[Refactor] Make _set_dispatch_td_nn_modules compatible with compile

[vmoens]

Stack from ghstack (oldest at bottom):

• -> [Refactor] Make _set_dispatch_td_nn_modules compatible with compile #1084

[github-actions]

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

Total Benchmarks: 217. Improved: $\large\color{#35bf28}33$. Worsened: $\large\color{#d91a1a}12$.

Expand to view detailed results

Name	Max	Mean	Ops	Ops on Repo HEAD	Change
test_plain_set_nested	56.1350μs	18.2674μs	54.7424 KOps/s	52.6051 KOps/s	$\color{#35bf28}+4.06\%$
test_plain_set_stack_nested	54.9320μs	18.2921μs	54.6685 KOps/s	52.2908 KOps/s	$\color{#35bf28}+4.55\%$
test_plain_set_nested_inplace	60.7840μs	19.5274μs	51.2102 KOps/s	47.7334 KOps/s	$\textbf{\color{#35bf28}+7.28\%}$
test_plain_set_stack_nested_inplace	60.5330μs	19.8163μs	50.4636 KOps/s	47.9882 KOps/s	$\textbf{\color{#35bf28}+5.16\%}$
test_items	31.8300μs	4.0950μs	244.1983 KOps/s	236.4556 KOps/s	$\color{#35bf28}+3.27\%$
test_items_nested	0.5016ms	0.3440ms	2.9071 KOps/s	2.9467 KOps/s	$\color{#d91a1a}-1.34\%$
test_items_nested_locked	0.5551ms	0.3446ms	2.9015 KOps/s	2.8863 KOps/s	$\color{#35bf28}+0.53\%$
test_items_nested_leaf	0.1401ms	71.5893μs	13.9686 KOps/s	13.9227 KOps/s	$\color{#35bf28}+0.33\%$
test_items_stack_nested	0.6900ms	0.3440ms	2.9066 KOps/s	2.9304 KOps/s	$\color{#d91a1a}-0.81\%$
test_items_stack_nested_leaf	0.1555ms	74.9646μs	13.3396 KOps/s	13.1065 KOps/s	$\color{#35bf28}+1.78\%$
test_items_stack_nested_locked	0.7159ms	0.3479ms	2.8744 KOps/s	2.9150 KOps/s	$\color{#d91a1a}-1.39\%$
test_keys	27.4010μs	3.4793μs	287.4106 KOps/s	285.9070 KOps/s	$\color{#35bf28}+0.53\%$
test_keys_nested	0.2308ms	0.1358ms	7.3626 KOps/s	7.2917 KOps/s	$\color{#35bf28}+0.97\%$
test_keys_nested_locked	1.9609ms	0.1418ms	7.0539 KOps/s	7.0262 KOps/s	$\color{#35bf28}+0.39\%$
test_keys_nested_leaf	0.2420ms	0.1172ms	8.5352 KOps/s	8.5083 KOps/s	$\color{#35bf28}+0.32\%$
test_keys_stack_nested	0.2693ms	0.1354ms	7.3870 KOps/s	7.2711 KOps/s	$\color{#35bf28}+1.59\%$
test_keys_stack_nested_leaf	0.1969ms	0.1147ms	8.7174 KOps/s	8.4358 KOps/s	$\color{#35bf28}+3.34\%$
test_keys_stack_nested_locked	0.2613ms	0.1399ms	7.1501 KOps/s	7.0895 KOps/s	$\color{#35bf28}+0.85\%$
test_values	6.1736μs	1.0282μs	972.5802 KOps/s	963.4700 KOps/s	$\color{#35bf28}+0.95\%$
test_values_nested	0.1185ms	55.5497μs	18.0019 KOps/s	18.0522 KOps/s	$\color{#d91a1a}-0.28\%$
test_values_nested_locked	0.1084ms	55.5991μs	17.9859 KOps/s	17.9743 KOps/s	$\color{#35bf28}+0.06\%$
test_values_nested_leaf	0.1171ms	60.9250μs	16.4136 KOps/s	16.3559 KOps/s	$\color{#35bf28}+0.35\%$
test_values_stack_nested	0.1140ms	57.6414μs	17.3486 KOps/s	17.5296 KOps/s	$\color{#d91a1a}-1.03\%$
test_values_stack_nested_leaf	0.1108ms	60.1675μs	16.6203 KOps/s	16.3607 KOps/s	$\color{#35bf28}+1.59\%$
test_values_stack_nested_locked	0.1121ms	57.0579μs	17.5261 KOps/s	16.8257 KOps/s	$\color{#35bf28}+4.16\%$
test_membership	15.0380μs	0.8620μs	1.1601 MOps/s	1.1394 MOps/s	$\color{#35bf28}+1.81\%$
test_membership_nested	16.7910μs	2.7555μs	362.9117 KOps/s	367.6993 KOps/s	$\color{#d91a1a}-1.30\%$
test_membership_nested_leaf	29.0840μs	2.7758μs	360.2601 KOps/s	364.2827 KOps/s	$\color{#d91a1a}-1.10\%$
test_membership_stacked_nested	21.7210μs	2.7583μs	362.5375 KOps/s	362.8547 KOps/s	$\color{#d91a1a}-0.09\%$
test_membership_stacked_nested_leaf	23.8340μs	2.7349μs	365.6438 KOps/s	358.8729 KOps/s	$\color{#35bf28}+1.89\%$
test_membership_nested_last	38.7520μs	4.1224μs	242.5782 KOps/s	246.9717 KOps/s	$\color{#d91a1a}-1.78\%$
test_membership_nested_leaf_last	29.4650μs	4.0682μs	245.8117 KOps/s	244.6397 KOps/s	$\color{#35bf28}+0.48\%$
test_membership_stacked_nested_last	49.6620μs	13.1804μs	75.8701 KOps/s	247.3790 KOps/s	$\textbf{\color{#d91a1a}-69.33\%}$
test_membership_stacked_nested_leaf_last	40.2850μs	12.9003μs	77.5175 KOps/s	247.4021 KOps/s	$\textbf{\color{#d91a1a}-68.67\%}$
test_nested_getleaf	35.6160μs	10.7434μs	93.0803 KOps/s	92.5089 KOps/s	$\color{#35bf28}+0.62\%$
test_nested_get	36.8980μs	10.1276μs	98.7397 KOps/s	97.9351 KOps/s	$\color{#35bf28}+0.82\%$
test_stacked_getleaf	36.8390μs	10.5862μs	94.4628 KOps/s	92.6034 KOps/s	$\color{#35bf28}+2.01\%$
test_stacked_get	30.5170μs	10.0931μs	99.0776 KOps/s	97.7631 KOps/s	$\color{#35bf28}+1.34\%$
test_nested_getitemleaf	37.3400μs	11.1577μs	89.6243 KOps/s	88.6915 KOps/s	$\color{#35bf28}+1.05\%$
test_nested_getitem	36.2370μs	10.5097μs	95.1506 KOps/s	95.4455 KOps/s	$\color{#d91a1a}-0.31\%$
test_stacked_getitemleaf	47.3890μs	11.2598μs	88.8114 KOps/s	86.6843 KOps/s	$\color{#35bf28}+2.45\%$
test_stacked_getitem	35.5360μs	10.3592μs	96.5322 KOps/s	94.3501 KOps/s	$\color{#35bf28}+2.31\%$
test_lock_nested	2.7781ms	0.4476ms	2.2344 KOps/s	1.8388 KOps/s	$\textbf{\color{#35bf28}+21.51\%}$
test_lock_stack_nested	0.7311ms	0.4038ms	2.4766 KOps/s	2.4102 KOps/s	$\color{#35bf28}+2.76\%$
test_unlock_nested	0.7440ms	0.3606ms	2.7732 KOps/s	2.7071 KOps/s	$\color{#35bf28}+2.44\%$
test_unlock_stack_nested	0.4721ms	0.3200ms	3.1246 KOps/s	2.9810 KOps/s	$\color{#35bf28}+4.82\%$
test_flatten_speed	0.1846ms	92.4299μs	10.8190 KOps/s	10.7926 KOps/s	$\color{#35bf28}+0.25\%$
test_unflatten_speed	0.6366ms	0.4812ms	2.0783 KOps/s	2.1347 KOps/s	$\color{#d91a1a}-2.65\%$
test_common_ops	1.5415ms	0.7629ms	1.3108 KOps/s	1.2126 KOps/s	$\textbf{\color{#35bf28}+8.09\%}$
test_creation	21.7110μs	2.1419μs	466.8822 KOps/s	481.7156 KOps/s	$\color{#d91a1a}-3.08\%$
test_creation_empty	36.5780μs	10.9703μs	91.1554 KOps/s	75.2768 KOps/s	$\textbf{\color{#35bf28}+21.09\%}$
test_creation_nested_1	42.8100μs	13.7493μs	72.7310 KOps/s	62.1495 KOps/s	$\textbf{\color{#35bf28}+17.03\%}$
test_creation_nested_2	60.7530μs	17.9737μs	55.6369 KOps/s	48.9896 KOps/s	$\textbf{\color{#35bf28}+13.57\%}$
test_clone	79.7580μs	13.1859μs	75.8387 KOps/s	76.0140 KOps/s	$\color{#d91a1a}-0.23\%$
