[Performance] Faster set

[vmoens]

cc @dubuqa

[github-actions]

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

Total Benchmarks: 120. Improved: $\large\color{#35bf28}13$. Worsened: $\large\color{#d91a1a}5$.

Expand to view detailed results

Name	Max	Mean	Ops	Ops on Repo HEAD	Change
test_plain_set_nested	0.1474ms	17.4049μs	57.4551 KOps/s	55.0453 KOps/s	$\color{#35bf28}+4.38\%$
test_plain_set_stack_nested	0.1987ms	0.1456ms	6.8668 KOps/s	6.6853 KOps/s	$\color{#35bf28}+2.72\%$
test_plain_set_nested_inplace	66.8750μs	19.6244μs	50.9571 KOps/s	49.4518 KOps/s	$\color{#35bf28}+3.04\%$
test_plain_set_stack_nested_inplace	0.3188ms	0.1799ms	5.5599 KOps/s	5.4782 KOps/s	$\color{#35bf28}+1.49\%$
test_items	18.5340μs	2.4148μs	414.1109 KOps/s	403.4641 KOps/s	$\color{#35bf28}+2.64\%$
test_items_nested	0.4429ms	0.2654ms	3.7676 KOps/s	3.6988 KOps/s	$\color{#35bf28}+1.86\%$
test_items_nested_locked	0.6331ms	0.2798ms	3.5737 KOps/s	3.6817 KOps/s	$\color{#d91a1a}-2.93\%$
test_items_nested_leaf	0.2909ms	0.1644ms	6.0812 KOps/s	5.9162 KOps/s	$\color{#35bf28}+2.79\%$
test_items_stack_nested	1.7300ms	1.3195ms	757.8714 Ops/s	705.8108 Ops/s	$\textbf{\color{#35bf28}+7.38\%}$
test_items_stack_nested_leaf	1.3097ms	1.1929ms	838.2958 Ops/s	828.1432 Ops/s	$\color{#35bf28}+1.23\%$
test_items_stack_nested_locked	1.1261ms	0.8722ms	1.1466 KOps/s	1.1441 KOps/s	$\color{#35bf28}+0.22\%$
test_keys	20.1780μs	3.8007μs	263.1101 KOps/s	261.8893 KOps/s	$\color{#35bf28}+0.47\%$
test_keys_nested	59.6221ms	0.1581ms	6.3252 KOps/s	6.8204 KOps/s	$\textbf{\color{#d91a1a}-7.26\%}$
test_keys_nested_locked	0.2666ms	0.1479ms	6.7625 KOps/s	6.8350 KOps/s	$\color{#d91a1a}-1.06\%$
test_keys_nested_leaf	0.2277ms	0.1304ms	7.6696 KOps/s	7.8440 KOps/s	$\color{#d91a1a}-2.22\%$
test_keys_stack_nested	1.7626ms	1.2749ms	784.3967 Ops/s	764.7685 Ops/s	$\color{#35bf28}+2.57\%$
test_keys_stack_nested_leaf	1.6094ms	1.2599ms	793.7245 Ops/s	763.9428 Ops/s	$\color{#35bf28}+3.90\%$
test_keys_stack_nested_locked	1.0252ms	0.7982ms	1.2528 KOps/s	1.2363 KOps/s	$\color{#35bf28}+1.33\%$
test_values	7.3716μs	1.0551μs	947.7851 KOps/s	880.9837 KOps/s	$\textbf{\color{#35bf28}+7.58\%}$
test_values_nested	99.4550μs	51.9548μs	19.2475 KOps/s	19.1889 KOps/s	$\color{#35bf28}+0.31\%$
test_values_nested_locked	97.4020μs	52.7308μs	18.9643 KOps/s	19.0777 KOps/s	$\color{#d91a1a}-0.59\%$
test_values_nested_leaf	0.1029ms	47.6777μs	20.9741 KOps/s	21.5859 KOps/s	$\color{#d91a1a}-2.83\%$
test_values_stack_nested	1.3010ms	1.0509ms	951.5260 Ops/s	942.7298 Ops/s	$\color{#35bf28}+0.93\%$
test_values_stack_nested_leaf	1.2065ms	1.0385ms	962.9303 Ops/s	957.4539 Ops/s	$\color{#35bf28}+0.57\%$
test_values_stack_nested_locked	3.4413ms	0.6157ms	1.6241 KOps/s	1.6339 KOps/s	$\color{#d91a1a}-0.60\%$
test_membership	37.5200μs	1.3521μs	739.5755 KOps/s	733.5604 KOps/s	$\color{#35bf28}+0.82\%$
test_membership_nested	35.4060μs	2.8890μs	346.1413 KOps/s	344.0494 KOps/s	$\color{#35bf28}+0.61\%$
test_membership_nested_leaf	39.3230μs	2.9555μs	338.3569 KOps/s	345.0321 KOps/s	$\color{#d91a1a}-1.93\%$
