[Feature] Subclass conservation in td ops

[vmoens]

Stack from ghstack (oldest at bottom):

• -> [Feature] Subclass conservation in td ops #1186

[github-actions]

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

Total Benchmarks: 217. Improved: $\large\color{#35bf28}41$. Worsened: $\large\color{#d91a1a}10$.

Expand to view detailed results

Name	Max	Mean	Ops	Ops on Repo HEAD	Change
test_plain_set_nested	75.8620μs	19.2901μs	51.8402 KOps/s	48.7280 KOps/s	$\textbf{\color{#35bf28}+6.39\%}$
test_plain_set_stack_nested	53.9110μs	19.4405μs	51.4389 KOps/s	48.1336 KOps/s	$\textbf{\color{#35bf28}+6.87\%}$
test_plain_set_nested_inplace	83.9870μs	21.2443μs	47.0714 KOps/s	44.5286 KOps/s	$\textbf{\color{#35bf28}+5.71\%}$
test_plain_set_stack_nested_inplace	53.0090μs	21.1288μs	47.3288 KOps/s	44.3081 KOps/s	$\textbf{\color{#35bf28}+6.82\%}$
test_items	44.5730μs	4.1749μs	239.5271 KOps/s	243.4163 KOps/s	$\color{#d91a1a}-1.60\%$
test_items_nested	0.6868ms	0.3944ms	2.5358 KOps/s	2.5407 KOps/s	$\color{#d91a1a}-0.19\%$
test_items_nested_locked	0.8076ms	0.3955ms	2.5287 KOps/s	2.4945 KOps/s	$\color{#35bf28}+1.37\%$
test_items_nested_leaf	0.1808ms	76.3298μs	13.1010 KOps/s	12.9339 KOps/s	$\color{#35bf28}+1.29\%$
test_items_stack_nested	0.6842ms	0.3968ms	2.5203 KOps/s	2.4985 KOps/s	$\color{#35bf28}+0.87\%$
test_items_stack_nested_leaf	0.1476ms	80.5904μs	12.4084 KOps/s	12.5911 KOps/s	$\color{#d91a1a}-1.45\%$
test_items_stack_nested_locked	0.5746ms	0.3973ms	2.5168 KOps/s	2.5080 KOps/s	$\color{#35bf28}+0.35\%$
test_keys	41.7480μs	3.6148μs	276.6369 KOps/s	281.9563 KOps/s	$\color{#d91a1a}-1.89\%$
test_keys_nested	0.3431ms	0.1652ms	6.0520 KOps/s	6.0619 KOps/s	$\color{#d91a1a}-0.16\%$
test_keys_nested_locked	1.5809ms	0.1690ms	5.9166 KOps/s	5.8109 KOps/s	$\color{#35bf28}+1.82\%$
test_keys_nested_leaf	0.2125ms	0.1423ms	7.0285 KOps/s	6.8054 KOps/s	$\color{#35bf28}+3.28\%$
test_keys_stack_nested	0.2460ms	0.1621ms	6.1680 KOps/s	5.9053 KOps/s	$\color{#35bf28}+4.45\%$
test_keys_stack_nested_leaf	0.2577ms	0.1381ms	7.2421 KOps/s	6.8006 KOps/s	$\textbf{\color{#35bf28}+6.49\%}$
test_keys_stack_nested_locked	0.3070ms	0.1669ms	5.9923 KOps/s	5.7725 KOps/s	$\color{#35bf28}+3.81\%$
test_values	8.2012μs	1.0339μs	967.2092 KOps/s	935.1485 KOps/s	$\color{#35bf28}+3.43\%$
test_values_nested	0.1055ms	62.1699μs	16.0850 KOps/s	15.9309 KOps/s	$\color{#35bf28}+0.97\%$
test_values_nested_locked	0.1234ms	61.6143μs	16.2300 KOps/s	16.1425 KOps/s	$\color{#35bf28}+0.54\%$
test_values_nested_leaf	0.1376ms	71.6461μs	13.9575 KOps/s	13.9544 KOps/s	$\color{#35bf28}+0.02\%$
test_values_stack_nested	0.1104ms	63.7615μs	15.6834 KOps/s	15.5741 KOps/s	$\color{#35bf28}+0.70\%$
test_values_stack_nested_leaf	0.1273ms	70.4078μs	14.2030 KOps/s	14.0395 KOps/s	$\color{#35bf28}+1.16\%$
test_values_stack_nested_locked	0.1143ms	63.8120μs	15.6710 KOps/s	16.0617 KOps/s	$\color{#d91a1a}-2.43\%$
test_membership	39.7640μs	0.8512μs	1.1749 MOps/s	1.1701 MOps/s	$\color{#35bf28}+0.41\%$
test_membership_nested	22.0710μs	2.9121μs	343.3908 KOps/s	345.1291 KOps/s	$\color{#d91a1a}-0.50\%$
test_membership_nested_leaf	25.5180μs	2.9135μs	343.2312 KOps/s	346.5557 KOps/s	$\color{#d91a1a}-0.96\%$
test_membership_stacked_nested	18.0240μs	2.8925μs	345.7206 KOps/s	352.1376 KOps/s	$\color{#d91a1a}-1.82\%$
test_membership_stacked_nested_leaf	42.0780μs	2.8887μs	346.1737 KOps/s	346.9598 KOps/s	$\color{#d91a1a}-0.23\%$
test_membership_nested_last	27.2510μs	4.3208μs	231.4369 KOps/s	230.9357 KOps/s	$\color{#35bf28}+0.22\%$
test_membership_nested_leaf_last	50.4240μs	4.3017μs	232.4658 KOps/s	227.4073 KOps/s	$\color{#35bf28}+2.22\%$
test_membership_stacked_nested_last	28.2620μs	5.4755μs	182.6309 KOps/s	231.7716 KOps/s	$\textbf{\color{#d91a1a}-21.20\%}$
test_membership_stacked_nested_leaf_last	38.1710μs	5.5448μs	180.3494 KOps/s	231.0055 KOps/s	$\textbf{\color{#d91a1a}-21.93\%}$
test_nested_getleaf	31.4680μs	10.4641μs	95.5645 KOps/s	93.5325 KOps/s	$\color{#35bf28}+2.17\%$
test_nested_get	32.0390μs	10.0610μs	99.3936 KOps/s	97.6411 KOps/s	$\color{#35bf28}+1.79\%$
test_stacked_getleaf	47.9700μs	10.5339μs	94.9320 KOps/s	94.6736 KOps/s	$\color{#35bf28}+0.27\%$
test_stacked_get	32.0200μs	9.9799μs	100.2019 KOps/s	97.9327 KOps/s	$\color{#35bf28}+2.32\%$
test_nested_getitemleaf	36.8290μs	11.1610μs	89.5977 KOps/s	89.1769 KOps/s	$\color{#35bf28}+0.47\%$
test_nested_getitem	61.7350μs	10.6077μs	94.2711 KOps/s	93.4859 KOps/s	$\color{#35bf28}+0.84\%$
test_stacked_getitemleaf	40.5760μs	11.2011μs	89.2768 KOps/s	88.3459 KOps/s	$\color{#35bf28}+1.05\%$
test_stacked_getitem	33.0120μs	10.4946μs	95.2869 KOps/s	93.5732 KOps/s	$\color{#35bf28}+1.83\%$
test_lock_nested	2.8349ms	0.4451ms	2.2465 KOps/s	1.7956 KOps/s	$\textbf{\color{#35bf28}+25.11\%}$
test_lock_stack_nested	0.5510ms	0.4146ms	2.4119 KOps/s	2.3685 KOps/s	$\color{#35bf28}+1.84\%$
test_unlock_nested	0.7380ms	0.3644ms	2.7443 KOps/s	2.6430 KOps/s	$\color{#35bf28}+3.83\%$
test_unlock_stack_nested	0.5265ms	0.3354ms	2.9812 KOps/s	2.8671 KOps/s	$\color{#35bf28}+3.98\%$
test_flatten_speed	0.1863ms	0.1005ms	9.9489 KOps/s	10.0295 KOps/s	$\color{#d91a1a}-0.80\%$
test_unflatten_speed	0.9152ms	0.5209ms	1.9198 KOps/s	1.9178 KOps/s	$\color{#35bf28}+0.10\%$
test_common_ops	4.1285ms	0.7250ms	1.3792 KOps/s	1.2448 KOps/s	$\textbf{\color{#35bf28}+10.79\%}$
test_creation	68.1070μs	2.4244μs	412.4793 KOps/s	407.1937 KOps/s	$\color{#35bf28}+1.30\%$
test_creation_empty	31.5390μs	9.2298μs	108.3446 KOps/s	82.9872 KOps/s	$\textbf{\color{#35bf28}+30.56\%}$
test_creation_nested_1	44.2620μs	12.0716μs	82.8393 KOps/s	68.0355 KOps/s	$\textbf{\color{#35bf28}+21.76\%}$
test_creation_nested_2	50.8140μs	16.4035μs	60.9627 KOps/s	51.5717 KOps/s	$\textbf{\color{#35bf28}+18.21\%}$
test_clone	62.6570μs	13.2463μs	75.4927 KOps/s	75.5107 KOps/s	$\color{#d91a1a}-0.02\%$