test_getitem[int]	1.3457ms	12.8439μs	77.8581 KOps/s	78.1964 KOps/s	$\color{#d91a1a}-0.43\%$
test_getitem[slice_int]	0.1393ms	24.1681μs	41.3769 KOps/s	41.0851 KOps/s	$\color{#35bf28}+0.71\%$
test_getitem[range]	0.1808ms	48.2985μs	20.7046 KOps/s	20.7213 KOps/s	$\color{#d91a1a}-0.08\%$
test_getitem[tuple]	0.1504ms	20.2769μs	49.3171 KOps/s	49.6313 KOps/s	$\color{#d91a1a}-0.63\%$
test_getitem[list]	0.1793ms	44.4660μs	22.4891 KOps/s	22.7358 KOps/s	$\color{#d91a1a}-1.09\%$
test_setitem_dim[int]	56.0640μs	26.2157μs	38.1451 KOps/s	37.5755 KOps/s	$\color{#35bf28}+1.52\%$
test_setitem_dim[slice_int]	86.6610μs	53.3631μs	18.7395 KOps/s	19.0240 KOps/s	$\color{#d91a1a}-1.50\%$
test_setitem_dim[range]	0.1205ms	73.9736μs	13.5183 KOps/s	13.4412 KOps/s	$\color{#35bf28}+0.57\%$
test_setitem_dim[tuple]	68.2670μs	41.0504μs	24.3603 KOps/s	23.6853 KOps/s	$\color{#35bf28}+2.85\%$
test_setitem	0.1313ms	20.7145μs	48.2754 KOps/s	45.9047 KOps/s	$\textbf{\color{#35bf28}+5.16\%}$
test_set	98.2330μs	19.7738μs	50.5720 KOps/s	47.1724 KOps/s	$\textbf{\color{#35bf28}+7.21\%}$
test_set_shared	3.2864ms	0.1687ms	5.9263 KOps/s	5.9412 KOps/s	$\color{#d91a1a}-0.25\%$
test_update	0.8096ms	22.4693μs	44.5052 KOps/s	40.3389 KOps/s	$\textbf{\color{#35bf28}+10.33\%}$
test_update_nested	0.1306ms	33.2430μs	30.0815 KOps/s	28.6179 KOps/s	$\textbf{\color{#35bf28}+5.11\%}$
test_update__nested	0.1774ms	33.7926μs	29.5923 KOps/s	30.9673 KOps/s	$\color{#d91a1a}-4.44\%$
test_set_nested	0.1017ms	21.9456μs	45.5673 KOps/s	42.4397 KOps/s	$\textbf{\color{#35bf28}+7.37\%}$
test_set_nested_new	81.6730μs	26.8578μs	37.2331 KOps/s	35.4875 KOps/s	$\color{#35bf28}+4.92\%$
test_select	0.2234ms	42.5996μs	23.4744 KOps/s	22.4723 KOps/s	$\color{#35bf28}+4.46\%$
test_select_nested	0.1349ms	61.3189μs	16.3082 KOps/s	16.6308 KOps/s	$\color{#d91a1a}-1.94\%$
test_exclude_nested	0.1929ms	77.2839μs	12.9393 KOps/s	13.2903 KOps/s	$\color{#d91a1a}-2.64\%$
test_empty[True]	0.5509ms	0.3529ms	2.8339 KOps/s	2.8619 KOps/s	$\color{#d91a1a}-0.98\%$
test_empty[False]	10.8277μs	1.3239μs	755.3193 KOps/s	755.1481 KOps/s	$\color{#35bf28}+0.02\%$
test_unbind_speed	0.3521ms	0.2661ms	3.7586 KOps/s	3.7083 KOps/s	$\color{#35bf28}+1.36\%$
test_unbind_speed_stack0	0.5181ms	0.2552ms	3.9188 KOps/s	3.8043 KOps/s	$\color{#35bf28}+3.01\%$
test_unbind_speed_stack1	0.1023s	0.7444ms	1.3434 KOps/s	1.4081 KOps/s	$\color{#d91a1a}-4.60\%$
test_split	2.5819ms	1.6026ms	623.9773 Ops/s	563.3246 Ops/s	$\textbf{\color{#35bf28}+10.77\%}$
test_chunk	0.1110s	1.9461ms	513.8536 Ops/s	566.6031 Ops/s	$\textbf{\color{#d91a1a}-9.31\%}$
test_consolidate_njt[False-None]	11.1408ms	8.1286ms	123.0220 Ops/s	123.2658 Ops/s	$\color{#d91a1a}-0.20\%$
test_creation[device0]	3.6947ms	93.2979μs	10.7184 KOps/s	10.9292 KOps/s	$\color{#d91a1a}-1.93\%$
test_creation_from_tensor	0.2786ms	92.6178μs	10.7971 KOps/s	10.7024 KOps/s	$\color{#35bf28}+0.88\%$
test_add_one[memmap_tensor0]	0.2361ms	4.7276μs	211.5233 KOps/s	203.4126 KOps/s	$\color{#35bf28}+3.99\%$
test_contiguous[memmap_tensor0]	15.3890μs	0.5157μs	1.9389 MOps/s	1.9576 MOps/s	$\color{#d91a1a}-0.95\%$
test_stack[memmap_tensor0]	29.7450μs	3.2717μs	305.6530 KOps/s	290.7916 KOps/s	$\textbf{\color{#35bf28}+5.11\%}$
test_memmaptd_index	1.0541ms	0.2386ms	4.1915 KOps/s	4.1696 KOps/s	$\color{#35bf28}+0.53\%$
test_memmaptd_index_astensor	0.5724ms	0.3174ms	3.1510 KOps/s	3.1644 KOps/s	$\color{#d91a1a}-0.42\%$
test_memmaptd_index_op	0.9333ms	0.5763ms	1.7352 KOps/s	1.5977 KOps/s	$\textbf{\color{#35bf28}+8.61\%}$
test_serialize_model	0.1220s	0.1167s	8.5684 Ops/s	7.4313 Ops/s	$\textbf{\color{#35bf28}+15.30\%}$
test_serialize_model_pickle	0.5029s	0.3903s	2.5624 Ops/s	2.5702 Ops/s	$\color{#d91a1a}-0.30\%$
test_serialize_weights	0.1175s	0.1120s	8.9286 Ops/s	8.6973 Ops/s	$\color{#35bf28}+2.66\%$
test_serialize_weights_returnearly	0.3521s	0.1842s	5.4294 Ops/s	6.2473 Ops/s	$\textbf{\color{#d91a1a}-13.09\%}$
test_serialize_weights_pickle	1.0420s	0.6750s	1.4816 Ops/s	2.4481 Ops/s	$\textbf{\color{#d91a1a}-39.48\%}$
test_serialize_weights_filesystem	0.1466s	0.1430s	6.9919 Ops/s	6.3830 Ops/s	$\textbf{\color{#35bf28}+9.54\%}$
test_serialize_model_filesystem	0.1487s	0.1435s	6.9677 Ops/s	6.6023 Ops/s	$\textbf{\color{#35bf28}+5.54\%}$
test_reshape_pytree	0.1188ms	27.5915μs	36.2430 KOps/s	37.8629 KOps/s	$\color{#d91a1a}-4.28\%$
test_reshape_td	70.2910μs	33.0757μs	30.2337 KOps/s	30.0614 KOps/s	$\color{#35bf28}+0.57\%$
test_view_pytree	62.1360μs	27.0018μs	37.0346 KOps/s	38.0179 KOps/s	$\color{#d91a1a}-2.59\%$
test_view_td	90.6390μs	38.4528μs	26.0059 KOps/s	26.4979 KOps/s	$\color{#d91a1a}-1.86\%$
test_unbind_pytree	99.0040μs	30.3155μs	32.9864 KOps/s	33.6151 KOps/s	$\color{#d91a1a}-1.87\%$
test_unbind_td	0.3525ms	39.2897μs	25.4520 KOps/s	25.6194 KOps/s	$\color{#d91a1a}-0.65\%$
test_split_pytree	74.8190μs	29.7768μs	33.5832 KOps/s	34.0818 KOps/s	$\color{#d91a1a}-1.46\%$
test_split_td	0.1093s	55.7086μs	17.9506 KOps/s	22.1238 KOps/s	$\textbf{\color{#d91a1a}-18.86\%}$
test_add_pytree	81.5620μs	34.9723μs	28.5941 KOps/s	27.8055 KOps/s	$\color{#35bf28}+2.84\%$
test_add_td	0.1144ms	53.0755μs	18.8411 KOps/s	16.7867 KOps/s	$\textbf{\color{#35bf28}+12.24\%}$
test_compile_add_one_nested[tensordict-compile]	0.1912ms	64.0043μs	15.6240 KOps/s	15.8475 KOps/s	$\color{#d91a1a}-1.41\%$
test_compile_add_one_nested[tensordict-eager]	0.3694ms	0.1623ms	6.1618 KOps/s	6.2463 KOps/s	$\color{#d91a1a}-1.35\%$
test_compile_add_one_nested[pytree-compile]	0.1019ms	46.7984μs	21.3682 KOps/s	21.8698 KOps/s	$\color{#d91a1a}-2.29\%$
test_compile_add_one_nested[pytree-eager]	0.2068ms	0.1173ms	8.5285 KOps/s	8.4053 KOps/s	$\color{#35bf28}+1.47\%$
test_compile_copy_nested[tensordict-compile]	58.1980μs	26.3172μs	37.9979 KOps/s	39.3175 KOps/s	$\color{#d91a1a}-3.36\%$
test_compile_copy_nested[tensordict-eager]	0.1027ms	54.0959μs	18.4857 KOps/s	18.6664 KOps/s	$\color{#d91a1a}-0.97\%$
