[Feature] TensorDict.separates

[vmoens]

Stack from ghstack (oldest at bottom):

• -> [Feature] TensorDict.separates #1120

[github-actions]

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

Total Benchmarks: 217. Improved: $\large\color{#35bf28}11$. Worsened: $\large\color{#d91a1a}6$.

Expand to view detailed results

Name	Max	Mean	Ops	Ops on Repo HEAD	Change
test_plain_set_nested	54.8730μs	18.2534μs	54.7844 KOps/s	53.4538 KOps/s	$\color{#35bf28}+2.49\%$
test_plain_set_stack_nested	44.6840μs	18.4487μs	54.2045 KOps/s	53.9841 KOps/s	$\color{#35bf28}+0.41\%$
test_plain_set_nested_inplace	52.5090μs	20.2055μs	49.4915 KOps/s	49.7396 KOps/s	$\color{#d91a1a}-0.50\%$
test_plain_set_stack_nested_inplace	57.0870μs	20.0926μs	49.7696 KOps/s	49.3802 KOps/s	$\color{#35bf28}+0.79\%$
test_items	32.6010μs	4.0366μs	247.7338 KOps/s	240.0875 KOps/s	$\color{#35bf28}+3.18\%$
test_items_nested	0.6958ms	0.4035ms	2.4781 KOps/s	2.4751 KOps/s	$\color{#35bf28}+0.12\%$
test_items_nested_locked	0.6421ms	0.4018ms	2.4890 KOps/s	2.4774 KOps/s	$\color{#35bf28}+0.47\%$
test_items_nested_leaf	0.1200ms	71.5049μs	13.9851 KOps/s	14.1024 KOps/s	$\color{#d91a1a}-0.83\%$
test_items_stack_nested	0.4758ms	0.4073ms	2.4553 KOps/s	2.4648 KOps/s	$\color{#d91a1a}-0.38\%$
test_items_stack_nested_leaf	0.1314ms	74.4352μs	13.4345 KOps/s	13.2399 KOps/s	$\color{#35bf28}+1.47\%$
test_items_stack_nested_locked	0.8710ms	0.4045ms	2.4719 KOps/s	2.4991 KOps/s	$\color{#d91a1a}-1.09\%$
test_keys	20.1880μs	3.5139μs	284.5836 KOps/s	283.4931 KOps/s	$\color{#35bf28}+0.38\%$
test_keys_nested	0.2211ms	0.1368ms	7.3082 KOps/s	7.0909 KOps/s	$\color{#35bf28}+3.06\%$
test_keys_nested_locked	0.7053ms	0.1427ms	7.0071 KOps/s	6.9528 KOps/s	$\color{#35bf28}+0.78\%$
test_keys_nested_leaf	0.2010ms	0.1186ms	8.4320 KOps/s	8.3591 KOps/s	$\color{#35bf28}+0.87\%$
test_keys_stack_nested	0.2220ms	0.1366ms	7.3184 KOps/s	7.2103 KOps/s	$\color{#35bf28}+1.50\%$
test_keys_stack_nested_leaf	0.2001ms	0.1175ms	8.5074 KOps/s	8.4359 KOps/s	$\color{#35bf28}+0.85\%$
test_keys_stack_nested_locked	0.2663ms	0.1434ms	6.9745 KOps/s	6.9141 KOps/s	$\color{#35bf28}+0.87\%$
test_values	7.2014μs	1.0522μs	950.3586 KOps/s	920.0374 KOps/s	$\color{#35bf28}+3.30\%$
test_values_nested	0.1022ms	56.1512μs	17.8091 KOps/s	18.2979 KOps/s	$\color{#d91a1a}-2.67\%$