test_membership_stacked_nested	30.0860μs	12.0192μs	83.2001 KOps/s	83.4028 KOps/s	$\color{#d91a1a}-0.24\%$
test_membership_stacked_nested_leaf	61.7650μs	11.9799μs	83.4735 KOps/s	83.2538 KOps/s	$\color{#35bf28}+0.26\%$
test_membership_nested_last	37.0980μs	6.1050μs	163.7989 KOps/s	167.1081 KOps/s	$\color{#d91a1a}-1.98\%$
test_membership_nested_leaf_last	39.3030μs	6.1398μs	162.8730 KOps/s	166.3374 KOps/s	$\color{#d91a1a}-2.08\%$
test_membership_stacked_nested_last	0.3059ms	0.1689ms	5.9212 KOps/s	5.9165 KOps/s	$\color{#35bf28}+0.08\%$
test_membership_stacked_nested_leaf_last	50.2740μs	14.0437μs	71.2063 KOps/s	72.1471 KOps/s	$\color{#d91a1a}-1.30\%$
test_nested_getleaf	55.2530μs	10.7147μs	93.3297 KOps/s	94.8429 KOps/s	$\color{#d91a1a}-1.60\%$
test_nested_get	42.1290μs	10.0794μs	99.2118 KOps/s	97.9070 KOps/s	$\color{#35bf28}+1.33\%$
test_stacked_getleaf	0.5442ms	0.4033ms	2.4795 KOps/s	2.4401 KOps/s	$\color{#35bf28}+1.61\%$
test_stacked_get	0.6758ms	0.3682ms	2.7159 KOps/s	2.6773 KOps/s	$\color{#35bf28}+1.44\%$
test_nested_getitemleaf	39.9640μs	10.6326μs	94.0500 KOps/s	94.5793 KOps/s	$\color{#d91a1a}-0.56\%$
test_nested_getitem	43.4310μs	10.0771μs	99.2351 KOps/s	100.0480 KOps/s	$\color{#d91a1a}-0.81\%$
test_stacked_getitemleaf	0.6179ms	0.4062ms	2.4616 KOps/s	2.4236 KOps/s	$\color{#35bf28}+1.57\%$
test_stacked_getitem	0.6182ms	0.3680ms	2.7177 KOps/s	2.6582 KOps/s	$\color{#35bf28}+2.24\%$
test_lock_nested	1.3040ms	0.4175ms	2.3951 KOps/s	2.4080 KOps/s	$\color{#d91a1a}-0.54\%$
test_lock_stack_nested	88.6757ms	6.9351ms	144.1935 Ops/s	144.7434 Ops/s	$\color{#d91a1a}-0.38\%$
test_unlock_nested	69.0728ms	0.4903ms	2.0395 KOps/s	2.3658 KOps/s	$\textbf{\color{#d91a1a}-13.79\%}$
test_unlock_stack_nested	82.2344ms	6.4046ms	156.1384 Ops/s	155.5763 Ops/s	$\color{#35bf28}+0.36\%$
test_flatten_speed	0.7273ms	0.3727ms	2.6833 KOps/s	2.6550 KOps/s	$\color{#35bf28}+1.07\%$
test_unflatten_speed	0.6560ms	0.4584ms	2.1815 KOps/s	2.1728 KOps/s	$\color{#35bf28}+0.40\%$
test_common_ops	1.4277ms	0.7007ms	1.4271 KOps/s	1.3917 KOps/s	$\color{#35bf28}+2.54\%$
test_creation	85.8590μs	2.0069μs	498.2833 KOps/s	494.4449 KOps/s	$\color{#35bf28}+0.78\%$
test_creation_empty	46.9070μs	11.0634μs	90.3885 KOps/s	85.2331 KOps/s	$\textbf{\color{#35bf28}+6.05\%}$
test_creation_nested_1	56.8650μs	14.1503μs	70.6698 KOps/s	67.9512 KOps/s	$\color{#35bf28}+4.00\%$
test_creation_nested_2	53.8000μs	17.3278μs	57.7107 KOps/s	50.7208 KOps/s	$\textbf{\color{#35bf28}+13.78\%}$
test_clone	0.2299ms	12.3308μs	81.0978 KOps/s	82.2104 KOps/s	$\color{#d91a1a}-1.35\%$
test_getitem[int]	43.3710μs	11.7848μs	84.8553 KOps/s	84.4453 KOps/s	$\color{#35bf28}+0.49\%$
test_getitem[slice_int]	55.8440μs	22.8609μs	43.7427 KOps/s	42.6405 KOps/s	$\color{#35bf28}+2.59\%$
test_getitem[range]	99.8860μs	44.4289μs	22.5079 KOps/s	23.9788 KOps/s	$\textbf{\color{#d91a1a}-6.13\%}$
test_getitem[tuple]	71.4930μs	18.8429μs	53.0705 KOps/s	52.6004 KOps/s	$\color{#35bf28}+0.89\%$
test_getitem[list]	0.4974ms	38.0287μs	26.2960 KOps/s	26.5253 KOps/s	$\color{#d91a1a}-0.86\%$