test_compile_copy_nested[pytree-compile]	0.1787ms	79.9248μs	12.5118 KOps/s	12.8657 KOps/s	$\color{#d91a1a}-2.75\%$
test_compile_copy_nested[pytree-eager]	0.1368ms	67.8369μs	14.7412 KOps/s	14.7272 KOps/s	$\color{#35bf28}+0.10\%$
test_compile_add_one_flat[tensordict-compile]	0.1860ms	0.1053ms	9.4957 KOps/s	9.5042 KOps/s	$\color{#d91a1a}-0.09\%$
test_compile_add_one_flat[tensordict-eager]	0.4292ms	0.2032ms	4.9223 KOps/s	5.0107 KOps/s	$\color{#d91a1a}-1.76\%$
test_compile_add_one_flat[tensorclass-compile]	0.1375ms	45.6190μs	21.9207 KOps/s	22.3365 KOps/s	$\color{#d91a1a}-1.86\%$
test_compile_add_one_flat[tensorclass-eager]	0.4702ms	63.7519μs	15.6858 KOps/s	16.2515 KOps/s	$\color{#d91a1a}-3.48\%$
test_compile_add_one_flat[pytree-compile]	0.1868ms	0.1039ms	9.6261 KOps/s	9.7459 KOps/s	$\color{#d91a1a}-1.23\%$
test_compile_add_one_flat[pytree-eager]	0.3832ms	0.2011ms	4.9735 KOps/s	4.8783 KOps/s	$\color{#35bf28}+1.95\%$
test_compile_add_self_flat[tensordict-eager]	0.3863ms	0.2111ms	4.7363 KOps/s	4.7126 KOps/s	$\color{#35bf28}+0.50\%$
test_compile_add_self_flat[tensordict-compile]	0.2396ms	0.1076ms	9.2896 KOps/s	9.5640 KOps/s	$\color{#d91a1a}-2.87\%$
test_compile_add_self_flat[tensorclass-eager]	0.3595ms	58.1955μs	17.1835 KOps/s	18.3894 KOps/s	$\textbf{\color{#d91a1a}-6.56\%}$
test_compile_add_self_flat[tensorclass-compile]	0.1240ms	48.1319μs	20.7762 KOps/s	21.6987 KOps/s	$\color{#d91a1a}-4.25\%$
test_compile_add_self_flat[pytree-eager]	0.6680ms	0.1592ms	6.2816 KOps/s	6.2741 KOps/s	$\color{#35bf28}+0.12\%$
test_compile_add_self_flat[pytree-compile]	0.1917ms	0.1056ms	9.4723 KOps/s	9.3446 KOps/s	$\color{#35bf28}+1.37\%$
test_compile_copy_flat[tensordict-compile]	84.7580μs	22.2648μs	44.9140 KOps/s	47.8046 KOps/s	$\textbf{\color{#d91a1a}-6.05\%}$
test_compile_copy_flat[tensordict-eager]	0.1196ms	59.4986μs	16.8071 KOps/s	17.3337 KOps/s	$\color{#d91a1a}-3.04\%$
test_compile_copy_flat[pytree-compile]	0.1619ms	81.3400μs	12.2941 KOps/s	12.5311 KOps/s	$\color{#d91a1a}-1.89\%$
test_compile_copy_flat[pytree-eager]	0.1275ms	68.4117μs	14.6174 KOps/s	14.7641 KOps/s	$\color{#d91a1a}-0.99\%$
test_compile_assign_and_add[tensordict-compile]	0.3037ms	0.2107ms	4.7457 KOps/s	4.7428 KOps/s	$\color{#35bf28}+0.06\%$
test_compile_assign_and_add[tensordict-eager]	1.5675ms	1.3099ms	763.4260 Ops/s	771.0764 Ops/s	$\color{#d91a1a}-0.99\%$
test_compile_assign_and_add[pytree-compile]	0.4097ms	0.2089ms	4.7872 KOps/s	4.8978 KOps/s	$\color{#d91a1a}-2.26\%$
test_compile_assign_and_add[pytree-eager]	1.4165ms	0.7843ms	1.2750 KOps/s	1.2764 KOps/s	$\color{#d91a1a}-0.11\%$
test_compile_assign_and_add_stack[compile]	0.5736ms	0.4607ms	2.1704 KOps/s	2.1695 KOps/s	$\color{#35bf28}+0.04\%$
test_compile_assign_and_add_stack[eager]	3.7087ms	2.5746ms	388.4051 Ops/s	360.9966 Ops/s	$\textbf{\color{#35bf28}+7.59\%}$
test_compile_indexing[tensor-tensordict-compile]	86.7710μs	37.2154μs	26.8706 KOps/s	27.7346 KOps/s	$\color{#d91a1a}-3.12\%$
test_compile_indexing[tensor-tensordict-eager]	0.5072ms	32.0443μs	31.2068 KOps/s	29.6975 KOps/s	$\textbf{\color{#35bf28}+5.08\%}$
test_compile_indexing[tensor-tensorclass-compile]	86.3110μs	30.7627μs	32.5069 KOps/s	33.7074 KOps/s	$\color{#d91a1a}-3.56\%$
test_compile_indexing[tensor-tensorclass-eager]	70.0300μs	22.7817μs	43.8949 KOps/s	41.9018 KOps/s	$\color{#35bf28}+4.76\%$
test_compile_indexing[tensor-pytree-compile]	82.3030μs	31.2648μs	31.9848 KOps/s	32.1741 KOps/s	$\color{#d91a1a}-0.59\%$
test_compile_indexing[tensor-pytree-eager]	84.7470μs	23.1040μs	43.2825 KOps/s	42.2989 KOps/s	$\color{#35bf28}+2.33\%$
test_compile_indexing[slice-tensordict-compile]	0.1138ms	51.6761μs	19.3513 KOps/s	19.3391 KOps/s	$\color{#35bf28}+0.06\%$
test_compile_indexing[slice-tensordict-eager]	0.5266ms	19.5174μs	51.2363 KOps/s	49.5024 KOps/s	$\color{#35bf28}+3.50\%$
test_compile_indexing[slice-tensorclass-compile]	97.8720μs	44.3731μs	22.5362 KOps/s	22.8456 KOps/s	$\color{#d91a1a}-1.35\%$
test_compile_indexing[slice-tensorclass-eager]	0.1001ms	18.9954μs	52.6444 KOps/s	52.8330 KOps/s	$\color{#d91a1a}-0.36\%$
test_compile_indexing[slice-pytree-compile]	0.1071ms	45.7547μs	21.8557 KOps/s	22.2612 KOps/s	$\color{#d91a1a}-1.82\%$
test_compile_indexing[slice-pytree-eager]	54.8020μs	18.8761μs	52.9770 KOps/s	53.0848 KOps/s	$\color{#d91a1a}-0.20\%$
test_compile_indexing[int-tensordict-compile]	0.1735ms	53.2082μs	18.7941 KOps/s	19.0330 KOps/s	$\color{#d91a1a}-1.26\%$
test_compile_indexing[int-tensordict-eager]	0.9424ms	19.3547μs	51.6671 KOps/s	50.4656 KOps/s	$\color{#35bf28}+2.38\%$
test_compile_indexing[int-tensorclass-compile]	0.1229ms	45.8359μs	21.8170 KOps/s	22.3963 KOps/s	$\color{#d91a1a}-2.59\%$
test_compile_indexing[int-tensorclass-eager]	85.2680μs	18.8996μs	52.9113 KOps/s	53.2338 KOps/s	$\color{#d91a1a}-0.61\%$
test_compile_indexing[int-pytree-compile]	0.1202ms	45.9164μs	21.7787 KOps/s	22.2282 KOps/s	$\color{#d91a1a}-2.02\%$
test_compile_indexing[int-pytree-eager]	0.1083ms	19.0111μs	52.6010 KOps/s	53.6024 KOps/s	$\color{#d91a1a}-1.87\%$
test_mod_add[eager]	73.1470μs	26.6568μs	37.5138 KOps/s	35.4504 KOps/s	$\textbf{\color{#35bf28}+5.82\%}$
test_mod_add[compile]	95.6080μs	45.2552μs	22.0969 KOps/s	21.5891 KOps/s	$\color{#35bf28}+2.35\%$
test_mod_add[compile-overhead]	0.1069ms	45.0004μs	22.2220 KOps/s	21.8358 KOps/s	$\color{#35bf28}+1.77\%$
test_mod_wrap[eager]	0.4248ms	0.2160ms	4.6300 KOps/s	4.5688 KOps/s	$\color{#35bf28}+1.34\%$
test_mod_wrap[compile]	1.7112ms	0.2014ms	4.9653 KOps/s	4.8223 KOps/s	$\color{#35bf28}+2.97\%$
test_mod_wrap[compile-overhead]	1.7755ms	0.2001ms	4.9963 KOps/s	4.8799 KOps/s	$\color{#35bf28}+2.39\%$
test_mod_wrap_and_backward[eager]	15.0048ms	11.1929ms	89.3423 Ops/s	77.7969 Ops/s	$\textbf{\color{#35bf28}+14.84\%}$