test_values_nested_locked	0.1098ms	56.5727μs	17.6764 KOps/s	18.3782 KOps/s	$\color{#d91a1a}-3.82\%$
test_values_nested_leaf	0.1080ms	60.6712μs	16.4823 KOps/s	16.7485 KOps/s	$\color{#d91a1a}-1.59\%$
test_values_stack_nested	0.1036ms	57.3260μs	17.4441 KOps/s	17.9595 KOps/s	$\color{#d91a1a}-2.87\%$
test_values_stack_nested_leaf	0.1084ms	60.9115μs	16.4173 KOps/s	16.6672 KOps/s	$\color{#d91a1a}-1.50\%$
test_values_stack_nested_locked	0.1137ms	57.0482μs	17.5290 KOps/s	17.9374 KOps/s	$\color{#d91a1a}-2.28\%$
test_membership	16.7920μs	0.8989μs	1.1124 MOps/s	1.1228 MOps/s	$\color{#d91a1a}-0.92\%$
test_membership_nested	22.1920μs	2.9378μs	340.3936 KOps/s	341.3718 KOps/s	$\color{#d91a1a}-0.29\%$
test_membership_nested_leaf	23.2130μs	2.9655μs	337.2141 KOps/s	340.3297 KOps/s	$\color{#d91a1a}-0.92\%$
test_membership_stacked_nested	25.2670μs	2.9695μs	336.7525 KOps/s	344.0093 KOps/s	$\color{#d91a1a}-2.11\%$
test_membership_stacked_nested_leaf	19.3870μs	2.9181μs	342.6880 KOps/s	341.8451 KOps/s	$\color{#35bf28}+0.25\%$
test_membership_nested_last	27.2010μs	4.2379μs	235.9634 KOps/s	234.6322 KOps/s	$\color{#35bf28}+0.57\%$
test_membership_nested_leaf_last	23.9550μs	4.2303μs	236.3905 KOps/s	236.8551 KOps/s	$\color{#d91a1a}-0.20\%$
test_membership_stacked_nested_last	28.9840μs	4.2266μs	236.5945 KOps/s	205.2153 KOps/s	$\textbf{\color{#35bf28}+15.29\%}$
test_membership_stacked_nested_leaf_last	25.1580μs	4.2446μs	235.5927 KOps/s	205.9137 KOps/s	$\textbf{\color{#35bf28}+14.41\%}$
test_nested_getleaf	39.4340μs	11.0945μs	90.1347 KOps/s	92.9675 KOps/s	$\color{#d91a1a}-3.05\%$
test_nested_get	57.1690μs	10.4648μs	95.5580 KOps/s	96.1941 KOps/s	$\color{#d91a1a}-0.66\%$
test_stacked_getleaf	37.6300μs	10.9525μs	91.3032 KOps/s	91.5084 KOps/s	$\color{#d91a1a}-0.22\%$
test_stacked_get	37.6200μs	10.4367μs	95.8156 KOps/s	98.9708 KOps/s	$\color{#d91a1a}-3.19\%$
test_nested_getitemleaf	31.4890μs	11.5220μs	86.7903 KOps/s	88.6199 KOps/s	$\color{#d91a1a}-2.06\%$
test_nested_getitem	35.5760μs	10.6559μs	93.8447 KOps/s	94.8312 KOps/s	$\color{#d91a1a}-1.04\%$
test_stacked_getitemleaf	46.3460μs	11.4298μs	87.4907 KOps/s	89.2929 KOps/s	$\color{#d91a1a}-2.02\%$
test_stacked_getitem	36.2380μs	10.5261μs	95.0015 KOps/s	95.6932 KOps/s	$\color{#d91a1a}-0.72\%$
test_lock_nested	0.9578ms	0.4407ms	2.2689 KOps/s	2.2566 KOps/s	$\color{#35bf28}+0.55\%$
test_lock_stack_nested	0.5081ms	0.4127ms	2.4230 KOps/s	2.4409 KOps/s	$\color{#d91a1a}-0.73\%$