test_setitem_dim[int]	82.3630μs	30.4845μs	32.8035 KOps/s	29.8512 KOps/s	$\textbf{\color{#35bf28}+9.89\%}$
test_setitem_dim[slice_int]	91.1290μs	56.9866μs	17.5480 KOps/s	16.9031 KOps/s	$\color{#35bf28}+3.82\%$
test_setitem_dim[range]	0.1229ms	78.5496μs	12.7308 KOps/s	12.9518 KOps/s	$\color{#d91a1a}-1.71\%$
test_setitem_dim[tuple]	98.2130μs	45.3928μs	22.0299 KOps/s	21.2960 KOps/s	$\color{#35bf28}+3.45\%$
test_setitem	0.2509ms	18.9518μs	52.7654 KOps/s	50.9169 KOps/s	$\color{#35bf28}+3.63\%$
test_set	0.2126ms	18.1843μs	54.9925 KOps/s	52.2237 KOps/s	$\textbf{\color{#35bf28}+5.30\%}$
test_set_shared	2.1551ms	0.1390ms	7.1919 KOps/s	7.0258 KOps/s	$\color{#35bf28}+2.36\%$
test_update	0.2128ms	21.8317μs	45.8051 KOps/s	44.5790 KOps/s	$\color{#35bf28}+2.75\%$
test_update_nested	0.2282ms	28.9964μs	34.4871 KOps/s	33.7546 KOps/s	$\color{#35bf28}+2.17\%$
test_set_nested	0.2172ms	19.9508μs	50.1233 KOps/s	48.1916 KOps/s	$\color{#35bf28}+4.01\%$
test_set_nested_new	0.2175ms	25.0202μs	39.9676 KOps/s	40.3067 KOps/s	$\color{#d91a1a}-0.84\%$
test_select	0.1202ms	47.9843μs	20.8402 KOps/s	20.4530 KOps/s	$\color{#35bf28}+1.89\%$
test_unbind_speed	0.5067ms	0.3423ms	2.9216 KOps/s	2.9574 KOps/s	$\color{#d91a1a}-1.21\%$
test_unbind_speed_stack0	69.9919ms	4.5807ms	218.3086 Ops/s	234.2013 Ops/s	$\textbf{\color{#d91a1a}-6.79\%}$
test_unbind_speed_stack1	2.5757μs	0.6200μs	1.6129 MOps/s	1.5195 MOps/s	$\textbf{\color{#35bf28}+6.15\%}$
test_split	2.3586ms	1.5571ms	642.2103 Ops/s	589.0819 Ops/s	$\textbf{\color{#35bf28}+9.02\%}$
test_chunk	69.4639ms	1.6630ms	601.3186 Ops/s	598.0948 Ops/s	$\color{#35bf28}+0.54\%$
test_creation[device0]	0.2261ms	0.1007ms	9.9288 KOps/s	9.9772 KOps/s	$\color{#d91a1a}-0.49\%$
test_creation_from_tensor	3.1541ms	80.7516μs	12.3837 KOps/s	12.2576 KOps/s	$\color{#35bf28}+1.03\%$
test_add_one[memmap_tensor0]	0.4102ms	5.1425μs	194.4593 KOps/s	194.2104 KOps/s	$\color{#35bf28}+0.13\%$
test_contiguous[memmap_tensor0]	13.0740μs	0.6407μs	1.5607 MOps/s	1.5950 MOps/s	$\color{#d91a1a}-2.15\%$
test_stack[memmap_tensor0]	71.1420μs	3.4960μs	286.0436 KOps/s	293.1548 KOps/s	$\color{#d91a1a}-2.43\%$
test_memmaptd_index	0.4047ms	0.1997ms	5.0064 KOps/s	5.0950 KOps/s	$\color{#d91a1a}-1.74\%$
test_memmaptd_index_astensor	0.9717ms	0.2614ms	3.8259 KOps/s	3.8652 KOps/s	$\color{#d91a1a}-1.02\%$
test_memmaptd_index_op	0.8038ms	0.5451ms	1.8346 KOps/s	1.7924 KOps/s	$\color{#35bf28}+2.36\%$
test_serialize_model	0.1021s	98.9067ms	10.1105 Ops/s	8.9330 Ops/s	$\textbf{\color{#35bf28}+13.18\%}$
test_serialize_model_pickle	0.4650s	0.3799s	2.6320 Ops/s	2.5846 Ops/s	$\color{#35bf28}+1.83\%$
test_serialize_weights	0.1731s	0.1036s	9.6551 Ops/s	9.3534 Ops/s	$\color{#35bf28}+3.22\%$
test_serialize_weights_returnearly	0.1798s	0.1288s	7.7658 Ops/s	7.3567 Ops/s	$\textbf{\color{#35bf28}+5.56\%}$
test_serialize_weights_pickle	1.0858s	0.6132s	1.6307 Ops/s	1.5945 Ops/s	$\color{#35bf28}+2.27\%$
test_serialize_weights_filesystem	0.1588s	95.6355ms	10.4564 Ops/s	10.6125 Ops/s	$\color{#d91a1a}-1.47\%$
test_serialize_model_filesystem	98.1990ms	91.4000ms	10.9409 Ops/s	10.1290 Ops/s	$\textbf{\color{#35bf28}+8.02\%}$
