaf 68.2620μs 20.2416μs 49.4031 KOps/s 49.7250 KOps/s $\color{#d91a1a}-0.65\%$
test_stacked_get 68.4020μs 19.2888μs 51.8436 KOps/s 51.2260 KOps/s $\color{#35bf28}+1.21\%$
test_nested_getitemleaf 56.1510μs 20.9453μs 47.7433 KOps/s 47.6525 KOps/s $\color{#35bf28}+0.19\%$
test_nested_getitem 72.3420μs 19.8155μs 50.4656 KOps/s 50.1998 KOps/s $\color{#35bf28}+0.53\%$
test_stacked_getitemleaf 0.2437ms 20.6670μs 48.3864 KOps/s 47.7555 KOps/s $\color{#35bf28}+1.32\%$
test_stacked_getitem 59.7110μs 19.7956μs 50.5163 KOps/s 49.4128 KOps/s $\color{#35bf28}+2.23\%$
test_lock_nested 8.0970ms 0.4571ms 2.1878 KOps/s 2.1943 KOps/s $\color{#d91a1a}-0.30\%$
test_lock_stack_nested 0.5177ms 0.4553ms 2.1965 KOps/s 2.1682 KOps/s $\color{#35bf28}+1.31\%$
test_unlock_nested 0.7638ms 0.3629ms 2.7554 KOps/s 2.7309 KOps/s $\color{#35bf28}+0.90\%$
test_unlock_stack_nested 0.4870ms 0.3660ms 2.7325 KOps/s 2.6713 KOps/s $\color{#35bf28}+2.29\%$
test_flatten_speed 0.5305ms 0.1161ms 8.6127 KOps/s 8.5269 KOps/s $\color{#35bf28}+1.01\%$
test_unflatten_speed 0.9637ms 0.5468ms 1.8287 KOps/s 1.8317 KOps/s $\color{#d91a1a}-0.17\%$
test_common_ops 0.8140ms 0.6648ms 1.5042 KOps/s 1.4982 KOps/s $\color{#35bf28}+0.40\%$
test_creation 0.4528ms 2.7220μs 367.3763 KOps/s 361.7426 KOps/s $\color{#35bf28}+1.56\%$
test_creation_empty 34.8500μs 5.7251μs 174.6692 KOps/s 173.4625 KOps/s $\color{#35bf28}+0.70\%$
test_creation_nested_1 41.9010μs 9.9479μs 100.5237 KOps/s 99.8800 KOps/s $\color{#35bf28}+0.64\%$
test_creation_nested_2 0.4340ms 11.0520μs 90.4813 KOps/s 88.5265 KOps/s $\color{#35bf28}+2.21\%$
test_creation_many_keys[10] 0.4247ms 16.8157μs 59.4681 KOps/s 58.9426 KOps/s $\color{#35bf28}+0.89\%$
test_creation_many_keys[50] 0.1070ms 72.1009μs 13.8695 KOps/s 13.7947 KOps/s $\color{#35bf28}+0.54\%$
test_creation_many_keys[100] 0.5736ms 0.1393ms 7.1806 KOps/s 7.0962 KOps/s $\color{#35bf28}+1.19\%$
test_creation_nested_many_keys[10] 0.4599ms 36.0710μs 27.7231 KOps/s 27.6690 KOps/s $\color{#35bf28}+0.20\%$
test_creation_nested_many_keys[50] 0.5559ms 0.1466ms 6.8228 KOps/s 6.7297 KOps/s $\color{#35bf28}+1.38\%$
test_clone 45.3210μs 12.9582μs 77.1711 KOps/s 77.5607 KOps/s $\color{#d91a1a}-0.50\%$
test_getitem[int] 1.7637ms 13.8308μs 72.3025 KOps/s 58.6803 KOps/s $\textbf{\color{#35bf28}+23.21\%}$
test_getitem[slice_int] 0.4464ms 23.5563μs 42.4515 KOps/s 42.0165 KOps/s $\color{#35bf28}+1.04\%$
test_getitem[range] 0.1786ms 60.3371μs 16.5735 KOps/s 16.3713 KOps/s $\color{#35bf28}+1.24\%$
test_getitem[tuple] 0.1429ms 22.9466μs 43.5794 KOps/s 43.4294 KOps/s $\color{#35bf28}+0.35\%$
test_getitem[list] 0.4942ms 55.7656μs 17.9322 KOps/s 17.8486 KOps/s $\color{#35bf28}+0.47\%$
test_setitem_dim[int] 51.2410μs 27.2866μs 36.6480 KOps/s 38.7064 KOps/s $\textbf{\color{#d91a1a}-5.32\%}$
test_setitem_dim[slice_int] 80.5710μs 44.7900μs 22.3264 KOps/s 23.2427 KOps/s $\color{#d91a1a}-3.94\%$
test_setitem_dim[range] 0.5165ms 95.6565μs 10.4541 KOps/s 10.7089 KOps/s $\color{#d91a1a}-2.38\%$
test_setitem_dim[tuple] 65.7910μs 41.8192μs 23.9125 KOps/s 24.5133 KOps/s $\color{#d91a1a}-2.45\%$
test_setitem 59.0110μs 17.5137μs 57.0980 KOps/s 57.7236 KOps/s $\color{#d91a1a}-1.08\%$
test_set 51.6210μs 16.6113μs 60.1999 KOps/s 60.1840 KOps/s $\color{#35bf28}+0.03\%$
test_set_shared 0.7068ms 0.2095ms 4.7738 KOps/s 4.9337 KOps/s $\color{#d91a1a}-3.24\%$
test_update 0.4439ms 20.8809μs 47.8906 KOps/s 47.9814 KOps/s $\color{#d91a1a}-0.19\%$
test_update_nested 79.6520μs 32.3667μs 30.8960 KOps/s 31.0025 KOps/s $\color{#d91a1a}-0.34\%$
test_update__nested 0.4878ms 33.3880μs 29.9508 KOps/s 29.7516 KOps/s $\color{#35bf28}+0.67\%$
test_set_nested 52.5910μs 18.4662μs 54.1531 KOps/s 53.9004 KOps/s $\color{#35bf28}+0.47\%$
test_set_nested_new 59.7010μs 22.7124μs 44.0289 KOps/s 42.7860 KOps/s $\color{#35bf28}+2.90\%$
test_select 75.1910μs 40.0867μs 24.9460 KOps/s 25.3473 KOps/s $\color{#d91a1a}-1.58\%$
test_select_nested 0.1317ms 71.1410μs 14.0566 KOps/s 14.3416 KOps/s $\color{#d91a1a}-1.99\%$
test_exclude_nested 0.1330ms 87.5507μs 11.4220 KOps/s 11.6107 KOps/s $\color{#d91a1a}-1.63\%$
test_empty[True] 0.4512ms 0.3771ms 2.6518 KOps/s 2.6742 KOps/s $\color{#d91a1a}-0.84\%$
test_empty[False] 6.9153μs 1.2539μs 797.4965 KOps/s 799.0047 KOps/s $\color{#d91a1a}-0.19\%$
test_to 0.1022ms 71.2639μs 14.0324 KOps/s 13.7836 KOps/s $\color{#35bf28}+1.80\%$
test_to_nonblocking 0.1126ms 64.3876μs 15.5309 KOps/s 15.7738 KOps/s $\color{#d91a1a}-1.54\%$
test_unbind_speed 0.3469ms 0.3130ms 3.1949 KOps/s 3.2524 KOps/s $\color{#d91a1a}-1.77\%$
test_unbind_speed_stack0 0.4528ms 0.3093ms 3.2327 KOps/s 3.2477 KOps/s $\color{#d91a1a}-0.46\%$
test_unbind_speed_stack1 0.1033s 0.8809ms 1.1352 KOps/s 1.1161 KOps/s $\color{#35bf28}+1.71\%$
test_split 1.2461ms 1.0904ms 917.0914 Ops/s 923.3898 Ops/s $\color{#d91a1a}-0.68\%$
test_chunk 0.1034s 1.1675ms 856.5426 Ops/s 966.0487 Ops/s $\textbf{\color{#d91a1a}-11.34\%}$
test_to_cpu_blocking 27.8507ms 23.9930ms 41.6789 Ops/s 45.2779 Ops/s $\textbf{\color{#d91a1a}-7.95\%}$
test_to_cpu_global_sync 11.2862ms 11.1799ms 89.4459 Ops/s 90.1406 Ops/s $\color{#d91a1a}-0.77\%$
test_to_cpu_event_sync 0.1151s 13.3949ms 74.6551 Ops/s 82.8123 Ops/s $\textbf{\color{#d91a1a}-9.85\%}$
test_to_cpu_default 12.3673ms 12.1745ms 82.1390 Ops/s 82.6242 Ops/s $\color{#d91a1a}-0.59\%$
test_consolidate[False-None] 4.2160ms 3.9782ms 251.3687 Ops/s 224.1985 Ops/s $\textbf{\color{#35bf28}+12.12\%}$
test_consolidate[default-None] 2.0519ms 1.9700ms 507.6061 Ops/s 496.1956 Ops/s $\color{#35bf28}+2.30\%$
test_consolidate[reduce-overhead-None] 1.9858ms 1.8907ms 528.9123 Ops/s 513.5559 Ops/s $\color{#35bf28}+2.99\%$
test_consolidate_njt[False-None] 8.5291ms 8.2159ms 121.7149 Ops/s 119.8197 Ops/s $\color{#35bf28}+1.58\%$
test_to[False-False-None] 2.2011ms 2.0471ms 488.4850 Ops/s 483.6971 Ops/s $\color{#35bf28}+0.99\%$