test_unlock_nested	0.7124ms	0.3568ms	2.8026 KOps/s	2.7502 KOps/s	$\color{#35bf28}+1.91\%$
test_unlock_stack_nested	0.7156ms	0.3331ms	3.0018 KOps/s	3.0229 KOps/s	$\color{#d91a1a}-0.70\%$
test_flatten_speed	0.1877ms	96.3639μs	10.3773 KOps/s	10.5936 KOps/s	$\color{#d91a1a}-2.04\%$
test_unflatten_speed	0.7136ms	0.5014ms	1.9944 KOps/s	2.0194 KOps/s	$\color{#d91a1a}-1.24\%$
test_common_ops	4.8915ms	0.7997ms	1.2505 KOps/s	1.2555 KOps/s	$\color{#d91a1a}-0.40\%$
test_creation	28.3030μs	2.2826μs	438.0910 KOps/s	477.9695 KOps/s	$\textbf{\color{#d91a1a}-8.34\%}$
test_creation_empty	41.4880μs	12.1974μs	81.9847 KOps/s	79.0993 KOps/s	$\color{#35bf28}+3.65\%$
test_creation_nested_1	72.7960μs	15.0652μs	66.3784 KOps/s	64.5231 KOps/s	$\color{#35bf28}+2.88\%$
test_creation_nested_2	54.0510μs	19.3344μs	51.7212 KOps/s	49.7255 KOps/s	$\color{#35bf28}+4.01\%$
test_clone	1.4731ms	13.3993μs	74.6310 KOps/s	74.5387 KOps/s	$\color{#35bf28}+0.12\%$
test_getitem[int]	0.8650ms	12.5395μs	79.7482 KOps/s	80.4626 KOps/s	$\color{#d91a1a}-0.89\%$
test_getitem[slice_int]	0.1367ms	24.9826μs	40.0278 KOps/s	40.8468 KOps/s	$\color{#d91a1a}-2.00\%$
test_getitem[range]	0.1736ms	49.6067μs	20.1586 KOps/s	20.2562 KOps/s	$\color{#d91a1a}-0.48\%$
test_getitem[tuple]	0.1315ms	20.3610μs	49.1136 KOps/s	50.0318 KOps/s	$\color{#d91a1a}-1.84\%$
test_getitem[list]	0.1661ms	45.2683μs	22.0905 KOps/s	22.7050 KOps/s	$\color{#d91a1a}-2.71\%$
test_setitem_dim[int]	59.1800μs	25.8309μs	38.7134 KOps/s	39.9616 KOps/s	$\color{#d91a1a}-3.12\%$
test_setitem_dim[slice_int]	0.1083ms	54.1535μs	18.4660 KOps/s	19.5224 KOps/s	$\textbf{\color{#d91a1a}-5.41\%}$
test_setitem_dim[range]	0.1146ms	73.6952μs	13.5694 KOps/s	13.5708 KOps/s	$\color{#d91a1a}-0.01\%$
test_setitem_dim[tuple]	69.6910μs	41.6073μs	24.0342 KOps/s	24.6045 KOps/s	$\color{#d91a1a}-2.32\%$
test_setitem	71.6740μs	20.9895μs	47.6428 KOps/s	47.4012 KOps/s	$\color{#35bf28}+0.51\%$
test_set	0.1140ms	20.3896μs	49.0447 KOps/s	48.3433 KOps/s	$\color{#35bf28}+1.45\%$
test_set_shared	1.1482ms	0.1678ms	5.9596 KOps/s	6.0010 KOps/s	$\color{#d91a1a}-0.69\%$
test_update	0.1381ms	23.2608μs	42.9908 KOps/s	41.9079 KOps/s	$\color{#35bf28}+2.58\%$
test_update_nested	0.1544ms	33.2612μs	30.0650 KOps/s	29.1453 KOps/s	$\color{#35bf28}+3.16\%$
test_update__nested	0.4045ms	31.9766μs	31.2728 KOps/s	31.0052 KOps/s	$\color{#35bf28}+0.86\%$