test_getitem[int]	1.1971ms	12.6741μs	78.9012 KOps/s	78.7566 KOps/s	$\color{#35bf28}+0.18\%$
test_getitem[slice_int]	0.1442ms	25.7148μs	38.8882 KOps/s	40.9765 KOps/s	$\textbf{\color{#d91a1a}-5.10\%}$
test_getitem[range]	0.1710ms	47.4933μs	21.0556 KOps/s	19.9796 KOps/s	$\textbf{\color{#35bf28}+5.39\%}$
test_getitem[tuple]	0.1247ms	19.9779μs	50.0552 KOps/s	50.3482 KOps/s	$\color{#d91a1a}-0.58\%$
test_getitem[list]	0.1647ms	42.2788μs	23.6525 KOps/s	22.0492 KOps/s	$\textbf{\color{#35bf28}+7.27\%}$
test_setitem_dim[int]	0.1061ms	29.4615μs	33.9426 KOps/s	39.8461 KOps/s	$\textbf{\color{#d91a1a}-14.82\%}$
test_setitem_dim[slice_int]	0.1285ms	51.2792μs	19.5011 KOps/s	19.6614 KOps/s	$\color{#d91a1a}-0.82\%$
test_setitem_dim[range]	0.1557ms	76.6388μs	13.0482 KOps/s	13.0616 KOps/s	$\color{#d91a1a}-0.10\%$
test_setitem_dim[tuple]	85.0490μs	41.0388μs	24.3672 KOps/s	24.7473 KOps/s	$\color{#d91a1a}-1.54\%$
test_setitem	0.2295ms	18.7805μs	53.2466 KOps/s	49.3131 KOps/s	$\textbf{\color{#35bf28}+7.98\%}$
test_set	0.2686ms	17.6882μs	56.5348 KOps/s	49.8732 KOps/s	$\textbf{\color{#35bf28}+13.36\%}$
test_set_shared	3.4013ms	0.1731ms	5.7759 KOps/s	5.8465 KOps/s	$\color{#d91a1a}-1.21\%$
test_update	0.2558ms	19.4309μs	51.4644 KOps/s	42.5604 KOps/s	$\textbf{\color{#35bf28}+20.92\%}$
test_update_nested	0.1779ms	29.6803μs	33.6924 KOps/s	30.2615 KOps/s	$\textbf{\color{#35bf28}+11.34\%}$
test_update__nested	0.5273ms	32.6023μs	30.6727 KOps/s	29.2223 KOps/s	$\color{#35bf28}+4.96\%$
test_set_nested	0.1131ms	20.0701μs	49.8255 KOps/s	44.7273 KOps/s	$\textbf{\color{#35bf28}+11.40\%}$
test_set_nested_new	0.1455ms	24.6838μs	40.5123 KOps/s	37.1776 KOps/s	$\textbf{\color{#35bf28}+8.97\%}$
test_select	0.1978ms	39.9437μs	25.0352 KOps/s	23.1433 KOps/s	$\textbf{\color{#35bf28}+8.17\%}$
test_select_nested	0.1282ms	61.8317μs	16.1729 KOps/s	15.9955 KOps/s	$\color{#35bf28}+1.11\%$
test_exclude_nested	0.1644ms	80.3432μs	12.4466 KOps/s	12.2263 KOps/s	$\color{#35bf28}+1.80\%$
test_empty[True]	0.7583ms	0.4033ms	2.4798 KOps/s	2.4697 KOps/s	$\color{#35bf28}+0.41\%$
test_empty[False]	15.6223μs	1.4503μs	689.5329 KOps/s	727.4528 KOps/s	$\textbf{\color{#d91a1a}-5.21\%}$
test_unbind_speed	0.4233ms	0.2654ms	3.7679 KOps/s	3.7069 KOps/s	$\color{#35bf28}+1.65\%$
test_unbind_speed_stack0	0.4125ms	0.2622ms	3.8144 KOps/s	3.7890 KOps/s	$\color{#35bf28}+0.67\%$
test_unbind_speed_stack1	0.1004s	0.7849ms	1.2741 KOps/s	1.3421 KOps/s	$\textbf{\color{#d91a1a}-5.07\%}$
test_split	1.7050ms	1.5911ms	628.4814 Ops/s	562.5200 Ops/s	$\textbf{\color{#35bf28}+11.73\%}$
test_chunk	0.1046s	1.7683ms	565.5289 Ops/s	553.5126 Ops/s	$\color{#35bf28}+2.17\%$
test_consolidate_njt[False-None]	0.1100s	8.9467ms	111.7726 Ops/s	116.8953 Ops/s	$\color{#d91a1a}-4.38\%$
test_creation[device0]	1.1356ms	98.7338μs	10.1282 KOps/s	10.7902 KOps/s	$\textbf{\color{#d91a1a}-6.14\%}$
test_creation_from_tensor	0.2626ms	94.0790μs	10.6294 KOps/s	10.4616 KOps/s	$\color{#35bf28}+1.60\%$
test_add_one[memmap_tensor0]	0.2423ms	4.9030μs	203.9551 KOps/s	204.3312 KOps/s	$\color{#d91a1a}-0.18\%$
test_contiguous[memmap_tensor0]	18.0340μs	0.5052μs	1.9794 MOps/s	1.9552 MOps/s	$\color{#35bf28}+1.24\%$
test_stack[memmap_tensor0]	62.2360μs	3.3721μs	296.5516 KOps/s	284.4937 KOps/s	$\color{#35bf28}+4.24\%$
test_memmaptd_index	0.9257ms	0.2407ms	4.1550 KOps/s	4.2151 KOps/s	$\color{#d91a1a}-1.43\%$
test_memmaptd_index_astensor	0.5776ms	0.3283ms	3.0457 KOps/s	3.1023 KOps/s	$\color{#d91a1a}-1.82\%$
test_memmaptd_index_op	0.9359ms	0.5497ms	1.8190 KOps/s	1.6794 KOps/s	$\textbf{\color{#35bf28}+8.31\%}$
test_serialize_model	0.1255s	0.1172s	8.5354 Ops/s	8.4530 Ops/s	$\color{#35bf28}+0.97\%$
test_serialize_model_pickle	0.4958s	0.4062s	2.4619 Ops/s	2.5283 Ops/s	$\color{#d91a1a}-2.63\%$
test_serialize_weights	0.1280s	0.1147s	8.7185 Ops/s	8.2904 Ops/s	$\textbf{\color{#35bf28}+5.16\%}$
test_serialize_weights_returnearly	0.2786s	0.1733s	5.7716 Ops/s	6.0828 Ops/s	$\textbf{\color{#d91a1a}-5.12\%}$
test_serialize_weights_pickle	0.9466s	0.6547s	1.5275 Ops/s	2.4645 Ops/s	$\textbf{\color{#d91a1a}-38.02\%}$
test_serialize_weights_filesystem	0.1494s	0.1417s	7.0586 Ops/s	6.7828 Ops/s	$\color{#35bf28}+4.07\%$
test_serialize_model_filesystem	0.1509s	0.1422s	7.0322 Ops/s	6.3100 Ops/s	$\textbf{\color{#35bf28}+11.44\%}$
test_reshape_pytree	71.4430μs	26.6937μs	37.4621 KOps/s	38.1798 KOps/s	$\color{#d91a1a}-1.88\%$
test_reshape_td	80.7410μs	32.3427μs	30.9189 KOps/s	29.6118 KOps/s	$\color{#35bf28}+4.41\%$
test_view_pytree	68.3680μs	26.5107μs	37.7207 KOps/s	38.3352 KOps/s	$\color{#d91a1a}-1.60\%$
test_view_td	76.0320μs	36.7952μs	27.1775 KOps/s	25.9760 KOps/s	$\color{#35bf28}+4.63\%$
test_unbind_pytree	72.7650μs	29.4928μs	33.9066 KOps/s	34.2937 KOps/s	$\color{#d91a1a}-1.13\%$
test_unbind_td	0.3240ms	39.2208μs	25.4967 KOps/s	24.9025 KOps/s	$\color{#35bf28}+2.39\%$
test_split_pytree	66.2140μs	29.1843μs	34.2650 KOps/s	34.7427 KOps/s	$\color{#d91a1a}-1.38\%$
test_split_td	0.2199ms	44.9353μs	22.2542 KOps/s	22.0428 KOps/s	$\color{#35bf28}+0.96\%$
test_add_pytree	75.4310μs	35.2184μs	28.3942 KOps/s	28.1077 KOps/s	$\color{#35bf28}+1.02\%$
test_add_td	0.2315ms	48.6175μs	20.5687 KOps/s	17.0883 KOps/s	$\textbf{\color{#35bf28}+20.37\%}$
test_compile_add_one_nested[tensordict-compile]	0.1403ms	63.9582μs	15.6352 KOps/s	15.5931 KOps/s	$\color{#35bf28}+0.27\%$
test_compile_add_one_nested[tensordict-eager]	0.3651ms	0.1734ms	5.7685 KOps/s	5.6747 KOps/s	$\color{#35bf28}+1.65\%$
test_compile_add_one_nested[pytree-compile]	0.1109ms	45.2270μs	22.1107 KOps/s	21.5888 KOps/s	$\color{#35bf28}+2.42\%$
test_compile_add_one_nested[pytree-eager]	0.2264ms	0.1180ms	8.4748 KOps/s	8.4129 KOps/s	$\color{#35bf28}+0.74\%$
test_compile_copy_nested[tensordict-compile]	65.9120μs	27.7026μs	36.0977 KOps/s	37.7600 KOps/s	$\color{#d91a1a}-4.40\%$
