[BugFix] TorchScript compat

[vmoens]

Stack from ghstack (oldest at bottom):

• -> [BugFix] TorchScript compat #1141

[github-actions]

$\color{#D29922}\textsf{\Large⚠\kern{0.2cm}\normalsize Warning}$ Result of CPU Benchmark Tests

Total Benchmarks: 217. Improved: $\large\color{#35bf28}12$. Worsened: $\large\color{#d91a1a}19$.

Expand to view detailed results

Name	Max	Mean	Ops	Ops on Repo HEAD	Change
test_plain_set_nested	45.0340μs	21.2680μs	47.0190 KOps/s	49.3471 KOps/s	$\color{#d91a1a}-4.72\%$
test_plain_set_stack_nested	72.8260μs	21.4210μs	46.6832 KOps/s	48.5349 KOps/s	$\color{#d91a1a}-3.82\%$
test_plain_set_nested_inplace	66.4430μs	23.1648μs	43.1690 KOps/s	44.5391 KOps/s	$\color{#d91a1a}-3.08\%$
test_plain_set_stack_nested_inplace	0.1033ms	22.8574μs	43.7496 KOps/s	44.8863 KOps/s	$\color{#d91a1a}-2.53\%$
test_items	23.5440μs	4.2501μs	235.2870 KOps/s	237.6643 KOps/s	$\color{#d91a1a}-1.00\%$
test_items_nested	0.5655ms	0.4046ms	2.4717 KOps/s	2.4629 KOps/s	$\color{#35bf28}+0.36\%$
test_items_nested_locked	0.7380ms	0.4105ms	2.4358 KOps/s	2.4542 KOps/s	$\color{#d91a1a}-0.75\%$
test_items_nested_leaf	0.1361ms	76.4867μs	13.0742 KOps/s	12.8085 KOps/s	$\color{#35bf28}+2.07\%$
test_items_stack_nested	0.8881ms	0.4080ms	2.4508 KOps/s	2.4283 KOps/s	$\color{#35bf28}+0.93\%$
test_items_stack_nested_leaf	0.1346ms	77.7326μs	12.8646 KOps/s	12.1708 KOps/s	$\textbf{\color{#35bf28}+5.70\%}$
test_items_stack_nested_locked	0.8524ms	0.4097ms	2.4407 KOps/s	2.4309 KOps/s	$\color{#35bf28}+0.40\%$
test_keys	21.6800μs	3.4776μs	287.5550 KOps/s	282.8066 KOps/s	$\color{#35bf28}+1.68\%$
test_keys_nested	0.2673ms	0.1650ms	6.0605 KOps/s	5.9278 KOps/s	$\color{#35bf28}+2.24\%$
test_keys_nested_locked	0.7157ms	0.1713ms	5.8381 KOps/s	5.6741 KOps/s	$\color{#35bf28}+2.89\%$
test_keys_nested_leaf	1.5820ms	0.1437ms	6.9579 KOps/s	6.7412 KOps/s	$\color{#35bf28}+3.22\%$
test_keys_stack_nested	0.3025ms	0.1657ms	6.0348 KOps/s	6.0176 KOps/s	$\color{#35bf28}+0.29\%$
test_keys_stack_nested_leaf	0.2596ms	0.1443ms	6.9313 KOps/s	6.9857 KOps/s	$\color{#d91a1a}-0.78\%$
test_keys_stack_nested_locked	0.3768ms	0.1718ms	5.8205 KOps/s	5.8190 KOps/s	$\color{#35bf28}+0.03\%$
test_values	9.6040μs	1.0361μs	965.1129 KOps/s	945.1036 KOps/s	$\color{#35bf28}+2.12\%$
test_values_nested	0.1315ms	62.4538μs	16.0118 KOps/s	15.7063 KOps/s	$\color{#35bf28}+1.95\%$
test_values_nested_locked	0.1200ms	62.9714μs	15.8802 KOps/s	15.8867 KOps/s	$\color{#d91a1a}-0.04\%$
test_values_nested_leaf	0.1293ms	72.3465μs	13.8224 KOps/s	13.6269 KOps/s	$\color{#35bf28}+1.43\%$
test_values_stack_nested	0.1157ms	63.9879μs	15.6280 KOps/s	15.4913 KOps/s	$\color{#35bf28}+0.88\%$
test_values_stack_nested_leaf	0.1374ms	72.7479μs	13.7461 KOps/s	13.8263 KOps/s	$\color{#d91a1a}-0.58\%$
test_values_stack_nested_locked	0.1471ms	63.1650μs	15.8316 KOps/s	15.4866 KOps/s	$\color{#35bf28}+2.23\%$
test_membership	40.1250μs	0.8970μs	1.1148 MOps/s	1.1453 MOps/s	$\color{#d91a1a}-2.66\%$
test_membership_nested	21.9510μs	3.0274μs	330.3189 KOps/s	337.9826 KOps/s	$\color{#d91a1a}-2.27\%$
test_membership_nested_leaf	42.0080μs	3.0863μs	324.0163 KOps/s	338.0672 KOps/s	$\color{#d91a1a}-4.16\%$
test_membership_stacked_nested	20.5980μs	3.0273μs	330.3231 KOps/s	345.1090 KOps/s	$\color{#d91a1a}-4.28\%$
test_membership_stacked_nested_leaf	33.4920μs	3.0426μs	328.6687 KOps/s	341.5634 KOps/s	$\color{#d91a1a}-3.78\%$
test_membership_nested_last	26.5300μs	4.5087μs	221.7919 KOps/s	221.6974 KOps/s	$\color{#35bf28}+0.04\%$
test_membership_nested_leaf_last	48.2100μs	4.5469μs	219.9301 KOps/s	224.2752 KOps/s	$\color{#d91a1a}-1.94\%$
test_membership_stacked_nested_last	28.0530μs	5.3623μs	186.4878 KOps/s	74.3643 KOps/s	$\textbf{\color{#35bf28}+150.78\%}$
test_membership_stacked_nested_leaf_last	47.8490μs	5.4287μs	184.2054 KOps/s	74.1934 KOps/s	$\textbf{\color{#35bf28}+148.28\%}$
test_nested_getleaf	33.2720μs	10.8366μs	92.2797 KOps/s	90.8610 KOps/s	$\color{#35bf28}+1.56\%$
test_nested_get	53.5000μs	10.1659μs	98.3682 KOps/s	98.2878 KOps/s	$\color{#35bf28}+0.08\%$
test_stacked_getleaf	49.9830μs	10.9067μs	91.6866 KOps/s	90.1873 KOps/s	$\color{#35bf28}+1.66\%$
test_stacked_get	32.9410μs	10.1343μs	98.6750 KOps/s	98.1655 KOps/s	$\color{#35bf28}+0.52\%$
test_nested_getitemleaf	53.2090μs	11.2182μs	89.1412 KOps/s	87.1073 KOps/s	$\color{#35bf28}+2.33\%$
test_nested_getitem	47.4190μs	10.4864μs	95.3618 KOps/s	94.1656 KOps/s	$\color{#35bf28}+1.27\%$
test_stacked_getitemleaf	35.5460μs	11.2636μs	88.7818 KOps/s	88.3804 KOps/s	$\color{#35bf28}+0.45\%$
test_stacked_getitem	52.5180μs	10.5692μs	94.6142 KOps/s	94.9250 KOps/s	$\color{#d91a1a}-0.33\%$
test_lock_nested	1.8963ms	0.4594ms	2.1765 KOps/s	2.1828 KOps/s	$\color{#d91a1a}-0.29\%$
test_lock_stack_nested	0.6311ms	0.4325ms	2.3120 KOps/s	2.3856 KOps/s	$\color{#d91a1a}-3.09\%$
test_unlock_nested	0.7650ms	0.3779ms	2.6463 KOps/s	2.6428 KOps/s	$\color{#35bf28}+0.14\%$
test_unlock_stack_nested	0.5448ms	0.3507ms	2.8515 KOps/s	2.9640 KOps/s	$\color{#d91a1a}-3.80\%$
test_flatten_speed	0.1992ms	0.1009ms	9.9127 KOps/s	9.8801 KOps/s	$\color{#35bf28}+0.33\%$
test_unflatten_speed	0.6909ms	0.5145ms	1.9438 KOps/s	1.8824 KOps/s	$\color{#35bf28}+3.26\%$
test_common_ops	4.8465ms	0.8335ms	1.1998 KOps/s	1.2853 KOps/s	$\textbf{\color{#d91a1a}-6.66\%}$
test_creation	17.2520μs	2.5566μs	391.1459 KOps/s	402.8764 KOps/s	$\color{#d91a1a}-2.91\%$
test_creation_empty	30.1560μs	12.6736μs	78.9042 KOps/s	94.8327 KOps/s	$\textbf{\color{#d91a1a}-16.80\%}$
test_creation_nested_1	73.7550μs	15.8245μs	63.1932 KOps/s	72.7796 KOps/s	$\textbf{\color{#d91a1a}-13.17\%}$
test_creation_nested_2	51.4760μs	20.6130μs	48.5130 KOps/s	55.4954 KOps/s	$\textbf{\color{#d91a1a}-12.58\%}$
test_clone	0.2141ms	14.0407μs	71.2217 KOps/s	72.5432 KOps/s	$\color{#d91a1a}-1.82\%$