test_to[True-False-None] 2.0310ms 1.8600ms 537.6254 Ops/s 534.3279 Ops/s $\color{#35bf28}+0.62\%$
test_to[within-False-None] 6.1776ms 5.9355ms 168.4768 Ops/s 167.5109 Ops/s $\color{#35bf28}+0.58\%$
test_to[True-default-None] 7.8184ms 7.6278ms 131.0990 Ops/s 133.2294 Ops/s $\color{#d91a1a}-1.60\%$
test_to_njt[False-False-None] 8.6725ms 8.3328ms 120.0077 Ops/s 119.4232 Ops/s $\color{#35bf28}+0.49\%$
test_to_njt[True-False-None] 7.1068ms 6.7759ms 147.5808 Ops/s 147.3237 Ops/s $\color{#35bf28}+0.17\%$
test_to_njt[within-False-None] 15.2227ms 14.8926ms 67.1475 Ops/s 65.4069 Ops/s $\color{#35bf28}+2.66\%$
test_creation[device0] 0.4196ms 0.1171ms 8.5404 KOps/s 8.7906 KOps/s $\color{#d91a1a}-2.85\%$
test_creation_from_tensor 0.4612ms 0.1161ms 8.6136 KOps/s 8.9854 KOps/s $\color{#d91a1a}-4.14\%$
test_add_one[memmap_tensor0] 0.2082ms 7.0400μs 142.0450 KOps/s 153.7823 KOps/s $\textbf{\color{#d91a1a}-7.63\%}$
test_contiguous[memmap_tensor0] 19.2410μs 0.6537μs 1.5297 MOps/s 2.3098 MOps/s $\textbf{\color{#d91a1a}-33.78\%}$
test_stack[memmap_tensor0] 35.4500μs 4.9100μs 203.6680 KOps/s 214.8281 KOps/s $\textbf{\color{#d91a1a}-5.19\%}$
test_memmaptd_index 0.9908ms 0.2557ms 3.9116 KOps/s 3.9306 KOps/s $\color{#d91a1a}-0.48\%$
test_memmaptd_index_astensor 0.4631ms 0.3457ms 2.8926 KOps/s 2.9151 KOps/s $\color{#d91a1a}-0.77\%$
test_memmaptd_index_op 0.8337ms 0.5906ms 1.6932 KOps/s 1.7003 KOps/s $\color{#d91a1a}-0.41\%$
test_serialize_model 0.1396s 0.1371s 7.2960 Ops/s 7.3204 Ops/s $\color{#d91a1a}-0.33\%$
test_serialize_model_pickle 1.3484s 1.2120s 0.8251 Ops/s 0.8259 Ops/s $\color{#d91a1a}-0.10\%$
test_serialize_weights 0.1392s 0.1357s 7.3675 Ops/s 7.3841 Ops/s $\color{#d91a1a}-0.23\%$
test_serialize_weights_returnearly 0.4411s 92.8939ms 10.7650 Ops/s 10.2828 Ops/s $\color{#35bf28}+4.69\%$
test_serialize_weights_pickle 1.3591s 1.1960s 0.8361 Ops/s 0.8183 Ops/s $\color{#35bf28}+2.18\%$
test_reshape_pytree 0.2008ms 32.3166μs 30.9439 KOps/s 31.0508 KOps/s $\color{#d91a1a}-0.34\%$
test_reshape_td 73.1110μs 43.1637μs 23.1676 KOps/s 23.0524 KOps/s $\color{#35bf28}+0.50\%$
test_view_pytree 0.2235ms 32.5984μs 30.6763 KOps/s 31.7570 KOps/s $\color{#d91a1a}-3.40\%$
test_view_td 96.9020μs 50.8348μs 19.6716 KOps/s 19.5421 KOps/s $\color{#35bf28}+0.66\%$
test_unbind_pytree 0.2296ms 35.5603μs 28.1212 KOps/s 27.9935 KOps/s $\color{#35bf28}+0.46\%$
test_unbind_td 0.1545ms 47.1081μs 21.2278 KOps/s 21.1792 KOps/s $\color{#35bf28}+0.23\%$
test_split_pytree 0.2543ms 41.3338μs 24.1933 KOps/s 24.4861 KOps/s $\color{#d91a1a}-1.20\%$
test_split_td 0.1956ms 62.4343μs 16.0168 KOps/s 15.9260 KOps/s $\color{#35bf28}+0.57\%$
test_add_pytree 0.2282ms 40.8432μs 24.4839 KOps/s 24.4341 KOps/s $\color{#35bf28}+0.20\%$
test_add_td 90.1220μs 53.0521μs 18.8494 KOps/s 19.1709 KOps/s $\color{#d91a1a}-1.68\%$
test_compile_add_one_nested[tensordict-compile] 0.1882ms 0.1352ms 7.3989 KOps/s 6.8105 KOps/s $\textbf{\color{#35bf28}+8.64\%}$
test_compile_add_one_nested[tensordict-eager] 0.3587ms 0.1874ms 5.3366 KOps/s 5.4430 KOps/s $\color{#d91a1a}-1.95\%$
test_compile_add_one_nested[pytree-compile] 0.1718ms 0.1056ms 9.4694 KOps/s 9.2882 KOps/s $\color{#35bf28}+1.95\%$
test_compile_add_one_nested[pytree-eager] 0.4303ms 0.1807ms 5.5355 KOps/s 5.6303 KOps/s $\color{#d91a1a}-1.68\%$
test_compile_copy_nested[tensordict-compile] 0.1401ms 29.5139μs 33.8824 KOps/s 29.3242 KOps/s $\textbf{\color{#35bf28}+15.54\%}$
test_compile_copy_nested[tensordict-eager] 97.2220μs 48.6432μs 20.5579 KOps/s 20.2920 KOps/s $\color{#35bf28}+1.31\%$
test_compile_copy_nested[pytree-compile] 42.3710μs 9.3407μs 107.0582 KOps/s 105.0844 KOps/s $\color{#35bf28}+1.88\%$
test_compile_copy_nested[pytree-eager] 0.4155ms 66.4275μs 15.0540 KOps/s 15.5443 KOps/s $\color{#d91a1a}-3.15\%$
test_compile_add_one_flat[tensordict-compile] 0.2221ms 0.1720ms 5.8141 KOps/s 5.4630 KOps/s $\textbf{\color{#35bf28}+6.43\%}$
test_compile_add_one_flat[tensordict-eager] 0.2972ms 0.2475ms 4.0412 KOps/s 3.9942 KOps/s $\color{#35bf28}+1.18\%$
test_compile_add_one_flat[tensorclass-compile] 0.1944ms 0.1128ms 8.8622 KOps/s 8.5114 KOps/s $\color{#35bf28}+4.12\%$
test_compile_add_one_flat[tensorclass-eager] 0.1143ms 67.2668μs 14.8662 KOps/s 14.8256 KOps/s $\color{#35bf28}+0.27\%$
test_compile_add_one_flat[pytree-compile] 0.2509ms 0.1546ms 6.4676 KOps/s 6.3089 KOps/s $\color{#35bf28}+2.51\%$
test_compile_add_one_flat[pytree-eager] 0.8113ms 0.5355ms 1.8675 KOps/s 1.8473 KOps/s $\color{#35bf28}+1.09\%$
test_compile_add_self_flat[tensordict-eager] 0.4535ms 0.2987ms 3.3480 KOps/s 3.2714 KOps/s $\color{#35bf28}+2.34\%$
test_compile_add_self_flat[tensordict-compile] 0.2347ms 0.1741ms 5.7430 KOps/s 5.3805 KOps/s $\textbf{\color{#35bf28}+6.74\%}$
test_compile_add_self_flat[tensorclass-eager] 0.1474ms 82.7098μs 12.0905 KOps/s 12.1385 KOps/s $\color{#d91a1a}-0.40\%$
test_compile_add_self_flat[tensorclass-compile] 0.1675ms 0.1152ms 8.6834 KOps/s 8.3142 KOps/s $\color{#35bf28}+4.44\%$
test_compile_add_self_flat[pytree-eager] 0.7286ms 0.4506ms 2.2190 KOps/s 2.2076 KOps/s $\color{#35bf28}+0.52\%$
test_compile_add_self_flat[pytree-compile] 0.1978ms 0.1547ms 6.4659 KOps/s 6.2564 KOps/s $\color{#35bf28}+3.35\%$
test_compile_copy_flat[tensordict-compile] 84.6810μs 22.9720μs 43.5313 KOps/s 38.8967 KOps/s $\textbf{\color{#35bf28}+11.92\%}$
test_compile_copy_flat[tensordict-eager] 78.1610μs 39.9605μs 25.0247 KOps/s 25.2277 KOps/s $\color{#d91a1a}-0.80\%$
test_compile_copy_flat[pytree-compile] 56.2110μs 10.4987μs 95.2496 KOps/s 96.5681 KOps/s $\color{#d91a1a}-1.37\%$
test_compile_copy_flat[pytree-eager] 0.1798s 60.7149μs 16.4704 KOps/s 19.7362 KOps/s $\textbf{\color{#d91a1a}-16.55\%}$
test_compile_assign_and_add[tensordict-compile] 1.9022ms 0.1684ms 5.9390 KOps/s 5.5055 KOps/s $\textbf{\color{#35bf28}+7.87\%}$
test_compile_assign_and_add[tensordict-eager] 3.4766ms 3.2405ms 308.5929 Ops/s 306.6134 Ops/s $\color{#35bf28}+0.65\%$