test_compile_copy_nested[tensordict-eager]	0.1140ms	57.9468μs	17.2572 KOps/s	17.2060 KOps/s	$\color{#35bf28}+0.30\%$
test_compile_copy_nested[pytree-compile]	0.1422ms	78.3828μs	12.7579 KOps/s	12.7352 KOps/s	$\color{#35bf28}+0.18\%$
test_compile_copy_nested[pytree-eager]	0.1447ms	66.8692μs	14.9546 KOps/s	15.0632 KOps/s	$\color{#d91a1a}-0.72\%$
test_compile_add_one_flat[tensordict-compile]	0.2039ms	0.1030ms	9.7114 KOps/s	9.5063 KOps/s	$\color{#35bf28}+2.16\%$
test_compile_add_one_flat[tensordict-eager]	0.3933ms	0.2130ms	4.6944 KOps/s	4.6113 KOps/s	$\color{#35bf28}+1.80\%$
test_compile_add_one_flat[tensorclass-compile]	93.7740μs	44.6216μs	22.4107 KOps/s	21.2872 KOps/s	$\textbf{\color{#35bf28}+5.28\%}$
test_compile_add_one_flat[tensorclass-eager]	0.5298ms	67.1349μs	14.8954 KOps/s	14.9340 KOps/s	$\color{#d91a1a}-0.26\%$
test_compile_add_one_flat[pytree-compile]	0.1858ms	0.1017ms	9.8370 KOps/s	9.7981 KOps/s	$\color{#35bf28}+0.40\%$
test_compile_add_one_flat[pytree-eager]	0.5344ms	0.2057ms	4.8626 KOps/s	4.9617 KOps/s	$\color{#d91a1a}-2.00\%$
test_compile_add_self_flat[tensordict-eager]	0.3461ms	0.2281ms	4.3836 KOps/s	4.3055 KOps/s	$\color{#35bf28}+1.81\%$
test_compile_add_self_flat[tensordict-compile]	0.2022ms	0.1038ms	9.6318 KOps/s	9.4643 KOps/s	$\color{#35bf28}+1.77\%$
test_compile_add_self_flat[tensorclass-eager]	0.1393ms	62.6789μs	15.9543 KOps/s	15.6304 KOps/s	$\color{#35bf28}+2.07\%$
test_compile_add_self_flat[tensorclass-compile]	0.1862ms	46.3100μs	21.5936 KOps/s	21.0279 KOps/s	$\color{#35bf28}+2.69\%$
test_compile_add_self_flat[pytree-eager]	0.3310ms	0.1607ms	6.2237 KOps/s	6.3481 KOps/s	$\color{#d91a1a}-1.96\%$
test_compile_add_self_flat[pytree-compile]	0.2380ms	0.1016ms	9.8459 KOps/s	9.2365 KOps/s	$\textbf{\color{#35bf28}+6.60\%}$
test_compile_copy_flat[tensordict-compile]	56.0840μs	20.8316μs	48.0041 KOps/s	46.4140 KOps/s	$\color{#35bf28}+3.43\%$
test_compile_copy_flat[tensordict-eager]	0.1471ms	66.4664μs	15.0452 KOps/s	15.1233 KOps/s	$\color{#d91a1a}-0.52\%$
test_compile_copy_flat[pytree-compile]	0.1921ms	78.3207μs	12.7680 KOps/s	12.8555 KOps/s	$\color{#d91a1a}-0.68\%$
test_compile_copy_flat[pytree-eager]	0.1507ms	70.0174μs	14.2822 KOps/s	15.0500 KOps/s	$\textbf{\color{#d91a1a}-5.10\%}$
test_compile_assign_and_add[tensordict-compile]	0.4258ms	0.2038ms	4.9076 KOps/s	4.7862 KOps/s	$\color{#35bf28}+2.54\%$
test_compile_assign_and_add[tensordict-eager]	1.4978ms	1.3228ms	755.9814 Ops/s	761.2707 Ops/s	$\color{#d91a1a}-0.69\%$
test_compile_assign_and_add[pytree-compile]	0.3550ms	0.2025ms	4.9380 KOps/s	4.9297 KOps/s	$\color{#35bf28}+0.17\%$
test_compile_assign_and_add[pytree-eager]	1.3675ms	0.7824ms	1.2781 KOps/s	1.2762 KOps/s	$\color{#35bf28}+0.15\%$
test_compile_assign_and_add_stack[compile]	0.9117ms	0.4538ms	2.2035 KOps/s	2.2219 KOps/s	$\color{#d91a1a}-0.83\%$
test_compile_assign_and_add_stack[eager]	2.8119ms	2.5317ms	394.9970 Ops/s	375.0284 Ops/s	$\textbf{\color{#35bf28}+5.32\%}$
test_compile_indexing[tensor-tensordict-compile]	0.1026ms	35.5789μs	28.1065 KOps/s	27.8206 KOps/s	$\color{#35bf28}+1.03\%$
test_compile_indexing[tensor-tensordict-eager]	0.7355ms	32.1890μs	31.0665 KOps/s	30.5857 KOps/s	$\color{#35bf28}+1.57\%$
test_compile_indexing[tensor-tensorclass-compile]	0.1133ms	28.8902μs	34.6138 KOps/s	34.4499 KOps/s	$\color{#35bf28}+0.48\%$
test_compile_indexing[tensor-tensorclass-eager]	0.1051ms	23.7953μs	42.0251 KOps/s	42.8636 KOps/s	$\color{#d91a1a}-1.96\%$
test_compile_indexing[tensor-pytree-compile]	0.1110ms	30.0826μs	33.2419 KOps/s	33.8588 KOps/s	$\color{#d91a1a}-1.82\%$
test_compile_indexing[tensor-pytree-eager]	88.8160μs	23.8272μs	41.9688 KOps/s	43.2496 KOps/s	$\color{#d91a1a}-2.96\%$
test_compile_indexing[slice-tensordict-compile]	0.1167ms	51.3922μs	19.4582 KOps/s	19.5767 KOps/s	$\color{#d91a1a}-0.61\%$
test_compile_indexing[slice-tensordict-eager]	0.6112ms	20.0957μs	49.7618 KOps/s	50.8750 KOps/s	$\color{#d91a1a}-2.19\%$
test_compile_indexing[slice-tensorclass-compile]	0.1523ms	43.4660μs	23.0065 KOps/s	22.8379 KOps/s	$\color{#35bf28}+0.74\%$
test_compile_indexing[slice-tensorclass-eager]	68.2670μs	18.6602μs	53.5901 KOps/s	54.5147 KOps/s	$\color{#d91a1a}-1.70\%$
test_compile_indexing[slice-pytree-compile]	0.1206ms	44.2922μs	22.5774 KOps/s	22.4712 KOps/s	$\color{#35bf28}+0.47\%$
test_compile_indexing[slice-pytree-eager]	77.4850μs	18.7374μs	53.3692 KOps/s	55.2233 KOps/s	$\color{#d91a1a}-3.36\%$
test_compile_indexing[int-tensordict-compile]	0.1204ms	51.5452μs	19.4004 KOps/s	19.3766 KOps/s	$\color{#35bf28}+0.12\%$
test_compile_indexing[int-tensordict-eager]	1.0121ms	19.8254μs	50.4403 KOps/s	51.4784 KOps/s	$\color{#d91a1a}-2.02\%$
test_compile_indexing[int-tensorclass-compile]	0.1251ms	44.5121μs	22.4658 KOps/s	22.4557 KOps/s	$\color{#35bf28}+0.05\%$
test_compile_indexing[int-tensorclass-eager]	70.3810μs	18.6548μs	53.6055 KOps/s	55.0752 KOps/s	$\color{#d91a1a}-2.67\%$
test_compile_indexing[int-pytree-compile]	0.1460ms	44.6778μs	22.3825 KOps/s	22.3916 KOps/s	$\color{#d91a1a}-0.04\%$
test_compile_indexing[int-pytree-eager]	91.2000μs	18.7183μs	53.4237 KOps/s	55.2872 KOps/s	$\color{#d91a1a}-3.37\%$
test_mod_add[eager]	0.1046ms	32.7530μs	30.5315 KOps/s	28.9307 KOps/s	$\textbf{\color{#35bf28}+5.53\%}$
test_mod_add[compile]	0.1713ms	48.3622μs	20.6773 KOps/s	20.8661 KOps/s	$\color{#d91a1a}-0.90\%$
test_mod_add[compile-overhead]	0.1071ms	47.5685μs	21.0223 KOps/s	20.8312 KOps/s	$\color{#35bf28}+0.92\%$
test_mod_wrap[eager]	0.4578ms	0.2235ms	4.4746 KOps/s	4.5807 KOps/s	$\color{#d91a1a}-2.32\%$
test_mod_wrap[compile]	0.4107ms	0.2054ms	4.8679 KOps/s	4.8043 KOps/s	$\color{#35bf28}+1.32\%$
test_mod_wrap[compile-overhead]	0.3845ms	0.2057ms	4.8607 KOps/s	4.9197 KOps/s	$\color{#d91a1a}-1.20\%$
test_mod_wrap_and_backward[eager]	14.5899ms	12.4329ms	80.4320 Ops/s	75.3865 Ops/s	$\textbf{\color{#35bf28}+6.69\%}$