test_getitem[int]	1.3211ms	13.4720μs	74.2280 KOps/s	77.0792 KOps/s	$\color{#d91a1a}-3.70\%$
test_getitem[slice_int]	0.1413ms	25.0608μs	39.9030 KOps/s	39.9383 KOps/s	$\color{#d91a1a}-0.09\%$
test_getitem[range]	0.1647ms	48.6879μs	20.5390 KOps/s	20.2495 KOps/s	$\color{#35bf28}+1.43\%$
test_getitem[tuple]	0.1374ms	21.0648μs	47.4725 KOps/s	47.9882 KOps/s	$\color{#d91a1a}-1.07\%$
test_getitem[list]	0.4916ms	44.2388μs	22.6046 KOps/s	22.1583 KOps/s	$\color{#35bf28}+2.01\%$
test_setitem_dim[int]	53.4400μs	25.7081μs	38.8983 KOps/s	37.1047 KOps/s	$\color{#35bf28}+4.83\%$
test_setitem_dim[slice_int]	87.0220μs	52.9879μs	18.8722 KOps/s	18.6698 KOps/s	$\color{#35bf28}+1.08\%$
test_setitem_dim[range]	0.1336ms	73.0143μs	13.6960 KOps/s	13.1681 KOps/s	$\color{#35bf28}+4.01\%$
test_setitem_dim[tuple]	66.7840μs	41.8817μs	23.8768 KOps/s	23.4973 KOps/s	$\color{#35bf28}+1.61\%$
test_setitem	0.3064ms	22.0978μs	45.2534 KOps/s	49.1714 KOps/s	$\textbf{\color{#d91a1a}-7.97\%}$
test_set	0.1297ms	21.6325μs	46.2267 KOps/s	50.9307 KOps/s	$\textbf{\color{#d91a1a}-9.24\%}$
test_set_shared	1.1903ms	0.1738ms	5.7533 KOps/s	5.8554 KOps/s	$\color{#d91a1a}-1.74\%$
test_update	0.2539ms	24.6671μs	40.5399 KOps/s	46.4001 KOps/s	$\textbf{\color{#d91a1a}-12.63\%}$
test_update_nested	0.2739ms	36.0650μs	27.7277 KOps/s	30.6573 KOps/s	$\textbf{\color{#d91a1a}-9.56\%}$
test_update__nested	0.8165ms	35.9231μs	27.8372 KOps/s	28.3888 KOps/s	$\color{#d91a1a}-1.94\%$
test_set_nested	0.2261ms	24.0307μs	41.6135 KOps/s	45.5799 KOps/s	$\textbf{\color{#d91a1a}-8.70\%}$
test_set_nested_new	0.2838ms	28.9885μs	34.4965 KOps/s	37.7929 KOps/s	$\textbf{\color{#d91a1a}-8.72\%}$
test_select	0.2242ms	46.6009μs	21.4588 KOps/s	22.8396 KOps/s	$\textbf{\color{#d91a1a}-6.05\%}$
test_select_nested	0.1281ms	65.6162μs	15.2401 KOps/s	15.6161 KOps/s	$\color{#d91a1a}-2.41\%$
test_exclude_nested	0.1616ms	86.2963μs	11.5880 KOps/s	11.9042 KOps/s	$\color{#d91a1a}-2.66\%$
test_empty[True]	0.9426ms	0.4207ms	2.3772 KOps/s	2.3813 KOps/s	$\color{#d91a1a}-0.17\%$
test_empty[False]	11.9072μs	1.4047μs	711.9060 KOps/s	688.9268 KOps/s	$\color{#35bf28}+3.34\%$
test_unbind_speed	0.6054ms	0.2797ms	3.5757 KOps/s	3.6190 KOps/s	$\color{#d91a1a}-1.20\%$
test_unbind_speed_stack0	0.4924ms	0.2769ms	3.6111 KOps/s	3.7888 KOps/s	$\color{#d91a1a}-4.69\%$
test_unbind_speed_stack1	0.1140s	0.8291ms	1.2061 KOps/s	1.3998 KOps/s	$\textbf{\color{#d91a1a}-13.83\%}$
test_split	0.1057s	1.7985ms	556.0139 Ops/s	566.3306 Ops/s	$\color{#d91a1a}-1.82\%$
test_chunk	0.1017s	1.7990ms	555.8730 Ops/s	559.8401 Ops/s	$\color{#d91a1a}-0.71\%$
test_consolidate_njt[False-None]	8.2465ms	7.9909ms	125.1427 Ops/s	119.2055 Ops/s	$\color{#35bf28}+4.98\%$
test_creation[device0]	0.2220ms	89.5886μs	11.1621 KOps/s	10.8339 KOps/s	$\color{#35bf28}+3.03\%$
test_creation_from_tensor	3.5658ms	93.7478μs	10.6669 KOps/s	10.4197 KOps/s	$\color{#35bf28}+2.37\%$
test_add_one[memmap_tensor0]	0.2137ms	5.0410μs	198.3735 KOps/s	204.1577 KOps/s	$\color{#d91a1a}-2.83\%$
test_contiguous[memmap_tensor0]	18.8050μs	0.5090μs	1.9648 MOps/s	1.9115 MOps/s	$\color{#35bf28}+2.79\%$
test_stack[memmap_tensor0]	49.7030μs	3.6047μs	277.4139 KOps/s	280.5899 KOps/s	$\color{#d91a1a}-1.13\%$
test_memmaptd_index	1.0275ms	0.2397ms	4.1714 KOps/s	3.9445 KOps/s	$\textbf{\color{#35bf28}+5.75\%}$
test_memmaptd_index_astensor	0.8110ms	0.3254ms	3.0731 KOps/s	2.8908 KOps/s	$\textbf{\color{#35bf28}+6.31\%}$
test_memmaptd_index_op	1.0184ms	0.6185ms	1.6168 KOps/s	1.6649 KOps/s	$\color{#d91a1a}-2.89\%$
test_serialize_model	0.1276s	0.1164s	8.5899 Ops/s	8.6100 Ops/s	$\color{#d91a1a}-0.23\%$
test_serialize_model_pickle	0.4622s	0.3842s	2.6030 Ops/s	2.5264 Ops/s	$\color{#35bf28}+3.04\%$
test_serialize_weights	0.1254s	0.1139s	8.7783 Ops/s	8.7413 Ops/s	$\color{#35bf28}+0.42\%$
test_serialize_weights_returnearly	0.3887s	0.2016s	4.9601 Ops/s	6.4815 Ops/s	$\textbf{\color{#d91a1a}-23.47\%}$
test_serialize_weights_pickle	0.4960s	0.4070s	2.4567 Ops/s	2.5994 Ops/s	$\textbf{\color{#d91a1a}-5.49\%}$
test_serialize_weights_filesystem	0.1477s	0.1438s	6.9559 Ops/s	7.0403 Ops/s	$\color{#d91a1a}-1.20\%$
test_serialize_model_filesystem	0.1554s	0.1451s	6.8907 Ops/s	5.9441 Ops/s	$\textbf{\color{#35bf28}+15.93\%}$
test_reshape_pytree	65.0110μs	27.3195μs	36.6039 KOps/s	36.7147 KOps/s	$\color{#d91a1a}-0.30\%$
test_reshape_td	67.1150μs	34.0547μs	29.3646 KOps/s	29.9077 KOps/s	$\color{#d91a1a}-1.82\%$
test_view_pytree	57.6780μs	27.4050μs	36.4897 KOps/s	36.6436 KOps/s	$\color{#d91a1a}-0.42\%$
test_view_td	84.6280μs	39.1340μs	25.5532 KOps/s	25.8203 KOps/s	$\color{#d91a1a}-1.03\%$
test_unbind_pytree	72.4250μs	30.5621μs	32.7203 KOps/s	33.3037 KOps/s	$\color{#d91a1a}-1.75\%$
test_unbind_td	0.4043ms	41.5165μs	24.0868 KOps/s	24.4180 KOps/s	$\color{#d91a1a}-1.36\%$
test_split_pytree	64.4710μs	30.0733μs	33.2521 KOps/s	33.2490 KOps/s	$+0.01\%$
test_split_td	0.4883ms	45.9476μs	21.7639 KOps/s	20.9301 KOps/s	$\color{#35bf28}+3.98\%$
test_add_pytree	87.7540μs	36.6934μs	27.2529 KOps/s	27.3551 KOps/s	$\color{#d91a1a}-0.37\%$
test_add_td	0.1333ms	59.7681μs	16.7313 KOps/s	16.3256 KOps/s	$\color{#35bf28}+2.49\%$
test_compile_add_one_nested[tensordict-compile]	0.1330ms	60.8345μs	16.4380 KOps/s	16.0861 KOps/s	$\color{#35bf28}+2.19\%$
test_compile_add_one_nested[tensordict-eager]	0.3995ms	0.1678ms	5.9612 KOps/s	5.6539 KOps/s	$\textbf{\color{#35bf28}+5.43\%}$
test_compile_add_one_nested[pytree-compile]	93.3240μs	45.7261μs	21.8694 KOps/s	21.6481 KOps/s	$\color{#35bf28}+1.02\%$
test_compile_add_one_nested[pytree-eager]	0.2144ms	0.1197ms	8.3546 KOps/s	8.4411 KOps/s	$\color{#d91a1a}-1.03\%$
test_compile_copy_nested[tensordict-compile]	63.2180μs	26.0528μs	38.3837 KOps/s	39.3020 KOps/s	$\color{#d91a1a}-2.34\%$
test_compile_copy_nested[tensordict-eager]	0.1306ms	60.2865μs	16.5875 KOps/s	16.8916 KOps/s	$\color{#d91a1a}-1.80\%$