test_compile_assign_and_add[pytree-compile] 1.8807ms 0.1559ms 6.4127 KOps/s 6.2587 KOps/s $\color{#35bf28}+2.46\%$
test_compile_assign_and_add[pytree-eager] 2.9383ms 2.7818ms 359.4758 Ops/s 354.2640 Ops/s $\color{#35bf28}+1.47\%$
test_compile_indexing[tensor-tensordict-compile] 0.1532ms 0.1057ms 9.4614 KOps/s 8.8299 KOps/s $\textbf{\color{#35bf28}+7.15\%}$
test_compile_indexing[tensor-tensordict-eager] 0.3104ms 76.0840μs 13.1434 KOps/s 13.1700 KOps/s $\color{#d91a1a}-0.20\%$
test_compile_indexing[tensor-tensorclass-compile] 0.1598ms 93.2209μs 10.7272 KOps/s 10.3320 KOps/s $\color{#35bf28}+3.82\%$
test_compile_indexing[tensor-tensorclass-eager] 0.2593ms 48.2792μs 20.7129 KOps/s 21.3014 KOps/s $\color{#d91a1a}-2.76\%$
test_compile_indexing[tensor-pytree-compile] 0.1439ms 94.2817μs 10.6065 KOps/s 10.0732 KOps/s $\textbf{\color{#35bf28}+5.29\%}$
test_compile_indexing[tensor-pytree-eager] 0.2474ms 47.4266μs 21.0852 KOps/s 21.0727 KOps/s $\color{#35bf28}+0.06\%$
test_compile_indexing[slice-tensordict-compile] 98.3820μs 54.2501μs 18.4332 KOps/s 17.5810 KOps/s $\color{#35bf28}+4.85\%$
test_compile_indexing[slice-tensordict-eager] 0.2271ms 26.9070μs 37.1651 KOps/s 36.6659 KOps/s $\color{#35bf28}+1.36\%$
test_compile_indexing[slice-tensorclass-compile] 87.4420μs 44.6303μs 22.4063 KOps/s 22.3955 KOps/s $\color{#35bf28}+0.05\%$
test_compile_indexing[slice-tensorclass-eager] 0.2527ms 22.1240μs 45.1997 KOps/s 45.8175 KOps/s $\color{#d91a1a}-1.35\%$
test_compile_indexing[slice-pytree-compile] 0.1067ms 46.1000μs 21.6920 KOps/s 22.5013 KOps/s $\color{#d91a1a}-3.60\%$
test_compile_indexing[slice-pytree-eager] 0.2964ms 21.9715μs 45.5135 KOps/s 45.6931 KOps/s $\color{#d91a1a}-0.39\%$
test_compile_indexing[int-tensordict-compile] 0.1011ms 57.4273μs 17.4133 KOps/s 17.1905 KOps/s $\color{#35bf28}+1.30\%$
test_compile_indexing[int-tensordict-eager] 0.3008ms 26.5486μs 37.6668 KOps/s 36.7680 KOps/s $\color{#35bf28}+2.44\%$
test_compile_indexing[int-tensorclass-compile] 0.1250ms 45.5456μs 21.9560 KOps/s 22.3541 KOps/s $\color{#d91a1a}-1.78\%$
test_compile_indexing[int-tensorclass-eager] 0.2643ms 21.9822μs 45.4913 KOps/s 46.1107 KOps/s $\color{#d91a1a}-1.34\%$
test_compile_indexing[int-pytree-compile] 0.1029ms 45.8772μs 21.7973 KOps/s 21.9905 KOps/s $\color{#d91a1a}-0.88\%$
test_compile_indexing[int-pytree-eager] 0.2797ms 21.7607μs 45.9544 KOps/s 45.8356 KOps/s $\color{#35bf28}+0.26\%$
test_mod_add[eager] 0.1192ms 51.0844μs 19.5754 KOps/s 20.4846 KOps/s $\color{#d91a1a}-4.44\%$
test_mod_add[compile] 0.1810ms 0.1027ms 9.7409 KOps/s 9.6010 KOps/s $\color{#35bf28}+1.46\%$
test_mod_add[compile-overhead] 0.2302ms 0.1446ms 6.9171 KOps/s 6.8798 KOps/s $\color{#35bf28}+0.54\%$
test_mod_wrap[eager] 0.3702ms 0.2860ms 3.4970 KOps/s 3.5054 KOps/s $\color{#d91a1a}-0.24\%$
test_mod_wrap[compile] 0.3913ms 0.3440ms 2.9071 KOps/s 2.8356 KOps/s $\color{#35bf28}+2.52\%$
test_mod_wrap[compile-overhead] 7.3097ms 4.0272ms 248.3129 Ops/s 249.4277 Ops/s $\color{#d91a1a}-0.45\%$
test_mod_wrap_and_backward[eager] 1.6539ms 1.5008ms 666.2931 Ops/s 668.5475 Ops/s $\color{#d91a1a}-0.34\%$
test_mod_wrap_and_backward[compile] 1.6376ms 1.4391ms 694.8907 Ops/s 699.0778 Ops/s $\color{#d91a1a}-0.60\%$
test_mod_wrap_and_backward[compile-overhead] 1.2301ms 0.8704ms 1.1489 KOps/s 1.1377 KOps/s $\color{#35bf28}+0.98\%$
test_seq_add[eager] 0.2436ms 0.1554ms 6.4359 KOps/s 6.3495 KOps/s $\color{#35bf28}+1.36\%$
test_seq_add[compile] 0.5461ms 0.1134ms 8.8147 KOps/s 8.2550 KOps/s $\textbf{\color{#35bf28}+6.78\%}$
test_seq_add[compile-overhead] 0.2085ms 0.1516ms 6.5953 KOps/s 6.5134 KOps/s $\color{#35bf28}+1.26\%$
test_seq_wrap[eager] 0.7273ms 0.5468ms 1.8289 KOps/s 1.9152 KOps/s $\color{#d91a1a}-4.51\%$
test_seq_wrap[compile] 0.4754ms 0.3787ms 2.6403 KOps/s 2.7641 KOps/s $\color{#d91a1a}-4.48\%$
test_seq_wrap[compile-overhead] 0.3425ms 0.2582ms 3.8729 KOps/s 3.8632 KOps/s $\color{#35bf28}+0.25\%$
test_func_call_runtime[False-eager] 1.0013ms 0.8843ms 1.1308 KOps/s 1.1460 KOps/s $\color{#d91a1a}-1.33\%$
test_func_call_runtime[False-compile] 1.3663ms 0.9500ms 1.0526 KOps/s 1.0882 KOps/s $\color{#d91a1a}-3.27\%$
test_func_call_runtime[False-compile-overhead] 0.5182ms 0.4403ms 2.2709 KOps/s 2.2442 KOps/s $\color{#35bf28}+1.19\%$
test_func_call_runtime[True-eager] 1.1474ms 1.0622ms 941.4860 Ops/s 939.3823 Ops/s $\color{#35bf28}+0.22\%$
test_func_call_runtime[True-compile] 1.0435ms 0.9097ms 1.0992 KOps/s 1.0757 KOps/s $\color{#35bf28}+2.18\%$
test_func_call_runtime[True-compile-overhead] 0.5190ms 0.4533ms 2.2059 KOps/s 2.1770 KOps/s $\color{#35bf28}+1.33\%$
test_func_call_cm_runtime[False-eager] 0.9881ms 0.8567ms 1.1673 KOps/s 1.2043 KOps/s $\color{#d91a1a}-3.07\%$
test_func_call_cm_runtime[False-compile] 1.0757ms 0.9407ms 1.0630 KOps/s 1.1201 KOps/s $\textbf{\color{#d91a1a}-5.10\%}$
test_func_call_cm_runtime[False-compile-overhead] 0.5750ms 0.4453ms 2.2454 KOps/s 2.2400 KOps/s $\color{#35bf28}+0.24\%$
test_func_call_cm_runtime[True-eager] 1.3102ms 1.2109ms 825.8645 Ops/s 831.7858 Ops/s $\color{#d91a1a}-0.71\%$
test_func_call_cm_runtime[True-compile] 1.0635ms 0.9581ms 1.0438 KOps/s 1.0692 KOps/s $\color{#d91a1a}-2.38\%$
test_func_call_cm_runtime[True-compile-overhead] 0.5669ms 0.4844ms 2.0646 KOps/s 2.0368 KOps/s $\color{#35bf28}+1.37\%$
test_vmap_func_call_cm_runtime[eager] 2.8733ms 2.3483ms 425.8363 Ops/s 424.1253 Ops/s $\color{#35bf28}+0.40\%$
test_vmap_func_call_cm_runtime[compile] 1.1942ms 0.9576ms 1.0442 KOps/s 1.0523 KOps/s $\color{#d91a1a}-0.77\%$
test_vmap_func_call_cm_runtime[compile-overhead] 0.5360ms 0.4920ms 2.0325 KOps/s 2.0096 KOps/s $\color{#35bf28}+1.14\%$
test_distributed 3.3076ms 0.1559ms 6.4136 KOps/s 6.5412 KOps/s $\color{#d91a1a}-1.95\%$
test_tdmodule 0.1621ms 27.4462μs 36.4349 KOps/s 35.6331 KOps/s $\color{#35bf28}+2.25\%$