test_mod_wrap_and_backward[compile]	19.2336ms	13.4903ms	74.1275 Ops/s	74.3224 Ops/s	$\color{#d91a1a}-0.26\%$
test_mod_wrap_and_backward[compile-overhead]	18.2843ms	12.9396ms	77.2821 Ops/s	74.6303 Ops/s	$\color{#35bf28}+3.55\%$
test_seq_add[eager]	0.2544ms	0.1136ms	8.8006 KOps/s	8.6231 KOps/s	$\color{#35bf28}+2.06\%$
test_seq_add[compile]	0.1491ms	63.0168μs	15.8688 KOps/s	15.5775 KOps/s	$\color{#35bf28}+1.87\%$
test_seq_add[compile-overhead]	0.1380ms	59.9933μs	16.6685 KOps/s	16.1987 KOps/s	$\color{#35bf28}+2.90\%$
test_seq_wrap[eager]	0.6997ms	0.4289ms	2.3314 KOps/s	2.2154 KOps/s	$\textbf{\color{#35bf28}+5.23\%}$
test_seq_wrap[compile]	0.6885ms	0.2270ms	4.4046 KOps/s	4.2894 KOps/s	$\color{#35bf28}+2.68\%$
test_seq_wrap[compile-overhead]	0.3584ms	0.2231ms	4.4821 KOps/s	4.3375 KOps/s	$\color{#35bf28}+3.33\%$
test_func_call_runtime[False-eager]	0.9674ms	0.5433ms	1.8405 KOps/s	1.8516 KOps/s	$\color{#d91a1a}-0.60\%$
test_func_call_runtime[False-compile]	0.7493ms	0.4207ms	2.3772 KOps/s	2.3098 KOps/s	$\color{#35bf28}+2.92\%$
test_func_call_runtime[False-compile-overhead]	0.5243ms	0.4202ms	2.3801 KOps/s	2.3776 KOps/s	$\color{#35bf28}+0.10\%$
test_func_call_runtime[True-eager]	0.9001ms	0.7526ms	1.3287 KOps/s	1.3114 KOps/s	$\color{#35bf28}+1.32\%$
test_func_call_runtime[True-compile]	0.6662ms	0.4582ms	2.1824 KOps/s	2.1514 KOps/s	$\color{#35bf28}+1.44\%$
test_func_call_runtime[True-compile-overhead]	0.6915ms	0.4621ms	2.1639 KOps/s	2.1551 KOps/s	$\color{#35bf28}+0.41\%$
test_func_call_cm_runtime[False-eager]	0.8352ms	0.5448ms	1.8356 KOps/s	1.8714 KOps/s	$\color{#d91a1a}-1.91\%$
test_func_call_cm_runtime[False-compile]	1.1342ms	0.4268ms	2.3430 KOps/s	2.3689 KOps/s	$\color{#d91a1a}-1.09\%$
test_func_call_cm_runtime[False-compile-overhead]	0.5704ms	0.4199ms	2.3817 KOps/s	2.3810 KOps/s	$\color{#35bf28}+0.03\%$
test_func_call_cm_runtime[True-eager]	1.4407ms	0.9011ms	1.1097 KOps/s	1.1180 KOps/s	$\color{#d91a1a}-0.74\%$
test_func_call_cm_runtime[True-compile]	0.5941ms	0.4835ms	2.0682 KOps/s	2.0697 KOps/s	$\color{#d91a1a}-0.07\%$
test_func_call_cm_runtime[True-compile-overhead]	0.6282ms	0.4826ms	2.0723 KOps/s	2.0830 KOps/s	$\color{#d91a1a}-0.51\%$
test_vmap_func_call_cm_runtime[eager]	2.6072ms	1.8896ms	529.2084 Ops/s	525.2757 Ops/s	$\color{#35bf28}+0.75\%$
test_vmap_func_call_cm_runtime[compile]	0.6862ms	0.5187ms	1.9279 KOps/s	1.9381 KOps/s	$\color{#d91a1a}-0.53\%$
test_vmap_func_call_cm_runtime[compile-overhead]	0.7174ms	0.5209ms	1.9198 KOps/s	1.9292 KOps/s	$\color{#d91a1a}-0.49\%$
test_distributed	0.3668ms	0.1254ms	7.9755 KOps/s	7.8023 KOps/s	$\color{#35bf28}+2.22\%$
test_tdmodule	75.8910μs	24.3432μs	41.0792 KOps/s	37.7461 KOps/s	$\textbf{\color{#35bf28}+8.83\%}$
test_tdmodule_dispatch	72.5150μs	43.5911μs	22.9404 KOps/s	20.9871 KOps/s	$\textbf{\color{#35bf28}+9.31\%}$
test_tdseq	47.0670μs	26.2869μs	38.0418 KOps/s	35.0551 KOps/s	$\textbf{\color{#35bf28}+8.52\%}$
test_tdseq_dispatch	95.4770μs	49.4920μs	20.2053 KOps/s	18.5958 KOps/s	$\textbf{\color{#35bf28}+8.66\%}$
test_instantiation_functorch	3.0774ms	1.5121ms	661.3475 Ops/s	655.8504 Ops/s	$\color{#35bf28}+0.84\%$
test_exec_functorch	0.3247ms	0.1810ms	5.5254 KOps/s	5.6397 KOps/s	$\color{#d91a1a}-2.03\%$
test_exec_functional_call	0.2956ms	0.1693ms	5.9078 KOps/s	5.8658 KOps/s	$\color{#35bf28}+0.72\%$
test_exec_td_decorator	0.5058ms	0.2276ms	4.3936 KOps/s	4.3938 KOps/s	$-0.00\%$
test_vmap_mlp_speed_decorator[True-True]	0.9553ms	0.6473ms	1.5449 KOps/s	1.4877 KOps/s	$\color{#35bf28}+3.85\%$
test_vmap_mlp_speed_decorator[True-False]	1.3060ms	0.6493ms	1.5402 KOps/s	1.5430 KOps/s	$\color{#d91a1a}-0.18\%$
test_vmap_mlp_speed_decorator[False-True]	0.8832ms	0.5303ms	1.8857 KOps/s	1.8986 KOps/s	$\color{#d91a1a}-0.68\%$
test_vmap_mlp_speed_decorator[False-False]	1.1113ms	0.5342ms	1.8721 KOps/s	1.9072 KOps/s	$\color{#d91a1a}-1.84\%$
test_to_module_speed[True]	1.7597ms	1.3357ms	748.6711 Ops/s	736.1793 Ops/s	$\color{#35bf28}+1.70\%$
test_to_module_speed[False]	1.9305ms	1.3032ms	767.3161 Ops/s	743.6382 Ops/s	$\color{#35bf28}+3.18\%$
test_tc_init	73.6770μs	41.5814μs	24.0492 KOps/s	21.4711 KOps/s	$\textbf{\color{#35bf28}+12.01\%}$
test_tc_init_nested	0.1656ms	81.6932μs	12.2409 KOps/s	10.2505 KOps/s	$\textbf{\color{#35bf28}+19.42\%}$
test_tc_first_layer_tensor	15.7790μs	1.5236μs	656.3321 KOps/s	661.4717 KOps/s	$\color{#d91a1a}-0.78\%$
test_tc_first_layer_nontensor	26.7690μs	4.6510μs	215.0060 KOps/s	217.0874 KOps/s	$\color{#d91a1a}-0.96\%$
test_tc_second_layer_tensor	43.4740μs	2.7661μs	361.5186 KOps/s	355.1725 KOps/s	$\color{#35bf28}+1.79\%$
test_tc_second_layer_nontensor	32.0900μs	5.9967μs	166.7587 KOps/s	168.2494 KOps/s	$\color{#d91a1a}-0.89\%$
test_unbind	0.2169s	13.3926ms	74.6682 Ops/s	78.2489 Ops/s	$\color{#d91a1a}-4.58\%$
test_full_like	8.4709ms	7.1128ms	140.5907 Ops/s	81.4874 Ops/s	$\textbf{\color{#35bf28}+72.53\%}$
test_zeros_like	3.2627ms	2.7222ms	367.3523 Ops/s	138.3856 Ops/s	$\textbf{\color{#35bf28}+165.46\%}$
test_ones_like	3.7025ms	3.2233ms	310.2372 Ops/s	133.2428 Ops/s	$\textbf{\color{#35bf28}+132.84\%}$
test_clone	8.5705ms	5.2735ms	189.6256 Ops/s	110.1729 Ops/s	$\textbf{\color{#35bf28}+72.12\%}$
test_squeeze	81.8930μs	11.9466μs	83.7061 KOps/s	83.7175 KOps/s	$\color{#d91a1a}-0.01\%$
test_unsqueeze	0.2361ms	88.2136μs	11.3361 KOps/s	10.7024 KOps/s	$\textbf{\color{#35bf28}+5.92\%}$
test_split	0.3615ms	0.1909ms	5.2392 KOps/s	5.1914 KOps/s	$\color{#35bf28}+0.92\%$
test_permute	0.3863ms	0.2003ms	4.9936 KOps/s	4.9049 KOps/s	$\color{#35bf28}+1.81\%$
test_stack	29.2729ms	25.3166ms	39.4997 Ops/s	39.8566 Ops/s	$\color{#d91a1a}-0.90\%$
test_cat	28.5507ms	24.7456ms	40.4112 Ops/s	40.0578 Ops/s	$\color{#35bf28}+0.88\%$

[github-actions]

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

Total Benchmarks: 229. Improved: $\large\color{#35bf28}33$. Worsened: $\large\color{#d91a1a}14$.