test_compile_copy_nested[pytree-compile]	0.1668ms	78.9877μs	12.6602 KOps/s	12.7229 KOps/s	$\color{#d91a1a}-0.49\%$
test_compile_copy_nested[pytree-eager]	0.1464ms	68.6359μs	14.5696 KOps/s	14.7662 KOps/s	$\color{#d91a1a}-1.33\%$
test_compile_add_one_flat[tensordict-compile]	0.1840ms	0.1037ms	9.6410 KOps/s	9.5041 KOps/s	$\color{#35bf28}+1.44\%$
test_compile_add_one_flat[tensordict-eager]	0.4328ms	0.2160ms	4.6295 KOps/s	4.5868 KOps/s	$\color{#35bf28}+0.93\%$
test_compile_add_one_flat[tensorclass-compile]	89.4770μs	44.8765μs	22.2834 KOps/s	21.4385 KOps/s	$\color{#35bf28}+3.94\%$
test_compile_add_one_flat[tensorclass-eager]	0.4986ms	65.0218μs	15.3794 KOps/s	15.3457 KOps/s	$\color{#35bf28}+0.22\%$
test_compile_add_one_flat[pytree-compile]	0.1753ms	0.1018ms	9.8274 KOps/s	9.7596 KOps/s	$\color{#35bf28}+0.70\%$
test_compile_add_one_flat[pytree-eager]	0.3670ms	0.2012ms	4.9712 KOps/s	4.9299 KOps/s	$\color{#35bf28}+0.84\%$
test_compile_add_self_flat[tensordict-eager]	0.4572ms	0.2330ms	4.2918 KOps/s	4.1632 KOps/s	$\color{#35bf28}+3.09\%$
test_compile_add_self_flat[tensordict-compile]	0.2215ms	0.1040ms	9.6158 KOps/s	9.6124 KOps/s	$\color{#35bf28}+0.04\%$
test_compile_add_self_flat[tensorclass-eager]	0.1352ms	58.3631μs	17.1341 KOps/s	16.5823 KOps/s	$\color{#35bf28}+3.33\%$
test_compile_add_self_flat[tensorclass-compile]	0.1156ms	46.2000μs	21.6450 KOps/s	22.3108 KOps/s	$\color{#d91a1a}-2.98\%$
test_compile_add_self_flat[pytree-eager]	0.6446ms	0.1617ms	6.1849 KOps/s	6.3233 KOps/s	$\color{#d91a1a}-2.19\%$
test_compile_add_self_flat[pytree-compile]	0.1841ms	0.1022ms	9.7828 KOps/s	9.7492 KOps/s	$\color{#35bf28}+0.34\%$
test_compile_copy_flat[tensordict-compile]	65.8930μs	21.4507μs	46.6185 KOps/s	48.7525 KOps/s	$\color{#d91a1a}-4.38\%$
test_compile_copy_flat[tensordict-eager]	0.1297ms	65.9643μs	15.1597 KOps/s	14.7932 KOps/s	$\color{#35bf28}+2.48\%$
test_compile_copy_flat[pytree-compile]	0.1546ms	80.8002μs	12.3762 KOps/s	12.1997 KOps/s	$\color{#35bf28}+1.45\%$
test_compile_copy_flat[pytree-eager]	0.1310ms	69.3193μs	14.4260 KOps/s	14.5074 KOps/s	$\color{#d91a1a}-0.56\%$
test_compile_assign_and_add[tensordict-compile]	0.3166ms	0.2037ms	4.9098 KOps/s	4.8771 KOps/s	$\color{#35bf28}+0.67\%$
test_compile_assign_and_add[tensordict-eager]	2.7809ms	1.3415ms	745.4429 Ops/s	744.0523 Ops/s	$\color{#35bf28}+0.19\%$
test_compile_assign_and_add[pytree-compile]	0.2930ms	0.2020ms	4.9507 KOps/s	5.0695 KOps/s	$\color{#d91a1a}-2.34\%$
test_compile_assign_and_add[pytree-eager]	0.9859ms	0.7766ms	1.2877 KOps/s	1.2838 KOps/s	$\color{#35bf28}+0.30\%$
test_compile_assign_and_add_stack[compile]	0.6939ms	0.4479ms	2.2328 KOps/s	2.2505 KOps/s	$\color{#d91a1a}-0.79\%$
test_compile_assign_and_add_stack[eager]	2.9978ms	2.8288ms	353.5055 Ops/s	372.2595 Ops/s	$\textbf{\color{#d91a1a}-5.04\%}$
test_compile_indexing[tensor-tensordict-compile]	0.1011ms	35.4355μs	28.2203 KOps/s	28.2461 KOps/s	$\color{#d91a1a}-0.09\%$
test_compile_indexing[tensor-tensordict-eager]	0.5251ms	33.0295μs	30.2760 KOps/s	29.1809 KOps/s	$\color{#35bf28}+3.75\%$
test_compile_indexing[tensor-tensorclass-compile]	67.5660μs	29.4807μs	33.9205 KOps/s	35.1603 KOps/s	$\color{#d91a1a}-3.53\%$
test_compile_indexing[tensor-tensorclass-eager]	0.1008ms	23.9674μs	41.7234 KOps/s	41.1343 KOps/s	$\color{#35bf28}+1.43\%$
test_compile_indexing[tensor-pytree-compile]	88.2750μs	29.5638μs	33.8252 KOps/s	33.3044 KOps/s	$\color{#35bf28}+1.56\%$
test_compile_indexing[tensor-pytree-eager]	91.1700μs	23.9546μs	41.7456 KOps/s	41.7546 KOps/s	$\color{#d91a1a}-0.02\%$
test_compile_indexing[slice-tensordict-compile]	0.1060ms	50.7354μs	19.7101 KOps/s	19.3261 KOps/s	$\color{#35bf28}+1.99\%$
test_compile_indexing[slice-tensordict-eager]	0.5543ms	20.0059μs	49.9853 KOps/s	48.4778 KOps/s	$\color{#35bf28}+3.11\%$
test_compile_indexing[slice-tensorclass-compile]	0.1307ms	43.1478μs	23.1761 KOps/s	22.4887 KOps/s	$\color{#35bf28}+3.06\%$
test_compile_indexing[slice-tensorclass-eager]	51.7360μs	18.8896μs	52.9393 KOps/s	51.6190 KOps/s	$\color{#35bf28}+2.56\%$
test_compile_indexing[slice-pytree-compile]	0.1066ms	43.9833μs	22.7359 KOps/s	21.5248 KOps/s	$\textbf{\color{#35bf28}+5.63\%}$
test_compile_indexing[slice-pytree-eager]	64.6200μs	18.9923μs	52.6528 KOps/s	52.0121 KOps/s	$\color{#35bf28}+1.23\%$
test_compile_indexing[int-tensordict-compile]	0.1203ms	52.1369μs	19.1803 KOps/s	18.8792 KOps/s	$\color{#35bf28}+1.60\%$
test_compile_indexing[int-tensordict-eager]	0.9476ms	20.0672μs	49.8326 KOps/s	49.8208 KOps/s	$\color{#35bf28}+0.02\%$
test_compile_indexing[int-tensorclass-compile]	0.1145ms	44.2137μs	22.6174 KOps/s	22.0388 KOps/s	$\color{#35bf28}+2.63\%$
test_compile_indexing[int-tensorclass-eager]	65.2700μs	18.8838μs	52.9554 KOps/s	52.2682 KOps/s	$\color{#35bf28}+1.31\%$
test_compile_indexing[int-pytree-compile]	0.1030ms	44.3701μs	22.5377 KOps/s	21.9309 KOps/s	$\color{#35bf28}+2.77\%$
test_compile_indexing[int-pytree-eager]	62.0060μs	18.8388μs	53.0819 KOps/s	51.8444 KOps/s	$\color{#35bf28}+2.39\%$
test_mod_add[eager]	83.4060μs	34.4392μs	29.0367 KOps/s	29.0540 KOps/s	$\color{#d91a1a}-0.06\%$
test_mod_add[compile]	0.1113ms	47.4561μs	21.0721 KOps/s	20.9776 KOps/s	$\color{#35bf28}+0.45\%$
test_mod_add[compile-overhead]	0.1094ms	46.4375μs	21.5343 KOps/s	20.4286 KOps/s	$\textbf{\color{#35bf28}+5.41\%}$
test_mod_wrap[eager]	0.4159ms	0.2222ms	4.5014 KOps/s	4.4494 KOps/s	$\color{#35bf28}+1.17\%$
test_mod_wrap[compile]	0.3546ms	0.2026ms	4.9364 KOps/s	4.8384 KOps/s	$\color{#35bf28}+2.03\%$
test_mod_wrap[compile-overhead]	0.2702ms	0.2026ms	4.9348 KOps/s	4.8850 KOps/s	$\color{#35bf28}+1.02\%$
test_mod_wrap_and_backward[eager]	14.1846ms	11.6884ms	85.5548 Ops/s	80.0720 Ops/s	$\textbf{\color{#35bf28}+6.85\%}$
test_mod_wrap_and_backward[compile]	17.7714ms	13.1284ms	76.1706 Ops/s	75.1930 Ops/s	$\color{#35bf28}+1.30\%$
