[Performance] Fewer syncs during calls to to

[vmoens]

No description provided.

[github-actions]

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

Total Benchmarks: 144. Improved: $\large\color{#35bf28}11$. Worsened: $\large\color{#d91a1a}16$.

Expand to view detailed results

Name	Max	Mean	Ops	Ops on Repo HEAD	Change
test_plain_set_nested	27.4520μs	16.9719μs	58.9210 KOps/s	60.1777 KOps/s	$\color{#d91a1a}-2.09\%$
test_plain_set_stack_nested	47.1480μs	17.9692μs	55.6508 KOps/s	58.2602 KOps/s	$\color{#d91a1a}-4.48\%$
test_plain_set_nested_inplace	48.0300μs	20.0397μs	49.9008 KOps/s	52.1245 KOps/s	$\color{#d91a1a}-4.27\%$
test_plain_set_stack_nested_inplace	45.1840μs	20.0824μs	49.7948 KOps/s	51.4836 KOps/s	$\color{#d91a1a}-3.28\%$
test_items	21.4900μs	2.6558μs	376.5401 KOps/s	370.7209 KOps/s	$\color{#35bf28}+1.57\%$
test_items_nested	1.2875ms	0.2697ms	3.7075 KOps/s	3.7934 KOps/s	$\color{#d91a1a}-2.26\%$
test_items_nested_locked	0.4232ms	0.2713ms	3.6857 KOps/s	3.7998 KOps/s	$\color{#d91a1a}-3.00\%$
test_items_nested_leaf	0.1498ms	77.5556μs	12.8940 KOps/s	13.0905 KOps/s	$\color{#d91a1a}-1.50\%$
test_items_stack_nested	0.3307ms	0.2733ms	3.6597 KOps/s	3.7117 KOps/s	$\color{#d91a1a}-1.40\%$
test_items_stack_nested_leaf	0.1516ms	79.5025μs	12.5782 KOps/s	12.6924 KOps/s	$\color{#d91a1a}-0.90\%$
test_items_stack_nested_locked	0.4243ms	0.2708ms	3.6926 KOps/s	3.7499 KOps/s	$\color{#d91a1a}-1.53\%$
test_keys	49.6840μs	3.7922μs	263.6985 KOps/s	257.4315 KOps/s	$\color{#35bf28}+2.43\%$
test_keys_nested	0.2848ms	0.1413ms	7.0784 KOps/s	7.2194 KOps/s	$\color{#d91a1a}-1.95\%$
test_keys_nested_locked	0.7179ms	0.1459ms	6.8548 KOps/s	7.0620 KOps/s	$\color{#d91a1a}-2.93\%$
test_keys_nested_leaf	0.2355ms	0.1205ms	8.2986 KOps/s	8.6306 KOps/s	$\color{#d91a1a}-3.85\%$
test_keys_stack_nested	0.2034ms	0.1399ms	7.1471 KOps/s	7.2552 KOps/s	$\color{#d91a1a}-1.49\%$
test_keys_stack_nested_leaf	0.2410ms	0.1202ms	8.3214 KOps/s	8.6069 KOps/s	$\color{#d91a1a}-3.32\%$
test_keys_stack_nested_locked	0.2553ms	0.1442ms	6.9342 KOps/s	7.0825 KOps/s	$\color{#d91a1a}-2.09\%$
test_values	8.7224μs	1.1674μs	856.5931 KOps/s	876.7103 KOps/s	$\color{#d91a1a}-2.29\%$
test_values_nested	94.7880μs	51.2814μs	19.5002 KOps/s	19.4146 KOps/s	$\color{#35bf28}+0.44\%$
test_values_nested_locked	0.1052ms	51.7041μs	19.3408 KOps/s	19.5007 KOps/s	$\color{#d91a1a}-0.82\%$
test_values_nested_leaf	93.4070μs	46.4710μs	21.5188 KOps/s	21.4449 KOps/s	$\color{#35bf28}+0.34\%$
test_values_stack_nested	94.8780μs	52.4486μs	19.0663 KOps/s	19.1127 KOps/s	$\color{#d91a1a}-0.24\%$
test_values_stack_nested_leaf	98.1440μs	46.6711μs	21.4265 KOps/s	21.2884 KOps/s	$\color{#35bf28}+0.65\%$
test_values_stack_nested_locked	0.1074ms	52.1143μs	19.1886 KOps/s	19.2780 KOps/s	$\color{#d91a1a}-0.46\%$
test_membership	30.3770μs	1.3611μs	734.6876 KOps/s	754.4887 KOps/s	$\color{#d91a1a}-2.62\%$
test_membership_nested	32.5510μs	3.4793μs	287.4120 KOps/s	294.1604 KOps/s	$\color{#d91a1a}-2.29\%$
test_membership_nested_leaf	23.0430μs	3.5253μs	283.6643 KOps/s	281.6909 KOps/s	$\color{#35bf28}+0.70\%$
test_membership_stacked_nested	31.8200μs	3.4933μs	286.2582 KOps/s	296.6883 KOps/s	$\color{#d91a1a}-3.52\%$
test_membership_stacked_nested_leaf	41.0370μs	3.4425μs	290.4903 KOps/s	295.0962 KOps/s	$\color{#d91a1a}-1.56\%$
test_membership_nested_last	45.0620μs	4.2040μs	237.8698 KOps/s	241.1007 KOps/s	$\color{#d91a1a}-1.34\%$
test_membership_nested_leaf_last	47.5250μs	4.2421μs	235.7331 KOps/s	237.8992 KOps/s	$\color{#d91a1a}-0.91\%$
test_membership_stacked_nested_last	40.3450μs	4.8098μs	207.9091 KOps/s	187.4677 KOps/s	$\textbf{\color{#35bf28}+10.90\%}$