Expand to view detailed results

Name	Max	Mean	Ops	Ops on Repo HEAD	Change
test_plain_set_nested	34.7220μs	11.2547μs	88.8519 KOps/s	79.5299 KOps/s	$\textbf{\color{#35bf28}+11.72\%}$
test_plain_set_stack_nested	35.5320μs	11.4421μs	87.3966 KOps/s	78.4291 KOps/s	$\textbf{\color{#35bf28}+11.43\%}$
test_plain_set_nested_inplace	40.2020μs	12.2643μs	81.5378 KOps/s	72.4528 KOps/s	$\textbf{\color{#35bf28}+12.54\%}$
test_plain_set_stack_nested_inplace	0.1095ms	12.3334μs	81.0804 KOps/s	73.2179 KOps/s	$\textbf{\color{#35bf28}+10.74\%}$
test_items	30.1120μs	2.8884μs	346.2111 KOps/s	344.0520 KOps/s	$\color{#35bf28}+0.63\%$
test_items_nested	0.4092ms	0.3599ms	2.7788 KOps/s	2.7740 KOps/s	$\color{#35bf28}+0.17\%$
test_items_nested_locked	0.4159ms	0.3599ms	2.7787 KOps/s	2.7717 KOps/s	$\color{#35bf28}+0.25\%$
test_items_nested_leaf	79.9440μs	57.8218μs	17.2945 KOps/s	17.2955 KOps/s	$-0.01\%$
test_items_stack_nested	0.4285ms	0.3617ms	2.7644 KOps/s	2.7583 KOps/s	$\color{#35bf28}+0.22\%$
test_items_stack_nested_leaf	86.8750μs	58.6112μs	17.0616 KOps/s	16.8370 KOps/s	$\color{#35bf28}+1.33\%$
test_items_stack_nested_locked	0.4368ms	0.3639ms	2.7483 KOps/s	2.7455 KOps/s	$\color{#35bf28}+0.10\%$
test_keys	32.3810μs	3.6075μs	277.2020 KOps/s	292.3105 KOps/s	$\textbf{\color{#d91a1a}-5.17\%}$
test_keys_nested	0.1185ms	88.1540μs	11.3438 KOps/s	11.4063 KOps/s	$\color{#d91a1a}-0.55\%$
test_keys_nested_locked	0.8121ms	93.7005μs	10.6723 KOps/s	10.7166 KOps/s	$\color{#d91a1a}-0.41\%$
test_keys_nested_leaf	0.1156ms	78.4276μs	12.7506 KOps/s	12.8466 KOps/s	$\color{#d91a1a}-0.75\%$
test_keys_stack_nested	0.1146ms	90.5310μs	11.0459 KOps/s	11.2482 KOps/s	$\color{#d91a1a}-1.80\%$
test_keys_stack_nested_leaf	0.1295ms	80.2567μs	12.4600 KOps/s	12.3777 KOps/s	$\color{#35bf28}+0.66\%$
test_keys_stack_nested_locked	0.1217ms	95.1842μs	10.5059 KOps/s	10.4785 KOps/s	$\color{#35bf28}+0.26\%$
test_values	7.9320μs	0.8489μs	1.1780 MOps/s	1.1798 MOps/s	$\color{#d91a1a}-0.16\%$
test_values_nested	71.9240μs	37.5900μs	26.6028 KOps/s	26.8613 KOps/s	$\color{#d91a1a}-0.96\%$
test_values_nested_locked	0.3354ms	39.3178μs	25.4338 KOps/s	25.6168 KOps/s	$\color{#d91a1a}-0.71\%$
test_values_nested_leaf	78.7540μs	41.9900μs	23.8152 KOps/s	24.0621 KOps/s	$\color{#d91a1a}-1.03\%$
test_values_stack_nested	73.0640μs	38.1301μs	26.2260 KOps/s	26.3485 KOps/s	$\color{#d91a1a}-0.46\%$
test_values_stack_nested_leaf	77.7640μs	42.3154μs	23.6321 KOps/s	23.7699 KOps/s	$\color{#d91a1a}-0.58\%$
test_values_stack_nested_locked	74.6140μs	40.0397μs	24.9752 KOps/s	25.0839 KOps/s	$\color{#d91a1a}-0.43\%$
test_membership	1.7366μs	0.5012μs	1.9951 MOps/s	1.9530 MOps/s	$\color{#35bf28}+2.16\%$
test_membership_nested	33.3320μs	2.0376μs	490.7641 KOps/s	505.7721 KOps/s	$\color{#d91a1a}-2.97\%$
test_membership_nested_leaf	15.2955μs	2.0040μs	499.0048 KOps/s	504.7272 KOps/s	$\color{#d91a1a}-1.13\%$
test_membership_stacked_nested	32.5120μs	2.0303μs	492.5351 KOps/s	485.0938 KOps/s	$\color{#35bf28}+1.53\%$
test_membership_stacked_nested_leaf	40.1220μs	2.0480μs	488.2922 KOps/s	488.7442 KOps/s	$\color{#d91a1a}-0.09\%$
test_membership_nested_last	35.7520μs	2.9939μs	334.0149 KOps/s	330.8566 KOps/s	$\color{#35bf28}+0.95\%$
test_membership_nested_leaf_last	27.6020μs	2.9744μs	336.2056 KOps/s	332.4787 KOps/s	$\color{#35bf28}+1.12\%$
test_membership_stacked_nested_last	28.8820μs	2.9894μs	334.5109 KOps/s	281.4104 KOps/s	$\textbf{\color{#35bf28}+18.87\%}$
test_membership_stacked_nested_leaf_last	39.5520μs	2.9966μs	333.7097 KOps/s	286.5069 KOps/s	$\textbf{\color{#35bf28}+16.48\%}$
test_nested_getleaf	42.1320μs	6.0478μs	165.3481 KOps/s	164.7907 KOps/s	$\color{#35bf28}+0.34\%$
test_nested_get	34.4620μs	5.7491μs	173.9409 KOps/s	174.0153 KOps/s	$\color{#d91a1a}-0.04\%$
test_stacked_getleaf	33.4820μs	6.0815μs	164.4337 KOps/s	163.6704 KOps/s	$\color{#35bf28}+0.47\%$
test_stacked_get	41.9020μs	5.7690μs	173.3393 KOps/s	172.5653 KOps/s	$\color{#35bf28}+0.45\%$
test_nested_getitemleaf	36.5120μs	6.4218μs	155.7201 KOps/s	155.6618 KOps/s	$\color{#35bf28}+0.04\%$
test_nested_getitem	31.7720μs	6.0743μs	164.6270 KOps/s	166.0551 KOps/s	$\color{#d91a1a}-0.86\%$
test_stacked_getitemleaf	45.7520μs	6.3758μs	156.8427 KOps/s	156.1297 KOps/s	$\color{#35bf28}+0.46\%$
test_stacked_getitem	38.5220μs	6.1000μs	163.9347 KOps/s	165.2739 KOps/s	$\color{#d91a1a}-0.81\%$
test_lock_nested	1.1250ms	0.3735ms	2.6772 KOps/s	2.6635 KOps/s	$\color{#35bf28}+0.52\%$
test_lock_stack_nested	0.4705ms	0.3474ms	2.8788 KOps/s	2.9346 KOps/s	$\color{#d91a1a}-1.90\%$
test_unlock_nested	0.6944ms	0.3183ms	3.1416 KOps/s	3.2518 KOps/s	$\color{#d91a1a}-3.39\%$
test_unlock_stack_nested	0.3262ms	0.2846ms	3.5140 KOps/s	3.5886 KOps/s	$\color{#d91a1a}-2.08\%$
test_flatten_speed	0.1157ms	74.7613μs	13.3759 KOps/s	13.4021 KOps/s	$\color{#d91a1a}-0.20\%$
test_unflatten_speed	0.3687ms	0.3195ms	3.1295 KOps/s	3.1639 KOps/s	$\color{#d91a1a}-1.09\%$
test_common_ops	1.6446ms	0.5585ms	1.7904 KOps/s	1.6355 KOps/s	$\textbf{\color{#35bf28}+9.47\%}$
test_creation	0.1715ms	1.7096μs	584.9290 KOps/s	590.7755 KOps/s	$\color{#d91a1a}-0.99\%$
test_creation_empty	38.5320μs	6.2687μs	159.5231 KOps/s	114.3177 KOps/s	$\textbf{\color{#35bf28}+39.54\%}$
test_creation_nested_1	35.2020μs	7.9421μs	125.9106 KOps/s	95.3676 KOps/s	$\textbf{\color{#35bf28}+32.03\%}$
test_creation_nested_2	46.3720μs	10.7329μs	93.1714 KOps/s	76.7714 KOps/s	$\textbf{\color{#35bf28}+21.36\%}$
test_clone	0.1243ms	10.2347μs	97.7065 KOps/s	100.3034 KOps/s	$\color{#d91a1a}-2.59\%$
test_getitem[int]	1.9867ms	11.0920μs	90.1549 KOps/s	93.2923 KOps/s	$\color{#d91a1a}-3.36\%$
test_getitem[slice_int]	0.1155ms	21.8072μs	45.8564 KOps/s	49.1642 KOps/s	$\textbf{\color{#d91a1a}-6.73\%}$
test_getitem[range]	0.1327ms	37.4109μs	26.7302 KOps/s	28.2197 KOps/s	$\textbf{\color{#d91a1a}-5.28\%}$
test_getitem[tuple]	0.1112ms	18.6891μs	53.5071 KOps/s	55.1818 KOps/s	$\color{#d91a1a}-3.03\%$
test_getitem[list]	0.1313ms	32.5971μs	30.6776 KOps/s	31.5977 KOps/s	$\color{#d91a1a}-2.91\%$