test_set_nested	0.2230ms	22.6279μs	44.1933 KOps/s	43.4991 KOps/s	$\color{#35bf28}+1.60\%$
test_set_nested_new	0.2118ms	27.4574μs	36.4201 KOps/s	36.6719 KOps/s	$\color{#d91a1a}-0.69\%$
test_select	0.2192ms	45.1976μs	22.1251 KOps/s	22.6826 KOps/s	$\color{#d91a1a}-2.46\%$
test_select_nested	0.1320ms	59.8592μs	16.7059 KOps/s	16.8160 KOps/s	$\color{#d91a1a}-0.65\%$
test_exclude_nested	0.1569ms	79.9248μs	12.5118 KOps/s	12.7405 KOps/s	$\color{#d91a1a}-1.80\%$
test_empty[True]	0.5662ms	0.3847ms	2.5992 KOps/s	2.5830 KOps/s	$\color{#35bf28}+0.63\%$
test_empty[False]	11.3235μs	1.2342μs	810.2670 KOps/s	821.3709 KOps/s	$\color{#d91a1a}-1.35\%$
test_unbind_speed	0.3770ms	0.2606ms	3.8366 KOps/s	3.7770 KOps/s	$\color{#35bf28}+1.58\%$
test_unbind_speed_stack0	0.6239ms	0.2591ms	3.8599 KOps/s	3.8929 KOps/s	$\color{#d91a1a}-0.85\%$
test_unbind_speed_stack1	95.2208ms	0.8180ms	1.2225 KOps/s	1.4590 KOps/s	$\textbf{\color{#d91a1a}-16.21\%}$
test_split	0.1024s	1.7751ms	563.3535 Ops/s	580.7394 Ops/s	$\color{#d91a1a}-2.99\%$
test_chunk	0.1078s	1.7779ms	562.4733 Ops/s	575.8273 Ops/s	$\color{#d91a1a}-2.32\%$
test_consolidate_njt[False-None]	8.9386ms	8.2526ms	121.1745 Ops/s	122.9759 Ops/s	$\color{#d91a1a}-1.46\%$
test_creation[device0]	0.2777ms	92.0276μs	10.8663 KOps/s	10.8229 KOps/s	$\color{#35bf28}+0.40\%$
test_creation_from_tensor	3.6470ms	95.2526μs	10.4984 KOps/s	10.4926 KOps/s	$\color{#35bf28}+0.05\%$
test_add_one[memmap_tensor0]	0.2235ms	4.9249μs	203.0487 KOps/s	210.3187 KOps/s	$\color{#d91a1a}-3.46\%$
test_contiguous[memmap_tensor0]	13.4850μs	0.5120μs	1.9532 MOps/s	1.8595 MOps/s	$\textbf{\color{#35bf28}+5.04\%}$
test_stack[memmap_tensor0]	74.1390μs	3.5333μs	283.0202 KOps/s	285.8714 KOps/s	$\color{#d91a1a}-1.00\%$
test_memmaptd_index	0.9957ms	0.2428ms	4.1185 KOps/s	4.2108 KOps/s	$\color{#d91a1a}-2.19\%$
test_memmaptd_index_astensor	0.5913ms	0.3225ms	3.1005 KOps/s	3.1670 KOps/s	$\color{#d91a1a}-2.10\%$
test_memmaptd_index_op	1.1363ms	0.6105ms	1.6379 KOps/s	1.6636 KOps/s	$\color{#d91a1a}-1.54\%$
test_serialize_model	0.1219s	0.1165s	8.5865 Ops/s	8.5185 Ops/s	$\color{#35bf28}+0.80\%$
test_serialize_model_pickle	0.4635s	0.3972s	2.5177 Ops/s	2.5773 Ops/s	$\color{#d91a1a}-2.31\%$
test_serialize_weights	0.1272s	0.1163s	8.6021 Ops/s	8.7961 Ops/s	$\color{#d91a1a}-2.21\%$