test_membership_stacked_nested_leaf_last	44.8740μs	4.8588μs	205.8114 KOps/s	189.6819 KOps/s	$\textbf{\color{#35bf28}+8.50\%}$
test_nested_getleaf	47.2690μs	11.3476μs	88.1244 KOps/s	95.5364 KOps/s	$\textbf{\color{#d91a1a}-7.76\%}$
test_nested_get	39.2930μs	10.7850μs	92.7211 KOps/s	99.5649 KOps/s	$\textbf{\color{#d91a1a}-6.87\%}$
test_stacked_getleaf	31.6500μs	11.3625μs	88.0085 KOps/s	97.0107 KOps/s	$\textbf{\color{#d91a1a}-9.28\%}$
test_stacked_get	50.3750μs	10.7035μs	93.4277 KOps/s	101.9007 KOps/s	$\textbf{\color{#d91a1a}-8.31\%}$
test_nested_getitemleaf	43.0710μs	11.7907μs	84.8124 KOps/s	89.7102 KOps/s	$\textbf{\color{#d91a1a}-5.46\%}$
test_nested_getitem	34.9760μs	11.0513μs	90.4868 KOps/s	97.1827 KOps/s	$\textbf{\color{#d91a1a}-6.89\%}$
test_stacked_getitemleaf	41.9290μs	11.7690μs	84.9689 KOps/s	90.7619 KOps/s	$\textbf{\color{#d91a1a}-6.38\%}$
test_stacked_getitem	36.9290μs	11.0143μs	90.7911 KOps/s	98.1992 KOps/s	$\textbf{\color{#d91a1a}-7.54\%}$
test_lock_nested	0.7709ms	0.3437ms	2.9096 KOps/s	2.9384 KOps/s	$\color{#d91a1a}-0.98\%$
test_lock_stack_nested	0.4281ms	0.3105ms	3.2207 KOps/s	3.2712 KOps/s	$\color{#d91a1a}-1.54\%$
test_unlock_nested	0.7538ms	0.3451ms	2.8974 KOps/s	2.9018 KOps/s	$\color{#d91a1a}-0.15\%$
test_unlock_stack_nested	0.3772ms	0.3177ms	3.1479 KOps/s	3.2020 KOps/s	$\color{#d91a1a}-1.69\%$
test_flatten_speed	0.5498ms	95.8806μs	10.4296 KOps/s	10.3721 KOps/s	$\color{#35bf28}+0.56\%$
test_unflatten_speed	0.6185ms	0.4246ms	2.3552 KOps/s	2.4576 KOps/s	$\color{#d91a1a}-4.17\%$
test_common_ops	3.3784ms	0.7144ms	1.3999 KOps/s	1.4228 KOps/s	$\color{#d91a1a}-1.61\%$
test_creation	15.5390μs	1.8912μs	528.7603 KOps/s	525.6850 KOps/s	$\color{#35bf28}+0.58\%$
test_creation_empty	34.7660μs	10.5216μs	95.0427 KOps/s	96.8261 KOps/s	$\color{#d91a1a}-1.84\%$
test_creation_nested_1	57.8490μs	13.3237μs	75.0544 KOps/s	76.5377 KOps/s	$\color{#d91a1a}-1.94\%$
test_creation_nested_2	37.5710μs	16.5537μs	60.4093 KOps/s	60.7456 KOps/s	$\color{#d91a1a}-0.55\%$
test_clone	97.1220μs	13.5201μs	73.9637 KOps/s	74.5563 KOps/s	$\color{#d91a1a}-0.79\%$
test_getitem[int]	35.1160μs	11.4210μs	87.5582 KOps/s	85.2826 KOps/s	$\color{#35bf28}+2.67\%$
test_getitem[slice_int]	64.0800μs	21.9760μs	45.5042 KOps/s	42.4712 KOps/s	$\textbf{\color{#35bf28}+7.14\%}$
test_getitem[range]	78.8080μs	58.8326μs	16.9974 KOps/s	16.7296 KOps/s	$\color{#35bf28}+1.60\%$
test_getitem[tuple]	48.0610μs	18.5085μs	54.0292 KOps/s	51.8040 KOps/s	$\color{#35bf28}+4.30\%$
test_getitem[list]	0.1202ms	41.0517μs	24.3595 KOps/s	23.8675 KOps/s	$\color{#35bf28}+2.06\%$
test_setitem_dim[int]	64.5610μs	34.4546μs	29.0237 KOps/s	27.0361 KOps/s	$\textbf{\color{#35bf28}+7.35\%}$
test_setitem_dim[slice_int]	99.0560μs	59.8697μs	16.7029 KOps/s	15.3170 KOps/s	$\textbf{\color{#35bf28}+9.05\%}$
test_setitem_dim[range]	0.1288ms	83.0458μs	12.0415 KOps/s	11.5278 KOps/s	$\color{#35bf28}+4.46\%$
test_setitem_dim[tuple]	91.8320μs	49.4098μs	20.2389 KOps/s	19.1029 KOps/s	$\textbf{\color{#35bf28}+5.95\%}$
test_setitem	58.9410μs	20.0971μs	49.7584 KOps/s	49.6156 KOps/s	$\color{#35bf28}+0.29\%$
test_set	62.3170μs	19.7327μs	50.6774 KOps/s	50.9618 KOps/s	$\color{#d91a1a}-0.56\%$
test_set_shared	1.1408ms	0.1413ms	7.0750 KOps/s	6.9247 KOps/s	$\color{#35bf28}+2.17\%$
test_update	0.1057ms	21.6336μs	46.2244 KOps/s	47.2356 KOps/s	$\color{#d91a1a}-2.14\%$
test_update_nested	67.6770μs	30.3228μs	32.9785 KOps/s	33.2471 KOps/s	$\color{#d91a1a}-0.81\%$
test_update__nested	55.8050μs	26.1034μs	38.3091 KOps/s	39.7347 KOps/s	$\color{#d91a1a}-3.59\%$
test_set_nested	54.1620μs	22.2873μs	44.8685 KOps/s	46.3313 KOps/s	$\color{#d91a1a}-3.16\%$
