[BugFix, Performance] Fewer imports at root

[vmoens]

cc @teopir

[github-actions]

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

Total Benchmarks: 126. Improved: $\large\color{#35bf28}7$. Worsened: $\large\color{#d91a1a}30$.

Expand to view detailed results

Name	Max	Mean	Ops	Ops on Repo HEAD	Change
test_plain_set_nested	37.2600μs	17.1645μs	58.2598 KOps/s	56.0834 KOps/s	$\color{#35bf28}+3.88\%$
test_plain_set_stack_nested	0.2833ms	0.1444ms	6.9245 KOps/s	55.7354 KOps/s	$\textbf{\color{#d91a1a}-87.58\%}$
test_plain_set_nested_inplace	46.5270μs	19.7451μs	50.6455 KOps/s	49.8567 KOps/s	$\color{#35bf28}+1.58\%$
test_plain_set_stack_nested_inplace	0.3161ms	0.1773ms	5.6403 KOps/s	49.5280 KOps/s	$\textbf{\color{#d91a1a}-88.61\%}$
test_items	16.2010μs	2.4126μs	414.4907 KOps/s	400.9563 KOps/s	$\color{#35bf28}+3.38\%$
test_items_nested	0.4615ms	0.2682ms	3.7288 KOps/s	3.6393 KOps/s	$\color{#35bf28}+2.46\%$
test_items_nested_locked	0.4157ms	0.2705ms	3.6967 KOps/s	3.7097 KOps/s	$\color{#d91a1a}-0.35\%$
test_items_nested_leaf	0.5102ms	0.1667ms	5.9988 KOps/s	6.0648 KOps/s	$\color{#d91a1a}-1.09\%$
test_items_stack_nested	2.9080ms	1.3740ms	727.8053 Ops/s	3.6584 KOps/s	$\textbf{\color{#d91a1a}-80.11\%}$
test_items_stack_nested_leaf	2.3515ms	1.1912ms	839.4659 Ops/s	6.0623 KOps/s	$\textbf{\color{#d91a1a}-86.15\%}$
test_items_stack_nested_locked	1.8360ms	0.8687ms	1.1511 KOps/s	3.6650 KOps/s	$\textbf{\color{#d91a1a}-68.59\%}$
test_keys	17.2920μs	3.8712μs	258.3148 KOps/s	260.5960 KOps/s	$\color{#d91a1a}-0.88\%$
test_keys_nested	0.5215ms	0.1467ms	6.8144 KOps/s	6.7342 KOps/s	$\color{#35bf28}+1.19\%$
test_keys_nested_locked	0.2484ms	0.1508ms	6.6315 KOps/s	6.5605 KOps/s	$\color{#35bf28}+1.08\%$
test_keys_nested_leaf	37.8776ms	0.1346ms	7.4291 KOps/s	7.6920 KOps/s	$\color{#d91a1a}-3.42\%$
test_keys_stack_nested	2.4397ms	1.2855ms	777.8943 Ops/s	6.5800 KOps/s	$\textbf{\color{#d91a1a}-88.18\%}$
test_keys_stack_nested_leaf	2.0499ms	1.2925ms	773.6872 Ops/s	7.5884 KOps/s	$\textbf{\color{#d91a1a}-89.80\%}$
test_keys_stack_nested_locked	0.9204ms	0.8077ms	1.2381 KOps/s	6.3888 KOps/s	$\textbf{\color{#d91a1a}-80.62\%}$
test_values	4.9365μs	1.1304μs	884.6652 KOps/s	859.6670 KOps/s	$\color{#35bf28}+2.91\%$
test_values_nested	0.1012ms	51.2759μs	19.5024 KOps/s	19.3285 KOps/s	$\color{#35bf28}+0.90\%$
test_values_nested_locked	0.2342ms	51.6226μs	19.3714 KOps/s	19.2511 KOps/s	$\color{#35bf28}+0.62\%$
test_values_nested_leaf	82.6950μs	45.5526μs	21.9526 KOps/s	21.6470 KOps/s	$\color{#35bf28}+1.41\%$
test_values_stack_nested	1.6547ms	1.0354ms	965.8531 Ops/s	19.0726 KOps/s	$\textbf{\color{#d91a1a}-94.94\%}$
test_values_stack_nested_leaf	1.2620ms	1.0251ms	975.5385 Ops/s	21.7910 KOps/s	$\textbf{\color{#d91a1a}-95.52\%}$
test_values_stack_nested_locked	1.1238ms	0.6035ms	1.6570 KOps/s	19.3248 KOps/s	$\textbf{\color{#d91a1a}-91.43\%}$
test_membership	9.6480μs	1.3721μs	728.7922 KOps/s	730.5592 KOps/s	$\color{#d91a1a}-0.24\%$
test_membership_nested	21.4410μs	3.4843μs	287.0029 KOps/s	284.6026 KOps/s	$\color{#35bf28}+0.84\%$
test_membership_nested_leaf	21.7110μs	3.5193μs	284.1471 KOps/s	292.1272 KOps/s	$\color{#d91a1a}-2.73\%$
test_membership_stacked_nested	38.1620μs	11.5301μs	86.7292 KOps/s	292.6972 KOps/s	$\textbf{\color{#d91a1a}-70.37\%}$