test_serialize_weights_returnearly	0.1678s	0.1597s	6.2629 Ops/s	6.4348 Ops/s	$\color{#d91a1a}-2.67\%$
test_serialize_weights_pickle	1.1998s	0.7098s	1.4089 Ops/s	2.5452 Ops/s	$\textbf{\color{#d91a1a}-44.64\%}$
test_serialize_weights_filesystem	0.1545s	0.1418s	7.0528 Ops/s	6.9516 Ops/s	$\color{#35bf28}+1.46\%$
test_serialize_model_filesystem	0.2544s	0.1575s	6.3496 Ops/s	5.9303 Ops/s	$\textbf{\color{#35bf28}+7.07\%}$
test_reshape_pytree	0.1077ms	26.5726μs	37.6328 KOps/s	35.7231 KOps/s	$\textbf{\color{#35bf28}+5.35\%}$
test_reshape_td	65.1710μs	32.7385μs	30.5451 KOps/s	29.7947 KOps/s	$\color{#35bf28}+2.52\%$
test_view_pytree	64.0600μs	26.7997μs	37.3138 KOps/s	37.1889 KOps/s	$\color{#35bf28}+0.34\%$
test_view_td	81.5120μs	37.5317μs	26.6441 KOps/s	25.5785 KOps/s	$\color{#35bf28}+4.17\%$
test_unbind_pytree	65.5730μs	29.7348μs	33.6306 KOps/s	33.1384 KOps/s	$\color{#35bf28}+1.49\%$
test_unbind_td	0.3495ms	38.4733μs	25.9921 KOps/s	25.5654 KOps/s	$\color{#35bf28}+1.67\%$
test_split_pytree	92.9740μs	29.5295μs	33.8644 KOps/s	33.7814 KOps/s	$\color{#35bf28}+0.25\%$
test_split_td	0.2065ms	44.5621μs	22.4406 KOps/s	22.4993 KOps/s	$\color{#d91a1a}-0.26\%$
test_add_pytree	85.7700μs	35.8553μs	27.8899 KOps/s	27.8485 KOps/s	$\color{#35bf28}+0.15\%$
test_add_td	0.1204ms	58.9386μs	16.9668 KOps/s	17.4549 KOps/s	$\color{#d91a1a}-2.80\%$
test_compile_add_one_nested[tensordict-compile]	0.1863ms	63.5420μs	15.7376 KOps/s	16.0664 KOps/s	$\color{#d91a1a}-2.05\%$
test_compile_add_one_nested[tensordict-eager]	0.4305ms	0.1618ms	6.1805 KOps/s	6.1799 KOps/s	$\color{#35bf28}+0.01\%$
test_compile_add_one_nested[pytree-compile]	0.1151ms	45.9278μs	21.7733 KOps/s	21.7236 KOps/s	$\color{#35bf28}+0.23\%$
test_compile_add_one_nested[pytree-eager]	0.2238ms	0.1200ms	8.3334 KOps/s	8.3978 KOps/s	$\color{#d91a1a}-0.77\%$
test_compile_copy_nested[tensordict-compile]	78.7670μs	26.7138μs	37.4339 KOps/s	37.9560 KOps/s	$\color{#d91a1a}-1.38\%$
test_compile_copy_nested[tensordict-eager]	0.1178ms	54.3315μs	18.4055 KOps/s	18.5156 KOps/s	$\color{#d91a1a}-0.59\%$
test_compile_copy_nested[pytree-compile]	0.1596ms	78.4112μs	12.7533 KOps/s	12.6968 KOps/s	$\color{#35bf28}+0.45\%$
test_compile_copy_nested[pytree-eager]	0.1294ms	67.4611μs	14.8233 KOps/s	14.7467 KOps/s	$\color{#35bf28}+0.52\%$
test_compile_add_one_flat[tensordict-compile]	0.1843ms	0.1050ms	9.5224 KOps/s	9.5247 KOps/s	$\color{#d91a1a}-0.02\%$