test_set_nested_new	61.1050μs	26.3505μs	37.9500 KOps/s	38.7554 KOps/s	$\color{#d91a1a}-2.08\%$
test_select	0.9548ms	41.9560μs	23.8345 KOps/s	23.4570 KOps/s	$\color{#35bf28}+1.61\%$
test_select_nested	0.1221ms	60.2343μs	16.6018 KOps/s	16.3883 KOps/s	$\color{#35bf28}+1.30\%$
test_exclude_nested	0.1954ms	0.1202ms	8.3216 KOps/s	8.1897 KOps/s	$\color{#35bf28}+1.61\%$
test_empty[True]	0.4738ms	0.3966ms	2.5212 KOps/s	2.5124 KOps/s	$\color{#35bf28}+0.35\%$
test_empty[False]	7.0732μs	1.1717μs	853.4628 KOps/s	871.1122 KOps/s	$\color{#d91a1a}-2.03\%$
test_unbind_speed	0.4084ms	0.2539ms	3.9392 KOps/s	3.9213 KOps/s	$\color{#35bf28}+0.46\%$
test_unbind_speed_stack0	0.3485ms	0.2527ms	3.9567 KOps/s	4.0123 KOps/s	$\color{#d91a1a}-1.39\%$
test_unbind_speed_stack1	68.6579ms	0.7236ms	1.3819 KOps/s	1.3934 KOps/s	$\color{#d91a1a}-0.82\%$
test_split	63.6140ms	1.5800ms	632.9043 Ops/s	619.8271 Ops/s	$\color{#35bf28}+2.11\%$
test_chunk	63.9552ms	1.5890ms	629.3255 Ops/s	620.1725 Ops/s	$\color{#35bf28}+1.48\%$
test_creation[device0]	0.1530ms	83.5224μs	11.9728 KOps/s	11.7932 KOps/s	$\color{#35bf28}+1.52\%$
test_creation_from_tensor	0.2163ms	84.0522μs	11.8974 KOps/s	11.6311 KOps/s	$\color{#35bf28}+2.29\%$
test_add_one[memmap_tensor0]	65.8630μs	5.3638μs	186.4352 KOps/s	187.9410 KOps/s	$\color{#d91a1a}-0.80\%$
test_contiguous[memmap_tensor0]	6.7820μs	0.6369μs	1.5701 MOps/s	1.6052 MOps/s	$\color{#d91a1a}-2.19\%$
test_stack[memmap_tensor0]	22.8120μs	3.6477μs	274.1436 KOps/s	281.5431 KOps/s	$\color{#d91a1a}-2.63\%$
test_memmaptd_index	0.9910ms	0.2554ms	3.9160 KOps/s	3.9734 KOps/s	$\color{#d91a1a}-1.44\%$
test_memmaptd_index_astensor	0.7441ms	0.3288ms	3.0415 KOps/s	3.0633 KOps/s	$\color{#d91a1a}-0.71\%$
test_memmaptd_index_op	1.1634ms	0.6124ms	1.6330 KOps/s	1.6728 KOps/s	$\color{#d91a1a}-2.38\%$
test_serialize_model	0.1723s	0.1135s	8.8127 Ops/s	8.6590 Ops/s	$\color{#35bf28}+1.77\%$
test_serialize_model_pickle	0.4471s	0.3796s	2.6344 Ops/s	2.6218 Ops/s	$\color{#35bf28}+0.48\%$
test_serialize_weights	0.1722s	0.1115s	8.9706 Ops/s	9.5646 Ops/s	$\textbf{\color{#d91a1a}-6.21\%}$
test_serialize_weights_returnearly	0.1833s	0.1301s	7.6849 Ops/s	7.3784 Ops/s	$\color{#35bf28}+4.15\%$
test_serialize_weights_pickle	0.6937s	0.4857s	2.0587 Ops/s	2.3532 Ops/s	$\textbf{\color{#d91a1a}-12.51\%}$
test_serialize_weights_filesystem	0.1598s	0.1011s	9.8893 Ops/s	9.9802 Ops/s	$\color{#d91a1a}-0.91\%$
test_serialize_model_filesystem	94.3306ms	92.0260ms	10.8665 Ops/s	10.7636 Ops/s	$\color{#35bf28}+0.96\%$
test_reshape_pytree	62.9380μs	25.2270μs	39.6401 KOps/s	39.7371 KOps/s	$\color{#d91a1a}-0.24\%$
test_reshape_td	95.4290μs	34.2551μs	29.1927 KOps/s	28.9931 KOps/s	$\color{#35bf28}+0.69\%$
test_view_pytree	63.5390μs	25.4574μs	39.2813 KOps/s	40.0870 KOps/s	$\color{#d91a1a}-2.01\%$
test_view_td	82.5960μs	38.1396μs	26.2195 KOps/s	25.7785 KOps/s	$\color{#35bf28}+1.71\%$
test_unbind_pytree	66.2050μs	28.9829μs	34.5031 KOps/s	34.4760 KOps/s	$\color{#35bf28}+0.08\%$
test_unbind_td	0.3982ms	37.9507μs	26.3500 KOps/s	26.3833 KOps/s	$\color{#d91a1a}-0.13\%$
test_split_pytree	66.5150μs	29.5963μs	33.7880 KOps/s	34.8065 KOps/s	$\color{#d91a1a}-2.93\%$
test_split_td	0.1227ms	40.7699μs	24.5279 KOps/s	24.2802 KOps/s	$\color{#35bf28}+1.02\%$
test_add_pytree	0.1099ms	37.1185μs	26.9407 KOps/s	28.8627 KOps/s	$\textbf{\color{#d91a1a}-6.66\%}$
test_add_td	0.1022ms	56.6162μs	17.6628 KOps/s	17.7298 KOps/s	$\color{#d91a1a}-0.38\%$
test_distributed	0.2282ms	0.1002ms	9.9763 KOps/s	9.8006 KOps/s	$\color{#35bf28}+1.79\%$
test_tdmodule	29.8860μs	18.0783μs	55.3149 KOps/s	50.2899 KOps/s	$\textbf{\color{#35bf28}+9.99\%}$
test_tdmodule_dispatch	65.8330μs	35.9529μs	27.8142 KOps/s	28.6405 KOps/s	$\color{#d91a1a}-2.89\%$