test_mod_wrap_and_backward[compile]	16.6407ms	11.3919ms	87.7814 Ops/s	76.1291 Ops/s	$\textbf{\color{#35bf28}+15.31\%}$
test_mod_wrap_and_backward[compile-overhead]	17.4138ms	11.9899ms	83.4039 Ops/s	73.9850 Ops/s	$\textbf{\color{#35bf28}+12.73\%}$
test_seq_add[eager]	0.2235ms	93.2954μs	10.7186 KOps/s	10.3658 KOps/s	$\color{#35bf28}+3.40\%$
test_seq_add[compile]	0.1271ms	59.6013μs	16.7782 KOps/s	16.9614 KOps/s	$\color{#d91a1a}-1.08\%$
test_seq_add[compile-overhead]	0.1122ms	57.4730μs	17.3995 KOps/s	17.0058 KOps/s	$\color{#35bf28}+2.32\%$
test_seq_wrap[eager]	0.7408ms	0.3968ms	2.5204 KOps/s	2.4835 KOps/s	$\color{#35bf28}+1.48\%$
test_seq_wrap[compile]	0.4460ms	0.2250ms	4.4452 KOps/s	4.3903 KOps/s	$\color{#35bf28}+1.25\%$
test_seq_wrap[compile-overhead]	0.3709ms	0.2254ms	4.4373 KOps/s	4.3980 KOps/s	$\color{#35bf28}+0.89\%$
test_func_call_runtime[False-eager]	0.7806ms	0.5580ms	1.7923 KOps/s	1.8341 KOps/s	$\color{#d91a1a}-2.28\%$
test_func_call_runtime[False-compile]	0.8699ms	0.4317ms	2.3162 KOps/s	2.3583 KOps/s	$\color{#d91a1a}-1.79\%$
test_func_call_runtime[False-compile-overhead]	0.8060ms	0.4287ms	2.3325 KOps/s	2.3393 KOps/s	$\color{#d91a1a}-0.29\%$
test_func_call_runtime[True-eager]	1.0603ms	0.7691ms	1.3002 KOps/s	1.3189 KOps/s	$\color{#d91a1a}-1.42\%$
test_func_call_runtime[True-compile]	0.8935ms	0.4663ms	2.1445 KOps/s	2.1440 KOps/s	$\color{#35bf28}+0.02\%$
test_func_call_runtime[True-compile-overhead]	0.8509ms	0.4645ms	2.1526 KOps/s	2.1524 KOps/s	$+0.01\%$
test_func_call_cm_runtime[False-eager]	0.8241ms	0.5513ms	1.8138 KOps/s	1.8469 KOps/s	$\color{#d91a1a}-1.79\%$
test_func_call_cm_runtime[False-compile]	0.9320ms	0.4270ms	2.3422 KOps/s	2.3356 KOps/s	$\color{#35bf28}+0.28\%$
test_func_call_cm_runtime[False-compile-overhead]	0.7998ms	0.4317ms	2.3162 KOps/s	2.3522 KOps/s	$\color{#d91a1a}-1.53\%$
test_func_call_cm_runtime[True-eager]	1.3085ms	0.9124ms	1.0960 KOps/s	1.1020 KOps/s	$\color{#d91a1a}-0.54\%$
test_func_call_cm_runtime[True-compile]	1.1178ms	0.4946ms	2.0220 KOps/s	2.0109 KOps/s	$\color{#35bf28}+0.55\%$
test_func_call_cm_runtime[True-compile-overhead]	0.6091ms	0.4956ms	2.0177 KOps/s	2.0393 KOps/s	$\color{#d91a1a}-1.06\%$
test_vmap_func_call_cm_runtime[eager]	2.5856ms	1.9125ms	522.8853 Ops/s	528.0477 Ops/s	$\color{#d91a1a}-0.98\%$
test_vmap_func_call_cm_runtime[compile]	0.7926ms	0.5224ms	1.9141 KOps/s	1.9018 KOps/s	$\color{#35bf28}+0.64\%$
test_vmap_func_call_cm_runtime[compile-overhead]	0.6947ms	0.5218ms	1.9165 KOps/s	1.8536 KOps/s	$\color{#35bf28}+3.39\%$
test_distributed	0.2734ms	0.1277ms	7.8295 KOps/s	7.7423 KOps/s	$\color{#35bf28}+1.13\%$
test_tdmodule	40.6350μs	18.8986μs	52.9139 KOps/s	50.2387 KOps/s	$\textbf{\color{#35bf28}+5.32\%}$
test_tdmodule_dispatch	64.0700μs	36.5632μs	27.3499 KOps/s	25.4335 KOps/s	$\textbf{\color{#35bf28}+7.54\%}$
test_tdseq	45.0950μs	21.4472μs	46.6261 KOps/s	43.3396 KOps/s	$\textbf{\color{#35bf28}+7.58\%}$
test_tdseq_dispatch	77.0140μs	42.3937μs	23.5884 KOps/s	21.8855 KOps/s	$\textbf{\color{#35bf28}+7.78\%}$
test_instantiation_functorch	1.7030ms	1.5536ms	643.6662 Ops/s	644.1281 Ops/s	$\color{#d91a1a}-0.07\%$
test_exec_functorch	0.3240ms	0.1791ms	5.5836 KOps/s	5.5623 KOps/s	$\color{#35bf28}+0.38\%$
test_exec_functional_call	0.3063ms	0.1774ms	5.6364 KOps/s	5.7575 KOps/s	$\color{#d91a1a}-2.10\%$
test_exec_td_decorator	0.4917ms	0.2328ms	4.2949 KOps/s	4.3695 KOps/s	$\color{#d91a1a}-1.71\%$
test_vmap_mlp_speed_decorator[True-True]	0.9946ms	0.6464ms	1.5471 KOps/s	1.5275 KOps/s	$\color{#35bf28}+1.29\%$
test_vmap_mlp_speed_decorator[True-False]	1.3531ms	0.6438ms	1.5533 KOps/s	1.5678 KOps/s	$\color{#d91a1a}-0.92\%$
test_vmap_mlp_speed_decorator[False-True]	0.9965ms	0.5348ms	1.8697 KOps/s	1.9271 KOps/s	$\color{#d91a1a}-2.98\%$
test_vmap_mlp_speed_decorator[False-False]	0.7328ms	0.5284ms	1.8927 KOps/s	1.9262 KOps/s	$\color{#d91a1a}-1.74\%$
test_to_module_speed[True]	2.1881ms	1.3058ms	765.8382 Ops/s	778.6525 Ops/s	$\color{#d91a1a}-1.65\%$
test_to_module_speed[False]	2.1386ms	1.2949ms	772.2884 Ops/s	803.8592 Ops/s	$\color{#d91a1a}-3.93\%$
test_tc_init	88.4950μs	44.2686μs	22.5894 KOps/s	19.7066 KOps/s	$\textbf{\color{#35bf28}+14.63\%}$
test_tc_init_nested	0.1651ms	87.9831μs	11.3658 KOps/s	9.8737 KOps/s	$\textbf{\color{#35bf28}+15.11\%}$
test_tc_first_layer_tensor	27.6620μs	1.5477μs	646.1076 KOps/s	646.4828 KOps/s	$\color{#d91a1a}-0.06\%$
test_tc_first_layer_nontensor	23.0830μs	4.8016μs	208.2627 KOps/s	206.7035 KOps/s	$\color{#35bf28}+0.75\%$
test_tc_second_layer_tensor	28.3730μs	2.8793μs	347.3042 KOps/s	343.4473 KOps/s	$\color{#35bf28}+1.12\%$
test_tc_second_layer_nontensor	57.3060μs	6.2294μs	160.5285 KOps/s	159.8004 KOps/s	$\color{#35bf28}+0.46\%$
test_unbind	0.2386s	12.9622ms	77.1476 Ops/s	75.0869 Ops/s	$\color{#35bf28}+2.74\%$
test_full_like	15.5450ms	12.2456ms	81.6620 Ops/s	130.8345 Ops/s	$\textbf{\color{#d91a1a}-37.58\%}$
test_zeros_like	15.2822ms	7.5922ms	131.7133 Ops/s	344.8787 Ops/s	$\textbf{\color{#d91a1a}-61.81\%}$
test_ones_like	12.1736ms	7.6456ms	130.7948 Ops/s	288.8177 Ops/s	$\textbf{\color{#d91a1a}-54.71\%}$
test_clone	12.6744ms	9.1825ms	108.9022 Ops/s	189.1269 Ops/s	$\textbf{\color{#d91a1a}-42.42\%}$
test_squeeze	57.2270μs	11.8140μs	84.6454 KOps/s	86.2258 KOps/s	$\color{#d91a1a}-1.83\%$
test_unsqueeze	0.1803ms	87.9719μs	11.3673 KOps/s	11.2393 KOps/s	$\color{#35bf28}+1.14\%$
test_split	0.5017ms	0.1936ms	5.1657 KOps/s	5.3684 KOps/s	$\color{#d91a1a}-3.78\%$
test_permute	0.3106ms	0.2228ms	4.4878 KOps/s	4.5901 KOps/s	$\color{#d91a1a}-2.23\%$
test_stack	29.7855ms	23.8272ms	41.9688 Ops/s	39.2103 Ops/s	$\textbf{\color{#35bf28}+7.04\%}$
test_cat	25.0331ms	23.5335ms	42.4926 Ops/s	39.5533 Ops/s	$\textbf{\color{#35bf28}+7.43\%}$

[github-actions]

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

Total Benchmarks: 229. Improved: $\large\color{#35bf28}9$. Worsened: $\large\color{#d91a1a}14$.