test_compile_add_one_flat[tensordict-eager]	0.3552ms	0.2005ms	4.9876 KOps/s	4.9941 KOps/s	$\color{#d91a1a}-0.13\%$
test_compile_add_one_flat[tensorclass-compile]	0.1257ms	45.2383μs	22.1052 KOps/s	22.1112 KOps/s	$\color{#d91a1a}-0.03\%$
test_compile_add_one_flat[tensorclass-eager]	0.4961ms	62.8109μs	15.9208 KOps/s	16.1116 KOps/s	$\color{#d91a1a}-1.18\%$
test_compile_add_one_flat[pytree-compile]	0.1713ms	0.1029ms	9.7206 KOps/s	9.8270 KOps/s	$\color{#d91a1a}-1.08\%$
test_compile_add_one_flat[pytree-eager]	0.3281ms	0.2028ms	4.9317 KOps/s	5.0014 KOps/s	$\color{#d91a1a}-1.39\%$
test_compile_add_self_flat[tensordict-eager]	0.2995ms	0.2111ms	4.7362 KOps/s	4.7499 KOps/s	$\color{#d91a1a}-0.29\%$
test_compile_add_self_flat[tensordict-compile]	0.2371ms	0.1079ms	9.2672 KOps/s	9.5197 KOps/s	$\color{#d91a1a}-2.65\%$
test_compile_add_self_flat[tensorclass-eager]	0.2094ms	56.6629μs	17.6482 KOps/s	18.3124 KOps/s	$\color{#d91a1a}-3.63\%$
test_compile_add_self_flat[tensorclass-compile]	0.2330ms	46.5809μs	21.4680 KOps/s	21.6182 KOps/s	$\color{#d91a1a}-0.69\%$
test_compile_add_self_flat[pytree-eager]	1.6912ms	0.1593ms	6.2793 KOps/s	6.2768 KOps/s	$\color{#35bf28}+0.04\%$
test_compile_add_self_flat[pytree-compile]	0.1844ms	0.1032ms	9.6921 KOps/s	9.7437 KOps/s	$\color{#d91a1a}-0.53\%$
test_compile_copy_flat[tensordict-compile]	62.1560μs	21.6756μs	46.1349 KOps/s	46.4232 KOps/s	$\color{#d91a1a}-0.62\%$
test_compile_copy_flat[tensordict-eager]	0.1540ms	60.1717μs	16.6191 KOps/s	16.9678 KOps/s	$\color{#d91a1a}-2.05\%$
test_compile_copy_flat[pytree-compile]	0.1745ms	80.8754μs	12.3647 KOps/s	12.4447 KOps/s	$\color{#d91a1a}-0.64\%$
test_compile_copy_flat[pytree-eager]	0.1580ms	69.2619μs	14.4379 KOps/s	14.6621 KOps/s	$\color{#d91a1a}-1.53\%$
test_compile_assign_and_add[tensordict-compile]	0.2899ms	0.2054ms	4.8692 KOps/s	4.8653 KOps/s	$\color{#35bf28}+0.08\%$
test_compile_assign_and_add[tensordict-eager]	2.2965ms	1.3213ms	756.8493 Ops/s	774.3006 Ops/s	$\color{#d91a1a}-2.25\%$
test_compile_assign_and_add[pytree-compile]	0.3476ms	0.2025ms	4.9371 KOps/s	4.9037 KOps/s	$\color{#35bf28}+0.68\%$
test_compile_assign_and_add[pytree-eager]	1.4116ms	0.7860ms	1.2723 KOps/s	1.2725 KOps/s	$\color{#d91a1a}-0.02\%$
test_compile_assign_and_add_stack[compile]	0.8104ms	0.4585ms	2.1812 KOps/s	2.2151 KOps/s	$\color{#d91a1a}-1.53\%$