test_membership_stacked_nested_leaf	38.3720μs	11.5929μs	86.2599 KOps/s	286.4749 KOps/s	$\textbf{\color{#d91a1a}-69.89\%}$
test_membership_nested_last	32.0200μs	6.6212μs	151.0295 KOps/s	149.7604 KOps/s	$\color{#35bf28}+0.85\%$
test_membership_nested_leaf_last	26.3790μs	6.6489μs	150.4012 KOps/s	149.3949 KOps/s	$\color{#35bf28}+0.67\%$
test_membership_stacked_nested_last	0.3214ms	0.1754ms	5.7026 KOps/s	62.9593 KOps/s	$\textbf{\color{#d91a1a}-90.94\%}$
test_membership_stacked_nested_leaf_last	65.4460μs	13.6715μs	73.1448 KOps/s	62.0608 KOps/s	$\textbf{\color{#35bf28}+17.86\%}$
test_nested_getleaf	41.2470μs	10.8812μs	91.9018 KOps/s	92.1120 KOps/s	$\color{#d91a1a}-0.23\%$
test_nested_get	43.4210μs	10.3476μs	96.6408 KOps/s	96.3422 KOps/s	$\color{#35bf28}+0.31\%$
test_stacked_getleaf	0.6262ms	0.4023ms	2.4859 KOps/s	93.9845 KOps/s	$\textbf{\color{#d91a1a}-97.36\%}$
test_stacked_get	0.5864ms	0.3715ms	2.6918 KOps/s	97.2587 KOps/s	$\textbf{\color{#d91a1a}-97.23\%}$
test_nested_getitemleaf	46.0760μs	12.2126μs	81.8826 KOps/s	80.0179 KOps/s	$\color{#35bf28}+2.33\%$
test_nested_getitem	45.7360μs	11.6681μs	85.7036 KOps/s	84.3607 KOps/s	$\color{#35bf28}+1.59\%$
test_stacked_getitemleaf	0.5467ms	0.4043ms	2.4735 KOps/s	81.2763 KOps/s	$\textbf{\color{#d91a1a}-96.96\%}$
test_stacked_getitem	0.6604ms	0.3738ms	2.6752 KOps/s	85.7487 KOps/s	$\textbf{\color{#d91a1a}-96.88\%}$
test_lock_nested	0.7426ms	0.3349ms	2.9861 KOps/s	2.4796 KOps/s	$\textbf{\color{#35bf28}+20.43\%}$
test_lock_stack_nested	65.4888ms	5.2028ms	192.2060 Ops/s	3.4723 KOps/s	$\textbf{\color{#d91a1a}-94.46\%}$
test_unlock_nested	49.9916ms	0.3868ms	2.5852 KOps/s	2.9766 KOps/s	$\textbf{\color{#d91a1a}-13.15\%}$
test_unlock_stack_nested	66.3088ms	5.4465ms	183.6046 Ops/s	3.3778 KOps/s	$\textbf{\color{#d91a1a}-94.56\%}$
test_flatten_speed	60.8896ms	0.3862ms	2.5894 KOps/s	2.6815 KOps/s	$\color{#d91a1a}-3.43\%$
test_unflatten_speed	0.6280ms	0.4661ms	2.1453 KOps/s	2.1542 KOps/s	$\color{#d91a1a}-0.41\%$
test_common_ops	6.0381ms	0.7069ms	1.4145 KOps/s	1.3996 KOps/s	$\color{#35bf28}+1.07\%$
test_creation	48.1200μs	1.8509μs	540.2662 KOps/s	544.4402 KOps/s	$\color{#d91a1a}-0.77\%$
test_creation_empty	28.2430μs	10.9707μs	91.1515 KOps/s	87.2378 KOps/s	$\color{#35bf28}+4.49\%$
test_creation_nested_1	33.3730μs	13.7002μs	72.9918 KOps/s	71.0783 KOps/s	$\color{#35bf28}+2.69\%$
test_creation_nested_2	44.8040μs	16.7643μs	59.6506 KOps/s	57.4865 KOps/s	$\color{#35bf28}+3.76\%$
test_clone	0.1214ms	13.5691μs	73.6968 KOps/s	77.3699 KOps/s	$\color{#d91a1a}-4.75\%$
test_getitem[int]	29.5950μs	11.0508μs	90.4911 KOps/s	88.5524 KOps/s	$\color{#35bf28}+2.19\%$
test_getitem[slice_int]	57.6980μs	22.3601μs	44.7226 KOps/s	44.3702 KOps/s	$\color{#35bf28}+0.79\%$
test_getitem[range]	85.0080μs	40.8823μs	24.4605 KOps/s	23.4104 KOps/s	$\color{#35bf28}+4.49\%$
test_getitem[tuple]	49.7330μs	18.3364μs	54.5363 KOps/s	54.3979 KOps/s	$\color{#35bf28}+0.25\%$
test_getitem[list]	0.2301ms	36.2807μs	27.5628 KOps/s	26.3589 KOps/s	$\color{#35bf28}+4.57\%$
test_setitem_dim[int]	52.9590μs	30.9581μs	32.3017 KOps/s	32.5500 KOps/s	$\color{#d91a1a}-0.76\%$
test_setitem_dim[slice_int]	93.2850μs	56.2724μs	17.7707 KOps/s	17.8757 KOps/s	$\color{#d91a1a}-0.59\%$