Expand to view detailed results

Name	Max	Mean	Ops	Ops on Repo HEAD	Change
test_plain_set_nested	32.5000μs	11.3883μs	87.8097 KOps/s	89.1362 KOps/s	$\color{#d91a1a}-1.49\%$
test_plain_set_stack_nested	0.1094ms	11.5582μs	86.5187 KOps/s	87.3994 KOps/s	$\color{#d91a1a}-1.01\%$
test_plain_set_nested_inplace	49.9210μs	12.3201μs	81.1679 KOps/s	81.2741 KOps/s	$\color{#d91a1a}-0.13\%$
test_plain_set_stack_nested_inplace	39.3200μs	12.3157μs	81.1972 KOps/s	81.7740 KOps/s	$\color{#d91a1a}-0.71\%$
test_items	40.9810μs	2.8940μs	345.5433 KOps/s	342.5475 KOps/s	$\color{#35bf28}+0.87\%$
test_items_nested	0.3697ms	0.3225ms	3.1004 KOps/s	3.0755 KOps/s	$\color{#35bf28}+0.81\%$
test_items_nested_locked	0.3947ms	0.3240ms	3.0865 KOps/s	3.0399 KOps/s	$\color{#35bf28}+1.53\%$
test_items_nested_leaf	92.6820μs	58.0128μs	17.2376 KOps/s	17.2242 KOps/s	$\color{#35bf28}+0.08\%$
test_items_stack_nested	0.4935ms	0.3233ms	3.0927 KOps/s	3.0097 KOps/s	$\color{#35bf28}+2.76\%$
test_items_stack_nested_leaf	0.2338ms	59.7639μs	16.7325 KOps/s	16.9906 KOps/s	$\color{#d91a1a}-1.52\%$
test_items_stack_nested_locked	0.5208ms	0.3272ms	3.0566 KOps/s	3.0694 KOps/s	$\color{#d91a1a}-0.42\%$
test_keys	29.1810μs	3.4657μs	288.5386 KOps/s	286.1515 KOps/s	$\color{#35bf28}+0.83\%$
test_keys_nested	0.1025ms	69.7941μs	14.3278 KOps/s	14.2681 KOps/s	$\color{#35bf28}+0.42\%$
test_keys_nested_locked	2.8663ms	75.2042μs	13.2971 KOps/s	13.0717 KOps/s	$\color{#35bf28}+1.72\%$
test_keys_nested_leaf	91.5720μs	60.8320μs	16.4387 KOps/s	16.3216 KOps/s	$\color{#35bf28}+0.72\%$
test_keys_stack_nested	0.1064ms	70.7589μs	14.1325 KOps/s	14.1295 KOps/s	$\color{#35bf28}+0.02\%$
test_keys_stack_nested_leaf	0.2260ms	61.6364μs	16.2242 KOps/s	16.1147 KOps/s	$\color{#35bf28}+0.68\%$
test_keys_stack_nested_locked	0.2528ms	75.8371μs	13.1862 KOps/s	13.1087 KOps/s	$\color{#35bf28}+0.59\%$
test_values	28.5588μs	0.8645μs	1.1567 MOps/s	1.1740 MOps/s	$\color{#d91a1a}-1.47\%$
test_values_nested	58.2020μs	31.0739μs	32.1814 KOps/s	32.1401 KOps/s	$\color{#35bf28}+0.13\%$
test_values_nested_locked	0.1163ms	32.6200μs	30.6560 KOps/s	30.6457 KOps/s	$\color{#35bf28}+0.03\%$
test_values_nested_leaf	0.1468ms	33.7138μs	29.6615 KOps/s	29.8668 KOps/s	$\color{#d91a1a}-0.69\%$
test_values_stack_nested	59.4810μs	31.5063μs	31.7397 KOps/s	31.6733 KOps/s	$\color{#35bf28}+0.21\%$
test_values_stack_nested_leaf	89.5410μs	34.4418μs	29.0345 KOps/s	29.3643 KOps/s	$\color{#d91a1a}-1.12\%$
test_values_stack_nested_locked	87.6510μs	33.0175μs	30.2869 KOps/s	30.3634 KOps/s	$\color{#d91a1a}-0.25\%$
test_membership	5.1046μs	0.5067μs	1.9737 MOps/s	1.9656 MOps/s	$\color{#35bf28}+0.41\%$
test_membership_nested	17.0205μs	1.8838μs	530.8398 KOps/s	540.8424 KOps/s	$\color{#d91a1a}-1.85\%$
test_membership_nested_leaf	16.7600μs	1.8935μs	528.1225 KOps/s	532.3517 KOps/s	$\color{#d91a1a}-0.79\%$
test_membership_stacked_nested	40.4500μs	2.0221μs	494.5318 KOps/s	508.8040 KOps/s	$\color{#d91a1a}-2.81\%$
test_membership_stacked_nested_leaf	29.0100μs	1.9932μs	501.7107 KOps/s	516.9342 KOps/s	$\color{#d91a1a}-2.94\%$
test_membership_nested_last	27.7300μs	2.7840μs	359.1937 KOps/s	360.7225 KOps/s	$\color{#d91a1a}-0.42\%$
test_membership_nested_leaf_last	45.7410μs	2.7934μs	357.9896 KOps/s	356.0740 KOps/s	$\color{#35bf28}+0.54\%$
test_membership_stacked_nested_last	26.5510μs	3.2699μs	305.8202 KOps/s	361.8897 KOps/s	$\textbf{\color{#d91a1a}-15.49\%}$
test_membership_stacked_nested_leaf_last	98.8920μs	3.2261μs	309.9762 KOps/s	361.3993 KOps/s	$\textbf{\color{#d91a1a}-14.23\%}$
test_nested_getleaf	1.6967ms	6.0433μs	165.4734 KOps/s	167.2056 KOps/s	$\color{#d91a1a}-1.04\%$
test_nested_get	88.4820μs	5.6562μs	176.7983 KOps/s	174.7884 KOps/s	$\color{#35bf28}+1.15\%$
test_stacked_getleaf	39.7110μs	5.9889μs	166.9761 KOps/s	167.0494 KOps/s	$\color{#d91a1a}-0.04\%$
test_stacked_get	38.0300μs	5.6442μs	177.1722 KOps/s	175.6811 KOps/s	$\color{#35bf28}+0.85\%$
test_nested_getitemleaf	34.7710μs	6.0985μs	163.9745 KOps/s	164.2013 KOps/s	$\color{#d91a1a}-0.14\%$
test_nested_getitem	30.7610μs	5.7551μs	173.7598 KOps/s	172.9363 KOps/s	$\color{#35bf28}+0.48\%$
test_stacked_getitemleaf	61.6810μs	6.1014μs	163.8973 KOps/s	164.5502 KOps/s	$\color{#d91a1a}-0.40\%$
test_stacked_getitem	32.9910μs	5.7812μs	172.9759 KOps/s	173.7539 KOps/s	$\color{#d91a1a}-0.45\%$
test_lock_nested	8.1975ms	0.3749ms	2.6671 KOps/s	2.7364 KOps/s	$\color{#d91a1a}-2.53\%$
test_lock_stack_nested	0.4625ms	0.3314ms	3.0178 KOps/s	2.9871 KOps/s	$\color{#35bf28}+1.03\%$
test_unlock_nested	0.6775ms	0.3056ms	3.2724 KOps/s	3.2891 KOps/s	$\color{#d91a1a}-0.51\%$
test_unlock_stack_nested	0.3908ms	0.2721ms	3.6757 KOps/s	3.6690 KOps/s	$\color{#35bf28}+0.18\%$
test_flatten_speed	0.1452ms	71.6516μs	13.9564 KOps/s	13.8196 KOps/s	$\color{#35bf28}+0.99\%$
test_unflatten_speed	0.4895ms	0.2929ms	3.4139 KOps/s	3.4324 KOps/s	$\color{#d91a1a}-0.54\%$
test_common_ops	1.8711ms	0.6168ms	1.6213 KOps/s	1.6195 KOps/s	$\color{#35bf28}+0.11\%$
test_creation	28.2300μs	1.4660μs	682.1208 KOps/s	682.9713 KOps/s	$\color{#d91a1a}-0.12\%$
test_creation_empty	35.7910μs	9.1000μs	109.8907 KOps/s	112.9340 KOps/s	$\color{#d91a1a}-2.69\%$
test_creation_nested_1	0.2039ms	10.7046μs	93.4178 KOps/s	97.2732 KOps/s	$\color{#d91a1a}-3.96\%$
test_creation_nested_2	0.2041ms	13.2008μs	75.7527 KOps/s	76.5733 KOps/s	$\color{#d91a1a}-1.07\%$
test_clone	55.2510μs	10.3212μs	96.8884 KOps/s	89.7223 KOps/s	$\textbf{\color{#35bf28}+7.99\%}$
test_getitem[int]	2.4140ms	11.0408μs	90.5730 KOps/s	91.8136 KOps/s	$\color{#d91a1a}-1.35\%$
test_getitem[slice_int]	0.1438ms	21.4285μs	46.6669 KOps/s	46.3757 KOps/s	$\color{#35bf28}+0.63\%$
test_getitem[range]	0.1823ms	38.4024μs	26.0400 KOps/s	26.0234 KOps/s	$\color{#35bf28}+0.06\%$
test_getitem[tuple]	0.1189ms	18.9865μs	52.6689 KOps/s	54.3388 KOps/s	$\color{#d91a1a}-3.07\%$
test_getitem[list]	0.1052s	43.0739μs	23.2159 KOps/s	29.2813 KOps/s	$\textbf{\color{#d91a1a}-20.71\%}$