test_compile_assign_and_add_stack[eager]	3.7010ms	2.7891ms	358.5372 Ops/s	359.2957 Ops/s	$\color{#d91a1a}-0.21\%$
test_compile_indexing[tensor-tensordict-compile]	87.5340μs	35.7266μs	27.9904 KOps/s	27.6130 KOps/s	$\color{#35bf28}+1.37\%$
test_compile_indexing[tensor-tensordict-eager]	0.5449ms	34.5334μs	28.9574 KOps/s	30.9649 KOps/s	$\textbf{\color{#d91a1a}-6.48\%}$
test_compile_indexing[tensor-tensorclass-compile]	82.4340μs	29.9334μs	33.4075 KOps/s	34.2046 KOps/s	$\color{#d91a1a}-2.33\%$
test_compile_indexing[tensor-tensorclass-eager]	61.6650μs	23.7634μs	42.0815 KOps/s	43.6802 KOps/s	$\color{#d91a1a}-3.66\%$
test_compile_indexing[tensor-pytree-compile]	89.4770μs	30.6937μs	32.5800 KOps/s	33.7693 KOps/s	$\color{#d91a1a}-3.52\%$
test_compile_indexing[tensor-pytree-eager]	0.1165ms	23.4789μs	42.5915 KOps/s	43.6007 KOps/s	$\color{#d91a1a}-2.31\%$
test_compile_indexing[slice-tensordict-compile]	0.1249ms	51.7991μs	19.3054 KOps/s	19.5317 KOps/s	$\color{#d91a1a}-1.16\%$
test_compile_indexing[slice-tensordict-eager]	0.6168ms	20.2795μs	49.3110 KOps/s	51.2073 KOps/s	$\color{#d91a1a}-3.70\%$
test_compile_indexing[slice-tensorclass-compile]	0.1227ms	43.7861μs	22.8383 KOps/s	22.6450 KOps/s	$\color{#35bf28}+0.85\%$
test_compile_indexing[slice-tensorclass-eager]	92.0520μs	18.7799μs	53.2484 KOps/s	53.4875 KOps/s	$\color{#d91a1a}-0.45\%$
test_compile_indexing[slice-pytree-compile]	0.1246ms	45.0920μs	22.1769 KOps/s	22.0837 KOps/s	$\color{#35bf28}+0.42\%$
test_compile_indexing[slice-pytree-eager]	82.1440μs	18.7534μs	53.3236 KOps/s	53.0449 KOps/s	$\color{#35bf28}+0.53\%$
test_compile_indexing[int-tensordict-compile]	0.1452ms	52.9564μs	18.8835 KOps/s	19.1258 KOps/s	$\color{#d91a1a}-1.27\%$
test_compile_indexing[int-tensordict-eager]	0.9503ms	20.2777μs	49.3152 KOps/s	51.0906 KOps/s	$\color{#d91a1a}-3.47\%$
test_compile_indexing[int-tensorclass-compile]	0.1383ms	44.8435μs	22.2998 KOps/s	22.0719 KOps/s	$\color{#35bf28}+1.03\%$
test_compile_indexing[int-tensorclass-eager]	70.3420μs	18.7543μs	53.3210 KOps/s	52.8427 KOps/s	$\color{#35bf28}+0.91\%$
test_compile_indexing[int-pytree-compile]	0.1135ms	44.9388μs	22.2525 KOps/s	22.1138 KOps/s	$\color{#35bf28}+0.63\%$
test_compile_indexing[int-pytree-eager]	79.1880μs	18.7520μs	53.3276 KOps/s	52.7335 KOps/s	$\color{#35bf28}+1.13\%$