test_tdmodule_dispatch 75.4810μs 44.9408μs 22.2515 KOps/s 22.1314 KOps/s $\color{#35bf28}+0.54\%$
test_tdseq 48.5110μs 26.5778μs 37.6254 KOps/s 37.5693 KOps/s $\color{#35bf28}+0.15\%$
test_tdseq_dispatch 85.1020μs 46.8697μs 21.3357 KOps/s 21.2914 KOps/s $\color{#35bf28}+0.21\%$
test_instantiation_functorch 2.0873ms 1.9891ms 502.7374 Ops/s 502.0651 Ops/s $\color{#35bf28}+0.13\%$
test_exec_functorch 0.3250ms 0.1763ms 5.6737 KOps/s 5.6822 KOps/s $\color{#d91a1a}-0.15\%$
test_exec_functional_call 0.2352ms 0.1594ms 6.2754 KOps/s 6.3492 KOps/s $\color{#d91a1a}-1.16\%$
test_exec_td_decorator 0.4534ms 0.2299ms 4.3490 KOps/s 4.3371 KOps/s $\color{#35bf28}+0.27\%$
test_vmap_mlp_speed_decorator[True-True] 1.0138ms 0.8159ms 1.2256 KOps/s 1.2222 KOps/s $\color{#35bf28}+0.28\%$
test_vmap_mlp_speed_decorator[True-False] 0.9921ms 0.8152ms 1.2267 KOps/s 1.2226 KOps/s $\color{#35bf28}+0.34\%$
test_vmap_mlp_speed_decorator[False-True] 0.9045ms 0.7156ms 1.3973 KOps/s 1.4098 KOps/s $\color{#d91a1a}-0.88\%$
test_vmap_mlp_speed_decorator[False-False] 0.9015ms 0.7088ms 1.4109 KOps/s 1.4089 KOps/s $\color{#35bf28}+0.14\%$
test_vmap_transformer_speed_decorator[True-True] 21.4338ms 20.5899ms 48.5674 Ops/s 48.6908 Ops/s $\color{#d91a1a}-0.25\%$
test_vmap_transformer_speed_decorator[True-False] 20.8576ms 20.5880ms 48.5720 Ops/s 48.5470 Ops/s $\color{#35bf28}+0.05\%$
test_vmap_transformer_speed_decorator[False-True] 20.5346ms 20.3653ms 49.1031 Ops/s 49.0943 Ops/s $\color{#35bf28}+0.02\%$
test_vmap_transformer_speed_decorator[False-False] 20.6114ms 20.3784ms 49.0716 Ops/s 49.2843 Ops/s $\color{#d91a1a}-0.43\%$
test_to_module_speed[True] 1.5097ms 1.4000ms 714.3106 Ops/s 708.8304 Ops/s $\color{#35bf28}+0.77\%$
test_to_module_speed[False] 1.4752ms 1.3895ms 719.6822 Ops/s 715.6778 Ops/s $\color{#35bf28}+0.56\%$
test_tc_init 68.8110μs 43.9680μs 22.7438 KOps/s 22.7849 KOps/s $\color{#d91a1a}-0.18\%$
test_tc_init_tensor_only 39.0900μs 9.3417μs 107.0467 KOps/s 108.0829 KOps/s $\color{#d91a1a}-0.96\%$
test_tc_init_nested 0.1446ms 87.5877μs 11.4171 KOps/s 11.5913 KOps/s $\color{#d91a1a}-1.50\%$
test_tc_init_many_fields 43.7810μs 15.6038μs 64.0869 KOps/s 64.5171 KOps/s $\color{#d91a1a}-0.67\%$
test_tc_first_layer_tensor 19.9300μs 1.6992μs 588.5057 KOps/s 585.1973 KOps/s $\color{#35bf28}+0.57\%$
test_tc_first_layer_tensor_only 4.7730μs 0.7074μs 1.4137 MOps/s 1.4060 MOps/s $\color{#35bf28}+0.55\%$
test_tc_first_layer_tensor_set 37.6210μs 3.7557μs 266.2652 KOps/s 267.6285 KOps/s $\color{#d91a1a}-0.51\%$
test_tc_first_layer_tensor_only_set 18.2200μs 2.9692μs 336.7923 KOps/s 331.4309 KOps/s $\color{#35bf28}+1.62\%$
test_tc_first_layer_nontensor 47.0910μs 5.8596μs 170.6590 KOps/s 161.6375 KOps/s $\textbf{\color{#35bf28}+5.58\%}$
test_tc_second_layer_tensor 26.5900μs 4.1285μs 242.2177 KOps/s 238.1644 KOps/s $\color{#35bf28}+1.70\%$
test_tc_second_layer_nontensor 43.5110μs 8.2767μs 120.8214 KOps/s 116.2426 KOps/s $\color{#35bf28}+3.94\%$
test_unbind 0.2445s 15.9412ms 62.7307 Ops/s 69.2608 Ops/s $\textbf{\color{#d91a1a}-9.43\%}$
test_full_like 4.4889ms 4.3707ms 228.7959 Ops/s 228.6874 Ops/s $\color{#35bf28}+0.05\%$
test_zeros_like 4.4998ms 4.3578ms 229.4713 Ops/s 228.9972 Ops/s $\color{#35bf28}+0.21\%$
test_ones_like 4.4745ms 4.3647ms 229.1129 Ops/s 228.6015 Ops/s $\color{#35bf28}+0.22\%$
test_clone 6.6929ms 6.4458ms 155.1387 Ops/s 154.6340 Ops/s $\color{#35bf28}+0.33\%$
test_squeeze 0.2163ms 13.6781μs 73.1098 KOps/s 67.0970 KOps/s $\textbf{\color{#35bf28}+8.96\%}$
test_unsqueeze 0.1979ms 0.1106ms 9.0440 KOps/s 8.9170 KOps/s $\color{#35bf28}+1.42\%$
test_split 0.2595ms 0.1821ms 5.4916 KOps/s 5.4343 KOps/s $\color{#35bf28}+1.05\%$
test_permute 0.2775ms 0.2091ms 4.7821 KOps/s 4.9535 KOps/s $\color{#d91a1a}-3.46\%$
test_stack 51.5781ms 51.2815ms 19.5002 Ops/s 19.4866 Ops/s $\color{#35bf28}+0.07\%$
test_cat 52.6899ms 51.0390ms 19.5929 Ops/s 19.4948 Ops/s $\color{#35bf28}+0.50\%$

@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: $\large\color{#35bf28}18$. Worsened: $\large\color{#d91a1a}5$.

Expand to view detailed results
Name Max Mean Ops Ops on Repo HEAD Change
test_plain_set_nested 28.5600μs 14.2742μs 70.0564 KOps/s 70.9501 KOps/s $\color{#d91a1a}-1.26\%$
test_plain_set_stack_nested 30.8300μs 14.2796μs 70.0299 KOps/s 69.9149 KOps/s $\color{#35bf28}+0.16\%$
test_plain_set_nested_inplace 45.6600μs 15.9062μs 62.8684 KOps/s 63.1051 KOps/s $\color{#d91a1a}-0.38\%$
test_plain_set_stack_nested_inplace 80.1600μs 15.2946μs 65.3824 KOps/s 64.1450 KOps/s $\color{#35bf28}+1.93\%$
test_items 26.0600μs 5.5659μs 179.6643 KOps/s 183.5566 KOps/s $\color{#d91a1a}-2.12\%$
test_items_nested 0.4917ms 0.4378ms 2.2840 KOps/s 2.2972 KOps/s $\color{#d91a1a}-0.57\%$
test_items_nested_locked 0.4888ms 0.4421ms 2.2619 KOps/s 2.2843 KOps/s $\color{#d91a1a}-0.98\%$
test_items_nested_leaf 0.1375ms 90.3894μs 11.0632 KOps/s 10.9981 KOps/s $\color{#35bf28}+0.59\%$
test_items_stack_nested 0.5090ms 0.4398ms 2.2737 KOps/s 2.2751 KOps/s $\color{#d91a1a}-0.06\%$
test_items_stack_nested_leaf 0.1472ms 91.0759μs 10.9799 KOps/s 10.8294 KOps/s $\color{#35bf28}+1.39\%$
test_items_stack_nested_locked 0.6667ms 0.4458ms 2.2432 KOps/s 2.2699 KOps/s $\color{#d91a1a}-1.18\%$
test_keys 33.6400μs 4.1119μs 243.1979 KOps/s 243.2334 KOps/s $\color{#d91a1a}-0.01\%$
test_keys_nested 0.1854ms 0.1179ms 8.4822 KOps/s 8.5634 KOps/s $\color{#d91a1a}-0.95\%$
test_keys_nested_locked 88.0764ms 0.1384ms 7.2258 KOps/s 8.0489 KOps/s $\textbf{\color{#d91a1a}-10.23\%}$
test_keys_nested_leaf 0.1491ms 0.1068ms 9.3636 KOps/s 9.3328 KOps/s $\color{#35bf28}+0.33\%$
test_keys_stack_nested 0.1977ms 0.1174ms 8.5164 KOps/s 8.5607 KOps/s $\color{#d91a1a}-0.52\%$
test_keys_stack_nested_leaf 0.1534ms 0.1086ms 9.2046 KOps/s 9.3149 KOps/s $\color{#d91a1a}-1.18\%$
test_keys_stack_nested_locked 0.1618ms 0.1267ms 7.8899 KOps/s 8.0448 KOps/s $\color{#d91a1a}-1.92\%$