test_setitem_dim[int]	43.6510μs	18.9388μs	52.8015 KOps/s	51.2410 KOps/s	$\color{#35bf28}+3.05\%$
test_setitem_dim[slice_int]	0.1445ms	38.0389μs	26.2889 KOps/s	25.8845 KOps/s	$\color{#35bf28}+1.56\%$
test_setitem_dim[range]	83.2220μs	52.7085μs	18.9723 KOps/s	18.2420 KOps/s	$\color{#35bf28}+4.00\%$
test_setitem_dim[tuple]	53.2010μs	31.3952μs	31.8521 KOps/s	30.2295 KOps/s	$\textbf{\color{#35bf28}+5.37\%}$
test_setitem	0.1844ms	15.8060μs	63.2673 KOps/s	61.5877 KOps/s	$\color{#35bf28}+2.73\%$
test_set	0.1522ms	14.9224μs	67.0135 KOps/s	64.4870 KOps/s	$\color{#35bf28}+3.92\%$
test_set_shared	1.6343ms	0.1476ms	6.7751 KOps/s	6.6528 KOps/s	$\color{#35bf28}+1.84\%$
test_update	0.3991ms	18.2917μs	54.6695 KOps/s	53.0100 KOps/s	$\color{#35bf28}+3.13\%$
test_update_nested	0.2200ms	23.6162μs	42.3438 KOps/s	42.0733 KOps/s	$\color{#35bf28}+0.64\%$
test_update__nested	0.7529ms	24.2966μs	41.1581 KOps/s	40.3456 KOps/s	$\color{#35bf28}+2.01\%$
test_set_nested	0.1861ms	16.0926μs	62.1402 KOps/s	59.2297 KOps/s	$\color{#35bf28}+4.91\%$
test_set_nested_new	0.2179ms	18.4091μs	54.3210 KOps/s	51.9380 KOps/s	$\color{#35bf28}+4.59\%$
test_select	0.2168ms	30.1267μs	33.1932 KOps/s	32.1104 KOps/s	$\color{#35bf28}+3.37\%$
test_select_nested	0.2290ms	41.2994μs	24.2134 KOps/s	23.8543 KOps/s	$\color{#35bf28}+1.51\%$
test_exclude_nested	84.3120μs	58.0883μs	17.2152 KOps/s	16.9448 KOps/s	$\color{#35bf28}+1.60\%$
test_empty[True]	0.3113ms	0.2515ms	3.9766 KOps/s	3.8838 KOps/s	$\color{#35bf28}+2.39\%$
test_empty[False]	7.3741μs	0.7421μs	1.3475 MOps/s	1.3376 MOps/s	$\color{#35bf28}+0.74\%$
test_to	86.6520μs	53.6085μs	18.6538 KOps/s	17.9424 KOps/s	$\color{#35bf28}+3.96\%$
test_to_nonblocking	0.2160ms	45.8219μs	21.8236 KOps/s	21.7702 KOps/s	$\color{#35bf28}+0.25\%$
test_unbind_speed	0.3101ms	0.2294ms	4.3597 KOps/s	4.3309 KOps/s	$\color{#35bf28}+0.67\%$
test_unbind_speed_stack0	0.2976ms	0.2279ms	4.3875 KOps/s	4.2895 KOps/s	$\color{#35bf28}+2.29\%$
test_unbind_speed_stack1	0.1065s	0.6482ms	1.5428 KOps/s	1.5196 KOps/s	$\color{#35bf28}+1.52\%$
test_split	0.1119s	1.6502ms	605.9972 Ops/s	554.8156 Ops/s	$\textbf{\color{#35bf28}+9.22\%}$
test_chunk	0.1084s	1.6418ms	609.1046 Ops/s	665.6010 Ops/s	$\textbf{\color{#d91a1a}-8.49\%}$
test_consolidate[False-None]	0.1128s	2.8957ms	345.3408 Ops/s	322.5240 Ops/s	$\textbf{\color{#35bf28}+7.07\%}$
test_consolidate[default-None]	1.9138ms	1.7063ms	586.0560 Ops/s	591.7415 Ops/s	$\color{#d91a1a}-0.96\%$
test_consolidate[reduce-overhead-None]	1.9124ms	1.7089ms	585.1662 Ops/s	585.8267 Ops/s	$\color{#d91a1a}-0.11\%$
test_consolidate_njt[False-None]	7.0360ms	6.5145ms	153.5040 Ops/s	150.9491 Ops/s	$\color{#35bf28}+1.69\%$
test_to[False-False-None]	1.9021ms	1.6288ms	613.9611 Ops/s	600.7517 Ops/s	$\color{#35bf28}+2.20\%$
test_to[True-False-None]	1.6549ms	1.3063ms	765.5494 Ops/s	777.5490 Ops/s	$\color{#d91a1a}-1.54\%$
test_to[within-False-None]	0.3925s	5.5652ms	179.6889 Ops/s	248.4570 Ops/s	$\textbf{\color{#d91a1a}-27.68\%}$
test_to[True-default-None]	5.3996ms	5.0235ms	199.0650 Ops/s	196.9447 Ops/s	$\color{#35bf28}+1.08\%$
test_to_njt[False-False-None]	7.3135ms	6.8848ms	145.2483 Ops/s	143.8930 Ops/s	$\color{#35bf28}+0.94\%$
test_to_njt[True-False-None]	5.9313ms	5.4898ms	182.1565 Ops/s	182.3257 Ops/s	$\color{#d91a1a}-0.09\%$
test_to_njt[within-False-None]	12.8135ms	12.2443ms	81.6709 Ops/s	82.7125 Ops/s	$\color{#d91a1a}-1.26\%$
test_creation[device0]	0.3767ms	79.9599μs	12.5063 KOps/s	11.8890 KOps/s	$\textbf{\color{#35bf28}+5.19\%}$
test_creation_from_tensor	0.4908ms	82.7540μs	12.0840 KOps/s	11.7804 KOps/s	$\color{#35bf28}+2.58\%$
test_add_one[memmap_tensor0]	0.7389ms	7.2326μs	138.2627 KOps/s	134.8497 KOps/s	$\color{#35bf28}+2.53\%$
test_contiguous[memmap_tensor0]	4.0631μs	0.4167μs	2.4001 MOps/s	2.3970 MOps/s	$\color{#35bf28}+0.13\%$
test_stack[memmap_tensor0]	0.1525ms	4.7008μs	212.7311 KOps/s	210.3067 KOps/s	$\color{#35bf28}+1.15\%$
test_memmaptd_index	2.1199ms	0.2540ms	3.9364 KOps/s	3.8637 KOps/s	$\color{#35bf28}+1.88\%$
test_memmaptd_index_astensor	0.8629ms	0.3089ms	3.2372 KOps/s	3.1490 KOps/s	$\color{#35bf28}+2.80\%$
test_memmaptd_index_op	1.1104ms	0.6163ms	1.6225 KOps/s	1.5789 KOps/s	$\color{#35bf28}+2.76\%$
test_serialize_model	0.1325s	0.1316s	7.5993 Ops/s	5.0754 Ops/s	$\textbf{\color{#35bf28}+49.73\%}$
test_serialize_model_pickle	1.4539s	1.2080s	0.8278 Ops/s	0.8387 Ops/s	$\color{#d91a1a}-1.29\%$
test_serialize_weights	0.1329s	0.1312s	7.6196 Ops/s	7.6034 Ops/s	$\color{#35bf28}+0.21\%$
test_serialize_weights_returnearly	0.4422s	72.9851ms	13.7014 Ops/s	23.6464 Ops/s	$\textbf{\color{#d91a1a}-42.06\%}$
test_serialize_weights_pickle	1.3756s	1.2167s	0.8219 Ops/s	0.8218 Ops/s	$\color{#35bf28}+0.01\%$
test_reshape_pytree	0.1644ms	22.6936μs	44.0652 KOps/s	43.6354 KOps/s	$\color{#35bf28}+0.98\%$
test_reshape_td	0.1020ms	26.8232μs	37.2811 KOps/s	37.6644 KOps/s	$\color{#d91a1a}-1.02\%$
test_view_pytree	0.1628ms	22.2938μs	44.8554 KOps/s	44.0888 KOps/s	$\color{#35bf28}+1.74\%$
test_view_td	0.1501ms	30.0140μs	33.3178 KOps/s	32.2702 KOps/s	$\color{#35bf28}+3.25\%$
test_unbind_pytree	0.1830ms	29.5896μs	33.7956 KOps/s	35.0241 KOps/s	$\color{#d91a1a}-3.51\%$
test_unbind_td	0.9932ms	35.5801μs	28.1056 KOps/s	28.0141 KOps/s	$\color{#35bf28}+0.33\%$
test_split_pytree	0.1284ms	31.0076μs	32.2502 KOps/s	32.5514 KOps/s	$\color{#d91a1a}-0.93\%$
test_split_td	0.1762ms	39.1792μs	25.5238 KOps/s	25.1869 KOps/s	$\color{#35bf28}+1.34\%$
test_add_pytree	0.1678ms	35.6635μs	28.0399 KOps/s	27.6133 KOps/s	$\color{#35bf28}+1.54\%$
test_add_td	0.1605ms	50.6014μs	19.7623 KOps/s	19.9620 KOps/s	$\color{#d91a1a}-1.00\%$
test_compile_add_one_nested[tensordict-compile]	0.2680ms	0.1175ms	8.5127 KOps/s	8.1249 KOps/s	$\color{#35bf28}+4.77\%$
test_compile_add_one_nested[tensordict-eager]	0.3104ms	0.1242ms	8.0485 KOps/s	7.8203 KOps/s	$\color{#35bf28}+2.92\%$
test_compile_add_one_nested[pytree-compile]	0.2758ms	97.8208μs	10.2228 KOps/s	10.3662 KOps/s	$\color{#d91a1a}-1.38\%$