test_reshape_pytree	66.3530μs	23.1842μs	43.1328 KOps/s	42.4549 KOps/s	$\color{#35bf28}+1.60\%$
test_reshape_td	62.5070μs	31.0382μs	32.2184 KOps/s	31.7562 KOps/s	$\color{#35bf28}+1.46\%$
test_view_pytree	55.7640μs	23.0998μs	43.2904 KOps/s	43.8069 KOps/s	$\color{#d91a1a}-1.18\%$
test_view_td	34.8550μs	4.8768μs	205.0533 KOps/s	201.5123 KOps/s	$\color{#35bf28}+1.76\%$
test_unbind_pytree	60.6330μs	26.2771μs	38.0560 KOps/s	37.6472 KOps/s	$\color{#35bf28}+1.09\%$
test_unbind_td	0.1181ms	54.8436μs	18.2337 KOps/s	18.1446 KOps/s	$\color{#35bf28}+0.49\%$
test_split_pytree	74.1380μs	26.6978μs	37.4563 KOps/s	38.2040 KOps/s	$\color{#d91a1a}-1.96\%$
test_split_td	0.5223ms	43.4233μs	23.0291 KOps/s	22.8524 KOps/s	$\color{#35bf28}+0.77\%$
test_add_pytree	93.4240μs	31.7574μs	31.4887 KOps/s	31.3653 KOps/s	$\color{#35bf28}+0.39\%$
test_add_td	0.1524ms	51.7880μs	19.3095 KOps/s	19.5074 KOps/s	$\color{#d91a1a}-1.01\%$
test_distributed	0.2183ms	97.4719μs	10.2594 KOps/s	9.6886 KOps/s	$\textbf{\color{#35bf28}+5.89\%}$
test_tdmodule	0.1122ms	23.1429μs	43.2099 KOps/s	41.8353 KOps/s	$\color{#35bf28}+3.29\%$
test_tdmodule_dispatch	0.2270ms	42.4961μs	23.5316 KOps/s	23.1495 KOps/s	$\color{#35bf28}+1.65\%$
test_tdseq	54.5520μs	26.4907μs	37.7491 KOps/s	36.7227 KOps/s	$\color{#35bf28}+2.79\%$
test_tdseq_dispatch	0.1443ms	46.8900μs	21.3265 KOps/s	20.8761 KOps/s	$\color{#35bf28}+2.16\%$
test_instantiation_functorch	1.5165ms	1.2998ms	769.3472 Ops/s	760.7691 Ops/s	$\color{#35bf28}+1.13\%$
test_instantiation_td	1.6187ms	1.0275ms	973.2514 Ops/s	977.2524 Ops/s	$\color{#d91a1a}-0.41\%$
test_exec_functorch	0.2939ms	0.1605ms	6.2293 KOps/s	6.3072 KOps/s	$\color{#d91a1a}-1.24\%$
test_exec_functional_call	0.2941ms	0.1487ms	6.7242 KOps/s	6.8808 KOps/s	$\color{#d91a1a}-2.28\%$
test_exec_td	0.2658ms	0.1433ms	6.9759 KOps/s	7.1467 KOps/s	$\color{#d91a1a}-2.39\%$
test_exec_td_decorator	0.7582ms	0.1818ms	5.4997 KOps/s	5.6534 KOps/s	$\color{#d91a1a}-2.72\%$
test_vmap_mlp_speed[True-True]	1.1650ms	0.8804ms	1.1359 KOps/s	1.1381 KOps/s	$\color{#d91a1a}-0.20\%$
test_vmap_mlp_speed[True-False]	0.6894ms	0.4711ms	2.1229 KOps/s	2.1433 KOps/s	$\color{#d91a1a}-0.95\%$
test_vmap_mlp_speed[False-True]	1.2347ms	0.7765ms	1.2878 KOps/s	1.3189 KOps/s	$\color{#d91a1a}-2.36\%$
test_vmap_mlp_speed[False-False]	0.5563ms	0.3863ms	2.5886 KOps/s	2.6602 KOps/s	$\color{#d91a1a}-2.69\%$
test_vmap_mlp_speed_decorator[True-True]	3.2518ms	2.4571ms	406.9840 Ops/s	377.5929 Ops/s	$\textbf{\color{#35bf28}+7.78\%}$
test_vmap_mlp_speed_decorator[True-False]	0.8686ms	0.5242ms	1.9078 KOps/s	1.9273 KOps/s	$\color{#d91a1a}-1.01\%$
test_vmap_mlp_speed_decorator[False-True]	79.1764ms	2.1393ms	467.4368 Ops/s	501.9978 Ops/s	$\textbf{\color{#d91a1a}-6.88\%}$
test_vmap_mlp_speed_decorator[False-False]	0.6989ms	0.3994ms	2.5040 KOps/s	2.5507 KOps/s	$\color{#d91a1a}-1.83\%$

[github-actions]

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

Total Benchmarks: 128. Improved: $\large\color{#35bf28}5$. Worsened: $\large\color{#d91a1a}19$.