test_setitem_dim[range]	0.1513ms	76.4379μs	13.0825 KOps/s	13.3145 KOps/s	$\color{#d91a1a}-1.74\%$
test_setitem_dim[tuple]	0.1033ms	46.1735μs	21.6574 KOps/s	22.2856 KOps/s	$\color{#d91a1a}-2.82\%$
test_setitem	0.1011ms	20.4171μs	48.9785 KOps/s	49.6486 KOps/s	$\color{#d91a1a}-1.35\%$
test_set	0.1038ms	19.7940μs	50.5202 KOps/s	50.3738 KOps/s	$\color{#35bf28}+0.29\%$
test_set_shared	4.0685ms	0.1398ms	7.1539 KOps/s	7.1859 KOps/s	$\color{#d91a1a}-0.45\%$
test_update	0.1170ms	23.0167μs	43.4466 KOps/s	43.2105 KOps/s	$\color{#35bf28}+0.55\%$
test_update_nested	0.1182ms	30.1338μs	33.1854 KOps/s	32.7050 KOps/s	$\color{#35bf28}+1.47\%$
test_set_nested	0.2050ms	21.7898μs	45.8930 KOps/s	46.7163 KOps/s	$\color{#d91a1a}-1.76\%$
test_set_nested_new	0.1475ms	25.6260μs	39.0229 KOps/s	39.4173 KOps/s	$\color{#d91a1a}-1.00\%$
test_select	0.1231ms	39.2524μs	25.4761 KOps/s	25.5374 KOps/s	$\color{#d91a1a}-0.24\%$
test_select_nested	0.1080ms	58.2067μs	17.1801 KOps/s	17.0042 KOps/s	$\color{#35bf28}+1.03\%$
test_exclude_nested	0.2249ms	0.1174ms	8.5197 KOps/s	8.4360 KOps/s	$\color{#35bf28}+0.99\%$
test_empty[True]	0.5789ms	0.4020ms	2.4876 KOps/s	2.4506 KOps/s	$\color{#35bf28}+1.51\%$
test_empty[False]	7.9168μs	1.0342μs	966.8847 KOps/s	940.5425 KOps/s	$\color{#35bf28}+2.80\%$
test_unbind_speed	0.3009ms	0.2478ms	4.0349 KOps/s	4.1269 KOps/s	$\color{#d91a1a}-2.23\%$
test_unbind_speed_stack0	57.4259ms	3.3123ms	301.9022 Ops/s	4.3187 KOps/s	$\textbf{\color{#d91a1a}-93.01\%}$
test_unbind_speed_stack1	24.2050μs	1.9678μs	508.1703 KOps/s	1.4883 KOps/s	$\textbf{\color{#35bf28}+34043.84\%}$
test_split	2.1844ms	1.4416ms	693.6564 Ops/s	681.0033 Ops/s	$\color{#35bf28}+1.86\%$
test_chunk	53.8952ms	1.6078ms	621.9677 Ops/s	678.2164 Ops/s	$\textbf{\color{#d91a1a}-8.29\%}$
test_creation[device0]	0.1787ms	0.1024ms	9.7632 KOps/s	9.9032 KOps/s	$\color{#d91a1a}-1.41\%$
test_creation_from_tensor	3.5432ms	83.7657μs	11.9381 KOps/s	12.2820 KOps/s	$\color{#d91a1a}-2.80\%$
test_add_one[memmap_tensor0]	0.2943ms	5.4757μs	182.6264 KOps/s	188.7538 KOps/s	$\color{#d91a1a}-3.25\%$
test_contiguous[memmap_tensor0]	9.0360μs	0.6290μs	1.5899 MOps/s	1.5390 MOps/s	$\color{#35bf28}+3.31\%$
test_stack[memmap_tensor0]	59.6520μs	3.6804μs	271.7109 KOps/s	278.8767 KOps/s	$\color{#d91a1a}-2.57\%$
test_memmaptd_index	0.9212ms	0.2360ms	4.2379 KOps/s	4.2195 KOps/s	$\color{#35bf28}+0.44\%$
test_memmaptd_index_astensor	0.5250ms	0.2986ms	3.3486 KOps/s	3.3227 KOps/s	$\color{#35bf28}+0.78\%$
test_memmaptd_index_op	1.2496ms	0.6146ms	1.6272 KOps/s	1.6403 KOps/s	$\color{#d91a1a}-0.80\%$
test_serialize_model	0.1606s	0.1109s	9.0134 Ops/s	9.5440 Ops/s	$\textbf{\color{#d91a1a}-5.56\%}$
test_serialize_model_pickle	0.4477s	0.3795s	2.6350 Ops/s	2.0607 Ops/s	$\textbf{\color{#35bf28}+27.87\%}$
test_serialize_weights	0.1525s	0.1060s	9.4322 Ops/s	10.0466 Ops/s	$\textbf{\color{#d91a1a}-6.12\%}$
test_serialize_weights_returnearly	0.1730s	0.1278s	7.8222 Ops/s	6.7299 Ops/s	$\textbf{\color{#35bf28}+16.23\%}$
test_serialize_weights_pickle	1.1521s	0.5959s	1.6781 Ops/s	2.4134 Ops/s	$\textbf{\color{#d91a1a}-30.47\%}$
test_serialize_weights_filesystem	0.1540s	97.2459ms	10.2832 Ops/s	10.6835 Ops/s	$\color{#d91a1a}-3.75\%$
test_serialize_model_filesystem	97.1307ms	92.9169ms	10.7623 Ops/s	10.2491 Ops/s	$\textbf{\color{#35bf28}+5.01\%}$