test_compile_add_one_nested[pytree-eager]	1.2083ms	0.1487ms	6.7233 KOps/s	6.6740 KOps/s	$\color{#35bf28}+0.74\%$
test_compile_copy_nested[tensordict-compile]	0.2399ms	29.5493μs	33.8418 KOps/s	47.1882 KOps/s	$\textbf{\color{#d91a1a}-28.28\%}$
test_compile_copy_nested[tensordict-eager]	0.1756ms	26.9844μs	37.0584 KOps/s	36.9096 KOps/s	$\color{#35bf28}+0.40\%$
test_compile_copy_nested[pytree-compile]	0.2505ms	64.8102μs	15.4297 KOps/s	15.2210 KOps/s	$\color{#35bf28}+1.37\%$
test_compile_copy_nested[pytree-eager]	0.1969ms	49.3607μs	20.2590 KOps/s	20.1023 KOps/s	$\color{#35bf28}+0.78\%$
test_compile_add_one_flat[tensordict-compile]	0.2872ms	0.1418ms	7.0516 KOps/s	6.8960 KOps/s	$\color{#35bf28}+2.26\%$
test_compile_add_one_flat[tensordict-eager]	0.3530ms	0.2094ms	4.7762 KOps/s	4.8665 KOps/s	$\color{#d91a1a}-1.85\%$
test_compile_add_one_flat[tensorclass-compile]	0.2776ms	99.5097μs	10.0493 KOps/s	10.2604 KOps/s	$\color{#d91a1a}-2.06\%$
test_compile_add_one_flat[tensorclass-eager]	0.2221ms	52.6053μs	19.0095 KOps/s	18.9408 KOps/s	$\color{#35bf28}+0.36\%$
test_compile_add_one_flat[pytree-compile]	0.2840ms	0.1360ms	7.3529 KOps/s	7.2734 KOps/s	$\color{#35bf28}+1.09\%$
test_compile_add_one_flat[pytree-eager]	0.6575ms	0.4793ms	2.0863 KOps/s	2.0723 KOps/s	$\color{#35bf28}+0.68\%$
test_compile_add_self_flat[tensordict-eager]	0.4277ms	0.2462ms	4.0613 KOps/s	4.0658 KOps/s	$\color{#d91a1a}-0.11\%$
test_compile_add_self_flat[tensordict-compile]	0.2803ms	0.1418ms	7.0531 KOps/s	6.9930 KOps/s	$\color{#35bf28}+0.86\%$
test_compile_add_self_flat[tensorclass-eager]	0.2287ms	62.9570μs	15.8839 KOps/s	15.7356 KOps/s	$\color{#35bf28}+0.94\%$
test_compile_add_self_flat[tensorclass-compile]	0.2437ms	98.2499μs	10.1781 KOps/s	10.1487 KOps/s	$\color{#35bf28}+0.29\%$
test_compile_add_self_flat[pytree-eager]	0.5595ms	0.3983ms	2.5107 KOps/s	2.5245 KOps/s	$\color{#d91a1a}-0.55\%$
test_compile_add_self_flat[pytree-compile]	0.2640ms	0.1363ms	7.3371 KOps/s	7.4219 KOps/s	$\color{#d91a1a}-1.14\%$
test_compile_copy_flat[tensordict-compile]	0.1227ms	17.9296μs	55.7737 KOps/s	55.1572 KOps/s	$\color{#35bf28}+1.12\%$
test_compile_copy_flat[tensordict-eager]	63.0010μs	27.2366μs	36.7152 KOps/s	37.0941 KOps/s	$\color{#d91a1a}-1.02\%$
test_compile_copy_flat[pytree-compile]	0.1727ms	69.8884μs	14.3085 KOps/s	14.3390 KOps/s	$\color{#d91a1a}-0.21\%$
test_compile_copy_flat[pytree-eager]	0.1093ms	51.6777μs	19.3507 KOps/s	19.3800 KOps/s	$\color{#d91a1a}-0.15\%$
test_compile_assign_and_add[tensordict-compile]	1.7104ms	0.4669ms	2.1420 KOps/s	2.1924 KOps/s	$\color{#d91a1a}-2.30\%$
test_compile_assign_and_add[tensordict-eager]	2.8783ms	2.5589ms	390.7871 Ops/s	387.8118 Ops/s	$\color{#35bf28}+0.77\%$
test_compile_assign_and_add[pytree-compile]	1.5960ms	0.4360ms	2.2934 KOps/s	2.2146 KOps/s	$\color{#35bf28}+3.56\%$
test_compile_assign_and_add[pytree-eager]	3.0246ms	2.6027ms	384.2122 Ops/s	379.3321 Ops/s	$\color{#35bf28}+1.29\%$
test_compile_indexing[tensor-tensordict-compile]	0.5426ms	0.1181ms	8.4697 KOps/s	8.4597 KOps/s	$\color{#35bf28}+0.12\%$
test_compile_indexing[tensor-tensordict-eager]	0.6013ms	82.4743μs	12.1250 KOps/s	12.4443 KOps/s	$\color{#d91a1a}-2.57\%$
test_compile_indexing[tensor-tensorclass-compile]	0.3029ms	0.1101ms	9.0830 KOps/s	9.2111 KOps/s	$\color{#d91a1a}-1.39\%$
test_compile_indexing[tensor-tensorclass-eager]	0.5063ms	70.3212μs	14.2205 KOps/s	13.6841 KOps/s	$\color{#35bf28}+3.92\%$
test_compile_indexing[tensor-pytree-compile]	0.3202ms	0.1119ms	8.9369 KOps/s	9.4773 KOps/s	$\textbf{\color{#d91a1a}-5.70\%}$
test_compile_indexing[tensor-pytree-eager]	0.2638ms	70.2705μs	14.2307 KOps/s	14.6060 KOps/s	$\color{#d91a1a}-2.57\%$
test_compile_indexing[slice-tensordict-compile]	0.2946ms	0.1049ms	9.5354 KOps/s	9.8343 KOps/s	$\color{#d91a1a}-3.04\%$
test_compile_indexing[slice-tensordict-eager]	0.2158ms	17.2269μs	58.0489 KOps/s	55.8557 KOps/s	$\color{#35bf28}+3.93\%$
test_compile_indexing[slice-tensorclass-compile]	0.2749ms	0.1014ms	9.8638 KOps/s	10.3116 KOps/s	$\color{#d91a1a}-4.34\%$
test_compile_indexing[slice-tensorclass-eager]	0.1435ms	16.1843μs	61.7882 KOps/s	61.2369 KOps/s	$\color{#35bf28}+0.90\%$
test_compile_indexing[slice-pytree-compile]	0.2738ms	0.1025ms	9.7606 KOps/s	9.9729 KOps/s	$\color{#d91a1a}-2.13\%$
test_compile_indexing[slice-pytree-eager]	0.1601ms	16.2736μs	61.4493 KOps/s	61.8302 KOps/s	$\color{#d91a1a}-0.62\%$
test_compile_indexing[int-tensordict-compile]	0.2831ms	0.1040ms	9.6125 KOps/s	9.6846 KOps/s	$\color{#d91a1a}-0.75\%$
test_compile_indexing[int-tensordict-eager]	0.6707ms	17.0136μs	58.7765 KOps/s	57.5651 KOps/s	$\color{#35bf28}+2.10\%$
test_compile_indexing[int-tensorclass-compile]	0.2398ms	98.2744μs	10.1756 KOps/s	9.8469 KOps/s	$\color{#35bf28}+3.34\%$
test_compile_indexing[int-tensorclass-eager]	0.1592ms	16.2935μs	61.3740 KOps/s	61.1839 KOps/s	$\color{#35bf28}+0.31\%$
test_compile_indexing[int-pytree-compile]	0.2743ms	0.1019ms	9.8100 KOps/s	10.2262 KOps/s	$\color{#d91a1a}-4.07\%$
test_compile_indexing[int-pytree-eager]	0.1553ms	16.1676μs	61.8520 KOps/s	61.9217 KOps/s	$\color{#d91a1a}-0.11\%$
test_mod_add[eager]	0.2091ms	35.0275μs	28.5490 KOps/s	30.4882 KOps/s	$\textbf{\color{#d91a1a}-6.36\%}$
test_mod_add[compile]	0.2206ms	77.0749μs	12.9744 KOps/s	12.9497 KOps/s	$\color{#35bf28}+0.19\%$
test_mod_add[compile-overhead]	0.3308ms	0.1755ms	5.6993 KOps/s	5.7671 KOps/s	$\color{#d91a1a}-1.18\%$
test_mod_wrap[eager]	0.3941ms	0.2447ms	4.0861 KOps/s	4.0513 KOps/s	$\color{#35bf28}+0.86\%$
test_mod_wrap[compile]	1.7697ms	0.2933ms	3.4095 KOps/s	3.4830 KOps/s	$\color{#d91a1a}-2.11\%$
test_mod_wrap[compile-overhead]	7.6743ms	4.0743ms	245.4398 Ops/s	238.1531 Ops/s	$\color{#35bf28}+3.06\%$
test_mod_wrap_and_backward[eager]	1.6147ms	1.3410ms	745.7308 Ops/s	695.7402 Ops/s	$\textbf{\color{#35bf28}+7.19\%}$
test_mod_wrap_and_backward[compile]	1.5978ms	1.2870ms	776.9765 Ops/s	723.7585 Ops/s	$\textbf{\color{#35bf28}+7.35\%}$