Expand to view detailed results

Name	Max	Mean	Ops	Ops on Repo HEAD	Change
test_plain_set_nested	0.1331ms	14.1728μs	70.5576 KOps/s	75.6005 KOps/s	$\textbf{\color{#d91a1a}-6.67\%}$
test_plain_set_stack_nested	0.1400ms	0.1182ms	8.4604 KOps/s	8.6574 KOps/s	$\color{#d91a1a}-2.28\%$
test_plain_set_nested_inplace	32.6000μs	15.3839μs	65.0032 KOps/s	68.8844 KOps/s	$\textbf{\color{#d91a1a}-5.63\%}$
test_plain_set_stack_nested_inplace	0.1862ms	0.1446ms	6.9146 KOps/s	6.9789 KOps/s	$\color{#d91a1a}-0.92\%$
test_items	24.0200μs	4.7008μs	212.7310 KOps/s	211.7866 KOps/s	$\color{#35bf28}+0.45\%$
test_items_nested	0.3785ms	0.3396ms	2.9450 KOps/s	2.9682 KOps/s	$\color{#d91a1a}-0.78\%$
test_items_nested_locked	0.3927ms	0.3398ms	2.9426 KOps/s	2.9510 KOps/s	$\color{#d91a1a}-0.29\%$
test_items_nested_leaf	0.2209ms	0.1984ms	5.0400 KOps/s	5.0811 KOps/s	$\color{#d91a1a}-0.81\%$
test_items_stack_nested	1.3855ms	1.2672ms	789.1410 Ops/s	768.6245 Ops/s	$\color{#35bf28}+2.67\%$
test_items_stack_nested_leaf	1.2710ms	1.1086ms	902.0275 Ops/s	886.6226 Ops/s	$\color{#35bf28}+1.74\%$
test_items_stack_nested_locked	1.0286ms	0.8871ms	1.1273 KOps/s	1.1050 KOps/s	$\color{#35bf28}+2.02\%$
test_keys	24.4110μs	4.5835μs	218.1762 KOps/s	216.9873 KOps/s	$\color{#35bf28}+0.55\%$
test_keys_nested	0.8489ms	95.2553μs	10.4981 KOps/s	10.6618 KOps/s	$\color{#d91a1a}-1.53\%$
test_keys_nested_locked	0.1416ms	94.1783μs	10.6182 KOps/s	10.6983 KOps/s	$\color{#d91a1a}-0.75\%$
test_keys_nested_leaf	0.1964ms	77.9798μs	12.8238 KOps/s	12.9729 KOps/s	$\color{#d91a1a}-1.15\%$
test_keys_stack_nested	1.1761ms	1.1141ms	897.6200 Ops/s	886.7466 Ops/s	$\color{#35bf28}+1.23\%$
test_keys_stack_nested_leaf	1.1421ms	1.0791ms	926.6861 Ops/s	903.3344 Ops/s	$\color{#35bf28}+2.59\%$
test_keys_stack_nested_locked	0.7951ms	0.7035ms	1.4215 KOps/s	1.3932 KOps/s	$\color{#35bf28}+2.03\%$
test_values	8.5003μs	1.8893μs	529.2931 KOps/s	529.8359 KOps/s	$\color{#d91a1a}-0.10\%$
test_values_nested	78.7910μs	45.1076μs	22.1692 KOps/s	22.0144 KOps/s	$\color{#35bf28}+0.70\%$
test_values_nested_locked	61.6910μs	47.4534μs	21.0733 KOps/s	20.9445 KOps/s	$\color{#35bf28}+0.61\%$
test_values_nested_leaf	53.2510μs	39.4725μs	25.3341 KOps/s	25.2484 KOps/s	$\color{#35bf28}+0.34\%$
test_values_stack_nested	0.9868ms	0.9263ms	1.0796 KOps/s	1.0493 KOps/s	$\color{#35bf28}+2.88\%$
test_values_stack_nested_leaf	1.0227ms	0.9212ms	1.0856 KOps/s	1.0592 KOps/s	$\color{#35bf28}+2.49\%$
test_values_stack_nested_locked	0.6559ms	0.5653ms	1.7691 KOps/s	1.7342 KOps/s	$\color{#35bf28}+2.01\%$
test_membership	3.7060μs	0.9419μs	1.0617 MOps/s	919.2465 KOps/s	$\textbf{\color{#35bf28}+15.49\%}$
test_membership_nested	11.1505μs	2.2367μs	447.0778 KOps/s	438.6079 KOps/s	$\color{#35bf28}+1.93\%$
test_membership_nested_leaf	13.1105μs	2.1953μs	455.5275 KOps/s	457.5116 KOps/s	$\color{#d91a1a}-0.43\%$
test_membership_stacked_nested	31.7210μs	11.0444μs	90.5434 KOps/s	92.2149 KOps/s	$\color{#d91a1a}-1.81\%$
test_membership_stacked_nested_leaf	30.1810μs	11.2039μs	89.2547 KOps/s	91.8081 KOps/s	$\color{#d91a1a}-2.78\%$
test_membership_nested_last	17.9490μs	4.7848μs	208.9971 KOps/s	211.5276 KOps/s	$\color{#d91a1a}-1.20\%$