test_reshape_pytree	48.9310μs	20.6341μs	48.4634 KOps/s	47.5101 KOps/s	$\color{#35bf28}+2.01\%$
test_reshape_td	94.7270μs	31.4950μs	31.7510 KOps/s	32.0923 KOps/s	$\color{#d91a1a}-1.06\%$
test_view_pytree	48.3600μs	20.6871μs	48.3392 KOps/s	47.7382 KOps/s	$\color{#35bf28}+1.26\%$
test_view_td	53.8560ms	10.4375μs	95.8087 KOps/s	15.2758 KOps/s	$\textbf{\color{#35bf28}+527.19\%}$
test_unbind_pytree	59.5020μs	23.8244μs	41.9738 KOps/s	41.6142 KOps/s	$\color{#35bf28}+0.86\%$
test_unbind_td	0.1121ms	35.5266μs	28.1479 KOps/s	28.2458 KOps/s	$\color{#d91a1a}-0.35\%$
test_split_pytree	54.3410μs	23.4214μs	42.6960 KOps/s	42.2287 KOps/s	$\color{#35bf28}+1.11\%$
test_split_td	1.0418ms	38.8435μs	25.7443 KOps/s	25.2804 KOps/s	$\color{#35bf28}+1.84\%$
test_add_pytree	67.7770μs	28.7999μs	34.7223 KOps/s	33.5719 KOps/s	$\color{#35bf28}+3.43\%$
test_add_td	0.1004ms	53.6694μs	18.6326 KOps/s	18.2311 KOps/s	$\color{#35bf28}+2.20\%$
test_distributed	0.1754ms	98.8539μs	10.1159 KOps/s	9.8774 KOps/s	$\color{#35bf28}+2.41\%$
test_tdmodule	0.2685ms	22.8553μs	43.7536 KOps/s	43.0832 KOps/s	$\color{#35bf28}+1.56\%$
test_tdmodule_dispatch	0.1477ms	44.1786μs	22.6354 KOps/s	21.9526 KOps/s	$\color{#35bf28}+3.11\%$
test_tdseq	0.1142ms	25.4935μs	39.2257 KOps/s	38.0109 KOps/s	$\color{#35bf28}+3.20\%$
test_tdseq_dispatch	0.3862ms	48.0756μs	20.8006 KOps/s	20.4326 KOps/s	$\color{#35bf28}+1.80\%$
test_instantiation_functorch	2.8284ms	1.3182ms	758.6058 Ops/s	771.7187 Ops/s	$\color{#d91a1a}-1.70\%$
test_instantiation_td	1.4747ms	0.9989ms	1.0011 KOps/s	995.7992 Ops/s	$\color{#35bf28}+0.53\%$
test_exec_functorch	0.5222ms	0.1594ms	6.2747 KOps/s	6.4054 KOps/s	$\color{#d91a1a}-2.04\%$
test_exec_functional_call	0.3823ms	0.1449ms	6.8991 KOps/s	7.0142 KOps/s	$\color{#d91a1a}-1.64\%$
test_exec_td	0.2869ms	0.1422ms	7.0330 KOps/s	7.1251 KOps/s	$\color{#d91a1a}-1.29\%$
test_exec_td_decorator	0.2543ms	0.1706ms	5.8610 KOps/s	5.9109 KOps/s	$\color{#d91a1a}-0.84\%$
test_vmap_mlp_speed[True-True]	1.2956ms	0.9092ms	1.0999 KOps/s	2.1428 KOps/s	$\textbf{\color{#d91a1a}-48.67\%}$
test_vmap_mlp_speed[True-False]	0.9144ms	0.4844ms	2.0644 KOps/s	2.1527 KOps/s	$\color{#d91a1a}-4.10\%$
test_vmap_mlp_speed[False-True]	1.1382ms	0.7940ms	1.2594 KOps/s	2.6485 KOps/s	$\textbf{\color{#d91a1a}-52.45\%}$
test_vmap_mlp_speed[False-False]	0.6903ms	0.3969ms	2.5196 KOps/s	2.6475 KOps/s	$\color{#d91a1a}-4.83\%$
test_vmap_mlp_speed_decorator[True-True]	2.1535ms	1.5603ms	640.8941 Ops/s	1.9534 KOps/s	$\textbf{\color{#d91a1a}-67.19\%}$
test_vmap_mlp_speed_decorator[True-False]	0.6402ms	0.5207ms	1.9207 KOps/s	1.9834 KOps/s	$\color{#d91a1a}-3.16\%$
test_vmap_mlp_speed_decorator[False-True]	2.1038ms	1.3166ms	759.5201 Ops/s	2.5856 KOps/s	$\textbf{\color{#d91a1a}-70.62\%}$
test_vmap_mlp_speed_decorator[False-False]	0.5703ms	0.4010ms	2.4938 KOps/s	2.6015 KOps/s	$\color{#d91a1a}-4.14\%$
test_to_module_speed[True]	1.7490ms	1.0990ms	909.9132 Ops/s	872.6712 Ops/s	$\color{#35bf28}+4.27\%$
test_to_module_speed[False]	1.1598ms	1.0789ms	926.8314 Ops/s	924.2349 Ops/s	$\color{#35bf28}+0.28\%$

[github-actions]

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

Total Benchmarks: 134. Improved: $\large\color{#35bf28}9$. Worsened: $\large\color{#d91a1a}40$.