test_tdseq	44.3530μs	21.4608μs	46.5965 KOps/s	47.7909 KOps/s	$\color{#d91a1a}-2.50\%$
test_tdseq_dispatch	64.1310μs	41.3412μs	24.1890 KOps/s	21.2054 KOps/s	$\textbf{\color{#35bf28}+14.07\%}$
test_instantiation_functorch	2.7877ms	1.3253ms	754.5553 Ops/s	770.1453 Ops/s	$\color{#d91a1a}-2.02\%$
test_instantiation_td	70.0751ms	1.1044ms	905.4358 Ops/s	989.6496 Ops/s	$\textbf{\color{#d91a1a}-8.51\%}$
test_exec_functorch	0.2249ms	0.1634ms	6.1188 KOps/s	6.1792 KOps/s	$\color{#d91a1a}-0.98\%$
test_exec_functional_call	0.3177ms	0.1503ms	6.6524 KOps/s	6.6214 KOps/s	$\color{#35bf28}+0.47\%$
test_exec_td	0.2556ms	0.1466ms	6.8201 KOps/s	6.8317 KOps/s	$\color{#d91a1a}-0.17\%$
test_exec_td_decorator	0.3500ms	0.2243ms	4.4581 KOps/s	4.5178 KOps/s	$\color{#d91a1a}-1.32\%$
test_vmap_mlp_speed[True-True]	0.9067ms	0.4888ms	2.0460 KOps/s	2.0489 KOps/s	$\color{#d91a1a}-0.14\%$
test_vmap_mlp_speed[True-False]	0.6940ms	0.4849ms	2.0624 KOps/s	2.0542 KOps/s	$\color{#35bf28}+0.40\%$
test_vmap_mlp_speed[False-True]	0.5284ms	0.3922ms	2.5498 KOps/s	2.5195 KOps/s	$\color{#35bf28}+1.20\%$
test_vmap_mlp_speed[False-False]	0.6521ms	0.3938ms	2.5392 KOps/s	2.5306 KOps/s	$\color{#35bf28}+0.34\%$
test_vmap_mlp_speed_decorator[True-True]	0.9865ms	0.5539ms	1.8053 KOps/s	1.7778 KOps/s	$\color{#35bf28}+1.55\%$
test_vmap_mlp_speed_decorator[True-False]	0.8642ms	0.5519ms	1.8120 KOps/s	1.6283 KOps/s	$\textbf{\color{#35bf28}+11.28\%}$
test_vmap_mlp_speed_decorator[False-True]	0.6293ms	0.4519ms	2.2127 KOps/s	2.1309 KOps/s	$\color{#35bf28}+3.83\%$
test_vmap_mlp_speed_decorator[False-False]	0.7777ms	0.4546ms	2.1999 KOps/s	2.0591 KOps/s	$\textbf{\color{#35bf28}+6.84\%}$
test_to_module_speed[True]	2.3973ms	1.7173ms	582.3160 Ops/s	596.3287 Ops/s	$\color{#d91a1a}-2.35\%$
test_to_module_speed[False]	2.3065ms	1.6822ms	594.4620 Ops/s	602.0977 Ops/s	$\color{#d91a1a}-1.27\%$
test_tc_init	60.4840μs	29.6378μs	33.7407 KOps/s	35.6483 KOps/s	$\textbf{\color{#d91a1a}-5.35\%}$
test_tc_init_nested	0.1459ms	61.4368μs	16.2769 KOps/s	18.2784 KOps/s	$\textbf{\color{#d91a1a}-10.95\%}$
test_tc_first_layer_tensor	4.7289μs	0.6984μs	1.4318 MOps/s	1.4685 MOps/s	$\color{#d91a1a}-2.50\%$
test_tc_first_layer_nontensor	1.8289μs	0.6743μs	1.4831 MOps/s	1.5083 MOps/s	$\color{#d91a1a}-1.67\%$
test_tc_second_layer_tensor	21.7410μs	1.8586μs	538.0396 KOps/s	541.7695 KOps/s	$\color{#d91a1a}-0.69\%$
test_tc_second_layer_nontensor	8.9970μs	1.5321μs	652.7104 KOps/s	657.7871 KOps/s	$\color{#d91a1a}-0.77\%$
test_unbind	80.5611ms	6.3890ms	156.5196 Ops/s	140.5268 Ops/s	$\textbf{\color{#35bf28}+11.38\%}$
test_full_like	16.4915ms	10.8209ms	92.4136 Ops/s	97.7183 Ops/s	$\textbf{\color{#d91a1a}-5.43\%}$
test_zeros_like	11.8999ms	5.5794ms	179.2316 Ops/s	171.0368 Ops/s	$\color{#35bf28}+4.79\%$
test_ones_like	14.0965ms	5.9528ms	167.9894 Ops/s	164.8356 Ops/s	$\color{#35bf28}+1.91\%$
test_clone	12.2555ms	7.4496ms	134.2349 Ops/s	129.5448 Ops/s	$\color{#35bf28}+3.62\%$
test_squeeze	68.2780μs	14.9362μs	66.9514 KOps/s	70.9312 KOps/s	$\textbf{\color{#d91a1a}-5.61\%}$
test_unsqueeze	0.1110ms	60.5221μs	16.5229 KOps/s	16.2168 KOps/s	$\color{#35bf28}+1.89\%$
test_split	0.1696ms	0.1120ms	8.9263 KOps/s	8.9043 KOps/s	$\color{#35bf28}+0.25\%$
test_permute	0.2093ms	0.1260ms	7.9338 KOps/s	7.8436 KOps/s	$\color{#35bf28}+1.15\%$
test_stack	24.7824ms	20.8681ms	47.9199 Ops/s	46.7340 Ops/s	$\color{#35bf28}+2.54\%$
test_cat	24.4549ms	20.8878ms	47.8748 Ops/s	46.3966 Ops/s	$\color{#35bf28}+3.19\%$

[github-actions]

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

Total Benchmarks: 152. Improved: $\large\color{#35bf28}6$. Worsened: $\large\color{#d91a1a}25$.