test_membership_nested_leaf_last	23.2500μs	4.8503μs	206.1742 KOps/s	209.9464 KOps/s	$\color{#d91a1a}-1.80\%$
test_membership_stacked_nested_last	0.1677ms	0.1350ms	7.4059 KOps/s	7.3920 KOps/s	$\color{#35bf28}+0.19\%$
test_membership_stacked_nested_leaf_last	36.6600μs	13.0503μs	76.6269 KOps/s	78.4232 KOps/s	$\color{#d91a1a}-2.29\%$
test_nested_getleaf	49.6410μs	8.4721μs	118.0346 KOps/s	120.3244 KOps/s	$\color{#d91a1a}-1.90\%$
test_nested_get	0.1950ms	7.9575μs	125.6674 KOps/s	127.2676 KOps/s	$\color{#d91a1a}-1.26\%$
test_stacked_getleaf	0.5176ms	0.3151ms	3.1741 KOps/s	3.1507 KOps/s	$\color{#35bf28}+0.74\%$
test_stacked_get	0.4920ms	0.2860ms	3.4960 KOps/s	3.4682 KOps/s	$\color{#35bf28}+0.80\%$
test_nested_getitemleaf	0.2056ms	8.4662μs	118.1171 KOps/s	119.3899 KOps/s	$\color{#d91a1a}-1.07\%$
test_nested_getitem	47.3920μs	7.9774μs	125.3547 KOps/s	126.2123 KOps/s	$\color{#d91a1a}-0.68\%$
test_stacked_getitemleaf	0.5251ms	0.3137ms	3.1873 KOps/s	3.1395 KOps/s	$\color{#35bf28}+1.52\%$
test_stacked_getitem	0.4905ms	0.2843ms	3.5171 KOps/s	3.5312 KOps/s	$\color{#d91a1a}-0.40\%$
test_lock_nested	4.6507ms	0.4211ms	2.3745 KOps/s	2.4201 KOps/s	$\color{#d91a1a}-1.88\%$
test_lock_stack_nested	84.1382ms	6.4732ms	154.4829 Ops/s	154.1364 Ops/s	$\color{#35bf28}+0.22\%$
test_unlock_nested	0.8045ms	0.4096ms	2.4413 KOps/s	2.4389 KOps/s	$\color{#35bf28}+0.10\%$
test_unlock_stack_nested	83.0853ms	6.8593ms	145.7884 Ops/s	145.4958 Ops/s	$\color{#35bf28}+0.20\%$
test_flatten_speed	0.4963ms	0.2631ms	3.8006 KOps/s	3.7987 KOps/s	$\color{#35bf28}+0.05\%$
test_unflatten_speed	0.5391ms	0.3516ms	2.8442 KOps/s	2.8206 KOps/s	$\color{#35bf28}+0.84\%$
test_common_ops	1.0392ms	0.5922ms	1.6886 KOps/s	1.6935 KOps/s	$\color{#d91a1a}-0.29\%$
test_creation	19.6600μs	1.5656μs	638.7213 KOps/s	629.4467 KOps/s	$\color{#35bf28}+1.47\%$
test_creation_empty	36.9810μs	9.2121μs	108.5528 KOps/s	131.9425 KOps/s	$\textbf{\color{#d91a1a}-17.73\%}$
test_creation_nested_1	0.2147ms	11.1678μs	89.5428 KOps/s	105.1924 KOps/s	$\textbf{\color{#d91a1a}-14.88\%}$
test_creation_nested_2	38.9500μs	13.6286μs	73.3753 KOps/s	70.6287 KOps/s	$\color{#35bf28}+3.89\%$
test_clone	0.1355ms	12.8384μs	77.8911 KOps/s	74.9827 KOps/s	$\color{#35bf28}+3.88\%$
test_getitem[int]	0.2224ms	11.1672μs	89.5483 KOps/s	89.2292 KOps/s	$\color{#35bf28}+0.36\%$
test_getitem[slice_int]	46.4210μs	20.7799μs	48.1234 KOps/s	47.9290 KOps/s	$\color{#35bf28}+0.41\%$
test_getitem[range]	61.5510μs	35.3104μs	28.3203 KOps/s	28.1072 KOps/s	$\color{#35bf28}+0.76\%$
test_getitem[tuple]	0.2290ms	18.6283μs	53.6818 KOps/s	53.5024 KOps/s	$\color{#35bf28}+0.34\%$
test_getitem[list]	57.5210μs	31.8430μs	31.4040 KOps/s	30.3369 KOps/s	$\color{#35bf28}+3.52\%$
test_setitem_dim[int]	43.7610μs	27.6357μs	36.1851 KOps/s	38.5296 KOps/s	$\textbf{\color{#d91a1a}-6.09\%}$
test_setitem_dim[slice_int]	73.7710μs	48.6710μs	20.5461 KOps/s	21.7423 KOps/s	$\textbf{\color{#d91a1a}-5.50\%}$
test_setitem_dim[range]	95.4320μs	61.7136μs	16.2039 KOps/s	16.7766 KOps/s	$\color{#d91a1a}-3.41\%$
test_setitem_dim[tuple]	61.5510μs	42.8583μs	23.3327 KOps/s	25.2647 KOps/s	$\textbf{\color{#d91a1a}-7.65\%}$
test_setitem	0.2142ms	18.2271μs	54.8633 KOps/s	58.3249 KOps/s	$\textbf{\color{#d91a1a}-5.94\%}$