Expand to view detailed results

Name	Max	Mean	Ops	Ops on Repo HEAD	Change
test_plain_set_nested	0.8577ms	14.1874μs	70.4849 KOps/s	68.9072 KOps/s	$\color{#35bf28}+2.29\%$
test_plain_set_stack_nested	0.1914ms	0.1209ms	8.2735 KOps/s	68.6553 KOps/s	$\textbf{\color{#d91a1a}-87.95\%}$
test_plain_set_nested_inplace	46.9710μs	15.5187μs	64.4385 KOps/s	63.4335 KOps/s	$\color{#35bf28}+1.58\%$
test_plain_set_stack_nested_inplace	0.2386ms	0.1509ms	6.6256 KOps/s	63.0898 KOps/s	$\textbf{\color{#d91a1a}-89.50\%}$
test_items	32.5520μs	4.7491μs	210.5651 KOps/s	209.4437 KOps/s	$\color{#35bf28}+0.54\%$
test_items_nested	0.3763ms	0.3393ms	2.9472 KOps/s	2.9365 KOps/s	$\color{#35bf28}+0.37\%$
test_items_nested_locked	0.4016ms	0.3439ms	2.9079 KOps/s	2.9156 KOps/s	$\color{#d91a1a}-0.27\%$
test_items_nested_leaf	0.2337ms	0.2004ms	4.9908 KOps/s	4.9707 KOps/s	$\color{#35bf28}+0.40\%$
test_items_stack_nested	1.5685ms	1.3235ms	755.5678 Ops/s	2.9213 KOps/s	$\textbf{\color{#d91a1a}-74.14\%}$
test_items_stack_nested_leaf	1.2586ms	1.1652ms	858.2354 Ops/s	4.9478 KOps/s	$\textbf{\color{#d91a1a}-82.65\%}$
test_items_stack_nested_locked	1.2094ms	0.9180ms	1.0893 KOps/s	2.8977 KOps/s	$\textbf{\color{#d91a1a}-62.41\%}$
test_keys	65.1610μs	4.7788μs	209.2555 KOps/s	219.9129 KOps/s	$\color{#d91a1a}-4.85\%$
test_keys_nested	44.5541ms	0.1000ms	9.9964 KOps/s	10.5529 KOps/s	$\textbf{\color{#d91a1a}-5.27\%}$
test_keys_nested_locked	0.1424ms	98.2525μs	10.1779 KOps/s	10.1862 KOps/s	$\color{#d91a1a}-0.08\%$
test_keys_nested_leaf	0.1156ms	77.6393μs	12.8801 KOps/s	12.8440 KOps/s	$\color{#35bf28}+0.28\%$
test_keys_stack_nested	1.2416ms	1.1518ms	868.2256 Ops/s	10.6360 KOps/s	$\textbf{\color{#d91a1a}-91.84\%}$
test_keys_stack_nested_leaf	1.1968ms	1.1386ms	878.2419 Ops/s	12.8791 KOps/s	$\textbf{\color{#d91a1a}-93.18\%}$
test_keys_stack_nested_locked	0.9195ms	0.7268ms	1.3760 KOps/s	10.2061 KOps/s	$\textbf{\color{#d91a1a}-86.52\%}$
test_values	9.2000μs	1.8913μs	528.7332 KOps/s	531.1263 KOps/s	$\color{#d91a1a}-0.45\%$
test_values_nested	67.4410μs	45.3661μs	22.0429 KOps/s	22.1450 KOps/s	$\color{#d91a1a}-0.46\%$
test_values_nested_locked	0.1030ms	47.5076μs	21.0492 KOps/s	20.9138 KOps/s	$\color{#35bf28}+0.65\%$
test_values_nested_leaf	61.7110μs	39.8719μs	25.0803 KOps/s	25.2976 KOps/s	$\color{#d91a1a}-0.86\%$
test_values_stack_nested	1.0658ms	0.9649ms	1.0364 KOps/s	21.5606 KOps/s	$\textbf{\color{#d91a1a}-95.19\%}$
test_values_stack_nested_leaf	1.0413ms	0.9596ms	1.0422 KOps/s	24.7079 KOps/s	$\textbf{\color{#d91a1a}-95.78\%}$
test_values_stack_nested_locked	0.6438ms	0.5788ms	1.7277 KOps/s	20.6538 KOps/s	$\textbf{\color{#d91a1a}-91.63\%}$
test_membership	4.9020μs	0.9456μs	1.0575 MOps/s	1.0443 MOps/s	$\color{#35bf28}+1.26\%$
test_membership_nested	22.0600μs	2.8957μs	345.3360 KOps/s	342.6186 KOps/s	$\color{#35bf28}+0.79\%$
test_membership_nested_leaf	23.9500μs	2.8684μs	348.6225 KOps/s	344.4675 KOps/s	$\color{#35bf28}+1.21\%$
test_membership_stacked_nested	33.0110μs	11.2395μs	88.9720 KOps/s	345.9554 KOps/s	$\textbf{\color{#d91a1a}-74.28\%}$
test_membership_stacked_nested_leaf	35.2310μs	11.3493μs	88.1115 KOps/s	344.8552 KOps/s	$\textbf{\color{#d91a1a}-74.45\%}$
test_membership_nested_last	66.9110μs	5.2677μs	189.8364 KOps/s	186.7205 KOps/s	$\color{#35bf28}+1.67\%$
test_membership_nested_leaf_last	20.4200μs	5.2975μs	188.7689 KOps/s	186.4738 KOps/s	$\color{#35bf28}+1.23\%$