test_mod_wrap_and_backward[compile-overhead]	17.0207ms	12.1146ms	82.5451 Ops/s	80.5837 Ops/s	$\color{#35bf28}+2.43\%$
test_seq_add[eager]	0.2137ms	0.1149ms	8.6999 KOps/s	8.5487 KOps/s	$\color{#35bf28}+1.77\%$
test_seq_add[compile]	0.1208ms	59.2461μs	16.8788 KOps/s	16.1237 KOps/s	$\color{#35bf28}+4.68\%$
test_seq_add[compile-overhead]	0.1531ms	58.3282μs	17.1444 KOps/s	16.2722 KOps/s	$\textbf{\color{#35bf28}+5.36\%}$
test_seq_wrap[eager]	0.6804ms	0.4598ms	2.1749 KOps/s	2.2163 KOps/s	$\color{#d91a1a}-1.87\%$
test_seq_wrap[compile]	0.3124ms	0.2232ms	4.4797 KOps/s	4.3046 KOps/s	$\color{#35bf28}+4.07\%$
test_seq_wrap[compile-overhead]	0.4241ms	0.2261ms	4.4222 KOps/s	4.3589 KOps/s	$\color{#35bf28}+1.45\%$
test_func_call_runtime[False-eager]	0.9408ms	0.5402ms	1.8511 KOps/s	1.8056 KOps/s	$\color{#35bf28}+2.52\%$
test_func_call_runtime[False-compile]	0.5528ms	0.4234ms	2.3618 KOps/s	2.3348 KOps/s	$\color{#35bf28}+1.15\%$
test_func_call_runtime[False-compile-overhead]	0.5933ms	0.4230ms	2.3643 KOps/s	2.3306 KOps/s	$\color{#35bf28}+1.44\%$
test_func_call_runtime[True-eager]	1.2382ms	0.7580ms	1.3192 KOps/s	1.2845 KOps/s	$\color{#35bf28}+2.70\%$
test_func_call_runtime[True-compile]	0.5751ms	0.4582ms	2.1823 KOps/s	2.1437 KOps/s	$\color{#35bf28}+1.80\%$
test_func_call_runtime[True-compile-overhead]	0.9738ms	0.4686ms	2.1340 KOps/s	2.1400 KOps/s	$\color{#d91a1a}-0.28\%$
test_func_call_cm_runtime[False-eager]	0.9763ms	0.5428ms	1.8424 KOps/s	1.8262 KOps/s	$\color{#35bf28}+0.89\%$
test_func_call_cm_runtime[False-compile]	0.5575ms	0.4220ms	2.3699 KOps/s	2.3445 KOps/s	$\color{#35bf28}+1.09\%$
test_func_call_cm_runtime[False-compile-overhead]	0.5405ms	0.4253ms	2.3515 KOps/s	2.3417 KOps/s	$\color{#35bf28}+0.42\%$
test_func_call_cm_runtime[True-eager]	1.0497ms	0.9130ms	1.0953 KOps/s	1.0927 KOps/s	$\color{#35bf28}+0.24\%$
test_func_call_cm_runtime[True-compile]	0.6359ms	0.4879ms	2.0497 KOps/s	2.0350 KOps/s	$\color{#35bf28}+0.72\%$
test_func_call_cm_runtime[True-compile-overhead]	0.9129ms	0.4891ms	2.0448 KOps/s	2.0540 KOps/s	$\color{#d91a1a}-0.45\%$
test_vmap_func_call_cm_runtime[eager]	2.8425ms	1.9079ms	524.1460 Ops/s	521.0721 Ops/s	$\color{#35bf28}+0.59\%$
test_vmap_func_call_cm_runtime[compile]	0.8437ms	0.5117ms	1.9544 KOps/s	1.9093 KOps/s	$\color{#35bf28}+2.36\%$
test_vmap_func_call_cm_runtime[compile-overhead]	0.6957ms	0.5127ms	1.9506 KOps/s	1.9301 KOps/s	$\color{#35bf28}+1.06\%$
test_distributed	0.2909ms	0.1253ms	7.9784 KOps/s	7.8208 KOps/s	$\color{#35bf28}+2.02\%$
test_tdmodule	47.7500μs	25.4048μs	39.3627 KOps/s	37.7593 KOps/s	$\color{#35bf28}+4.25\%$
test_tdmodule_dispatch	73.5880μs	48.3171μs	20.6966 KOps/s	20.9042 KOps/s	$\color{#d91a1a}-0.99\%$
test_tdseq	47.9990μs	26.1682μs	38.2144 KOps/s	38.2082 KOps/s	$\color{#35bf28}+0.02\%$
test_tdseq_dispatch	77.8760μs	50.8112μs	19.6807 KOps/s	20.0790 KOps/s	$\color{#d91a1a}-1.98\%$
test_instantiation_functorch	1.7616ms	1.5115ms	661.5833 Ops/s	638.5717 Ops/s	$\color{#35bf28}+3.60\%$
test_exec_functorch	0.7974ms	0.1832ms	5.4573 KOps/s	5.5852 KOps/s	$\color{#d91a1a}-2.29\%$
test_exec_functional_call	0.3513ms	0.1717ms	5.8253 KOps/s	5.7878 KOps/s	$\color{#35bf28}+0.65\%$
test_exec_td_decorator	0.4888ms	0.2299ms	4.3502 KOps/s	4.2835 KOps/s	$\color{#35bf28}+1.56\%$
test_vmap_mlp_speed_decorator[True-True]	0.9915ms	0.6557ms	1.5251 KOps/s	1.5368 KOps/s	$\color{#d91a1a}-0.76\%$
test_vmap_mlp_speed_decorator[True-False]	0.9140ms	0.6561ms	1.5241 KOps/s	1.5299 KOps/s	$\color{#d91a1a}-0.38\%$
test_vmap_mlp_speed_decorator[False-True]	1.0345ms	0.5300ms	1.8868 KOps/s	1.8954 KOps/s	$\color{#d91a1a}-0.46\%$
test_vmap_mlp_speed_decorator[False-False]	1.0791ms	0.5273ms	1.8965 KOps/s	1.8857 KOps/s	$\color{#35bf28}+0.57\%$
test_to_module_speed[True]	2.1904ms	1.3530ms	739.0771 Ops/s	738.4737 Ops/s	$\color{#35bf28}+0.08\%$
test_to_module_speed[False]	1.8023ms	1.3092ms	763.8455 Ops/s	756.5758 Ops/s	$\color{#35bf28}+0.96\%$
test_tc_init	83.3050μs	47.7584μs	20.9387 KOps/s	21.5224 KOps/s	$\color{#d91a1a}-2.71\%$
test_tc_init_nested	0.1763ms	98.0669μs	10.1971 KOps/s	10.6196 KOps/s	$\color{#d91a1a}-3.98\%$
test_tc_first_layer_tensor	42.8600μs	1.5073μs	663.4343 KOps/s	648.0631 KOps/s	$\color{#35bf28}+2.37\%$
test_tc_first_layer_nontensor	33.5330μs	4.6590μs	214.6401 KOps/s	208.2979 KOps/s	$\color{#35bf28}+3.04\%$
test_tc_second_layer_tensor	33.0190μs	2.7441μs	364.4196 KOps/s	352.9297 KOps/s	$\color{#35bf28}+3.26\%$
test_tc_second_layer_nontensor	28.9540μs	6.0074μs	166.4620 KOps/s	162.4829 KOps/s	$\color{#35bf28}+2.45\%$
test_unbind	0.2132s	13.7627ms	72.6600 Ops/s	78.5207 Ops/s	$\textbf{\color{#d91a1a}-7.46\%}$
test_full_like	8.9633ms	7.2674ms	137.6017 Ops/s	136.3050 Ops/s	$\color{#35bf28}+0.95\%$
test_zeros_like	12.4368ms	7.5541ms	132.3778 Ops/s	359.9212 Ops/s	$\textbf{\color{#d91a1a}-63.22\%}$
test_ones_like	10.7830ms	7.7576ms	128.9059 Ops/s	315.0060 Ops/s	$\textbf{\color{#d91a1a}-59.08\%}$
test_clone	17.0024ms	9.7336ms	102.7372 Ops/s	199.2713 Ops/s	$\textbf{\color{#d91a1a}-48.44\%}$
test_squeeze	57.7880μs	12.5087μs	79.9445 KOps/s	79.3434 KOps/s	$\color{#35bf28}+0.76\%$
test_unsqueeze	0.1721ms	91.6715μs	10.9085 KOps/s	10.0343 KOps/s	$\textbf{\color{#35bf28}+8.71\%}$
test_split	0.4953ms	0.1964ms	5.0925 KOps/s	5.0207 KOps/s	$\color{#35bf28}+1.43\%$
test_permute	0.3569ms	0.2052ms	4.8742 KOps/s	4.6878 KOps/s	$\color{#35bf28}+3.98\%$
test_stack	34.6889ms	25.2601ms	39.5881 Ops/s	39.8501 Ops/s	$\color{#d91a1a}-0.66\%$
test_cat	30.5922ms	25.1086ms	39.8270 Ops/s	38.8148 Ops/s	$\color{#35bf28}+2.61\%$

[github-actions]

$\color{#D29922}\textsf{\Large⚠\kern{0.2cm}\normalsize Warning}$ Result of GPU Benchmark Tests

Total Benchmarks: 229. Improved: $\large\color{#35bf28}7$. Worsened: $\large\color{#d91a1a}26$.