test_setitem_dim[int]	40.8120μs	19.1747μs	52.1520 KOps/s	54.4386 KOps/s	$\color{#d91a1a}-4.20\%$
test_setitem_dim[slice_int]	61.5030μs	38.0382μs	26.2893 KOps/s	27.2815 KOps/s	$\color{#d91a1a}-3.64\%$
test_setitem_dim[range]	77.2740μs	51.9584μs	19.2462 KOps/s	19.7108 KOps/s	$\color{#d91a1a}-2.36\%$
test_setitem_dim[tuple]	65.8040μs	32.4320μs	30.8338 KOps/s	32.2009 KOps/s	$\color{#d91a1a}-4.25\%$
test_setitem	45.3520μs	13.5758μs	73.6605 KOps/s	67.7661 KOps/s	$\textbf{\color{#35bf28}+8.70\%}$
test_set	0.1326ms	12.9779μs	77.0540 KOps/s	70.0116 KOps/s	$\textbf{\color{#35bf28}+10.06\%}$
test_set_shared	1.4527ms	0.1489ms	6.7160 KOps/s	6.6982 KOps/s	$\color{#35bf28}+0.27\%$
test_update	0.8622ms	15.0018μs	66.6586 KOps/s	57.0089 KOps/s	$\textbf{\color{#35bf28}+16.93\%}$
test_update_nested	0.1321ms	20.3521μs	49.1350 KOps/s	43.3763 KOps/s	$\textbf{\color{#35bf28}+13.28\%}$
test_update__nested	1.3175ms	24.9415μs	40.0939 KOps/s	41.2102 KOps/s	$\color{#d91a1a}-2.71\%$
test_set_nested	0.1255ms	14.5429μs	68.7622 KOps/s	65.2803 KOps/s	$\textbf{\color{#35bf28}+5.33\%}$
test_set_nested_new	0.1246ms	16.9654μs	58.9436 KOps/s	56.5187 KOps/s	$\color{#35bf28}+4.29\%$
test_select	0.1413ms	28.6981μs	34.8455 KOps/s	33.5196 KOps/s	$\color{#35bf28}+3.96\%$
test_select_nested	75.3140μs	43.6877μs	22.8897 KOps/s	22.7982 KOps/s	$\color{#35bf28}+0.40\%$
test_exclude_nested	92.6050μs	61.5834μs	16.2381 KOps/s	16.0411 KOps/s	$\color{#35bf28}+1.23\%$
test_empty[True]	0.3491ms	0.2932ms	3.4107 KOps/s	3.4376 KOps/s	$\color{#d91a1a}-0.78\%$
test_empty[False]	4.5782μs	0.8221μs	1.2164 MOps/s	1.2179 MOps/s	$\color{#d91a1a}-0.12\%$
test_to	84.7050μs	56.2844μs	17.7669 KOps/s	17.1846 KOps/s	$\color{#35bf28}+3.39\%$
test_to_nonblocking	0.1200ms	46.2809μs	21.6072 KOps/s	21.1275 KOps/s	$\color{#35bf28}+2.27\%$
test_unbind_speed	0.8247ms	0.2412ms	4.1468 KOps/s	4.2411 KOps/s	$\color{#d91a1a}-2.22\%$
test_unbind_speed_stack0	0.2915ms	0.2393ms	4.1780 KOps/s	4.2398 KOps/s	$\color{#d91a1a}-1.46\%$
test_unbind_speed_stack1	0.6728ms	0.6182ms	1.6176 KOps/s	1.5049 KOps/s	$\textbf{\color{#35bf28}+7.49\%}$
test_split	95.1560ms	1.6468ms	607.2350 Ops/s	622.4164 Ops/s	$\color{#d91a1a}-2.44\%$
test_chunk	97.0090ms	1.6411ms	609.3517 Ops/s	620.3660 Ops/s	$\color{#d91a1a}-1.78\%$
test_consolidate[False-None]	97.0805ms	2.8843ms	346.7095 Ops/s	341.4397 Ops/s	$\color{#35bf28}+1.54\%$
test_consolidate[default-None]	1.8007ms	1.7126ms	583.8972 Ops/s	601.5385 Ops/s	$\color{#d91a1a}-2.93\%$
test_consolidate[reduce-overhead-None]	1.8650ms	1.7555ms	569.6533 Ops/s	590.4789 Ops/s	$\color{#d91a1a}-3.53\%$
test_consolidate_njt[False-None]	6.7159ms	6.4852ms	154.1962 Ops/s	156.0760 Ops/s	$\color{#d91a1a}-1.20\%$
test_to[False-False-None]	1.8143ms	1.7010ms	587.8758 Ops/s	589.0270 Ops/s	$\color{#d91a1a}-0.20\%$
test_to[True-False-None]	1.4956ms	1.2986ms	770.0760 Ops/s	796.2186 Ops/s	$\color{#d91a1a}-3.28\%$
test_to[within-False-None]	4.3647ms	4.1011ms	243.8373 Ops/s	178.4825 Ops/s	$\textbf{\color{#35bf28}+36.62\%}$
test_to[True-default-None]	5.5659ms	5.3393ms	187.2896 Ops/s	193.9373 Ops/s	$\color{#d91a1a}-3.43\%$
test_to_njt[False-False-None]	7.0973ms	6.7955ms	147.1556 Ops/s	143.1369 Ops/s	$\color{#35bf28}+2.81\%$
test_to_njt[True-False-None]	5.6958ms	5.4001ms	185.1833 Ops/s	181.9411 Ops/s	$\color{#35bf28}+1.78\%$
test_to_njt[within-False-None]	12.4818ms	12.1266ms	82.4631 Ops/s	82.8275 Ops/s	$\color{#d91a1a}-0.44\%$
test_creation[device0]	0.6185ms	80.3841μs	12.4403 KOps/s	12.1822 KOps/s	$\color{#35bf28}+2.12\%$
test_creation_from_tensor	0.5955ms	84.5659μs	11.8251 KOps/s	11.6993 KOps/s	$\color{#35bf28}+1.07\%$
test_add_one[memmap_tensor0]	0.4075ms	6.3254μs	158.0937 KOps/s	160.5520 KOps/s	$\color{#d91a1a}-1.53\%$
test_contiguous[memmap_tensor0]	2.1076μs	0.3992μs	2.5049 MOps/s	2.4666 MOps/s	$\color{#35bf28}+1.55\%$
test_stack[memmap_tensor0]	52.9120μs	4.6050μs	217.1559 KOps/s	219.7225 KOps/s	$\color{#d91a1a}-1.17\%$
test_memmaptd_index	1.5123ms	0.2597ms	3.8507 KOps/s	4.0186 KOps/s	$\color{#d91a1a}-4.18\%$
test_memmaptd_index_astensor	0.5880ms	0.3237ms	3.0889 KOps/s	3.2148 KOps/s	$\color{#d91a1a}-3.92\%$
test_memmaptd_index_op	1.0530ms	0.5554ms	1.8004 KOps/s	1.7232 KOps/s	$\color{#35bf28}+4.48\%$
test_serialize_model	0.1318s	0.1312s	7.6243 Ops/s	7.6426 Ops/s	$\color{#d91a1a}-0.24\%$
test_serialize_model_pickle	1.3776s	1.2190s	0.8204 Ops/s	0.8443 Ops/s	$\color{#d91a1a}-2.83\%$
test_serialize_weights	0.1310s	0.1304s	7.6716 Ops/s	7.6867 Ops/s	$\color{#d91a1a}-0.20\%$
test_serialize_weights_returnearly	0.5131s	72.7543ms	13.7449 Ops/s	15.3446 Ops/s	$\textbf{\color{#d91a1a}-10.43\%}$
test_serialize_weights_pickle	1.3772s	1.2161s	0.8223 Ops/s	0.8222 Ops/s	$+0.00\%$
test_reshape_pytree	55.1830μs	21.8910μs	45.6808 KOps/s	45.2079 KOps/s	$\color{#35bf28}+1.05\%$
test_reshape_td	59.2630μs	26.4187μs	37.8520 KOps/s	37.4825 KOps/s	$\color{#35bf28}+0.99\%$
test_view_pytree	47.1220μs	21.8535μs	45.7592 KOps/s	45.8195 KOps/s	$\color{#d91a1a}-0.13\%$
test_view_td	74.5140μs	31.6835μs	31.5622 KOps/s	33.9184 KOps/s	$\textbf{\color{#d91a1a}-6.95\%}$
test_unbind_pytree	56.4530μs	27.9367μs	35.7952 KOps/s	35.8382 KOps/s	$\color{#d91a1a}-0.12\%$
test_unbind_td	1.0009ms	36.3647μs	27.4992 KOps/s	27.8305 KOps/s	$\color{#d91a1a}-1.19\%$
test_split_pytree	64.1740μs	29.4431μs	33.9638 KOps/s	32.7569 KOps/s	$\color{#35bf28}+3.68\%$
test_split_td	0.1818ms	39.3721μs	25.3987 KOps/s	25.9971 KOps/s	$\color{#d91a1a}-2.30\%$
test_add_pytree	69.1840μs	32.9244μs	30.3726 KOps/s	30.7799 KOps/s	$\color{#d91a1a}-1.32\%$
test_add_td	84.2440μs	43.2222μs	23.1362 KOps/s	20.8011 KOps/s	$\textbf{\color{#35bf28}+11.23\%}$
test_compile_add_one_nested[tensordict-compile]	0.1765ms	0.1214ms	8.2367 KOps/s	7.7906 KOps/s	$\textbf{\color{#35bf28}+5.73\%}$
test_compile_add_one_nested[tensordict-eager]	0.2271ms	0.1307ms	7.6524 KOps/s	7.6110 KOps/s	$\color{#35bf28}+0.54\%$
test_compile_add_one_nested[pytree-compile]	0.1327ms	93.9959μs	10.6388 KOps/s	10.7203 KOps/s	$\color{#d91a1a}-0.76\%$