Expand to view detailed results

Name	Max	Mean	Ops	Ops on Repo HEAD	Change
test_plain_set_nested	0.1141ms	13.6671μs	73.1682 KOps/s	78.0721 KOps/s	$\textbf{\color{#d91a1a}-6.28\%}$
test_plain_set_stack_nested	28.9510μs	13.9635μs	71.6152 KOps/s	76.6288 KOps/s	$\textbf{\color{#d91a1a}-6.54\%}$
test_plain_set_nested_inplace	39.1300μs	15.0363μs	66.5056 KOps/s	70.9863 KOps/s	$\textbf{\color{#d91a1a}-6.31\%}$
test_plain_set_stack_nested_inplace	39.7810μs	15.0733μs	66.3425 KOps/s	70.2583 KOps/s	$\textbf{\color{#d91a1a}-5.57\%}$
test_items	23.0100μs	4.6166μs	216.6110 KOps/s	217.5550 KOps/s	$\color{#d91a1a}-0.43\%$
test_items_nested	0.3831ms	0.3360ms	2.9759 KOps/s	2.9795 KOps/s	$\color{#d91a1a}-0.12\%$
test_items_nested_locked	0.3931ms	0.3386ms	2.9534 KOps/s	2.9827 KOps/s	$\color{#d91a1a}-0.98\%$
test_items_nested_leaf	0.1028ms	82.4014μs	12.1357 KOps/s	12.1664 KOps/s	$\color{#d91a1a}-0.25\%$
test_items_stack_nested	0.4040ms	0.3428ms	2.9170 KOps/s	2.9640 KOps/s	$\color{#d91a1a}-1.59\%$
test_items_stack_nested_leaf	0.1104ms	84.0762μs	11.8940 KOps/s	12.1617 KOps/s	$\color{#d91a1a}-2.20\%$
test_items_stack_nested_locked	0.4155ms	0.3434ms	2.9124 KOps/s	2.9696 KOps/s	$\color{#d91a1a}-1.92\%$
test_keys	31.5310μs	4.2978μs	232.6765 KOps/s	230.2486 KOps/s	$\color{#35bf28}+1.05\%$
test_keys_nested	91.2810μs	67.7513μs	14.7599 KOps/s	14.9879 KOps/s	$\color{#d91a1a}-1.52\%$
test_keys_nested_locked	2.0737ms	72.9738μs	13.7036 KOps/s	13.9437 KOps/s	$\color{#d91a1a}-1.72\%$
test_keys_nested_leaf	86.9510μs	57.7867μs	17.3050 KOps/s	17.5431 KOps/s	$\color{#d91a1a}-1.36\%$
test_keys_stack_nested	96.6420μs	67.8984μs	14.7279 KOps/s	14.9547 KOps/s	$\color{#d91a1a}-1.52\%$
test_keys_stack_nested_leaf	95.0210μs	58.4006μs	17.1231 KOps/s	17.4023 KOps/s	$\color{#d91a1a}-1.60\%$
test_keys_stack_nested_locked	0.1030ms	72.3150μs	13.8284 KOps/s	13.9753 KOps/s	$\color{#d91a1a}-1.05\%$
test_values	12.3037μs	1.8059μs	553.7391 KOps/s	549.4052 KOps/s	$\color{#35bf28}+0.79\%$
test_values_nested	65.2100μs	35.1922μs	28.4153 KOps/s	28.2309 KOps/s	$\color{#35bf28}+0.65\%$
test_values_nested_locked	61.5910μs	36.8270μs	27.1540 KOps/s	27.1737 KOps/s	$\color{#d91a1a}-0.07\%$
test_values_nested_leaf	48.3710μs	31.0585μs	32.1973 KOps/s	31.9263 KOps/s	$\color{#35bf28}+0.85\%$
test_values_stack_nested	62.2110μs	35.8252μs	27.9133 KOps/s	27.8208 KOps/s	$\color{#35bf28}+0.33\%$
test_values_stack_nested_leaf	71.9810μs	31.7500μs	31.4961 KOps/s	31.0021 KOps/s	$\color{#35bf28}+1.59\%$
test_values_stack_nested_locked	65.1310μs	37.7592μs	26.4836 KOps/s	26.5730 KOps/s	$\color{#d91a1a}-0.34\%$
test_membership	1.6405μs	0.7002μs	1.4281 MOps/s	1.4289 MOps/s	$\color{#d91a1a}-0.05\%$
test_membership_nested	13.9910μs	2.4741μs	404.1912 KOps/s	403.0279 KOps/s	$\color{#35bf28}+0.29\%$
test_membership_nested_leaf	15.8400μs	2.4679μs	405.2100 KOps/s	401.8616 KOps/s	$\color{#35bf28}+0.83\%$
test_membership_stacked_nested	26.8310μs	2.5236μs	396.2577 KOps/s	397.7349 KOps/s	$\color{#d91a1a}-0.37\%$
test_membership_stacked_nested_leaf	32.8710μs	2.4729μs	404.3821 KOps/s	400.9003 KOps/s	$\color{#35bf28}+0.87\%$
test_membership_nested_last	16.3190μs	2.9954μs	333.8439 KOps/s	333.1456 KOps/s	$\color{#35bf28}+0.21\%$
test_membership_nested_leaf_last	43.1010μs	3.0109μs	332.1263 KOps/s	334.2476 KOps/s	$\color{#d91a1a}-0.63\%$
test_membership_stacked_nested_last	15.9800μs	3.0028μs	333.0225 KOps/s	334.4661 KOps/s	$\color{#d91a1a}-0.43\%$
test_membership_stacked_nested_leaf_last	26.5500μs	3.0054μs	332.7296 KOps/s	334.7012 KOps/s	$\color{#d91a1a}-0.59\%$
test_nested_getleaf	38.9300μs	8.3285μs	120.0695 KOps/s	120.3894 KOps/s	$\color{#d91a1a}-0.27\%$
test_nested_get	31.5500μs	7.7818μs	128.5045 KOps/s	127.2977 KOps/s	$\color{#35bf28}+0.95\%$
test_stacked_getleaf	31.4000μs	8.3107μs	120.3263 KOps/s	119.6821 KOps/s	$\color{#35bf28}+0.54\%$