test_values 6.4382μs 0.9875μs 1.0127 MOps/s 1.0021 MOps/s $\color{#35bf28}+1.06\%$
test_values_nested 76.3800μs 46.5098μs 21.5008 KOps/s 21.4968 KOps/s $\color{#35bf28}+0.02\%$
test_values_nested_locked 90.5300μs 49.3952μs 20.2449 KOps/s 20.2962 KOps/s $\color{#d91a1a}-0.25\%$
test_values_nested_leaf 85.9810μs 52.6515μs 18.9928 KOps/s 18.9129 KOps/s $\color{#35bf28}+0.42\%$
test_values_stack_nested 74.5910μs 46.6376μs 21.4419 KOps/s 21.5835 KOps/s $\color{#d91a1a}-0.66\%$
test_values_stack_nested_leaf 91.4200μs 53.0960μs 18.8338 KOps/s 19.1366 KOps/s $\color{#d91a1a}-1.58\%$
test_values_stack_nested_locked 88.2110μs 49.5871μs 20.1665 KOps/s 20.2565 KOps/s $\color{#d91a1a}-0.44\%$
test_membership 4.7933μs 0.8054μs 1.2416 MOps/s 1.2761 MOps/s $\color{#d91a1a}-2.71\%$
test_membership_nested 28.0500μs 2.7320μs 366.0311 KOps/s 369.2657 KOps/s $\color{#d91a1a}-0.88\%$
test_membership_nested_leaf 15.2350μs 2.6698μs 374.5611 KOps/s 366.7795 KOps/s $\color{#35bf28}+2.12\%$
test_membership_stacked_nested 24.0900μs 2.7282μs 366.5369 KOps/s 362.4633 KOps/s $\color{#35bf28}+1.12\%$
test_membership_stacked_nested_leaf 32.8600μs 2.7457μs 364.2041 KOps/s 365.9633 KOps/s $\color{#d91a1a}-0.48\%$
test_membership_nested_last 31.9900μs 4.1029μs 243.7273 KOps/s 242.7725 KOps/s $\color{#35bf28}+0.39\%$
test_membership_nested_leaf_last 32.4210μs 4.0899μs 244.5043 KOps/s 242.5869 KOps/s $\color{#35bf28}+0.79\%$
test_membership_stacked_nested_last 31.1800μs 4.0296μs 248.1634 KOps/s 244.1694 KOps/s $\color{#35bf28}+1.64\%$
test_membership_stacked_nested_leaf_last 23.7400μs 4.0878μs 244.6294 KOps/s 243.5370 KOps/s $\color{#35bf28}+0.45\%$
test_nested_getleaf 50.0100μs 20.1954μs 49.5162 KOps/s 49.9058 KOps/s $\color{#d91a1a}-0.78\%$
test_nested_get 47.1200μs 19.6127μs 50.9875 KOps/s 53.0332 KOps/s $\color{#d91a1a}-3.86\%$
test_stacked_getleaf 64.8010μs 20.2254μs 49.4428 KOps/s 50.0684 KOps/s $\color{#d91a1a}-1.25\%$
test_stacked_get 50.9700μs 19.2559μs 51.9321 KOps/s 52.5576 KOps/s $\color{#d91a1a}-1.19\%$
test_nested_getitemleaf 48.1910μs 20.9110μs 47.8216 KOps/s 48.3083 KOps/s $\color{#d91a1a}-1.01\%$
test_nested_getitem 55.1000μs 19.7643μs 50.5963 KOps/s 50.3776 KOps/s $\color{#35bf28}+0.43\%$
test_stacked_getitemleaf 59.7500μs 20.9411μs 47.7529 KOps/s 48.1318 KOps/s $\color{#d91a1a}-0.79\%$
test_stacked_getitem 60.3400μs 19.7071μs 50.7431 KOps/s 51.0908 KOps/s $\color{#d91a1a}-0.68\%$
test_lock_nested 8.0514ms 0.4591ms 2.1782 KOps/s 2.1819 KOps/s $\color{#d91a1a}-0.17\%$
test_lock_stack_nested 0.5063ms 0.4552ms 2.1967 KOps/s 2.1700 KOps/s $\color{#35bf28}+1.23\%$
test_unlock_nested 0.4315ms 0.3648ms 2.7416 KOps/s 2.7322 KOps/s $\color{#35bf28}+0.34\%$
test_unlock_stack_nested 0.4045ms 0.3664ms 2.7295 KOps/s 2.6927 KOps/s $\color{#35bf28}+1.37\%$
test_flatten_speed 0.1612ms 0.1165ms 8.5864 KOps/s 8.5310 KOps/s $\color{#35bf28}+0.65\%$
test_unflatten_speed 0.6098ms 0.5485ms 1.8233 KOps/s 1.8516 KOps/s $\color{#d91a1a}-1.53\%$
test_common_ops 0.7858ms 0.6611ms 1.5127 KOps/s 1.4981 KOps/s $\color{#35bf28}+0.97\%$
test_creation 0.1235ms 2.7231μs 367.2309 KOps/s 357.8159 KOps/s $\color{#35bf28}+2.63\%$
test_creation_empty 38.0200μs 5.7770μs 173.0989 KOps/s 171.7527 KOps/s $\color{#35bf28}+0.78\%$
test_creation_nested_1 47.0710μs 10.0055μs 99.9454 KOps/s 98.2883 KOps/s $\color{#35bf28}+1.69\%$
test_creation_nested_2 36.5200μs 11.2037μs 89.2566 KOps/s 88.7832 KOps/s $\color{#35bf28}+0.53\%$
test_creation_many_keys[10] 45.7510μs 17.0403μs 58.6844 KOps/s 58.8699 KOps/s $\color{#d91a1a}-0.32\%$
test_creation_many_keys[50] 0.1055ms 72.3978μs 13.8126 KOps/s 13.7089 KOps/s $\color{#35bf28}+0.76\%$
test_creation_many_keys[100] 0.1766ms 0.1415ms 7.0664 KOps/s 7.0694 KOps/s $\color{#d91a1a}-0.04\%$
test_creation_nested_many_keys[10] 63.5000μs 36.7012μs 27.2470 KOps/s 27.2835 KOps/s $\color{#d91a1a}-0.13\%$
test_creation_nested_many_keys[50] 0.1825ms 0.1499ms 6.6694 KOps/s 6.7356 KOps/s $\color{#d91a1a}-0.98\%$
test_clone 42.6600μs 13.0327μs 76.7301 KOps/s 76.8654 KOps/s $\color{#d91a1a}-0.18\%$
test_getitem[int] 1.7850ms 14.1041μs 70.9016 KOps/s 59.1116 KOps/s $\textbf{\color{#35bf28}+19.95\%}$
test_getitem[slice_int] 0.1344ms 23.3554μs 42.8167 KOps/s 42.2889 KOps/s $\color{#35bf28}+1.25\%$
test_getitem[range] 0.1879ms 59.5805μs 16.7840 KOps/s 16.3950 KOps/s $\color{#35bf28}+2.37\%$
test_getitem[tuple] 0.1415ms 23.3473μs 42.8315 KOps/s 43.2983 KOps/s $\color{#d91a1a}-1.08\%$
test_getitem[list] 0.1803ms 55.7535μs 17.9361 KOps/s 17.8372 KOps/s $\color{#35bf28}+0.55\%$
test_setitem_dim[int] 43.9910μs 25.2524μs 39.6002 KOps/s 37.8951 KOps/s $\color{#35bf28}+4.50\%$
test_setitem_dim[slice_int] 69.6100μs 42.4042μs 23.5826 KOps/s 23.6955 KOps/s $\color{#d91a1a}-0.48\%$
test_setitem_dim[range] 0.1135ms 90.7855μs 11.0150 KOps/s 10.9570 KOps/s $\color{#35bf28}+0.53\%$
test_setitem_dim[tuple] 61.2110μs 39.5875μs 25.2605 KOps/s 24.9198 KOps/s $\color{#35bf28}+1.37\%$
test_setitem 52.2400μs 17.2514μs 57.9662 KOps/s 57.8104 KOps/s $\color{#35bf28}+0.27\%$
test_set 53.8310μs 16.5924μs 60.2686 KOps/s 59.7114 KOps/s $\color{#35bf28}+0.93\%$
test_set_shared 0.4907ms 0.2024ms 4.9417 KOps/s 4.8319 KOps/s $\color{#35bf28}+2.27\%$
test_update 0.3301ms 21.1195μs 47.3497 KOps/s 47.5982 KOps/s $\color{#d91a1a}-0.52\%$
test_update_nested 68.3810μs 32.4427μs 30.8236 KOps/s 30.9067 KOps/s $\color{#d91a1a}-0.27\%$
test_update__nested 0.4736ms 33.4566μs 29.8894 KOps/s 29.4572 KOps/s $\color{#35bf28}+1.47\%$
test_set_nested 53.1410μs 18.2627μs 54.7564 KOps/s 54.1384 KOps/s $\color{#35bf28}+1.14\%$
test_set_nested_new 59.1500μs 23.0099μs 43.4596 KOps/s 42.5226 KOps/s $\color{#35bf28}+2.20\%$
test_select 79.7100μs 40.6065μs 24.6266 KOps/s 24.8924 KOps/s $\color{#d91a1a}-1.07\%$
test_select_nested 0.1013ms 71.3638μs 14.0127 KOps/s 14.1231 KOps/s $\color{#d91a1a}-0.78\%$