test_membership_stacked_nested_last	0.1970ms	0.1560ms	6.4094 KOps/s	146.8461 KOps/s	$\textbf{\color{#d91a1a}-95.64\%}$
test_membership_stacked_nested_leaf_last	45.3110μs	13.1356μs	76.1288 KOps/s	146.9113 KOps/s	$\textbf{\color{#d91a1a}-48.18\%}$
test_nested_getleaf	38.0110μs	8.3893μs	119.1991 KOps/s	118.9444 KOps/s	$\color{#35bf28}+0.21\%$
test_nested_get	67.3410μs	7.9778μs	125.3486 KOps/s	125.9248 KOps/s	$\color{#d91a1a}-0.46\%$
test_stacked_getleaf	0.3742ms	0.3299ms	3.0309 KOps/s	118.9930 KOps/s	$\textbf{\color{#d91a1a}-97.45\%}$
test_stacked_get	0.3343ms	0.2973ms	3.3637 KOps/s	126.1084 KOps/s	$\textbf{\color{#d91a1a}-97.33\%}$
test_nested_getitemleaf	28.2410μs	9.7800μs	102.2493 KOps/s	102.2508 KOps/s	$-0.00\%$
test_nested_getitem	24.4410μs	9.3170μs	107.3309 KOps/s	107.5794 KOps/s	$\color{#d91a1a}-0.23\%$
test_stacked_getitemleaf	0.3958ms	0.3323ms	3.0093 KOps/s	102.7099 KOps/s	$\textbf{\color{#d91a1a}-97.07\%}$
test_stacked_getitem	0.3257ms	0.2973ms	3.3632 KOps/s	107.5553 KOps/s	$\textbf{\color{#d91a1a}-96.87\%}$
test_lock_nested	1.2810ms	0.3575ms	2.7970 KOps/s	2.3220 KOps/s	$\textbf{\color{#35bf28}+20.46\%}$
test_lock_stack_nested	72.1149ms	5.7819ms	172.9531 Ops/s	3.2812 KOps/s	$\textbf{\color{#d91a1a}-94.73\%}$
test_unlock_nested	0.7214ms	0.3541ms	2.8245 KOps/s	2.8848 KOps/s	$\color{#d91a1a}-2.09\%$
test_unlock_stack_nested	75.0942ms	6.0423ms	165.5004 Ops/s	3.1785 KOps/s	$\textbf{\color{#d91a1a}-94.79\%}$
test_flatten_speed	0.4854ms	0.2599ms	3.8470 KOps/s	3.8634 KOps/s	$\color{#d91a1a}-0.42\%$
test_unflatten_speed	0.3805ms	0.3589ms	2.7865 KOps/s	2.7742 KOps/s	$\color{#35bf28}+0.44\%$
test_common_ops	1.1050ms	0.6195ms	1.6141 KOps/s	1.5861 KOps/s	$\color{#35bf28}+1.76\%$
test_creation	37.1000μs	1.5600μs	641.0288 KOps/s	636.8811 KOps/s	$\color{#35bf28}+0.65\%$
test_creation_empty	39.5310μs	9.3099μs	107.4124 KOps/s	99.6868 KOps/s	$\textbf{\color{#35bf28}+7.75\%}$
test_creation_nested_1	29.9100μs	11.1624μs	89.5863 KOps/s	85.0510 KOps/s	$\textbf{\color{#35bf28}+5.33\%}$
test_creation_nested_2	37.8100μs	13.7066μs	72.9576 KOps/s	69.9322 KOps/s	$\color{#35bf28}+4.33\%$
test_clone	80.9110μs	14.2733μs	70.0610 KOps/s	74.4649 KOps/s	$\textbf{\color{#d91a1a}-5.91\%}$
test_getitem[int]	23.7510μs	10.6628μs	93.7836 KOps/s	92.5668 KOps/s	$\color{#35bf28}+1.31\%$
test_getitem[slice_int]	38.3800μs	21.0256μs	47.5610 KOps/s	48.0907 KOps/s	$\color{#d91a1a}-1.10\%$
test_getitem[range]	69.5810μs	51.0628μs	19.5837 KOps/s	25.7268 KOps/s	$\textbf{\color{#d91a1a}-23.88\%}$
test_getitem[tuple]	41.4510μs	18.9320μs	52.8206 KOps/s	54.3796 KOps/s	$\color{#d91a1a}-2.87\%$
test_getitem[list]	0.1371ms	36.4524μs	27.4330 KOps/s	27.9522 KOps/s	$\color{#d91a1a}-1.86\%$
test_setitem_dim[int]	49.1010μs	29.4311μs	33.9776 KOps/s	34.3142 KOps/s	$\color{#d91a1a}-0.98\%$
test_setitem_dim[slice_int]	72.2410μs	49.4907μs	20.2058 KOps/s	20.6954 KOps/s	$\color{#d91a1a}-2.37\%$
test_setitem_dim[range]	95.5410μs	68.7816μs	14.5388 KOps/s	14.5314 KOps/s	$\color{#35bf28}+0.05\%$
test_setitem_dim[tuple]	66.6010μs	43.6062μs	22.9325 KOps/s	23.3897 KOps/s	$\color{#d91a1a}-1.95\%$
test_setitem	47.5500μs	19.9756μs	50.0610 KOps/s	52.9906 KOps/s	$\textbf{\color{#d91a1a}-5.53\%}$
test_set	49.5200μs	19.6964μs	50.7706 KOps/s	54.2296 KOps/s	$\textbf{\color{#d91a1a}-6.38\%}$
test_set_shared	1.6430ms	0.1037ms	9.6441 KOps/s	9.8943 KOps/s	$\color{#d91a1a}-2.53\%$
test_update	84.5120μs	22.0021μs	45.4503 KOps/s	44.9152 KOps/s	$\color{#35bf28}+1.19\%$