test_stacked_get	36.0310μs	7.8422μs	127.5148 KOps/s	126.8039 KOps/s	$\color{#35bf28}+0.56\%$
test_nested_getitemleaf	25.0410μs	8.4593μs	118.2132 KOps/s	116.8286 KOps/s	$\color{#35bf28}+1.19\%$
test_nested_getitem	31.7800μs	8.0956μs	123.5244 KOps/s	124.2038 KOps/s	$\color{#d91a1a}-0.55\%$
test_stacked_getitemleaf	34.8710μs	8.7148μs	114.7473 KOps/s	117.2304 KOps/s	$\color{#d91a1a}-2.12\%$
test_stacked_getitem	23.9600μs	8.0673μs	123.9571 KOps/s	124.8569 KOps/s	$\color{#d91a1a}-0.72\%$
test_lock_nested	58.8258ms	0.3960ms	2.5254 KOps/s	2.4835 KOps/s	$\color{#35bf28}+1.69\%$
test_lock_stack_nested	0.3650ms	0.2976ms	3.3599 KOps/s	3.3498 KOps/s	$\color{#35bf28}+0.30\%$
test_unlock_nested	60.9245ms	0.4036ms	2.4778 KOps/s	2.4605 KOps/s	$\color{#35bf28}+0.70\%$
test_unlock_stack_nested	0.3616ms	0.3087ms	3.2389 KOps/s	3.2517 KOps/s	$\color{#d91a1a}-0.39\%$
test_flatten_speed	0.3420ms	0.1008ms	9.9226 KOps/s	9.9772 KOps/s	$\color{#d91a1a}-0.55\%$
test_unflatten_speed	0.3455ms	0.2915ms	3.4300 KOps/s	3.4453 KOps/s	$\color{#d91a1a}-0.44\%$
test_common_ops	1.0860ms	0.6060ms	1.6501 KOps/s	1.7184 KOps/s	$\color{#d91a1a}-3.97\%$
test_creation	36.5210μs	1.6044μs	623.2779 KOps/s	625.1654 KOps/s	$\color{#d91a1a}-0.30\%$
test_creation_empty	24.6610μs	10.3911μs	96.2358 KOps/s	113.1088 KOps/s	$\textbf{\color{#d91a1a}-14.92\%}$
test_creation_nested_1	30.5510μs	12.0518μs	82.9755 KOps/s	93.4381 KOps/s	$\textbf{\color{#d91a1a}-11.20\%}$
test_creation_nested_2	41.8610μs	14.3056μs	69.9029 KOps/s	78.2383 KOps/s	$\textbf{\color{#d91a1a}-10.65\%}$
test_clone	71.6810μs	11.3747μs	87.9146 KOps/s	88.3223 KOps/s	$\color{#d91a1a}-0.46\%$
test_getitem[int]	35.8600μs	10.5382μs	94.8926 KOps/s	91.9311 KOps/s	$\color{#35bf28}+3.22\%$
test_getitem[slice_int]	36.6000μs	20.0597μs	49.8513 KOps/s	48.7748 KOps/s	$\color{#35bf28}+2.21\%$
test_getitem[range]	64.7200μs	45.7488μs	21.8585 KOps/s	21.7729 KOps/s	$\color{#35bf28}+0.39\%$
test_getitem[tuple]	42.1000μs	18.2530μs	54.7856 KOps/s	54.9595 KOps/s	$\color{#d91a1a}-0.32\%$
test_getitem[list]	0.1592ms	33.8962μs	29.5019 KOps/s	30.6659 KOps/s	$\color{#d91a1a}-3.80\%$
test_setitem_dim[int]	49.5610μs	32.4252μs	30.8402 KOps/s	35.7706 KOps/s	$\textbf{\color{#d91a1a}-13.78\%}$
test_setitem_dim[slice_int]	73.9610μs	52.9934μs	18.8703 KOps/s	20.4952 KOps/s	$\textbf{\color{#d91a1a}-7.93\%}$
test_setitem_dim[range]	96.7500μs	71.6504μs	13.9566 KOps/s	15.4394 KOps/s	$\textbf{\color{#d91a1a}-9.60\%}$
test_setitem_dim[tuple]	67.0110μs	45.6037μs	21.9280 KOps/s	24.3818 KOps/s	$\textbf{\color{#d91a1a}-10.06\%}$
test_setitem	57.4510μs	16.8511μs	59.3432 KOps/s	61.3303 KOps/s	$\color{#d91a1a}-3.24\%$
test_set	47.7010μs	16.3718μs	61.0808 KOps/s	64.7284 KOps/s	$\textbf{\color{#d91a1a}-5.64\%}$
test_set_shared	1.3730ms	99.0246μs	10.0985 KOps/s	10.2984 KOps/s	$\color{#d91a1a}-1.94\%$
test_update	95.5210μs	19.9502μs	50.1247 KOps/s	56.3009 KOps/s	$\textbf{\color{#d91a1a}-10.97\%}$
test_update_nested	87.0720μs	24.5887μs	40.6690 KOps/s	43.7156 KOps/s	$\textbf{\color{#d91a1a}-6.97\%}$
test_update__nested	52.3510μs	21.5634μs	46.3748 KOps/s	45.9103 KOps/s	$\color{#35bf28}+1.01\%$
test_set_nested	61.5710μs	17.1380μs	58.3499 KOps/s	59.6343 KOps/s	$\color{#d91a1a}-2.15\%$
test_set_nested_new	95.5720μs	20.0822μs	49.7952 KOps/s	52.0651 KOps/s	$\color{#d91a1a}-4.36\%$
test_select	67.1610μs	32.4807μs	30.7875 KOps/s	31.1908 KOps/s	$\color{#d91a1a}-1.29\%$
test_select_nested	89.3320μs	54.9990μs	18.1822 KOps/s	18.2351 KOps/s	$\color{#d91a1a}-0.29\%$
test_exclude_nested	0.1563ms	0.1104ms	9.0576 KOps/s	9.0741 KOps/s	$\color{#d91a1a}-0.18\%$
test_empty[True]	0.4027ms	0.3397ms	2.9436 KOps/s	2.9382 KOps/s	$\color{#35bf28}+0.18\%$
test_empty[False]	2.8911μs	0.9237μs	1.0826 MOps/s	1.0901 MOps/s	$\color{#d91a1a}-0.69\%$