test_set	0.1280ms	17.5148μs	57.0947 KOps/s	60.4818 KOps/s	$\textbf{\color{#d91a1a}-5.60\%}$
test_set_shared	2.6933ms	0.1001ms	9.9948 KOps/s	9.9486 KOps/s	$\color{#35bf28}+0.46\%$
test_update	0.2447ms	20.1657μs	49.5891 KOps/s	52.6379 KOps/s	$\textbf{\color{#d91a1a}-5.79\%}$
test_update_nested	0.1453ms	26.7801μs	37.3412 KOps/s	40.3925 KOps/s	$\textbf{\color{#d91a1a}-7.55\%}$
test_set_nested	0.2367ms	18.6114μs	53.7306 KOps/s	55.5180 KOps/s	$\color{#d91a1a}-3.22\%$
test_set_nested_new	0.1254ms	21.9051μs	45.6516 KOps/s	48.0688 KOps/s	$\textbf{\color{#d91a1a}-5.03\%}$
test_select	0.2500ms	42.6711μs	23.4351 KOps/s	24.5786 KOps/s	$\color{#d91a1a}-4.65\%$
test_to	93.6010μs	56.5308μs	17.6895 KOps/s	18.7841 KOps/s	$\textbf{\color{#d91a1a}-5.83\%}$
test_to_nonblocking	67.7810μs	32.2609μs	30.9973 KOps/s	30.4862 KOps/s	$\color{#35bf28}+1.68\%$
test_unbind_speed	0.3777ms	0.3240ms	3.0860 KOps/s	3.0472 KOps/s	$\color{#35bf28}+1.27\%$
test_unbind_speed_stack0	79.5677ms	3.7558ms	266.2564 Ops/s	263.0733 Ops/s	$\color{#35bf28}+1.21\%$
test_unbind_speed_stack1	1.4930μs	0.5300μs	1.8869 MOps/s	1.8728 MOps/s	$\color{#35bf28}+0.75\%$
test_split	74.6590ms	1.7087ms	585.2569 Ops/s	576.9215 Ops/s	$\color{#35bf28}+1.44\%$
test_chunk	1.6166ms	1.5644ms	639.2424 Ops/s	587.3949 Ops/s	$\textbf{\color{#35bf28}+8.83\%}$
test_creation[device0]	0.1438ms	71.5275μs	13.9806 KOps/s	14.5282 KOps/s	$\color{#d91a1a}-3.77\%$
test_creation_from_tensor	0.1303ms	55.6750μs	17.9614 KOps/s	19.3336 KOps/s	$\textbf{\color{#d91a1a}-7.10\%}$
test_add_one[memmap_tensor0]	0.1202ms	6.6279μs	150.8783 KOps/s	149.2507 KOps/s	$\color{#35bf28}+1.09\%$
test_contiguous[memmap_tensor0]	25.3810μs	0.6302μs	1.5867 MOps/s	1.5950 MOps/s	$\color{#d91a1a}-0.52\%$
test_stack[memmap_tensor0]	37.8720μs	4.3904μs	227.7697 KOps/s	234.7277 KOps/s	$\color{#d91a1a}-2.96\%$
test_memmaptd_index	0.2632ms	0.2301ms	4.3464 KOps/s	4.3325 KOps/s	$\color{#35bf28}+0.32\%$
test_memmaptd_index_astensor	0.3197ms	0.2880ms	3.4722 KOps/s	3.4627 KOps/s	$\color{#35bf28}+0.27\%$
test_memmaptd_index_op	0.7643ms	0.5802ms	1.7235 KOps/s	1.7946 KOps/s	$\color{#d91a1a}-3.96\%$
test_serialize_model	0.1713s	96.2412ms	10.3906 Ops/s	9.6232 Ops/s	$\textbf{\color{#35bf28}+7.97\%}$
test_serialize_model_pickle	1.3488s	1.2368s	0.8086 Ops/s	0.8065 Ops/s	$\color{#35bf28}+0.26\%$
test_serialize_weights	88.9779ms	86.1719ms	11.6047 Ops/s	9.8866 Ops/s	$\textbf{\color{#35bf28}+17.38\%}$
test_serialize_weights_returnearly	0.2500s	77.6090ms	12.8851 Ops/s	14.8368 Ops/s	$\textbf{\color{#d91a1a}-13.15\%}$
test_serialize_weights_pickle	1.3519s	1.2451s	0.8032 Ops/s	0.8088 Ops/s	$\color{#d91a1a}-0.69\%$
test_reshape_pytree	50.7810μs	23.9955μs	41.6746 KOps/s	41.9168 KOps/s	$\color{#d91a1a}-0.58\%$
test_reshape_td	46.2310μs	27.7701μs	36.0099 KOps/s	35.1486 KOps/s	$\color{#35bf28}+2.45\%$
test_view_pytree	45.7910μs	23.4807μs	42.5881 KOps/s	42.6964 KOps/s	$\color{#d91a1a}-0.25\%$
test_view_td	21.5800μs	4.0309μs	248.0866 KOps/s	242.5299 KOps/s	$\color{#35bf28}+2.29\%$
test_unbind_pytree	58.2110μs	29.7987μs	33.5585 KOps/s	33.2560 KOps/s	$\color{#35bf28}+0.91\%$
test_unbind_td	76.1600μs	50.5159μs	19.7957 KOps/s	19.1499 KOps/s	$\color{#35bf28}+3.37\%$
test_split_pytree	42.6610μs	27.3592μs	36.5507 KOps/s	36.0397 KOps/s	$\color{#35bf28}+1.42\%$