test_exclude_nested 0.1160ms 88.4301μs 11.3084 KOps/s 11.4363 KOps/s $\color{#d91a1a}-1.12\%$
test_empty[True] 0.4298ms 0.3768ms 2.6541 KOps/s 2.6388 KOps/s $\color{#35bf28}+0.58\%$
test_empty[False] 8.1525μs 1.2554μs 796.5497 KOps/s 790.3976 KOps/s $\color{#35bf28}+0.78\%$
test_to 0.1016ms 70.2281μs 14.2393 KOps/s 14.1596 KOps/s $\color{#35bf28}+0.56\%$
test_to_nonblocking 0.1049ms 63.4520μs 15.7600 KOps/s 15.3554 KOps/s $\color{#35bf28}+2.63\%$
test_unbind_speed 0.3458ms 0.3134ms 3.1913 KOps/s 3.1867 KOps/s $\color{#35bf28}+0.14\%$
test_unbind_speed_stack0 0.3784ms 0.3087ms 3.2391 KOps/s 3.2312 KOps/s $\color{#35bf28}+0.25\%$
test_unbind_speed_stack1 0.1035s 0.8813ms 1.1347 KOps/s 1.1147 KOps/s $\color{#35bf28}+1.79\%$
test_split 1.1734ms 1.0981ms 910.6525 Ops/s 911.6260 Ops/s $\color{#d91a1a}-0.11\%$
test_chunk 0.1034s 1.1632ms 859.6896 Ops/s 956.1808 Ops/s $\textbf{\color{#d91a1a}-10.09\%}$
test_to_cpu_blocking 19.5360ms 19.4480ms 51.4192 Ops/s 31.8989 Ops/s $\textbf{\color{#35bf28}+61.19\%}$
test_to_cpu_global_sync 11.1944ms 11.0936ms 90.1419 Ops/s 90.2407 Ops/s $\color{#d91a1a}-0.11\%$
test_to_cpu_event_sync 0.1154s 13.3157ms 75.0995 Ops/s 82.9582 Ops/s $\textbf{\color{#d91a1a}-9.47\%}$
test_to_cpu_default 12.3538ms 12.0547ms 82.9550 Ops/s 83.1476 Ops/s $\color{#d91a1a}-0.23\%$
test_consolidate[False-None] 4.1657ms 3.9666ms 252.1056 Ops/s 227.7500 Ops/s $\textbf{\color{#35bf28}+10.69\%}$
test_consolidate[default-None] 2.4626ms 1.9425ms 514.8091 Ops/s 499.6564 Ops/s $\color{#35bf28}+3.03\%$
test_consolidate[reduce-overhead-None] 1.9823ms 1.8715ms 534.3216 Ops/s 520.8757 Ops/s $\color{#35bf28}+2.58\%$
test_consolidate_njt[False-None] 8.3651ms 8.1433ms 122.8002 Ops/s 121.7620 Ops/s $\color{#35bf28}+0.85\%$
test_to[False-False-None] 2.1024ms 2.0244ms 493.9720 Ops/s 488.7688 Ops/s $\color{#35bf28}+1.06\%$
test_to[True-False-None] 1.9877ms 1.8619ms 537.0930 Ops/s 539.2482 Ops/s $\color{#d91a1a}-0.40\%$
test_to[within-False-None] 6.1598ms 5.9302ms 168.6294 Ops/s 168.7865 Ops/s $\color{#d91a1a}-0.09\%$
test_to[True-default-None] 7.9604ms 7.6360ms 130.9588 Ops/s 131.9401 Ops/s $\color{#d91a1a}-0.74\%$
test_to_njt[False-False-None] 8.4312ms 8.3200ms 120.1925 Ops/s 119.4379 Ops/s $\color{#35bf28}+0.63\%$
test_to_njt[True-False-None] 6.8647ms 6.7321ms 148.5430 Ops/s 147.6597 Ops/s $\color{#35bf28}+0.60\%$
test_to_njt[within-False-None] 15.1545ms 15.0404ms 66.4876 Ops/s 66.2399 Ops/s $\color{#35bf28}+0.37\%$
test_creation[device0] 0.3976ms 0.1123ms 8.9036 KOps/s 8.5785 KOps/s $\color{#35bf28}+3.79\%$
test_creation_from_tensor 0.3632ms 0.1113ms 8.9859 KOps/s 8.6752 KOps/s $\color{#35bf28}+3.58\%$
test_add_one[memmap_tensor0] 0.2356ms 6.3663μs 157.0768 KOps/s 152.0291 KOps/s $\color{#35bf28}+3.32\%$
test_contiguous[memmap_tensor0] 38.7400μs 0.6185μs 1.6168 MOps/s 2.3176 MOps/s $\textbf{\color{#d91a1a}-30.24\%}$
test_stack[memmap_tensor0] 21.6000μs 4.5798μs 218.3518 KOps/s 215.9906 KOps/s $\color{#35bf28}+1.09\%$
test_memmaptd_index 1.0258ms 0.2517ms 3.9722 KOps/s 3.8981 KOps/s $\color{#35bf28}+1.90\%$
test_memmaptd_index_astensor 0.5057ms 0.3427ms 2.9184 KOps/s 2.9087 KOps/s $\color{#35bf28}+0.33\%$
test_memmaptd_index_op 0.7575ms 0.5783ms 1.7291 KOps/s 1.6977 KOps/s $\color{#35bf28}+1.85\%$
test_serialize_model 0.1369s 0.1361s 7.3496 Ops/s 7.3395 Ops/s $\color{#35bf28}+0.14\%$
test_serialize_model_pickle 1.3493s 1.1921s 0.8389 Ops/s 0.8220 Ops/s $\color{#35bf28}+2.06\%$
test_serialize_weights 0.1375s 0.1345s 7.4369 Ops/s 7.4405 Ops/s $\color{#d91a1a}-0.05\%$
test_serialize_weights_returnearly 0.4595s 92.1191ms 10.8555 Ops/s 5.6540 Ops/s $\textbf{\color{#35bf28}+92.00\%}$
test_serialize_weights_pickle 1.3639s 1.2143s 0.8235 Ops/s 0.8237 Ops/s $\color{#d91a1a}-0.02\%$
test_reshape_pytree 0.2162ms 31.6419μs 31.6037 KOps/s 30.8184 KOps/s $\color{#35bf28}+2.55\%$
test_reshape_td 76.6310μs 42.8581μs 23.3328 KOps/s 23.0528 KOps/s $\color{#35bf28}+1.21\%$
test_view_pytree 0.2223ms 31.5124μs 31.7336 KOps/s 31.3523 KOps/s $\color{#35bf28}+1.22\%$
test_view_td 0.1016ms 49.9331μs 20.0268 KOps/s 19.8698 KOps/s $\color{#35bf28}+0.79\%$
test_unbind_pytree 0.2502ms 34.9657μs 28.5995 KOps/s 27.7785 KOps/s $\color{#35bf28}+2.96\%$
test_unbind_td 0.1115ms 46.1940μs 21.6478 KOps/s 20.8098 KOps/s $\color{#35bf28}+4.03\%$
test_split_pytree 0.2508ms 40.1178μs 24.9266 KOps/s 24.3961 KOps/s $\color{#35bf28}+2.17\%$
test_split_td 0.1934ms 62.7943μs 15.9250 KOps/s 16.0672 KOps/s $\color{#d91a1a}-0.88\%$
test_add_pytree 0.2306ms 40.2865μs 24.8222 KOps/s 24.5846 KOps/s $\color{#35bf28}+0.97\%$
test_add_td 98.7810μs 52.7898μs 18.9430 KOps/s 18.8700 KOps/s $\color{#35bf28}+0.39\%$
test_compile_add_one_nested[tensordict-compile] 0.2208ms 0.1387ms 7.2115 KOps/s 6.6257 KOps/s $\textbf{\color{#35bf28}+8.84\%}$
test_compile_add_one_nested[tensordict-eager] 0.4010ms 0.1899ms 5.2659 KOps/s 5.4628 KOps/s $\color{#d91a1a}-3.60\%$
test_compile_add_one_nested[pytree-compile] 0.1586ms 0.1098ms 9.1098 KOps/s 9.3944 KOps/s $\color{#d91a1a}-3.03\%$
test_compile_add_one_nested[pytree-eager] 0.4343ms 0.1737ms 5.7557 KOps/s 5.6542 KOps/s $\color{#35bf28}+1.80\%$
test_compile_copy_nested[tensordict-compile] 82.9110μs 29.4786μs 33.9230 KOps/s 31.3523 KOps/s $\textbf{\color{#35bf28}+8.20\%}$
test_compile_copy_nested[tensordict-eager] 90.8410μs 49.1683μs 20.3383 KOps/s 20.0578 KOps/s $\color{#35bf28}+1.40\%$
test_compile_copy_nested[pytree-compile] 46.2610μs 9.4532μs 105.7842 KOps/s 103.8834 KOps/s $\color{#35bf28}+1.83\%$
test_compile_copy_nested[pytree-eager] 0.4319ms 64.5232μs 15.4983 KOps/s 15.0906 KOps/s $\color{#35bf28}+2.70\%$
test_compile_add_one_flat[tensordict-compile] 0.2134ms 0.1734ms 5.7665 KOps/s 5.5841 KOps/s $\color{#35bf28}+3.27\%$
test_compile_add_one_flat[tensordict-eager] 0.3045ms 0.2511ms 3.9818 KOps/s 4.0140 KOps/s $\color{#d91a1a}-0.80\%$