Expand to view detailed results

Name	Max	Mean	Ops	Ops on Repo HEAD	Change
test_plain_set_nested	39.0210μs	13.1368μs	76.1223 KOps/s	78.5362 KOps/s	$\color{#d91a1a}-3.07\%$
test_plain_set_stack_nested	39.9310μs	13.2767μs	75.3197 KOps/s	77.2398 KOps/s	$\color{#d91a1a}-2.49\%$
test_plain_set_nested_inplace	67.4920μs	14.2499μs	70.1758 KOps/s	71.9214 KOps/s	$\color{#d91a1a}-2.43\%$
test_plain_set_stack_nested_inplace	44.0110μs	14.3972μs	69.4579 KOps/s	72.0576 KOps/s	$\color{#d91a1a}-3.61\%$
test_items	38.2300μs	2.8864μs	346.4478 KOps/s	345.4498 KOps/s	$\color{#35bf28}+0.29\%$
test_items_nested	0.4159ms	0.3621ms	2.7616 KOps/s	2.7589 KOps/s	$\color{#35bf28}+0.10\%$
test_items_nested_locked	0.4225ms	0.3647ms	2.7423 KOps/s	2.7520 KOps/s	$\color{#d91a1a}-0.35\%$
test_items_nested_leaf	80.2310μs	57.9860μs	17.2455 KOps/s	17.1704 KOps/s	$\color{#35bf28}+0.44\%$
test_items_stack_nested	0.4160ms	0.3636ms	2.7500 KOps/s	2.7413 KOps/s	$\color{#35bf28}+0.32\%$
test_items_stack_nested_leaf	90.2920μs	58.9419μs	16.9658 KOps/s	16.7178 KOps/s	$\color{#35bf28}+1.48\%$
test_items_stack_nested_locked	0.4168ms	0.3622ms	2.7610 KOps/s	2.7390 KOps/s	$\color{#35bf28}+0.80\%$
test_keys	27.5000μs	3.4704μs	288.1521 KOps/s	289.8236 KOps/s	$\color{#d91a1a}-0.58\%$
test_keys_nested	0.1249ms	81.4685μs	12.2747 KOps/s	12.2531 KOps/s	$\color{#35bf28}+0.18\%$
test_keys_nested_locked	0.7460ms	87.1770μs	11.4709 KOps/s	11.4734 KOps/s	$\color{#d91a1a}-0.02\%$
test_keys_nested_leaf	0.1079ms	71.9067μs	13.9069 KOps/s	13.8625 KOps/s	$\color{#35bf28}+0.32\%$
test_keys_stack_nested	0.1193ms	81.6794μs	12.2430 KOps/s	12.0175 KOps/s	$\color{#35bf28}+1.88\%$
test_keys_stack_nested_leaf	0.1078ms	72.9616μs	13.7058 KOps/s	13.5073 KOps/s	$\color{#35bf28}+1.47\%$
test_keys_stack_nested_locked	0.1254ms	87.2869μs	11.4565 KOps/s	11.2424 KOps/s	$\color{#35bf28}+1.90\%$
test_values	5.7802μs	0.8529μs	1.1724 MOps/s	1.1684 MOps/s	$\color{#35bf28}+0.35\%$
test_values_nested	66.6310μs	34.9781μs	28.5894 KOps/s	29.3290 KOps/s	$\color{#d91a1a}-2.52\%$
test_values_nested_locked	77.2020μs	36.1706μs	27.6468 KOps/s	28.0696 KOps/s	$\color{#d91a1a}-1.51\%$
test_values_nested_leaf	69.0610μs	39.8218μs	25.1119 KOps/s	25.4402 KOps/s	$\color{#d91a1a}-1.29\%$
test_values_stack_nested	68.1520μs	35.1500μs	28.4495 KOps/s	28.5073 KOps/s	$\color{#d91a1a}-0.20\%$
test_values_stack_nested_leaf	91.3420μs	39.8444μs	25.0976 KOps/s	25.4924 KOps/s	$\color{#d91a1a}-1.55\%$
test_values_stack_nested_locked	64.0310μs	36.3664μs	27.4979 KOps/s	27.6883 KOps/s	$\color{#d91a1a}-0.69\%$
test_membership	2.1490μs	0.5214μs	1.9180 MOps/s	1.9786 MOps/s	$\color{#d91a1a}-3.06\%$
test_membership_nested	22.6205μs	1.9670μs	508.3944 KOps/s	476.8773 KOps/s	$\textbf{\color{#35bf28}+6.61\%}$
test_membership_nested_leaf	18.3905μs	1.9887μs	502.8528 KOps/s	495.6430 KOps/s	$\color{#35bf28}+1.45\%$
test_membership_stacked_nested	39.4310μs	2.0818μs	480.3635 KOps/s	491.8342 KOps/s	$\color{#d91a1a}-2.33\%$
test_membership_stacked_nested_leaf	30.2510μs	2.0371μs	490.9001 KOps/s	492.6623 KOps/s	$\color{#d91a1a}-0.36\%$
test_membership_nested_last	40.1910μs	3.0753μs	325.1758 KOps/s	329.1931 KOps/s	$\color{#d91a1a}-1.22\%$
test_membership_nested_leaf_last	31.4710μs	3.0414μs	328.7954 KOps/s	329.6375 KOps/s	$\color{#d91a1a}-0.26\%$
test_membership_stacked_nested_last	35.2110μs	3.0894μs	323.6845 KOps/s	328.9488 KOps/s	$\color{#d91a1a}-1.60\%$
test_membership_stacked_nested_leaf_last	39.8510μs	3.0432μs	328.6052 KOps/s	325.0405 KOps/s	$\color{#35bf28}+1.10\%$
test_nested_getleaf	38.9110μs	6.1446μs	162.7434 KOps/s	162.4166 KOps/s	$\color{#35bf28}+0.20\%$
test_nested_get	40.3110μs	5.7521μs	173.8505 KOps/s	169.4204 KOps/s	$\color{#35bf28}+2.61\%$
test_stacked_getleaf	40.6500μs	6.1267μs	163.2190 KOps/s	162.7967 KOps/s	$\color{#35bf28}+0.26\%$
test_stacked_get	51.1210μs	5.8516μs	170.8943 KOps/s	171.5759 KOps/s	$\color{#d91a1a}-0.40\%$
test_nested_getitemleaf	38.8510μs	6.2935μs	158.8940 KOps/s	158.7326 KOps/s	$\color{#35bf28}+0.10\%$
test_nested_getitem	41.2310μs	5.9778μs	167.2855 KOps/s	167.0766 KOps/s	$\color{#35bf28}+0.13\%$
test_stacked_getitemleaf	43.6710μs	6.3465μs	157.5674 KOps/s	160.0806 KOps/s	$\color{#d91a1a}-1.57\%$
test_stacked_getitem	49.7710μs	5.9711μs	167.4737 KOps/s	168.7500 KOps/s	$\color{#d91a1a}-0.76\%$
test_lock_nested	0.8315ms	0.3744ms	2.6706 KOps/s	2.6603 KOps/s	$\color{#35bf28}+0.39\%$
test_lock_stack_nested	0.4145ms	0.3511ms	2.8486 KOps/s	2.9262 KOps/s	$\color{#d91a1a}-2.65\%$
test_unlock_nested	0.6575ms	0.3161ms	3.1637 KOps/s	3.2621 KOps/s	$\color{#d91a1a}-3.02\%$
test_unlock_stack_nested	0.3422ms	0.2886ms	3.4648 KOps/s	3.5749 KOps/s	$\color{#d91a1a}-3.08\%$
test_flatten_speed	0.1123ms	74.7777μs	13.3730 KOps/s	13.3524 KOps/s	$\color{#35bf28}+0.15\%$
test_unflatten_speed	0.3766ms	0.3200ms	3.1249 KOps/s	3.1088 KOps/s	$\color{#35bf28}+0.52\%$
test_common_ops	1.5365ms	0.6323ms	1.5815 KOps/s	1.6407 KOps/s	$\color{#d91a1a}-3.61\%$
test_creation	0.1125ms	1.7211μs	581.0333 KOps/s	579.2806 KOps/s	$\color{#35bf28}+0.30\%$
test_creation_empty	37.5810μs	9.8303μs	101.7260 KOps/s	106.6056 KOps/s	$\color{#d91a1a}-4.58\%$
test_creation_nested_1	48.3500μs	11.4494μs	87.3409 KOps/s	91.7236 KOps/s	$\color{#d91a1a}-4.78\%$
test_creation_nested_2	37.0200μs	14.1443μs	70.7001 KOps/s	73.9672 KOps/s	$\color{#d91a1a}-4.42\%$
test_clone	0.1201ms	10.5226μs	95.0339 KOps/s	98.1424 KOps/s	$\color{#d91a1a}-3.17\%$
test_getitem[int]	1.8181ms	10.8425μs	92.2293 KOps/s	93.6355 KOps/s	$\color{#d91a1a}-1.50\%$
test_getitem[slice_int]	0.1094ms	21.9156μs	45.6295 KOps/s	49.2116 KOps/s	$\textbf{\color{#d91a1a}-7.28\%}$
test_getitem[range]	0.1353ms	39.6526μs	25.2190 KOps/s	27.9139 KOps/s	$\textbf{\color{#d91a1a}-9.65\%}$
test_getitem[tuple]	0.1126ms	18.9765μs	52.6969 KOps/s	54.7156 KOps/s	$\color{#d91a1a}-3.69\%$
test_getitem[list]	0.1257ms	33.8624μs	29.5313 KOps/s	30.7141 KOps/s	$\color{#d91a1a}-3.85\%$
test_setitem_dim[int]	30.5010μs	19.3136μs	51.7769 KOps/s	57.3985 KOps/s	$\textbf{\color{#d91a1a}-9.79\%}$