test_to	97.2020μs	71.1886μs	14.0472 KOps/s	12.9805 KOps/s	$\textbf{\color{#35bf28}+8.22\%}$
test_to_nonblocking	0.1143ms	64.4849μs	15.5075 KOps/s	16.5345 KOps/s	$\textbf{\color{#d91a1a}-6.21\%}$
test_unbind_speed	0.2930ms	0.2598ms	3.8491 KOps/s	3.8159 KOps/s	$\color{#35bf28}+0.87\%$
test_unbind_speed_stack0	0.3705ms	0.2617ms	3.8210 KOps/s	3.8346 KOps/s	$\color{#d91a1a}-0.35\%$
test_unbind_speed_stack1	75.8264ms	0.7970ms	1.2547 KOps/s	1.2563 KOps/s	$\color{#d91a1a}-0.13\%$
test_split	76.2312ms	1.6960ms	589.6111 Ops/s	590.5487 Ops/s	$\color{#d91a1a}-0.16\%$
test_chunk	76.1631ms	1.7091ms	585.1034 Ops/s	591.2135 Ops/s	$\color{#d91a1a}-1.03\%$
test_creation[device0]	0.1319ms	56.3764μs	17.7379 KOps/s	17.8680 KOps/s	$\color{#d91a1a}-0.73\%$
test_creation_from_tensor	0.1340ms	52.8389μs	18.9255 KOps/s	18.8382 KOps/s	$\color{#35bf28}+0.46\%$
test_add_one[memmap_tensor0]	79.1610μs	6.8268μs	146.4806 KOps/s	149.4470 KOps/s	$\color{#d91a1a}-1.98\%$
test_contiguous[memmap_tensor0]	10.2010μs	0.6282μs	1.5919 MOps/s	1.5684 MOps/s	$\color{#35bf28}+1.50\%$
test_stack[memmap_tensor0]	28.8700μs	4.7401μs	210.9675 KOps/s	215.7849 KOps/s	$\color{#d91a1a}-2.23\%$
test_memmaptd_index	1.0617ms	0.2816ms	3.5512 KOps/s	3.5211 KOps/s	$\color{#35bf28}+0.85\%$
test_memmaptd_index_astensor	0.6121ms	0.3506ms	2.8519 KOps/s	2.8241 KOps/s	$\color{#35bf28}+0.98\%$
test_memmaptd_index_op	0.9681ms	0.6836ms	1.4629 KOps/s	1.5446 KOps/s	$\textbf{\color{#d91a1a}-5.29\%}$
test_serialize_model	0.1830s	0.1109s	9.0183 Ops/s	8.5146 Ops/s	$\textbf{\color{#35bf28}+5.92\%}$
test_serialize_model_pickle	1.3579s	1.2357s	0.8093 Ops/s	0.8085 Ops/s	$\color{#35bf28}+0.10\%$
test_serialize_weights	0.1806s	0.1080s	9.2550 Ops/s	8.7874 Ops/s	$\textbf{\color{#35bf28}+5.32\%}$
test_serialize_weights_returnearly	0.3002s	0.1043s	9.5919 Ops/s	10.3566 Ops/s	$\textbf{\color{#d91a1a}-7.38\%}$
test_serialize_weights_pickle	1.3539s	1.2485s	0.8009 Ops/s	0.8087 Ops/s	$\color{#d91a1a}-0.96\%$
test_reshape_pytree	48.8610μs	25.6972μs	38.9148 KOps/s	38.3664 KOps/s	$\color{#35bf28}+1.43\%$
test_reshape_td	74.6610μs	30.1510μs	33.1664 KOps/s	32.2733 KOps/s	$\color{#35bf28}+2.77\%$
test_view_pytree	90.6910μs	25.8604μs	38.6692 KOps/s	39.2348 KOps/s	$\color{#d91a1a}-1.44\%$
test_view_td	61.2610μs	35.5705μs	28.1132 KOps/s	28.2579 KOps/s	$\color{#d91a1a}-0.51\%$
test_unbind_pytree	54.3710μs	31.4364μs	31.8102 KOps/s	31.9740 KOps/s	$\color{#d91a1a}-0.51\%$
test_unbind_td	0.4767ms	40.3232μs	24.7996 KOps/s	24.1878 KOps/s	$\color{#35bf28}+2.53\%$
test_split_pytree	59.3710μs	34.3811μs	29.0857 KOps/s	28.1578 KOps/s	$\color{#35bf28}+3.30\%$
test_split_td	0.1061ms	38.7872μs	25.7817 KOps/s	25.5367 KOps/s	$\color{#35bf28}+0.96\%$
test_add_pytree	74.3510μs	37.9485μs	26.3515 KOps/s	26.4985 KOps/s	$\color{#d91a1a}-0.55\%$
test_add_td	85.1420μs	57.3453μs	17.4382 KOps/s	19.7347 KOps/s	$\textbf{\color{#d91a1a}-11.64\%}$
test_distributed	2.8447ms	88.4177μs	11.3100 KOps/s	14.9865 KOps/s	$\textbf{\color{#d91a1a}-24.53\%}$
test_tdmodule	30.9410μs	15.5608μs	64.2640 KOps/s	67.5866 KOps/s	$\color{#d91a1a}-4.92\%$
test_tdmodule_dispatch	52.9200μs	30.8563μs	32.4083 KOps/s	34.7692 KOps/s	$\textbf{\color{#d91a1a}-6.79\%}$
test_tdseq	33.4100μs	17.4805μs	57.2067 KOps/s	60.3959 KOps/s	$\textbf{\color{#d91a1a}-5.28\%}$
test_tdseq_dispatch	51.5920μs	34.3819μs	29.0851 KOps/s	31.1933 KOps/s	$\textbf{\color{#d91a1a}-6.76\%}$
test_instantiation_functorch	1.6502ms	1.5217ms	657.1749 Ops/s	657.5885 Ops/s	$\color{#d91a1a}-0.06\%$
test_instantiation_td	1.5694ms	1.0524ms	950.2055 Ops/s	961.1444 Ops/s	$\color{#d91a1a}-1.14\%$
test_exec_functorch	0.2351ms	0.1488ms	6.7196 KOps/s	6.6723 KOps/s	$\color{#35bf28}+0.71\%$
test_exec_functional_call	0.1832ms	0.1349ms	7.4104 KOps/s	7.3576 KOps/s	$\color{#35bf28}+0.72\%$