test_compile_add_one_flat[tensorclass-compile] 0.1649ms 0.1137ms 8.7959 KOps/s 8.3489 KOps/s $\textbf{\color{#35bf28}+5.35\%}$
test_compile_add_one_flat[tensorclass-eager] 0.1275ms 69.4304μs 14.4029 KOps/s 14.5246 KOps/s $\color{#d91a1a}-0.84\%$
test_compile_add_one_flat[pytree-compile] 0.2458ms 0.1563ms 6.3999 KOps/s 4.0314 KOps/s $\textbf{\color{#35bf28}+58.75\%}$
test_compile_add_one_flat[pytree-eager] 0.8033ms 0.5221ms 1.9152 KOps/s 1.9087 KOps/s $\color{#35bf28}+0.34\%$
test_compile_add_self_flat[tensordict-eager] 0.3516ms 0.3024ms 3.3067 KOps/s 3.2806 KOps/s $\color{#35bf28}+0.79\%$
test_compile_add_self_flat[tensordict-compile] 0.2388ms 0.1756ms 5.6952 KOps/s 5.2744 KOps/s $\textbf{\color{#35bf28}+7.98\%}$
test_compile_add_self_flat[tensorclass-eager] 0.1423ms 83.0355μs 12.0430 KOps/s 12.1295 KOps/s $\color{#d91a1a}-0.71\%$
test_compile_add_self_flat[tensorclass-compile] 0.1819ms 0.1162ms 8.6066 KOps/s 8.2713 KOps/s $\color{#35bf28}+4.05\%$
test_compile_add_self_flat[pytree-eager] 0.6608ms 0.4413ms 2.2658 KOps/s 2.2480 KOps/s $\color{#35bf28}+0.79\%$
test_compile_add_self_flat[pytree-compile] 0.2024ms 0.1558ms 6.4203 KOps/s 6.2618 KOps/s $\color{#35bf28}+2.53\%$
test_compile_copy_flat[tensordict-compile] 51.7410μs 23.3962μs 42.7419 KOps/s 39.0600 KOps/s $\textbf{\color{#35bf28}+9.43\%}$
test_compile_copy_flat[tensordict-eager] 78.6710μs 39.8832μs 25.0732 KOps/s 24.7171 KOps/s $\color{#35bf28}+1.44\%$
test_compile_copy_flat[pytree-compile] 50.2000μs 10.4051μs 96.1068 KOps/s 93.5315 KOps/s $\color{#35bf28}+2.75\%$
test_compile_copy_flat[pytree-eager] 0.3970ms 51.2047μs 19.5295 KOps/s 19.4470 KOps/s $\color{#35bf28}+0.42\%$
test_compile_assign_and_add[tensordict-compile] 1.9528ms 0.1701ms 5.8781 KOps/s 5.5661 KOps/s $\textbf{\color{#35bf28}+5.61\%}$
test_compile_assign_and_add[tensordict-eager] 3.5159ms 3.2686ms 305.9383 Ops/s 309.9926 Ops/s $\color{#d91a1a}-1.31\%$
test_compile_assign_and_add[pytree-compile] 1.9000ms 0.1579ms 6.3346 KOps/s 6.1689 KOps/s $\color{#35bf28}+2.69\%$
test_compile_assign_and_add[pytree-eager] 2.8987ms 2.7491ms 363.7551 Ops/s 360.0438 Ops/s $\color{#35bf28}+1.03\%$
test_compile_indexing[tensor-tensordict-compile] 0.1586ms 0.1059ms 9.4395 KOps/s 9.1659 KOps/s $\color{#35bf28}+2.99\%$
test_compile_indexing[tensor-tensordict-eager] 0.3144ms 71.1967μs 14.0456 KOps/s 13.8042 KOps/s $\color{#35bf28}+1.75\%$
test_compile_indexing[tensor-tensorclass-compile] 0.1913ms 94.5066μs 10.5813 KOps/s 10.5641 KOps/s $\color{#35bf28}+0.16\%$
test_compile_indexing[tensor-tensorclass-eager] 0.2549ms 43.8653μs 22.7971 KOps/s 22.3727 KOps/s $\color{#35bf28}+1.90\%$
test_compile_indexing[tensor-pytree-compile] 0.1376ms 94.6680μs 10.5632 KOps/s 10.5462 KOps/s $\color{#35bf28}+0.16\%$
test_compile_indexing[tensor-pytree-eager] 0.2693ms 45.2956μs 22.0772 KOps/s 22.3890 KOps/s $\color{#d91a1a}-1.39\%$
test_compile_indexing[slice-tensordict-compile] 0.1172ms 54.6368μs 18.3027 KOps/s 17.5794 KOps/s $\color{#35bf28}+4.11\%$
test_compile_indexing[slice-tensordict-eager] 0.2329ms 27.1418μs 36.8435 KOps/s 36.9756 KOps/s $\color{#d91a1a}-0.36\%$
test_compile_indexing[slice-tensorclass-compile] 96.4700μs 45.7684μs 21.8491 KOps/s 22.1612 KOps/s $\color{#d91a1a}-1.41\%$
test_compile_indexing[slice-tensorclass-eager] 0.2595ms 21.7225μs 46.0352 KOps/s 45.2774 KOps/s $\color{#35bf28}+1.67\%$
test_compile_indexing[slice-pytree-compile] 0.1352ms 44.5839μs 22.4296 KOps/s 21.7459 KOps/s $\color{#35bf28}+3.14\%$
test_compile_indexing[slice-pytree-eager] 0.2765ms 21.4236μs 46.6774 KOps/s 45.3739 KOps/s $\color{#35bf28}+2.87\%$
test_compile_indexing[int-tensordict-compile] 95.9710μs 54.4138μs 18.3777 KOps/s 17.3452 KOps/s $\textbf{\color{#35bf28}+5.95\%}$
test_compile_indexing[int-tensordict-eager] 0.2482ms 26.4892μs 37.7512 KOps/s 37.3982 KOps/s $\color{#35bf28}+0.94\%$
test_compile_indexing[int-tensorclass-compile] 0.1087ms 44.8159μs 22.3135 KOps/s 21.9887 KOps/s $\color{#35bf28}+1.48\%$
test_compile_indexing[int-tensorclass-eager] 0.2730ms 21.5639μs 46.3737 KOps/s 45.4474 KOps/s $\color{#35bf28}+2.04\%$
test_compile_indexing[int-pytree-compile] 87.4110μs 44.3585μs 22.5436 KOps/s 21.9623 KOps/s $\color{#35bf28}+2.65\%$
test_compile_indexing[int-pytree-eager] 0.2659ms 21.6193μs 46.2551 KOps/s 45.1744 KOps/s $\color{#35bf28}+2.39\%$
test_mod_add[eager] 0.1233ms 48.2906μs 20.7080 KOps/s 20.4082 KOps/s $\color{#35bf28}+1.47\%$
test_mod_add[compile] 0.1695ms 0.1012ms 9.8856 KOps/s 9.5573 KOps/s $\color{#35bf28}+3.44\%$
test_mod_add[compile-overhead] 0.2276ms 0.1444ms 6.9260 KOps/s 6.7960 KOps/s $\color{#35bf28}+1.91\%$
test_mod_wrap[eager] 0.3594ms 0.2818ms 3.5484 KOps/s 3.4937 KOps/s $\color{#35bf28}+1.57\%$
test_mod_wrap[compile] 0.3960ms 0.3389ms 2.9506 KOps/s 2.8922 KOps/s $\color{#35bf28}+2.02\%$
test_mod_wrap[compile-overhead] 7.9398ms 4.0935ms 244.2909 Ops/s 253.5831 Ops/s $\color{#d91a1a}-3.66\%$
test_mod_wrap_and_backward[eager] 1.5891ms 1.4651ms 682.5463 Ops/s 676.5763 Ops/s $\color{#35bf28}+0.88\%$
test_mod_wrap_and_backward[compile] 1.5114ms 1.4167ms 705.8853 Ops/s 699.3496 Ops/s $\color{#35bf28}+0.93\%$
test_mod_wrap_and_backward[compile-overhead] 1.2042ms 0.8609ms 1.1615 KOps/s 1.1340 KOps/s $\color{#35bf28}+2.43\%$
test_seq_add[eager] 0.2023ms 0.1515ms 6.6025 KOps/s 6.3502 KOps/s $\color{#35bf28}+3.97\%$
test_seq_add[compile] 0.5721ms 0.1105ms 9.0505 KOps/s 8.5347 KOps/s $\textbf{\color{#35bf28}+6.04\%}$
test_seq_add[compile-overhead] 0.3003ms 0.1496ms 6.6826 KOps/s 6.1528 KOps/s $\textbf{\color{#35bf28}+8.61\%}$
test_seq_wrap[eager] 0.7070ms 0.5092ms 1.9641 KOps/s 1.8617 KOps/s $\textbf{\color{#35bf28}+5.50\%}$
test_seq_wrap[compile] 0.4001ms 0.3567ms 2.8034 KOps/s 2.6367 KOps/s $\textbf{\color{#35bf28}+6.32\%}$
test_seq_wrap[compile-overhead] 0.3387ms 0.2563ms 3.9013 KOps/s 3.7080 KOps/s $\textbf{\color{#35bf28}+5.21\%}$
test_func_call_runtime[False-eager] 0.9052ms 0.8228ms 1.2153 KOps/s 1.1389 KOps/s $\textbf{\color{#35bf28}+6.71\%}$