test_setitem_dim[slice_int]	75.2810μs	38.4953μs	25.9772 KOps/s	27.6449 KOps/s	$\textbf{\color{#d91a1a}-6.03\%}$
test_setitem_dim[range]	80.7820μs	54.6874μs	18.2857 KOps/s	19.8046 KOps/s	$\textbf{\color{#d91a1a}-7.67\%}$
test_setitem_dim[tuple]	54.1320μs	32.7763μs	30.5098 KOps/s	31.3742 KOps/s	$\color{#d91a1a}-2.75\%$
test_setitem	51.5010μs	16.0629μs	62.2552 KOps/s	65.7967 KOps/s	$\textbf{\color{#d91a1a}-5.38\%}$
test_set	0.1234ms	15.5228μs	64.4213 KOps/s	68.0955 KOps/s	$\textbf{\color{#d91a1a}-5.40\%}$
test_set_shared	1.6327ms	0.1487ms	6.7237 KOps/s	6.7540 KOps/s	$\color{#d91a1a}-0.45\%$
test_update	0.5557ms	19.3580μs	51.6582 KOps/s	56.1929 KOps/s	$\textbf{\color{#d91a1a}-8.07\%}$
test_update_nested	0.1244ms	24.7366μs	40.4260 KOps/s	42.5051 KOps/s	$\color{#d91a1a}-4.89\%$
test_update__nested	0.9100ms	25.5775μs	39.0968 KOps/s	42.5405 KOps/s	$\textbf{\color{#d91a1a}-8.10\%}$
test_set_nested	0.1325ms	16.6832μs	59.9404 KOps/s	63.4724 KOps/s	$\textbf{\color{#d91a1a}-5.56\%}$
test_set_nested_new	0.1223ms	19.2677μs	51.9002 KOps/s	55.0538 KOps/s	$\textbf{\color{#d91a1a}-5.73\%}$
test_select	0.1354ms	31.8629μs	31.3845 KOps/s	33.9238 KOps/s	$\textbf{\color{#d91a1a}-7.49\%}$
test_select_nested	73.9620μs	43.8381μs	22.8112 KOps/s	23.2607 KOps/s	$\color{#d91a1a}-1.93\%$
test_exclude_nested	0.1031ms	62.8792μs	15.9035 KOps/s	15.8571 KOps/s	$\color{#35bf28}+0.29\%$
test_empty[True]	0.3558ms	0.2877ms	3.4754 KOps/s	3.4785 KOps/s	$\color{#d91a1a}-0.09\%$
test_empty[False]	3.7731μs	0.8273μs	1.2088 MOps/s	1.1981 MOps/s	$\color{#35bf28}+0.89\%$
test_to	82.7920μs	55.5583μs	17.9991 KOps/s	17.8567 KOps/s	$\color{#35bf28}+0.80\%$
test_to_nonblocking	0.9683ms	47.8102μs	20.9160 KOps/s	21.3313 KOps/s	$\color{#d91a1a}-1.95\%$
test_unbind_speed	1.8433ms	0.2358ms	4.2407 KOps/s	4.2753 KOps/s	$\color{#d91a1a}-0.81\%$
test_unbind_speed_stack0	0.3617ms	0.2401ms	4.1656 KOps/s	4.2621 KOps/s	$\color{#d91a1a}-2.26\%$
test_unbind_speed_stack1	91.7154ms	0.6702ms	1.4921 KOps/s	1.4973 KOps/s	$\color{#d91a1a}-0.34\%$
test_split	92.0432ms	1.6179ms	618.0773 Ops/s	637.8002 Ops/s	$\color{#d91a1a}-3.09\%$
test_chunk	94.5223ms	1.7670ms	565.9293 Ops/s	582.6468 Ops/s	$\color{#d91a1a}-2.87\%$
test_consolidate[False-None]	2.9947ms	2.6742ms	373.9450 Ops/s	381.1775 Ops/s	$\color{#d91a1a}-1.90\%$
test_consolidate[default-None]	1.8384ms	1.7251ms	579.6824 Ops/s	590.2473 Ops/s	$\color{#d91a1a}-1.79\%$
test_consolidate[reduce-overhead-None]	1.9102ms	1.7683ms	565.5114 Ops/s	577.1500 Ops/s	$\color{#d91a1a}-2.02\%$
test_consolidate_njt[False-None]	6.7341ms	6.5582ms	152.4805 Ops/s	154.2526 Ops/s	$\color{#d91a1a}-1.15\%$
test_to[False-False-None]	1.8064ms	1.7104ms	584.6473 Ops/s	579.3490 Ops/s	$\color{#35bf28}+0.91\%$
test_to[True-False-None]	1.5575ms	1.3296ms	752.0819 Ops/s	763.1031 Ops/s	$\color{#d91a1a}-1.44\%$
test_to[within-False-None]	4.4892ms	4.1858ms	238.9034 Ops/s	245.9609 Ops/s	$\color{#d91a1a}-2.87\%$
test_to[True-default-None]	5.5606ms	5.2929ms	188.9316 Ops/s	178.9869 Ops/s	$\textbf{\color{#35bf28}+5.56\%}$
test_to_njt[False-False-None]	7.0154ms	6.8603ms	145.7653 Ops/s	145.8504 Ops/s	$\color{#d91a1a}-0.06\%$
test_to_njt[True-False-None]	5.8832ms	5.4582ms	183.2112 Ops/s	185.0032 Ops/s	$\color{#d91a1a}-0.97\%$
test_to_njt[within-False-None]	12.4625ms	12.1718ms	82.1571 Ops/s	82.7965 Ops/s	$\color{#d91a1a}-0.77\%$
test_creation[device0]	0.7147ms	78.6515μs	12.7143 KOps/s	12.7696 KOps/s	$\color{#d91a1a}-0.43\%$
test_creation_from_tensor	0.6045ms	82.6553μs	12.0984 KOps/s	12.0795 KOps/s	$\color{#35bf28}+0.16\%$
test_add_one[memmap_tensor0]	0.4753ms	6.3516μs	157.4414 KOps/s	163.6476 KOps/s	$\color{#d91a1a}-3.79\%$
test_contiguous[memmap_tensor0]	1.8505μs	0.4071μs	2.4564 MOps/s	2.5096 MOps/s	$\color{#d91a1a}-2.12\%$
test_stack[memmap_tensor0]	46.5310μs	4.6191μs	216.4909 KOps/s	223.9520 KOps/s	$\color{#d91a1a}-3.33\%$
test_memmaptd_index	1.8130ms	0.2640ms	3.7873 KOps/s	4.0084 KOps/s	$\textbf{\color{#d91a1a}-5.51\%}$
test_memmaptd_index_astensor	0.9737ms	0.3306ms	3.0248 KOps/s	3.2014 KOps/s	$\textbf{\color{#d91a1a}-5.52\%}$
test_memmaptd_index_op	1.0555ms	0.6106ms	1.6376 KOps/s	1.7122 KOps/s	$\color{#d91a1a}-4.36\%$
test_serialize_model	0.1309s	0.1298s	7.7022 Ops/s	7.6661 Ops/s	$\color{#35bf28}+0.47\%$
test_serialize_model_pickle	1.3494s	1.2152s	0.8229 Ops/s	0.8222 Ops/s	$\color{#35bf28}+0.08\%$
test_serialize_weights	0.4113s	0.1697s	5.8914 Ops/s	7.6890 Ops/s	$\textbf{\color{#d91a1a}-23.38\%}$
test_serialize_weights_returnearly	0.3140s	53.5523ms	18.6733 Ops/s	14.7766 Ops/s	$\textbf{\color{#35bf28}+26.37\%}$
test_serialize_weights_pickle	1.3729s	1.2155s	0.8227 Ops/s	0.8222 Ops/s	$\color{#35bf28}+0.06\%$
test_reshape_pytree	83.2420μs	22.0522μs	45.3469 KOps/s	45.9656 KOps/s	$\color{#d91a1a}-1.35\%$
test_reshape_td	52.5510μs	26.5055μs	37.7280 KOps/s	36.5794 KOps/s	$\color{#35bf28}+3.14\%$
test_view_pytree	47.8710μs	22.1203μs	45.2074 KOps/s	46.6570 KOps/s	$\color{#d91a1a}-3.11\%$
test_view_td	60.3610μs	30.5680μs	32.7140 KOps/s	33.0339 KOps/s	$\color{#d91a1a}-0.97\%$
test_unbind_pytree	55.4810μs	28.1236μs	35.5573 KOps/s	36.7003 KOps/s	$\color{#d91a1a}-3.11\%$
test_unbind_td	0.8392ms	36.3838μs	27.4848 KOps/s	27.6954 KOps/s	$\color{#d91a1a}-0.76\%$
test_split_pytree	61.4110μs	30.8405μs	32.4249 KOps/s	33.9817 KOps/s	$\color{#d91a1a}-4.58\%$
test_split_td	1.0917ms	39.6647μs	25.2113 KOps/s	27.0273 KOps/s	$\textbf{\color{#d91a1a}-6.72\%}$
test_add_pytree	63.2710μs	33.7783μs	29.6048 KOps/s	31.1241 KOps/s	$\color{#d91a1a}-4.88\%$
test_add_td	0.1872ms	50.2621μs	19.8957 KOps/s	21.3675 KOps/s	$\textbf{\color{#d91a1a}-6.89\%}$
test_compile_add_one_nested[tensordict-compile]	0.1720ms	0.1192ms	8.3881 KOps/s	8.3989 KOps/s	$\color{#d91a1a}-0.13\%$
test_compile_add_one_nested[tensordict-eager]	0.2159ms	0.1281ms	7.8048 KOps/s	7.6568 KOps/s	$\color{#35bf28}+1.93\%$
test_compile_add_one_nested[pytree-compile]	0.1584ms	94.5265μs	10.5790 KOps/s	10.6926 KOps/s	$\color{#d91a1a}-1.06\%$