test_update_nested	72.0310μs	28.6898μs	34.8555 KOps/s	35.1884 KOps/s	$\color{#d91a1a}-0.95\%$
test_set_nested	66.3100μs	20.5205μs	48.7318 KOps/s	51.2676 KOps/s	$\color{#d91a1a}-4.95\%$
test_set_nested_new	67.2320μs	23.3532μs	42.8207 KOps/s	44.5667 KOps/s	$\color{#d91a1a}-3.92\%$
test_select	72.9010μs	37.0420μs	26.9964 KOps/s	28.5709 KOps/s	$\textbf{\color{#d91a1a}-5.51\%}$
test_select_nested	84.6820μs	53.5621μs	18.6699 KOps/s	18.8240 KOps/s	$\color{#d91a1a}-0.82\%$
test_exclude_nested	0.1609ms	0.1139ms	8.7826 KOps/s	8.8573 KOps/s	$\color{#d91a1a}-0.84\%$
test_empty[True]	0.9041ms	0.3878ms	2.5785 KOps/s	2.5974 KOps/s	$\color{#d91a1a}-0.73\%$
test_empty[False]	3.0271μs	0.8562μs	1.1679 MOps/s	1.1756 MOps/s	$\color{#d91a1a}-0.65\%$
test_to	74.5520μs	54.0528μs	18.5004 KOps/s	18.3787 KOps/s	$\color{#35bf28}+0.66\%$
test_to_nonblocking	69.8510μs	34.5352μs	28.9559 KOps/s	27.4972 KOps/s	$\textbf{\color{#35bf28}+5.31\%}$
test_unbind_speed	0.3204ms	0.2690ms	3.7173 KOps/s	3.7484 KOps/s	$\color{#d91a1a}-0.83\%$
test_unbind_speed_stack0	79.8090ms	3.5498ms	281.7095 Ops/s	3.7564 KOps/s	$\textbf{\color{#d91a1a}-92.50\%}$
test_unbind_speed_stack1	16.9500μs	1.8433μs	542.5076 KOps/s	1.2632 KOps/s	$\textbf{\color{#35bf28}+42846.64\%}$
test_split	77.0626ms	1.6609ms	602.0663 Ops/s	668.2552 Ops/s	$\textbf{\color{#d91a1a}-9.90\%}$
test_chunk	1.8936ms	1.4875ms	672.2768 Ops/s	586.6866 Ops/s	$\textbf{\color{#35bf28}+14.59\%}$
test_creation[device0]	0.1704ms	71.9045μs	13.9073 KOps/s	13.8709 KOps/s	$\color{#35bf28}+0.26\%$
test_creation_from_tensor	0.1811ms	54.4236μs	18.3744 KOps/s	18.1479 KOps/s	$\color{#35bf28}+1.25\%$
test_add_one[memmap_tensor0]	87.7710μs	7.2142μs	138.6159 KOps/s	151.5799 KOps/s	$\textbf{\color{#d91a1a}-8.55\%}$
test_contiguous[memmap_tensor0]	25.5800μs	0.6359μs	1.5726 MOps/s	1.6075 MOps/s	$\color{#d91a1a}-2.18\%$
test_stack[memmap_tensor0]	19.5010μs	4.4844μs	222.9953 KOps/s	239.7819 KOps/s	$\textbf{\color{#d91a1a}-7.00\%}$
test_memmaptd_index	0.9994ms	0.2606ms	3.8375 KOps/s	3.8622 KOps/s	$\color{#d91a1a}-0.64\%$
test_memmaptd_index_astensor	0.5666ms	0.3195ms	3.1295 KOps/s	3.1692 KOps/s	$\color{#d91a1a}-1.25\%$
test_memmaptd_index_op	0.9820ms	0.6379ms	1.5676 KOps/s	1.5853 KOps/s	$\color{#d91a1a}-1.12\%$
test_serialize_model	92.9466ms	90.0860ms	11.1005 Ops/s	10.6910 Ops/s	$\color{#35bf28}+3.83\%$
test_serialize_model_pickle	1.3506s	1.2354s	0.8095 Ops/s	0.8084 Ops/s	$\color{#35bf28}+0.13\%$
test_serialize_weights	0.1779s	95.6430ms	10.4555 Ops/s	10.9571 Ops/s	$\color{#d91a1a}-4.58\%$
test_serialize_weights_returnearly	0.3014s	79.5324ms	12.5735 Ops/s	12.5700 Ops/s	$\color{#35bf28}+0.03\%$
test_serialize_weights_pickle	1.4123s	1.2553s	0.7966 Ops/s	0.8036 Ops/s	$\color{#d91a1a}-0.87\%$
test_reshape_pytree	46.0710μs	24.6141μs	40.6271 KOps/s	40.7952 KOps/s	$\color{#d91a1a}-0.41\%$
test_reshape_td	69.7210μs	30.7051μs	32.5679 KOps/s	31.8591 KOps/s	$\color{#35bf28}+2.22\%$
test_view_pytree	43.0100μs	24.3987μs	40.9857 KOps/s	41.2593 KOps/s	$\color{#d91a1a}-0.66\%$
test_view_td	91.6331ms	10.7538μs	92.9908 KOps/s	16.1062 KOps/s	$\textbf{\color{#35bf28}+477.36\%}$
test_unbind_pytree	0.2963ms	30.2183μs	33.0925 KOps/s	33.3883 KOps/s	$\color{#d91a1a}-0.89\%$
test_unbind_td	72.6810μs	40.4580μs	24.7170 KOps/s	25.2779 KOps/s	$\color{#d91a1a}-2.22\%$
test_split_pytree	45.6200μs	28.4853μs	35.1058 KOps/s	35.8822 KOps/s	$\color{#d91a1a}-2.16\%$