test_func_call_runtime[False-compile] 1.0706ms 0.8811ms 1.1350 KOps/s 1.1092 KOps/s $\color{#35bf28}+2.32\%$
test_func_call_runtime[False-compile-overhead] 0.5014ms 0.4425ms 2.2596 KOps/s 2.2428 KOps/s $\color{#35bf28}+0.75\%$
test_func_call_runtime[True-eager] 1.1107ms 1.0554ms 947.5033 Ops/s 938.7233 Ops/s $\color{#35bf28}+0.94\%$
test_func_call_runtime[True-compile] 1.0171ms 0.8969ms 1.1150 KOps/s 1.1058 KOps/s $\color{#35bf28}+0.83\%$
test_func_call_runtime[True-compile-overhead] 0.5197ms 0.4542ms 2.2019 KOps/s 2.1643 KOps/s $\color{#35bf28}+1.74\%$
test_func_call_cm_runtime[False-eager] 0.8932ms 0.8201ms 1.2193 KOps/s 1.1636 KOps/s $\color{#35bf28}+4.79\%$
test_func_call_cm_runtime[False-compile] 1.0584ms 0.8908ms 1.1226 KOps/s 1.1013 KOps/s $\color{#35bf28}+1.93\%$
test_func_call_cm_runtime[False-compile-overhead] 0.5073ms 0.4448ms 2.2482 KOps/s 2.2282 KOps/s $\color{#35bf28}+0.90\%$
test_func_call_cm_runtime[True-eager] 1.2808ms 1.1887ms 841.2635 Ops/s 825.1927 Ops/s $\color{#35bf28}+1.95\%$
test_func_call_cm_runtime[True-compile] 0.9732ms 0.9272ms 1.0785 KOps/s 1.0609 KOps/s $\color{#35bf28}+1.65\%$
test_func_call_cm_runtime[True-compile-overhead] 0.5281ms 0.4862ms 2.0570 KOps/s 1.9980 KOps/s $\color{#35bf28}+2.95\%$
test_vmap_func_call_cm_runtime[eager] 2.8170ms 2.3121ms 432.5073 Ops/s 423.1005 Ops/s $\color{#35bf28}+2.22\%$
test_vmap_func_call_cm_runtime[compile] 1.0075ms 0.9446ms 1.0587 KOps/s 1.0245 KOps/s $\color{#35bf28}+3.34\%$
test_vmap_func_call_cm_runtime[compile-overhead] 0.5414ms 0.4908ms 2.0377 KOps/s 1.9931 KOps/s $\color{#35bf28}+2.24\%$
test_distributed 2.9094ms 0.1656ms 6.0379 KOps/s 6.4812 KOps/s $\textbf{\color{#d91a1a}-6.84\%}$
test_tdmodule 49.6500μs 27.2469μs 36.7014 KOps/s 36.3603 KOps/s $\color{#35bf28}+0.94\%$
test_tdmodule_dispatch 76.4100μs 44.1266μs 22.6621 KOps/s 22.5196 KOps/s $\color{#35bf28}+0.63\%$
test_tdseq 47.3100μs 26.2750μs 38.0589 KOps/s 37.9073 KOps/s $\color{#35bf28}+0.40\%$
test_tdseq_dispatch 66.9410μs 46.1537μs 21.6668 KOps/s 21.4726 KOps/s $\color{#35bf28}+0.90\%$
test_instantiation_functorch 2.0604ms 1.9845ms 503.8979 Ops/s 504.2883 Ops/s $\color{#d91a1a}-0.08\%$
test_exec_functorch 0.2194ms 0.1756ms 5.6938 KOps/s 5.6681 KOps/s $\color{#35bf28}+0.45\%$
test_exec_functional_call 0.2029ms 0.1569ms 6.3718 KOps/s 6.3689 KOps/s $\color{#35bf28}+0.05\%$
test_exec_td_decorator 0.4376ms 0.2284ms 4.3778 KOps/s 4.3430 KOps/s $\color{#35bf28}+0.80\%$
test_vmap_mlp_speed_decorator[True-True] 1.0239ms 0.8017ms 1.2474 KOps/s 1.2324 KOps/s $\color{#35bf28}+1.22\%$
test_vmap_mlp_speed_decorator[True-False] 0.9795ms 0.8047ms 1.2428 KOps/s 1.2298 KOps/s $\color{#35bf28}+1.06\%$
test_vmap_mlp_speed_decorator[False-True] 0.9112ms 0.6941ms 1.4406 KOps/s 1.4165 KOps/s $\color{#35bf28}+1.70\%$
test_vmap_mlp_speed_decorator[False-False] 0.8539ms 0.6943ms 1.4403 KOps/s 1.4193 KOps/s $\color{#35bf28}+1.48\%$
test_vmap_transformer_speed_decorator[True-True] 20.2330ms 20.1184ms 49.7058 Ops/s 49.4175 Ops/s $\color{#35bf28}+0.58\%$
test_vmap_transformer_speed_decorator[True-False] 20.2409ms 20.1289ms 49.6799 Ops/s 49.3416 Ops/s $\color{#35bf28}+0.69\%$
test_vmap_transformer_speed_decorator[False-True] 20.4440ms 19.9438ms 50.1409 Ops/s 49.7801 Ops/s $\color{#35bf28}+0.72\%$
test_vmap_transformer_speed_decorator[False-False] 20.6112ms 19.9542ms 50.1147 Ops/s 49.8126 Ops/s $\color{#35bf28}+0.61\%$
test_to_module_speed[True] 2.0029ms 1.4111ms 708.6671 Ops/s 708.8523 Ops/s $\color{#d91a1a}-0.03\%$
test_to_module_speed[False] 1.8816ms 1.3775ms 725.9282 Ops/s 720.6599 Ops/s $\color{#35bf28}+0.73\%$
test_tc_init 71.5710μs 44.5356μs 22.4539 KOps/s 22.3684 KOps/s $\color{#35bf28}+0.38\%$
test_tc_init_tensor_only 46.2010μs 9.3942μs 106.4489 KOps/s 106.8625 KOps/s $\color{#d91a1a}-0.39\%$
test_tc_init_nested 0.1272ms 89.0561μs 11.2289 KOps/s 11.2762 KOps/s $\color{#d91a1a}-0.42\%$
test_tc_init_many_fields 39.2210μs 15.6112μs 64.0565 KOps/s 63.5755 KOps/s $\color{#35bf28}+0.76\%$
test_tc_first_layer_tensor 18.0000μs 1.7325μs 577.2087 KOps/s 572.8015 KOps/s $\color{#35bf28}+0.77\%$
test_tc_first_layer_tensor_only 5.9071μs 0.7002μs 1.4282 MOps/s 1.3989 MOps/s $\color{#35bf28}+2.10\%$
test_tc_first_layer_tensor_set 26.0510μs 3.7328μs 267.8950 KOps/s 264.7982 KOps/s $\color{#35bf28}+1.17\%$
test_tc_first_layer_tensor_only_set 18.1410μs 3.1224μs 320.2711 KOps/s 329.7414 KOps/s $\color{#d91a1a}-2.87\%$
test_tc_first_layer_nontensor 30.6500μs 5.8047μs 172.2750 KOps/s 170.8178 KOps/s $\color{#35bf28}+0.85\%$
test_tc_second_layer_tensor 25.1000μs 4.1238μs 242.4960 KOps/s 236.1437 KOps/s $\color{#35bf28}+2.69\%$
test_tc_second_layer_nontensor 39.7500μs 8.2578μs 121.0970 KOps/s 118.1727 KOps/s $\color{#35bf28}+2.47\%$
test_unbind 0.2594s 15.6458ms 63.9150 Ops/s 64.4928 Ops/s $\color{#d91a1a}-0.90\%$
test_full_like 4.7382ms 4.3637ms 229.1625 Ops/s 228.4654 Ops/s $\color{#35bf28}+0.31\%$
test_zeros_like 4.6763ms 4.3532ms 229.7152 Ops/s 228.2477 Ops/s $\color{#35bf28}+0.64\%$
test_ones_like 4.5154ms 4.3593ms 229.3928 Ops/s 229.0494 Ops/s $\color{#35bf28}+0.15\%$
test_clone 6.6022ms 6.4467ms 155.1180 Ops/s 155.2733 Ops/s $\color{#d91a1a}-0.10\%$
test_squeeze 65.1600μs 13.9064μs 71.9095 KOps/s 72.8937 KOps/s $\color{#d91a1a}-1.35\%$
test_unsqueeze 0.1612ms 0.1073ms 9.3208 KOps/s 9.2323 KOps/s $\color{#35bf28}+0.96\%$
test_split 0.2925ms 0.1762ms 5.6762 KOps/s 5.5547 KOps/s $\color{#35bf28}+2.19\%$
test_permute 0.2440ms 0.1981ms 5.0473 KOps/s 4.9845 KOps/s $\color{#35bf28}+1.26\%$
test_stack 51.5213ms 51.2936ms 19.4956 Ops/s 19.5653 Ops/s $\color{#d91a1a}-0.36\%$
test_cat 51.7249ms 51.3488ms 19.4747 Ops/s 19.5707 Ops/s $\color{#d91a1a}-0.49\%$

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.

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

1 participant

@vmoens