test_compile_add_one_nested[pytree-eager]	1.3785ms	0.1493ms	6.6959 KOps/s	6.6505 KOps/s	$\color{#35bf28}+0.68\%$
test_compile_copy_nested[tensordict-compile]	52.3410μs	22.8747μs	43.7163 KOps/s	47.0566 KOps/s	$\textbf{\color{#d91a1a}-7.10\%}$
test_compile_copy_nested[tensordict-eager]	55.2320μs	29.3456μs	34.0766 KOps/s	34.0377 KOps/s	$\color{#35bf28}+0.11\%$
test_compile_copy_nested[pytree-compile]	99.4320μs	64.1058μs	15.5992 KOps/s	15.4898 KOps/s	$\color{#35bf28}+0.71\%$
test_compile_copy_nested[pytree-eager]	79.5310μs	48.7973μs	20.4929 KOps/s	20.1268 KOps/s	$\color{#35bf28}+1.82\%$
test_compile_add_one_flat[tensordict-compile]	0.1812ms	0.1406ms	7.1146 KOps/s	7.3294 KOps/s	$\color{#d91a1a}-2.93\%$
test_compile_add_one_flat[tensordict-eager]	0.3117ms	0.2149ms	4.6534 KOps/s	4.6823 KOps/s	$\color{#d91a1a}-0.62\%$
test_compile_add_one_flat[tensorclass-compile]	0.1562ms	99.9890μs	10.0011 KOps/s	10.5223 KOps/s	$\color{#d91a1a}-4.95\%$
test_compile_add_one_flat[tensorclass-eager]	0.1142ms	54.0755μs	18.4927 KOps/s	18.0204 KOps/s	$\color{#35bf28}+2.62\%$
test_compile_add_one_flat[pytree-compile]	0.1850ms	0.1369ms	7.3036 KOps/s	7.4143 KOps/s	$\color{#d91a1a}-1.49\%$
test_compile_add_one_flat[pytree-eager]	0.6008ms	0.4867ms	2.0548 KOps/s	2.0688 KOps/s	$\color{#d91a1a}-0.68\%$
test_compile_add_self_flat[tensordict-eager]	0.3722ms	0.2583ms	3.8713 KOps/s	3.8739 KOps/s	$\color{#d91a1a}-0.07\%$
test_compile_add_self_flat[tensordict-compile]	0.2100ms	0.1463ms	6.8335 KOps/s	7.1968 KOps/s	$\textbf{\color{#d91a1a}-5.05\%}$
test_compile_add_self_flat[tensorclass-eager]	0.1622ms	64.4133μs	15.5247 KOps/s	15.4355 KOps/s	$\color{#35bf28}+0.58\%$
test_compile_add_self_flat[tensorclass-compile]	0.1482ms	99.9866μs	10.0013 KOps/s	10.0682 KOps/s	$\color{#d91a1a}-0.66\%$
test_compile_add_self_flat[pytree-eager]	0.5094ms	0.4159ms	2.4044 KOps/s	2.4484 KOps/s	$\color{#d91a1a}-1.80\%$
test_compile_add_self_flat[pytree-compile]	0.1836ms	0.1348ms	7.4206 KOps/s	7.2278 KOps/s	$\color{#35bf28}+2.67\%$
test_compile_copy_flat[tensordict-compile]	51.4510μs	18.7097μs	53.4482 KOps/s	56.3523 KOps/s	$\textbf{\color{#d91a1a}-5.15\%}$
test_compile_copy_flat[tensordict-eager]	58.8610μs	30.5439μs	32.7398 KOps/s	32.1659 KOps/s	$\color{#35bf28}+1.78\%$
test_compile_copy_flat[pytree-compile]	0.1772ms	70.0578μs	14.2739 KOps/s	14.4246 KOps/s	$\color{#d91a1a}-1.04\%$
test_compile_copy_flat[pytree-eager]	80.8420μs	51.1734μs	19.5414 KOps/s	19.4725 KOps/s	$\color{#35bf28}+0.35\%$
test_compile_assign_and_add[tensordict-compile]	1.7169ms	0.4200ms	2.3809 KOps/s	2.2721 KOps/s	$\color{#35bf28}+4.79\%$
test_compile_assign_and_add[tensordict-eager]	2.9484ms	2.6414ms	378.5803 Ops/s	385.0892 Ops/s	$\color{#d91a1a}-1.69\%$
test_compile_assign_and_add[pytree-compile]	1.5727ms	0.4272ms	2.3406 KOps/s	2.3157 KOps/s	$\color{#35bf28}+1.07\%$
test_compile_assign_and_add[pytree-eager]	3.0413ms	2.7183ms	367.8785 Ops/s	383.8057 Ops/s	$\color{#d91a1a}-4.15\%$
test_compile_indexing[tensor-tensordict-compile]	0.2319ms	0.1196ms	8.3645 KOps/s	8.8019 KOps/s	$\color{#d91a1a}-4.97\%$
test_compile_indexing[tensor-tensordict-eager]	0.5477ms	77.2885μs	12.9385 KOps/s	12.6805 KOps/s	$\color{#35bf28}+2.04\%$
test_compile_indexing[tensor-tensorclass-compile]	0.1666ms	0.1049ms	9.5292 KOps/s	9.5207 KOps/s	$\color{#35bf28}+0.09\%$
test_compile_indexing[tensor-tensorclass-eager]	0.1092ms	66.6579μs	15.0020 KOps/s	14.8689 KOps/s	$\color{#35bf28}+0.89\%$
test_compile_indexing[tensor-pytree-compile]	0.1630ms	0.1055ms	9.4819 KOps/s	9.4286 KOps/s	$\color{#35bf28}+0.56\%$
test_compile_indexing[tensor-pytree-eager]	0.1485ms	70.3738μs	14.2098 KOps/s	14.8917 KOps/s	$\color{#d91a1a}-4.58\%$
test_compile_indexing[slice-tensordict-compile]	0.1490ms	0.1034ms	9.6756 KOps/s	9.7133 KOps/s	$\color{#d91a1a}-0.39\%$
test_compile_indexing[slice-tensordict-eager]	0.1487ms	17.3434μs	57.6589 KOps/s	56.6672 KOps/s	$\color{#35bf28}+1.75\%$
test_compile_indexing[slice-tensorclass-compile]	0.1455ms	0.1006ms	9.9408 KOps/s	10.3922 KOps/s	$\color{#d91a1a}-4.34\%$
test_compile_indexing[slice-tensorclass-eager]	42.7700μs	15.6720μs	63.8080 KOps/s	64.7889 KOps/s	$\color{#d91a1a}-1.51\%$
test_compile_indexing[slice-pytree-compile]	0.1528ms	0.1014ms	9.8639 KOps/s	10.1063 KOps/s	$\color{#d91a1a}-2.40\%$
test_compile_indexing[slice-pytree-eager]	79.1920μs	15.4517μs	64.7178 KOps/s	65.1692 KOps/s	$\color{#d91a1a}-0.69\%$
test_compile_indexing[int-tensordict-compile]	0.1516ms	0.1005ms	9.9466 KOps/s	9.6231 KOps/s	$\color{#35bf28}+3.36\%$
test_compile_indexing[int-tensordict-eager]	0.5634ms	16.8814μs	59.2367 KOps/s	59.9622 KOps/s	$\color{#d91a1a}-1.21\%$
test_compile_indexing[int-tensorclass-compile]	0.1391ms	96.7940μs	10.3312 KOps/s	10.2800 KOps/s	$\color{#35bf28}+0.50\%$
test_compile_indexing[int-tensorclass-eager]	49.8710μs	15.6690μs	63.8201 KOps/s	65.0500 KOps/s	$\color{#d91a1a}-1.89\%$
test_compile_indexing[int-pytree-compile]	0.1486ms	96.8750μs	10.3226 KOps/s	10.2172 KOps/s	$\color{#35bf28}+1.03\%$
test_compile_indexing[int-pytree-eager]	41.7310μs	15.7740μs	63.3955 KOps/s	65.2235 KOps/s	$\color{#d91a1a}-2.80\%$
test_mod_add[eager]	0.1163ms	38.3374μs	26.0842 KOps/s	26.0311 KOps/s	$\color{#35bf28}+0.20\%$
test_mod_add[compile]	0.1296ms	81.3895μs	12.2866 KOps/s	12.7114 KOps/s	$\color{#d91a1a}-3.34\%$
test_mod_add[compile-overhead]	0.3247ms	0.1766ms	5.6632 KOps/s	5.8441 KOps/s	$\color{#d91a1a}-3.10\%$
test_mod_wrap[eager]	0.3313ms	0.2540ms	3.9368 KOps/s	4.0739 KOps/s	$\color{#d91a1a}-3.36\%$
test_mod_wrap[compile]	0.3481ms	0.2797ms	3.5750 KOps/s	3.5349 KOps/s	$\color{#35bf28}+1.14\%$
test_mod_wrap[compile-overhead]	7.1096ms	3.7643ms	265.6540 Ops/s	266.9544 Ops/s	$\color{#d91a1a}-0.49\%$
test_mod_wrap_and_backward[eager]	1.4252ms	1.3200ms	757.5848 Ops/s	704.3921 Ops/s	$\textbf{\color{#35bf28}+7.55\%}$
test_mod_wrap_and_backward[compile]	1.6854ms	1.2507ms	799.5617 Ops/s	727.4414 Ops/s	$\textbf{\color{#35bf28}+9.91\%}$