test_compile_add_one_nested[pytree-eager]	0.2029ms	0.1483ms	6.7420 KOps/s	6.7362 KOps/s	$\color{#35bf28}+0.09\%$
test_compile_copy_nested[tensordict-compile]	59.5930μs	25.2350μs	39.6275 KOps/s	44.3595 KOps/s	$\textbf{\color{#d91a1a}-10.67\%}$
test_compile_copy_nested[tensordict-eager]	57.4630μs	28.7693μs	34.7593 KOps/s	34.0518 KOps/s	$\color{#35bf28}+2.08\%$
test_compile_copy_nested[pytree-compile]	0.3343ms	64.6168μs	15.4759 KOps/s	15.6510 KOps/s	$\color{#d91a1a}-1.12\%$
test_compile_copy_nested[pytree-eager]	73.7740μs	48.8258μs	20.4810 KOps/s	20.1801 KOps/s	$\color{#35bf28}+1.49\%$
test_compile_add_one_flat[tensordict-compile]	0.1831ms	0.1407ms	7.1058 KOps/s	7.1557 KOps/s	$\color{#d91a1a}-0.70\%$
test_compile_add_one_flat[tensordict-eager]	0.3188ms	0.2176ms	4.5952 KOps/s	4.6463 KOps/s	$\color{#d91a1a}-1.10\%$
test_compile_add_one_flat[tensorclass-compile]	0.1580ms	0.1011ms	9.8878 KOps/s	10.3629 KOps/s	$\color{#d91a1a}-4.58\%$
test_compile_add_one_flat[tensorclass-eager]	0.1213ms	57.6444μs	17.3477 KOps/s	18.0375 KOps/s	$\color{#d91a1a}-3.82\%$
test_compile_add_one_flat[pytree-compile]	0.1750ms	0.1346ms	7.4294 KOps/s	7.2501 KOps/s	$\color{#35bf28}+2.47\%$
test_compile_add_one_flat[pytree-eager]	0.5871ms	0.4809ms	2.0795 KOps/s	2.0813 KOps/s	$\color{#d91a1a}-0.09\%$
test_compile_add_self_flat[tensordict-eager]	0.4004ms	0.2610ms	3.8318 KOps/s	3.8235 KOps/s	$\color{#35bf28}+0.22\%$
test_compile_add_self_flat[tensordict-compile]	0.1824ms	0.1425ms	7.0158 KOps/s	6.8394 KOps/s	$\color{#35bf28}+2.58\%$
test_compile_add_self_flat[tensorclass-eager]	0.1644ms	66.0051μs	15.1503 KOps/s	15.0853 KOps/s	$\color{#35bf28}+0.43\%$
test_compile_add_self_flat[tensorclass-compile]	0.1371ms	98.9696μs	10.1041 KOps/s	10.2648 KOps/s	$\color{#d91a1a}-1.57\%$
test_compile_add_self_flat[pytree-eager]	0.4570ms	0.4099ms	2.4397 KOps/s	2.4690 KOps/s	$\color{#d91a1a}-1.19\%$
test_compile_add_self_flat[pytree-compile]	0.1762ms	0.1343ms	7.4479 KOps/s	7.5053 KOps/s	$\color{#d91a1a}-0.77\%$
test_compile_copy_flat[tensordict-compile]	50.7020μs	19.2374μs	51.9821 KOps/s	56.2192 KOps/s	$\textbf{\color{#d91a1a}-7.54\%}$
test_compile_copy_flat[tensordict-eager]	67.4140μs	30.7192μs	32.5530 KOps/s	31.9528 KOps/s	$\color{#35bf28}+1.88\%$
test_compile_copy_flat[pytree-compile]	0.1071ms	69.5336μs	14.3815 KOps/s	14.3566 KOps/s	$\color{#35bf28}+0.17\%$
test_compile_copy_flat[pytree-eager]	0.1908ms	50.6273μs	19.7522 KOps/s	19.7803 KOps/s	$\color{#d91a1a}-0.14\%$
test_compile_assign_and_add[tensordict-compile]	1.6005ms	0.3855ms	2.5943 KOps/s	2.2404 KOps/s	$\textbf{\color{#35bf28}+15.80\%}$
test_compile_assign_and_add[tensordict-eager]	2.7965ms	2.5825ms	387.2158 Ops/s	379.6917 Ops/s	$\color{#35bf28}+1.98\%$
test_compile_assign_and_add[pytree-compile]	1.5750ms	0.4291ms	2.3302 KOps/s	2.3499 KOps/s	$\color{#d91a1a}-0.84\%$
test_compile_assign_and_add[pytree-eager]	2.7035ms	2.6046ms	383.9422 Ops/s	386.8415 Ops/s	$\color{#d91a1a}-0.75\%$
test_compile_indexing[tensor-tensordict-compile]	0.1664ms	0.1159ms	8.6292 KOps/s	9.1939 KOps/s	$\textbf{\color{#d91a1a}-6.14\%}$
test_compile_indexing[tensor-tensordict-eager]	0.5551ms	80.4090μs	12.4364 KOps/s	13.1836 KOps/s	$\textbf{\color{#d91a1a}-5.67\%}$
test_compile_indexing[tensor-tensorclass-compile]	0.1633ms	0.1067ms	9.3694 KOps/s	9.8955 KOps/s	$\textbf{\color{#d91a1a}-5.32\%}$
test_compile_indexing[tensor-tensorclass-eager]	0.1296ms	69.4867μs	14.3912 KOps/s	15.2716 KOps/s	$\textbf{\color{#d91a1a}-5.76\%}$
test_compile_indexing[tensor-pytree-compile]	0.1636ms	0.1084ms	9.2267 KOps/s	9.7992 KOps/s	$\textbf{\color{#d91a1a}-5.84\%}$
test_compile_indexing[tensor-pytree-eager]	0.1847ms	67.1319μs	14.8961 KOps/s	15.2948 KOps/s	$\color{#d91a1a}-2.61\%$
test_compile_indexing[slice-tensordict-compile]	0.1597ms	0.1001ms	9.9881 KOps/s	10.1860 KOps/s	$\color{#d91a1a}-1.94\%$
test_compile_indexing[slice-tensordict-eager]	0.1418ms	17.3407μs	57.6679 KOps/s	58.3530 KOps/s	$\color{#d91a1a}-1.17\%$
test_compile_indexing[slice-tensorclass-compile]	0.1454ms	95.9372μs	10.4235 KOps/s	10.5737 KOps/s	$\color{#d91a1a}-1.42\%$
test_compile_indexing[slice-tensorclass-eager]	56.3730μs	15.8989μs	62.8974 KOps/s	63.7575 KOps/s	$\color{#d91a1a}-1.35\%$
test_compile_indexing[slice-pytree-compile]	0.1352ms	95.5633μs	10.4643 KOps/s	10.5376 KOps/s	$\color{#d91a1a}-0.70\%$
test_compile_indexing[slice-pytree-eager]	49.2830μs	15.7897μs	63.3322 KOps/s	64.6620 KOps/s	$\color{#d91a1a}-2.06\%$
test_compile_indexing[int-tensordict-compile]	0.1416ms	0.1010ms	9.9002 KOps/s	10.0684 KOps/s	$\color{#d91a1a}-1.67\%$
test_compile_indexing[int-tensordict-eager]	0.5765ms	16.7003μs	59.8792 KOps/s	58.9535 KOps/s	$\color{#35bf28}+1.57\%$
test_compile_indexing[int-tensorclass-compile]	0.1571ms	95.5840μs	10.4620 KOps/s	10.5284 KOps/s	$\color{#d91a1a}-0.63\%$
test_compile_indexing[int-tensorclass-eager]	63.2030μs	15.8193μs	63.2141 KOps/s	64.2480 KOps/s	$\color{#d91a1a}-1.61\%$
test_compile_indexing[int-pytree-compile]	0.1351ms	93.8926μs	10.6505 KOps/s	10.5200 KOps/s	$\color{#35bf28}+1.24\%$
test_compile_indexing[int-pytree-eager]	49.0530μs	15.7693μs	63.4145 KOps/s	63.9852 KOps/s	$\color{#d91a1a}-0.89\%$
test_mod_add[eager]	0.1453ms	35.5214μs	28.1520 KOps/s	25.7210 KOps/s	$\textbf{\color{#35bf28}+9.45\%}$
test_mod_add[compile]	0.1321ms	79.4973μs	12.5790 KOps/s	12.6481 KOps/s	$\color{#d91a1a}-0.55\%$
test_mod_add[compile-overhead]	0.3257ms	0.1656ms	6.0399 KOps/s	5.6686 KOps/s	$\textbf{\color{#35bf28}+6.55\%}$
test_mod_wrap[eager]	0.3126ms	0.2360ms	4.2375 KOps/s	4.0491 KOps/s	$\color{#35bf28}+4.65\%$
test_mod_wrap[compile]	0.6765ms	0.2752ms	3.6342 KOps/s	3.6157 KOps/s	$\color{#35bf28}+0.51\%$
test_mod_wrap[compile-overhead]	6.9577ms	3.6533ms	273.7225 Ops/s	263.4296 Ops/s	$\color{#35bf28}+3.91\%$
test_mod_wrap_and_backward[eager]	1.7785ms	1.3383ms	747.2186 Ops/s	711.8402 Ops/s	$\color{#35bf28}+4.97\%$
test_mod_wrap_and_backward[compile]	1.3172ms	1.2297ms	813.2384 Ops/s	789.2293 Ops/s	$\color{#35bf28}+3.04\%$