test_split_td	0.6566ms	38.9585μs	25.6683 KOps/s	25.0382 KOps/s	$\color{#35bf28}+2.52\%$
test_add_pytree	59.7820μs	34.9105μs	28.6447 KOps/s	28.3057 KOps/s	$\color{#35bf28}+1.20\%$
test_add_td	67.7920μs	49.1326μs	20.3531 KOps/s	23.2013 KOps/s	$\textbf{\color{#d91a1a}-12.28\%}$
test_distributed	0.2641ms	70.6544μs	14.1534 KOps/s	14.6141 KOps/s	$\color{#d91a1a}-3.15\%$
test_tdmodule	0.1077ms	18.3946μs	54.3639 KOps/s	57.8624 KOps/s	$\textbf{\color{#d91a1a}-6.05\%}$
test_tdmodule_dispatch	0.2680ms	35.5234μs	28.1505 KOps/s	30.1569 KOps/s	$\textbf{\color{#d91a1a}-6.65\%}$
test_tdseq	0.1815ms	21.4833μs	46.5479 KOps/s	49.5362 KOps/s	$\textbf{\color{#d91a1a}-6.03\%}$
test_tdseq_dispatch	0.1623ms	37.2535μs	26.8431 KOps/s	27.4810 KOps/s	$\color{#d91a1a}-2.32\%$
test_instantiation_functorch	1.6961ms	1.6353ms	611.5133 Ops/s	609.3928 Ops/s	$\color{#35bf28}+0.35\%$
test_instantiation_td	1.7098ms	1.1524ms	867.7469 Ops/s	865.4654 Ops/s	$\color{#35bf28}+0.26\%$
test_exec_functorch	0.1864ms	0.1550ms	6.4504 KOps/s	6.4701 KOps/s	$\color{#d91a1a}-0.30\%$
test_exec_functional_call	0.1803ms	0.1541ms	6.4880 KOps/s	6.4784 KOps/s	$\color{#35bf28}+0.15\%$
test_exec_td	0.1767ms	0.1447ms	6.9127 KOps/s	6.7361 KOps/s	$\color{#35bf28}+2.62\%$
test_exec_td_decorator	0.8928ms	0.1864ms	5.3640 KOps/s	5.3831 KOps/s	$\color{#d91a1a}-0.36\%$
test_vmap_mlp_speed[True-True]	1.1581ms	1.0714ms	933.3856 Ops/s	925.8897 Ops/s	$\color{#35bf28}+0.81\%$
test_vmap_mlp_speed[True-False]	0.7640ms	0.6435ms	1.5540 KOps/s	1.5340 KOps/s	$\color{#35bf28}+1.30\%$
test_vmap_mlp_speed[False-True]	1.0979ms	0.9925ms	1.0076 KOps/s	993.7144 Ops/s	$\color{#35bf28}+1.40\%$
test_vmap_mlp_speed[False-False]	0.6189ms	0.5748ms	1.7396 KOps/s	1.6832 KOps/s	$\color{#35bf28}+3.35\%$
test_vmap_mlp_speed_decorator[True-True]	3.1181ms	2.4689ms	405.0393 Ops/s	406.2466 Ops/s	$\color{#d91a1a}-0.30\%$
test_vmap_mlp_speed_decorator[True-False]	1.0199ms	0.6933ms	1.4423 KOps/s	1.4300 KOps/s	$\color{#35bf28}+0.86\%$
test_vmap_mlp_speed_decorator[False-True]	2.5513ms	2.1267ms	470.2117 Ops/s	485.8885 Ops/s	$\color{#d91a1a}-3.23\%$
test_vmap_mlp_speed_decorator[False-False]	0.9752ms	0.6150ms	1.6261 KOps/s	1.6670 KOps/s	$\color{#d91a1a}-2.45\%$
test_vmap_transformer_speed[True-True]	12.7093ms	12.2856ms	81.3960 Ops/s	82.3253 Ops/s	$\color{#d91a1a}-1.13\%$
test_vmap_transformer_speed[True-False]	8.4520ms	8.1461ms	122.7579 Ops/s	123.7864 Ops/s	$\color{#d91a1a}-0.83\%$
test_vmap_transformer_speed[False-True]	13.3532ms	12.3299ms	81.1034 Ops/s	83.0202 Ops/s	$\color{#d91a1a}-2.31\%$
test_vmap_transformer_speed[False-False]	8.4429ms	8.1203ms	123.1486 Ops/s	125.2735 Ops/s	$\color{#d91a1a}-1.70\%$
test_vmap_transformer_speed_decorator[True-True]	0.1654s	82.4073ms	12.1348 Ops/s	12.4329 Ops/s	$\color{#d91a1a}-2.40\%$
test_vmap_transformer_speed_decorator[True-False]	21.4881ms	19.7849ms	50.5437 Ops/s	51.4946 Ops/s	$\color{#d91a1a}-1.85\%$
test_vmap_transformer_speed_decorator[False-True]	70.8819ms	68.6744ms	14.5615 Ops/s	14.9704 Ops/s	$\color{#d91a1a}-2.73\%$
test_vmap_transformer_speed_decorator[False-False]	21.0174ms	19.3594ms	51.6546 Ops/s	47.9706 Ops/s	$\textbf{\color{#35bf28}+7.68\%}$