test_mod_wrap_and_backward[compile-overhead]	1.3743ms	0.9219ms	1.0847 KOps/s	980.2974 Ops/s	$\textbf{\color{#35bf28}+10.65\%}$
test_seq_add[eager]	0.5212ms	0.1147ms	8.7165 KOps/s	8.6547 KOps/s	$\color{#35bf28}+0.71\%$
test_seq_add[compile]	0.4794ms	88.2753μs	11.3282 KOps/s	11.4213 KOps/s	$\color{#d91a1a}-0.82\%$
test_seq_add[compile-overhead]	0.5181ms	0.1264ms	7.9112 KOps/s	7.8903 KOps/s	$\color{#35bf28}+0.27\%$
test_seq_wrap[eager]	0.7927ms	0.4153ms	2.4081 KOps/s	2.4022 KOps/s	$\color{#35bf28}+0.24\%$
test_seq_wrap[compile]	0.7395ms	0.3056ms	3.2718 KOps/s	3.3311 KOps/s	$\color{#d91a1a}-1.78\%$
test_seq_wrap[compile-overhead]	0.6157ms	0.2205ms	4.5353 KOps/s	4.5601 KOps/s	$\color{#d91a1a}-0.54\%$
test_func_call_runtime[False-eager]	1.1031ms	0.7067ms	1.4150 KOps/s	1.3494 KOps/s	$\color{#35bf28}+4.87\%$
test_func_call_runtime[False-compile]	1.1579ms	0.7372ms	1.3565 KOps/s	1.2864 KOps/s	$\textbf{\color{#35bf28}+5.44\%}$
test_func_call_runtime[False-compile-overhead]	0.4381ms	0.3564ms	2.8057 KOps/s	2.8238 KOps/s	$\color{#d91a1a}-0.64\%$
test_func_call_runtime[True-eager]	1.2841ms	0.8734ms	1.1450 KOps/s	1.1379 KOps/s	$\color{#35bf28}+0.62\%$
test_func_call_runtime[True-compile]	1.1619ms	0.7622ms	1.3121 KOps/s	1.2974 KOps/s	$\color{#35bf28}+1.13\%$
test_func_call_runtime[True-compile-overhead]	0.7925ms	0.3794ms	2.6356 KOps/s	2.6552 KOps/s	$\color{#d91a1a}-0.74\%$
test_func_call_cm_runtime[False-eager]	1.1240ms	0.7082ms	1.4120 KOps/s	1.3864 KOps/s	$\color{#35bf28}+1.85\%$
test_func_call_cm_runtime[False-compile]	1.1291ms	0.7392ms	1.3529 KOps/s	1.3283 KOps/s	$\color{#35bf28}+1.85\%$
test_func_call_cm_runtime[False-compile-overhead]	0.4265ms	0.3605ms	2.7738 KOps/s	2.7839 KOps/s	$\color{#d91a1a}-0.36\%$
test_func_call_cm_runtime[True-eager]	1.3810ms	0.9860ms	1.0142 KOps/s	1.0059 KOps/s	$\color{#35bf28}+0.82\%$
test_func_call_cm_runtime[True-compile]	1.1214ms	0.8141ms	1.2283 KOps/s	1.2596 KOps/s	$\color{#d91a1a}-2.49\%$
test_func_call_cm_runtime[True-compile-overhead]	0.4626ms	0.4037ms	2.4772 KOps/s	2.4599 KOps/s	$\color{#35bf28}+0.70\%$
test_vmap_func_call_cm_runtime[eager]	2.5309ms	2.0057ms	498.5880 Ops/s	493.0859 Ops/s	$\color{#35bf28}+1.12\%$
test_vmap_func_call_cm_runtime[compile]	0.9003ms	0.7949ms	1.2581 KOps/s	1.2303 KOps/s	$\color{#35bf28}+2.26\%$
test_vmap_func_call_cm_runtime[compile-overhead]	0.4657ms	0.4048ms	2.4701 KOps/s	2.4597 KOps/s	$\color{#35bf28}+0.42\%$
test_distributed	2.8632ms	0.3489ms	2.8660 KOps/s	8.1271 KOps/s	$\textbf{\color{#d91a1a}-64.73\%}$
test_tdmodule	77.7520μs	20.0186μs	49.9537 KOps/s	47.6033 KOps/s	$\color{#35bf28}+4.94\%$
test_tdmodule_dispatch	64.9110μs	37.0159μs	27.0154 KOps/s	26.9646 KOps/s	$\color{#35bf28}+0.19\%$
test_tdseq	41.7810μs	20.1526μs	49.6214 KOps/s	48.9263 KOps/s	$\color{#35bf28}+1.42\%$
test_tdseq_dispatch	57.0920μs	38.5958μs	25.9096 KOps/s	25.5012 KOps/s	$\color{#35bf28}+1.60\%$
test_instantiation_functorch	1.6411ms	1.5250ms	655.7187 Ops/s	651.9999 Ops/s	$\color{#35bf28}+0.57\%$
test_exec_functorch	0.2093ms	0.1400ms	7.1426 KOps/s	7.1566 KOps/s	$\color{#d91a1a}-0.20\%$
test_exec_functional_call	0.1834ms	0.1315ms	7.6066 KOps/s	7.4481 KOps/s	$\color{#35bf28}+2.13\%$
test_exec_td_decorator	0.3654ms	0.1798ms	5.5631 KOps/s	5.5375 KOps/s	$\color{#35bf28}+0.46\%$
test_vmap_mlp_speed_decorator[True-True]	0.8534ms	0.6790ms	1.4727 KOps/s	1.4529 KOps/s	$\color{#35bf28}+1.36\%$
test_vmap_mlp_speed_decorator[True-False]	0.8872ms	0.6908ms	1.4475 KOps/s	1.4565 KOps/s	$\color{#d91a1a}-0.61\%$
test_vmap_mlp_speed_decorator[False-True]	0.7238ms	0.6006ms	1.6650 KOps/s	1.7314 KOps/s	$\color{#d91a1a}-3.83\%$
test_vmap_mlp_speed_decorator[False-False]	0.7372ms	0.6030ms	1.6584 KOps/s	1.7306 KOps/s	$\color{#d91a1a}-4.17\%$
test_vmap_transformer_speed_decorator[True-True]	19.9346ms	18.9257ms	52.8381 Ops/s	53.6828 Ops/s	$\color{#d91a1a}-1.57\%$
test_vmap_transformer_speed_decorator[True-False]	19.3842ms	18.6755ms	53.5461 Ops/s	53.4498 Ops/s	$\color{#35bf28}+0.18\%$
test_vmap_transformer_speed_decorator[False-True]	18.7925ms	18.5658ms	53.8624 Ops/s	54.1120 Ops/s	$\color{#d91a1a}-0.46\%$
test_vmap_transformer_speed_decorator[False-False]	18.7233ms	18.5171ms	54.0042 Ops/s	54.1277 Ops/s	$\color{#d91a1a}-0.23\%$
test_to_module_speed[True]	1.0569ms	0.9596ms	1.0421 KOps/s	1.0276 KOps/s	$\color{#35bf28}+1.42\%$
test_to_module_speed[False]	1.3282ms	0.9437ms	1.0597 KOps/s	1.0520 KOps/s	$\color{#35bf28}+0.73\%$
test_tc_init	72.0710μs	39.0386μs	25.6157 KOps/s	26.5280 KOps/s	$\color{#d91a1a}-3.44\%$
test_tc_init_nested	0.1192ms	79.6364μs	12.5571 KOps/s	13.5484 KOps/s	$\textbf{\color{#d91a1a}-7.32\%}$
test_tc_first_layer_tensor	26.7110μs	0.8054μs	1.2417 MOps/s	1.4419 MOps/s	$\textbf{\color{#d91a1a}-13.89\%}$
test_tc_first_layer_nontensor	25.4710μs	2.2934μs	436.0337 KOps/s	435.9161 KOps/s	$\color{#35bf28}+0.03\%$
test_tc_second_layer_tensor	36.2282μs	1.4101μs	709.1483 KOps/s	714.0697 KOps/s	$\color{#d91a1a}-0.69\%$
test_tc_second_layer_nontensor	26.2710μs	3.0424μs	328.6830 KOps/s	332.5238 KOps/s	$\color{#d91a1a}-1.16\%$
test_unbind	0.2227s	10.3147ms	96.9491 Ops/s	144.8519 Ops/s	$\textbf{\color{#d91a1a}-33.07\%}$
test_full_like	10.0962ms	9.1059ms	109.8190 Ops/s	110.2935 Ops/s	$\color{#d91a1a}-0.43\%$
test_zeros_like	9.1652ms	7.1783ms	139.3083 Ops/s	235.5167 Ops/s	$\textbf{\color{#d91a1a}-40.85\%}$
test_ones_like	4.4809ms	4.3037ms	232.3560 Ops/s	237.9432 Ops/s	$\color{#d91a1a}-2.35\%$
test_clone	13.6095ms	10.3209ms	96.8906 Ops/s	111.1751 Ops/s	$\textbf{\color{#d91a1a}-12.85\%}$
test_squeeze	57.4710μs	9.5322μs	104.9075 KOps/s	107.1155 KOps/s	$\color{#d91a1a}-2.06\%$
test_unsqueeze	0.2117ms	72.5303μs	13.7873 KOps/s	13.5943 KOps/s	$\color{#35bf28}+1.42\%$
test_split	0.2669ms	0.1586ms	6.3034 KOps/s	6.1041 KOps/s	$\color{#35bf28}+3.26\%$
test_permute	0.2598ms	0.1770ms	5.6482 KOps/s	5.4102 KOps/s	$\color{#35bf28}+4.40\%$
test_stack	50.6017ms	50.2868ms	19.8859 Ops/s	19.9289 Ops/s	$\color{#d91a1a}-0.22\%$
test_cat	50.4578ms	50.1391ms	19.9445 Ops/s	19.8803 Ops/s	$\color{#35bf28}+0.32\%$