test_split_td	0.1023ms	38.3155μs	26.0991 KOps/s	26.3170 KOps/s	$\color{#d91a1a}-0.83\%$
test_add_pytree	66.9310μs	36.6114μs	27.3139 KOps/s	28.9226 KOps/s	$\textbf{\color{#d91a1a}-5.56\%}$
test_add_td	0.1297ms	51.7435μs	19.3261 KOps/s	18.8914 KOps/s	$\color{#35bf28}+2.30\%$
test_distributed	0.1621ms	69.8398μs	14.3185 KOps/s	12.2467 KOps/s	$\textbf{\color{#35bf28}+16.92\%}$
test_tdmodule	50.6800μs	18.5119μs	54.0192 KOps/s	53.6378 KOps/s	$\color{#35bf28}+0.71\%$
test_tdmodule_dispatch	0.1470ms	37.9391μs	26.3580 KOps/s	25.6148 KOps/s	$\color{#35bf28}+2.90\%$
test_tdseq	41.2200μs	21.4014μs	46.7260 KOps/s	46.2533 KOps/s	$\color{#35bf28}+1.02\%$
test_tdseq_dispatch	60.9110μs	40.8167μs	24.4998 KOps/s	23.9831 KOps/s	$\color{#35bf28}+2.15\%$
test_instantiation_functorch	2.5199ms	1.6847ms	593.5888 Ops/s	602.4006 Ops/s	$\color{#d91a1a}-1.46\%$
test_instantiation_td	1.6718ms	1.1548ms	865.9856 Ops/s	722.5366 Ops/s	$\textbf{\color{#35bf28}+19.85\%}$
test_exec_functorch	0.3859ms	0.1625ms	6.1557 KOps/s	6.4168 KOps/s	$\color{#d91a1a}-4.07\%$
test_exec_functional_call	0.3655ms	0.1596ms	6.2668 KOps/s	6.4451 KOps/s	$\color{#d91a1a}-2.77\%$
test_exec_td	0.1821ms	0.1506ms	6.6384 KOps/s	6.9758 KOps/s	$\color{#d91a1a}-4.84\%$
test_exec_td_decorator	0.6551ms	0.1788ms	5.5931 KOps/s	5.7512 KOps/s	$\color{#d91a1a}-2.75\%$
test_vmap_mlp_speed[True-True]	1.2470ms	1.0394ms	962.0847 Ops/s	1.6406 KOps/s	$\textbf{\color{#d91a1a}-41.36\%}$
test_vmap_mlp_speed[True-False]	0.8214ms	0.6100ms	1.6393 KOps/s	1.6601 KOps/s	$\color{#d91a1a}-1.25\%$
test_vmap_mlp_speed[False-True]	1.1851ms	0.9486ms	1.0542 KOps/s	1.8913 KOps/s	$\textbf{\color{#d91a1a}-44.26\%}$
test_vmap_mlp_speed[False-False]	0.7665ms	0.5328ms	1.8767 KOps/s	1.8576 KOps/s	$\color{#35bf28}+1.03\%$
test_vmap_mlp_speed_decorator[True-True]	2.0326ms	1.7980ms	556.1725 Ops/s	1.5668 KOps/s	$\textbf{\color{#d91a1a}-64.50\%}$
test_vmap_mlp_speed_decorator[True-False]	0.8346ms	0.6410ms	1.5601 KOps/s	1.5734 KOps/s	$\color{#d91a1a}-0.85\%$
test_vmap_mlp_speed_decorator[False-True]	1.8066ms	1.5859ms	630.5736 Ops/s	1.8587 KOps/s	$\textbf{\color{#d91a1a}-66.07\%}$
test_vmap_mlp_speed_decorator[False-False]	0.8506ms	0.5473ms	1.8271 KOps/s	1.8645 KOps/s	$\color{#d91a1a}-2.00\%$
test_vmap_transformer_speed[True-True]	12.4155ms	12.1786ms	82.1113 Ops/s	123.3384 Ops/s	$\textbf{\color{#d91a1a}-33.43\%}$
test_vmap_transformer_speed[True-False]	8.3124ms	8.0302ms	124.5294 Ops/s	124.2484 Ops/s	$\color{#35bf28}+0.23\%$
test_vmap_transformer_speed[False-True]	12.3206ms	12.0447ms	83.0238 Ops/s	124.7350 Ops/s	$\textbf{\color{#d91a1a}-33.44\%}$
test_vmap_transformer_speed[False-False]	8.3884ms	7.9806ms	125.3044 Ops/s	125.1664 Ops/s	$\color{#35bf28}+0.11\%$
test_vmap_transformer_speed_decorator[True-True]	58.7971ms	58.5729ms	17.0728 Ops/s	50.0057 Ops/s	$\textbf{\color{#d91a1a}-65.86\%}$
test_vmap_transformer_speed_decorator[True-False]	19.3554ms	19.0717ms	52.4338 Ops/s	49.9880 Ops/s	$\color{#35bf28}+4.89\%$
test_vmap_transformer_speed_decorator[False-True]	56.0120ms	53.4675ms	18.7030 Ops/s	51.3182 Ops/s	$\textbf{\color{#d91a1a}-63.55\%}$
test_vmap_transformer_speed_decorator[False-False]	18.9157ms	18.6926ms	53.4972 Ops/s	51.2119 Ops/s	$\color{#35bf28}+4.46\%$
test_to_module_speed[True]	2.6158ms	1.0138ms	986.3693 Ops/s	1.0007 KOps/s	$\color{#d91a1a}-1.43\%$
test_to_module_speed[False]	1.4720ms	0.9745ms	1.0262 KOps/s	1.0317 KOps/s	$\color{#d91a1a}-0.53\%$