test_mod_wrap_and_backward[compile-overhead]	1.4096ms	0.9536ms	1.0487 KOps/s	949.4089 Ops/s	$\textbf{\color{#35bf28}+10.46\%}$
test_seq_add[eager]	0.2911ms	99.6050μs	10.0397 KOps/s	10.0313 KOps/s	$\color{#35bf28}+0.08\%$
test_seq_add[compile]	0.2932ms	87.2108μs	11.4665 KOps/s	11.1815 KOps/s	$\color{#35bf28}+2.55\%$
test_seq_add[compile-overhead]	0.2770ms	0.1277ms	7.8298 KOps/s	7.7505 KOps/s	$\color{#35bf28}+1.02\%$
test_seq_wrap[eager]	0.5636ms	0.3887ms	2.5729 KOps/s	2.5652 KOps/s	$\color{#35bf28}+0.30\%$
test_seq_wrap[compile]	0.5118ms	0.3070ms	3.2576 KOps/s	3.1396 KOps/s	$\color{#35bf28}+3.76\%$
test_seq_wrap[compile-overhead]	0.3725ms	0.2247ms	4.4508 KOps/s	4.3020 KOps/s	$\color{#35bf28}+3.46\%$
test_func_call_runtime[False-eager]	0.9010ms	0.7374ms	1.3561 KOps/s	1.3236 KOps/s	$\color{#35bf28}+2.46\%$
test_func_call_runtime[False-compile]	0.9146ms	0.7506ms	1.3324 KOps/s	1.3203 KOps/s	$\color{#35bf28}+0.92\%$
test_func_call_runtime[False-compile-overhead]	0.5126ms	0.3644ms	2.7439 KOps/s	2.7241 KOps/s	$\color{#35bf28}+0.73\%$
test_func_call_runtime[True-eager]	1.2079ms	0.9069ms	1.1027 KOps/s	1.0989 KOps/s	$\color{#35bf28}+0.34\%$
test_func_call_runtime[True-compile]	0.9648ms	0.7772ms	1.2866 KOps/s	1.2845 KOps/s	$\color{#35bf28}+0.16\%$
test_func_call_runtime[True-compile-overhead]	0.5319ms	0.3856ms	2.5934 KOps/s	2.6090 KOps/s	$\color{#d91a1a}-0.60\%$
test_func_call_cm_runtime[False-eager]	0.9782ms	0.7913ms	1.2638 KOps/s	1.3509 KOps/s	$\textbf{\color{#d91a1a}-6.45\%}$
test_func_call_cm_runtime[False-compile]	1.0223ms	0.7706ms	1.2978 KOps/s	1.3192 KOps/s	$\color{#d91a1a}-1.63\%$
test_func_call_cm_runtime[False-compile-overhead]	0.5093ms	0.3660ms	2.7324 KOps/s	2.7032 KOps/s	$\color{#35bf28}+1.08\%$
test_func_call_cm_runtime[True-eager]	1.1619ms	1.0000ms	1.0000 KOps/s	989.7729 Ops/s	$\color{#35bf28}+1.04\%$
test_func_call_cm_runtime[True-compile]	1.0028ms	0.8065ms	1.2400 KOps/s	1.2313 KOps/s	$\color{#35bf28}+0.71\%$
test_func_call_cm_runtime[True-compile-overhead]	0.6001ms	0.4113ms	2.4312 KOps/s	2.4091 KOps/s	$\color{#35bf28}+0.92\%$
test_vmap_func_call_cm_runtime[eager]	2.5033ms	2.0557ms	486.4499 Ops/s	478.5895 Ops/s	$\color{#35bf28}+1.64\%$
test_vmap_func_call_cm_runtime[compile]	0.9882ms	0.8112ms	1.2328 KOps/s	1.2236 KOps/s	$\color{#35bf28}+0.75\%$
test_vmap_func_call_cm_runtime[compile-overhead]	0.5745ms	0.4142ms	2.4142 KOps/s	2.3688 KOps/s	$\color{#35bf28}+1.91\%$
test_distributed	1.7962ms	0.1261ms	7.9298 KOps/s	8.3300 KOps/s	$\color{#d91a1a}-4.80\%$
test_tdmodule	0.1286ms	15.7193μs	63.6160 KOps/s	69.5759 KOps/s	$\textbf{\color{#d91a1a}-8.57\%}$
test_tdmodule_dispatch	0.1348ms	30.4377μs	32.8540 KOps/s	35.8400 KOps/s	$\textbf{\color{#d91a1a}-8.33\%}$
test_tdseq	36.5910μs	16.6913μs	59.9114 KOps/s	63.8164 KOps/s	$\textbf{\color{#d91a1a}-6.12\%}$
test_tdseq_dispatch	56.4110μs	33.3502μs	29.9848 KOps/s	31.4026 KOps/s	$\color{#d91a1a}-4.51\%$
test_instantiation_functorch	1.7225ms	1.5626ms	639.9490 Ops/s	641.5820 Ops/s	$\color{#d91a1a}-0.25\%$
test_exec_functorch	0.2374ms	0.1465ms	6.8252 KOps/s	6.6015 KOps/s	$\color{#35bf28}+3.39\%$
test_exec_functional_call	0.3119ms	0.1448ms	6.9072 KOps/s	6.7344 KOps/s	$\color{#35bf28}+2.57\%$
test_exec_td_decorator	0.3776ms	0.1871ms	5.3460 KOps/s	5.2221 KOps/s	$\color{#35bf28}+2.37\%$
test_vmap_mlp_speed_decorator[True-True]	0.8811ms	0.6915ms	1.4460 KOps/s	1.4844 KOps/s	$\color{#d91a1a}-2.58\%$
test_vmap_mlp_speed_decorator[True-False]	0.8963ms	0.6898ms	1.4497 KOps/s	1.4792 KOps/s	$\color{#d91a1a}-2.00\%$
test_vmap_mlp_speed_decorator[False-True]	0.7787ms	0.6050ms	1.6530 KOps/s	1.6947 KOps/s	$\color{#d91a1a}-2.46\%$
test_vmap_mlp_speed_decorator[False-False]	0.8053ms	0.6090ms	1.6420 KOps/s	1.6950 KOps/s	$\color{#d91a1a}-3.12\%$
test_vmap_transformer_speed_decorator[True-True]	19.6310ms	19.0673ms	52.4457 Ops/s	52.3863 Ops/s	$\color{#35bf28}+0.11\%$
test_vmap_transformer_speed_decorator[True-False]	19.5216ms	19.1161ms	52.3119 Ops/s	52.3085 Ops/s	$+0.01\%$
test_vmap_transformer_speed_decorator[False-True]	19.5443ms	19.0206ms	52.5746 Ops/s	52.8920 Ops/s	$\color{#d91a1a}-0.60\%$
test_vmap_transformer_speed_decorator[False-False]	19.6738ms	18.9483ms	52.7753 Ops/s	52.5685 Ops/s	$\color{#35bf28}+0.39\%$
test_to_module_speed[True]	2.2845ms	0.9297ms	1.0756 KOps/s	1.0799 KOps/s	$\color{#d91a1a}-0.40\%$
test_to_module_speed[False]	1.0255ms	0.9047ms	1.1053 KOps/s	1.1051 KOps/s	$\color{#35bf28}+0.02\%$
test_tc_init	0.1829ms	35.8820μs	27.8691 KOps/s	27.3924 KOps/s	$\color{#35bf28}+1.74\%$
test_tc_init_nested	0.2379ms	72.4052μs	13.8112 KOps/s	13.4962 KOps/s	$\color{#35bf28}+2.33\%$
test_tc_first_layer_tensor	11.9373μs	0.7015μs	1.4256 MOps/s	1.4374 MOps/s	$\color{#d91a1a}-0.82\%$
test_tc_first_layer_nontensor	34.1910μs	2.2974μs	435.2687 KOps/s	427.0781 KOps/s	$\color{#35bf28}+1.92\%$
test_tc_second_layer_tensor	9.0103μs	1.4068μs	710.8218 KOps/s	713.0397 KOps/s	$\color{#d91a1a}-0.31\%$
test_tc_second_layer_nontensor	25.0510μs	3.0231μs	330.7841 KOps/s	328.0502 KOps/s	$\color{#35bf28}+0.83\%$
test_unbind	6.8783ms	6.6514ms	150.3449 Ops/s	150.6938 Ops/s	$\color{#d91a1a}-0.23\%$
test_full_like	13.4280ms	10.9989ms	90.9182 Ops/s	89.3423 Ops/s	$\color{#35bf28}+1.76\%$
test_zeros_like	6.2730ms	4.9244ms	203.0706 Ops/s	209.4616 Ops/s	$\color{#d91a1a}-3.05\%$
test_ones_like	6.1918ms	4.8955ms	204.2675 Ops/s	202.6660 Ops/s	$\color{#35bf28}+0.79\%$
test_clone	9.5622ms	7.9227ms	126.2203 Ops/s	124.6901 Ops/s	$\color{#35bf28}+1.23\%$
test_squeeze	67.6610μs	9.2297μs	108.3460 KOps/s	106.3564 KOps/s	$\color{#35bf28}+1.87\%$
test_unsqueeze	0.1218ms	72.1058μs	13.8685 KOps/s	14.2932 KOps/s	$\color{#d91a1a}-2.97\%$
test_split	0.2674s	0.2338ms	4.2766 KOps/s	6.2824 KOps/s	$\textbf{\color{#d91a1a}-31.93\%}$
test_permute	0.3614ms	0.1803ms	5.5464 KOps/s	5.6320 KOps/s	$\color{#d91a1a}-1.52\%$
test_stack	55.2714ms	53.7744ms	18.5962 Ops/s	18.7371 Ops/s	$\color{#d91a1a}-0.75\%$
test_cat	58.4928ms	53.9091ms	18.5497 Ops/s	18.6760 Ops/s	$\color{#d91a1a}-0.68\%$