test_mod_wrap_and_backward[compile-overhead]	1.3366ms	0.8989ms	1.1125 KOps/s	1.1009 KOps/s	$\color{#35bf28}+1.05\%$
test_seq_add[eager]	0.1460ms	0.1098ms	9.1108 KOps/s	8.7351 KOps/s	$\color{#35bf28}+4.30\%$
test_seq_add[compile]	0.1409ms	86.6636μs	11.5389 KOps/s	11.6046 KOps/s	$\color{#d91a1a}-0.57\%$
test_seq_add[compile-overhead]	0.1820ms	0.1276ms	7.8392 KOps/s	7.7668 KOps/s	$\color{#35bf28}+0.93\%$
test_seq_wrap[eager]	0.4791ms	0.3983ms	2.5107 KOps/s	2.4342 KOps/s	$\color{#35bf28}+3.14\%$
test_seq_wrap[compile]	0.3584ms	0.2902ms	3.4453 KOps/s	3.4004 KOps/s	$\color{#35bf28}+1.32\%$
test_seq_wrap[compile-overhead]	0.2687ms	0.2197ms	4.5515 KOps/s	4.5381 KOps/s	$\color{#35bf28}+0.29\%$
test_func_call_runtime[False-eager]	0.7675ms	0.7020ms	1.4246 KOps/s	1.3456 KOps/s	$\textbf{\color{#35bf28}+5.87\%}$
test_func_call_runtime[False-compile]	0.9005ms	0.7183ms	1.3921 KOps/s	1.3920 KOps/s	$+0.01\%$
test_func_call_runtime[False-compile-overhead]	0.3995ms	0.3503ms	2.8546 KOps/s	2.8633 KOps/s	$\color{#d91a1a}-0.30\%$
test_func_call_runtime[True-eager]	0.9612ms	0.8586ms	1.1647 KOps/s	1.1434 KOps/s	$\color{#35bf28}+1.87\%$
test_func_call_runtime[True-compile]	0.7887ms	0.7354ms	1.3598 KOps/s	1.3603 KOps/s	$\color{#d91a1a}-0.04\%$
test_func_call_runtime[True-compile-overhead]	0.4307ms	0.3709ms	2.6964 KOps/s	2.7079 KOps/s	$\color{#d91a1a}-0.42\%$
test_func_call_cm_runtime[False-eager]	0.7517ms	0.7010ms	1.4265 KOps/s	1.4252 KOps/s	$\color{#35bf28}+0.09\%$
test_func_call_cm_runtime[False-compile]	0.7813ms	0.7209ms	1.3871 KOps/s	1.3861 KOps/s	$\color{#35bf28}+0.07\%$
test_func_call_cm_runtime[False-compile-overhead]	0.4019ms	0.3544ms	2.8216 KOps/s	2.8409 KOps/s	$\color{#d91a1a}-0.68\%$
test_func_call_cm_runtime[True-eager]	1.0639ms	0.9702ms	1.0307 KOps/s	1.0169 KOps/s	$\color{#35bf28}+1.36\%$
test_func_call_cm_runtime[True-compile]	0.8184ms	0.7685ms	1.3012 KOps/s	1.3016 KOps/s	$\color{#d91a1a}-0.03\%$
test_func_call_cm_runtime[True-compile-overhead]	0.4791ms	0.3956ms	2.5279 KOps/s	2.5188 KOps/s	$\color{#35bf28}+0.36\%$
test_vmap_func_call_cm_runtime[eager]	2.4242ms	1.9867ms	503.3438 Ops/s	498.5650 Ops/s	$\color{#35bf28}+0.96\%$
test_vmap_func_call_cm_runtime[compile]	0.8716ms	0.7982ms	1.2528 KOps/s	1.2748 KOps/s	$\color{#d91a1a}-1.72\%$
test_vmap_func_call_cm_runtime[compile-overhead]	0.4621ms	0.4011ms	2.4931 KOps/s	2.4959 KOps/s	$\color{#d91a1a}-0.11\%$
test_distributed	2.8675ms	0.2493ms	4.0116 KOps/s	7.9003 KOps/s	$\textbf{\color{#d91a1a}-49.22\%}$
test_tdmodule	0.2731ms	18.9941μs	52.6479 KOps/s	49.4999 KOps/s	$\textbf{\color{#35bf28}+6.36\%}$
test_tdmodule_dispatch	62.8130μs	33.2606μs	30.0656 KOps/s	27.5137 KOps/s	$\textbf{\color{#35bf28}+9.28\%}$
test_tdseq	26.8510μs	19.3759μs	51.6104 KOps/s	46.9755 KOps/s	$\textbf{\color{#35bf28}+9.87\%}$
test_tdseq_dispatch	64.9330μs	35.5195μs	28.1536 KOps/s	25.1208 KOps/s	$\textbf{\color{#35bf28}+12.07\%}$
test_instantiation_functorch	1.6917ms	1.5088ms	662.7574 Ops/s	658.6560 Ops/s	$\color{#35bf28}+0.62\%$
test_exec_functorch	0.2027ms	0.1427ms	7.0080 KOps/s	7.1140 KOps/s	$\color{#d91a1a}-1.49\%$
test_exec_functional_call	0.1766ms	0.1320ms	7.5782 KOps/s	7.6413 KOps/s	$\color{#d91a1a}-0.83\%$
test_exec_td_decorator	0.3721ms	0.1800ms	5.5565 KOps/s	5.5423 KOps/s	$\color{#35bf28}+0.26\%$
test_vmap_mlp_speed_decorator[True-True]	0.7982ms	0.6566ms	1.5229 KOps/s	1.4978 KOps/s	$\color{#35bf28}+1.67\%$
test_vmap_mlp_speed_decorator[True-False]	0.7733ms	0.6596ms	1.5160 KOps/s	1.4884 KOps/s	$\color{#35bf28}+1.85\%$
test_vmap_mlp_speed_decorator[False-True]	0.7653ms	0.5714ms	1.7501 KOps/s	1.6478 KOps/s	$\textbf{\color{#35bf28}+6.20\%}$
test_vmap_mlp_speed_decorator[False-False]	0.7021ms	0.5759ms	1.7364 KOps/s	1.6488 KOps/s	$\textbf{\color{#35bf28}+5.32\%}$
test_vmap_transformer_speed_decorator[True-True]	18.5657ms	18.4701ms	54.1415 Ops/s	53.8538 Ops/s	$\color{#35bf28}+0.53\%$
test_vmap_transformer_speed_decorator[True-False]	18.9438ms	18.4736ms	54.1313 Ops/s	53.1730 Ops/s	$\color{#35bf28}+1.80\%$
test_vmap_transformer_speed_decorator[False-True]	19.5586ms	18.3743ms	54.4237 Ops/s	54.1865 Ops/s	$\color{#35bf28}+0.44\%$
test_vmap_transformer_speed_decorator[False-False]	19.2815ms	18.4060ms	54.3300 Ops/s	53.9413 Ops/s	$\color{#35bf28}+0.72\%$
test_to_module_speed[True]	1.3219ms	0.9630ms	1.0384 KOps/s	1.0168 KOps/s	$\color{#35bf28}+2.12\%$
test_to_module_speed[False]	1.4717ms	0.9498ms	1.0529 KOps/s	1.0251 KOps/s	$\color{#35bf28}+2.71\%$
test_tc_init	84.6540μs	33.4077μs	29.9332 KOps/s	26.9109 KOps/s	$\textbf{\color{#35bf28}+11.23\%}$
test_tc_init_nested	99.3760μs	65.3180μs	15.3097 KOps/s	13.3784 KOps/s	$\textbf{\color{#35bf28}+14.44\%}$
test_tc_first_layer_tensor	6.3161μs	0.6870μs	1.4556 MOps/s	1.2415 MOps/s	$\textbf{\color{#35bf28}+17.24\%}$
test_tc_first_layer_nontensor	23.6810μs	2.2373μs	446.9742 KOps/s	443.8622 KOps/s	$\color{#35bf28}+0.70\%$
test_tc_second_layer_tensor	63.2467μs	1.3990μs	714.8042 KOps/s	689.0311 KOps/s	$\color{#35bf28}+3.74\%$
test_tc_second_layer_nontensor	47.4930μs	2.9217μs	342.2635 KOps/s	330.2787 KOps/s	$\color{#35bf28}+3.63\%$
test_unbind	0.2338s	10.1977ms	98.0615 Ops/s	142.8387 Ops/s	$\textbf{\color{#d91a1a}-31.35\%}$
test_full_like	10.1944ms	9.1677ms	109.0790 Ops/s	109.1917 Ops/s	$\color{#d91a1a}-0.10\%$
test_zeros_like	5.2727ms	4.3256ms	231.1814 Ops/s	231.7920 Ops/s	$\color{#d91a1a}-0.26\%$
test_ones_like	4.9001ms	4.3278ms	231.0653 Ops/s	231.7467 Ops/s	$\color{#d91a1a}-0.29\%$
test_clone	6.7279ms	6.3067ms	158.5625 Ops/s	158.1899 Ops/s	$\color{#35bf28}+0.24\%$
test_squeeze	57.3030μs	9.1812μs	108.9182 KOps/s	108.1743 KOps/s	$\color{#35bf28}+0.69\%$
test_unsqueeze	0.1462ms	72.8202μs	13.7325 KOps/s	14.1147 KOps/s	$\color{#d91a1a}-2.71\%$
test_split	0.5813ms	0.1664ms	6.0086 KOps/s	6.2868 KOps/s	$\color{#d91a1a}-4.42\%$
test_permute	0.2464ms	0.1682ms	5.9440 KOps/s	5.5992 KOps/s	$\textbf{\color{#35bf28}+6.16\%}$
test_stack	50.9293ms	50.2353ms	19.9063 Ops/s	19.9300 Ops/s	$\color{#d91a1a}-0.12\%$
test_cat	50.6232ms	50.1617ms	19.9355 Ops/s	19.9800 Ops/s	$\color{#d91a1a}-0.22\%$