test_exec_td	0.1677ms	0.1348ms	7.4169 KOps/s	7.4480 KOps/s	$\color{#d91a1a}-0.42\%$
test_exec_td_decorator	0.7006ms	0.2072ms	4.8258 KOps/s	4.8106 KOps/s	$\color{#35bf28}+0.32\%$
test_vmap_mlp_speed[True-True]	0.7542ms	0.5755ms	1.7376 KOps/s	1.7722 KOps/s	$\color{#d91a1a}-1.95\%$
test_vmap_mlp_speed[True-False]	0.7149ms	0.5770ms	1.7331 KOps/s	1.7481 KOps/s	$\color{#d91a1a}-0.86\%$
test_vmap_mlp_speed[False-True]	0.5952ms	0.5217ms	1.9167 KOps/s	2.0136 KOps/s	$\color{#d91a1a}-4.81\%$
test_vmap_mlp_speed[False-False]	0.6070ms	0.5186ms	1.9283 KOps/s	2.0294 KOps/s	$\color{#d91a1a}-4.99\%$
test_vmap_mlp_speed_decorator[True-True]	1.3773ms	0.6363ms	1.5716 KOps/s	1.5967 KOps/s	$\color{#d91a1a}-1.57\%$
test_vmap_mlp_speed_decorator[True-False]	0.7556ms	0.6324ms	1.5813 KOps/s	1.5985 KOps/s	$\color{#d91a1a}-1.08\%$
test_vmap_mlp_speed_decorator[False-True]	0.7134ms	0.5618ms	1.7801 KOps/s	1.8157 KOps/s	$\color{#d91a1a}-1.96\%$
test_vmap_mlp_speed_decorator[False-False]	0.7254ms	0.5674ms	1.7625 KOps/s	1.8021 KOps/s	$\color{#d91a1a}-2.20\%$
test_vmap_transformer_speed[True-True]	7.6819ms	7.3985ms	135.1632 Ops/s	134.2127 Ops/s	$\color{#35bf28}+0.71\%$
test_vmap_transformer_speed[True-False]	7.6557ms	7.3623ms	135.8265 Ops/s	129.0384 Ops/s	$\textbf{\color{#35bf28}+5.26\%}$
test_vmap_transformer_speed[False-True]	7.8032ms	7.4004ms	135.1281 Ops/s	133.1076 Ops/s	$\color{#35bf28}+1.52\%$
test_vmap_transformer_speed[False-False]	7.7914ms	7.4244ms	134.6906 Ops/s	133.4651 Ops/s	$\color{#35bf28}+0.92\%$
test_vmap_transformer_speed_decorator[True-True]	18.7091ms	18.1782ms	55.0111 Ops/s	54.3071 Ops/s	$\color{#35bf28}+1.30\%$
test_vmap_transformer_speed_decorator[True-False]	18.4415ms	18.0661ms	55.3522 Ops/s	53.9662 Ops/s	$\color{#35bf28}+2.57\%$
test_vmap_transformer_speed_decorator[False-True]	18.5552ms	18.0265ms	55.4739 Ops/s	54.5605 Ops/s	$\color{#35bf28}+1.67\%$
test_vmap_transformer_speed_decorator[False-False]	18.8151ms	18.0841ms	55.2972 Ops/s	54.8361 Ops/s	$\color{#35bf28}+0.84\%$
test_to_module_speed[True]	1.6920ms	1.5440ms	647.6740 Ops/s	630.1736 Ops/s	$\color{#35bf28}+2.78\%$
test_to_module_speed[False]	2.0284ms	1.5262ms	655.2077 Ops/s	647.1554 Ops/s	$\color{#35bf28}+1.24\%$
test_tc_init	50.9910μs	29.1908μs	34.2574 KOps/s	38.1760 KOps/s	$\textbf{\color{#d91a1a}-10.26\%}$
test_tc_init_nested	88.6620μs	57.4431μs	17.4085 KOps/s	18.1597 KOps/s	$\color{#d91a1a}-4.14\%$
test_tc_first_layer_tensor	3.2050μs	0.3965μs	2.5218 MOps/s	2.8059 MOps/s	$\textbf{\color{#d91a1a}-10.12\%}$
test_tc_first_layer_nontensor	1.4724μs	0.3897μs	2.5662 MOps/s	2.5836 MOps/s	$\color{#d91a1a}-0.67\%$
test_tc_second_layer_tensor	3.8680μs	0.9687μs	1.0324 MOps/s	936.8412 KOps/s	$\textbf{\color{#35bf28}+10.20\%}$
test_tc_second_layer_nontensor	2.6539μs	0.8064μs	1.2400 MOps/s	1.2162 MOps/s	$\color{#35bf28}+1.95\%$
test_unbind	0.1102s	6.6968ms	149.3243 Ops/s	204.3132 Ops/s	$\textbf{\color{#d91a1a}-26.91\%}$
test_full_like	13.7695ms	13.1701ms	75.9298 Ops/s	75.3013 Ops/s	$\color{#35bf28}+0.83\%$
test_zeros_like	7.9992ms	7.8265ms	127.7714 Ops/s	128.3861 Ops/s	$\color{#d91a1a}-0.48\%$
test_ones_like	8.0176ms	7.8113ms	128.0200 Ops/s	126.9401 Ops/s	$\color{#35bf28}+0.85\%$
test_clone	9.5457ms	9.3962ms	106.4257 Ops/s	107.3349 Ops/s	$\color{#d91a1a}-0.85\%$
test_squeeze	71.5220μs	10.5980μs	94.3570 KOps/s	94.1865 KOps/s	$\color{#35bf28}+0.18\%$
test_unsqueeze	0.1196ms	49.8651μs	20.0541 KOps/s	20.0138 KOps/s	$\color{#35bf28}+0.20\%$
test_split	0.1478ms	98.2883μs	10.1742 KOps/s	9.9351 KOps/s	$\color{#35bf28}+2.41\%$
test_permute	0.1622ms	0.1096ms	9.1279 KOps/s	8.6173 KOps/s	$\textbf{\color{#35bf28}+5.93\%}$
test_stack	27.6940ms	27.3483ms	36.5653 Ops/s	36.7886 Ops/s	$\color{#d91a1a}-0.61\%$
test_cat	27.3653ms	27.1452ms	36.8389 Ops/s	36.8376 Ops/s	$+0.00\%$