test_mod_add[eager]	0.1171ms	34.9610μs	28.6033 KOps/s	27.4818 KOps/s	$\color{#35bf28}+4.08\%$
test_mod_add[compile]	0.1049ms	47.2363μs	21.1702 KOps/s	21.0926 KOps/s	$\color{#35bf28}+0.37\%$
test_mod_add[compile-overhead]	0.1029ms	47.0607μs	21.2492 KOps/s	21.0630 KOps/s	$\color{#35bf28}+0.88\%$
test_mod_wrap[eager]	0.3543ms	0.2219ms	4.5061 KOps/s	4.3319 KOps/s	$\color{#35bf28}+4.02\%$
test_mod_wrap[compile]	0.3504ms	0.2043ms	4.8944 KOps/s	4.8240 KOps/s	$\color{#35bf28}+1.46\%$
test_mod_wrap[compile-overhead]	0.3981ms	0.2063ms	4.8473 KOps/s	4.8308 KOps/s	$\color{#35bf28}+0.34\%$
test_mod_wrap_and_backward[eager]	12.8880ms	11.1583ms	89.6192 Ops/s	85.8502 Ops/s	$\color{#35bf28}+4.39\%$
test_mod_wrap_and_backward[compile]	12.4491ms	11.0243ms	90.7086 Ops/s	89.7290 Ops/s	$\color{#35bf28}+1.09\%$
test_mod_wrap_and_backward[compile-overhead]	12.0655ms	11.1437ms	89.7365 Ops/s	90.5023 Ops/s	$\color{#d91a1a}-0.85\%$
test_seq_add[eager]	0.2225ms	0.1148ms	8.7110 KOps/s	8.1896 KOps/s	$\textbf{\color{#35bf28}+6.37\%}$
test_seq_add[compile]	0.1381ms	61.7229μs	16.2014 KOps/s	16.0074 KOps/s	$\color{#35bf28}+1.21\%$
test_seq_add[compile-overhead]	0.1359ms	60.0812μs	16.6441 KOps/s	16.5018 KOps/s	$\color{#35bf28}+0.86\%$
test_seq_wrap[eager]	0.7163ms	0.4377ms	2.2848 KOps/s	2.1806 KOps/s	$\color{#35bf28}+4.78\%$
test_seq_wrap[compile]	0.4267ms	0.2259ms	4.4276 KOps/s	4.2571 KOps/s	$\color{#35bf28}+4.01\%$
test_seq_wrap[compile-overhead]	0.4065ms	0.2257ms	4.4309 KOps/s	4.3337 KOps/s	$\color{#35bf28}+2.24\%$
test_func_call_runtime[False-eager]	0.9429ms	0.5521ms	1.8114 KOps/s	1.7833 KOps/s	$\color{#35bf28}+1.57\%$
test_func_call_runtime[False-compile]	0.5252ms	0.4259ms	2.3482 KOps/s	2.3019 KOps/s	$\color{#35bf28}+2.01\%$
test_func_call_runtime[False-compile-overhead]	0.8218ms	0.4266ms	2.3442 KOps/s	2.3147 KOps/s	$\color{#35bf28}+1.28\%$
test_func_call_runtime[True-eager]	0.8892ms	0.7627ms	1.3112 KOps/s	1.2930 KOps/s	$\color{#35bf28}+1.41\%$
test_func_call_runtime[True-compile]	0.5769ms	0.4678ms	2.1376 KOps/s	2.1320 KOps/s	$\color{#35bf28}+0.26\%$
test_func_call_runtime[True-compile-overhead]	0.5714ms	0.4705ms	2.1253 KOps/s	2.1411 KOps/s	$\color{#d91a1a}-0.74\%$
test_func_call_cm_runtime[False-eager]	1.1115ms	0.5555ms	1.8003 KOps/s	1.8240 KOps/s	$\color{#d91a1a}-1.30\%$
test_func_call_cm_runtime[False-compile]	0.5322ms	0.4242ms	2.3572 KOps/s	2.3452 KOps/s	$\color{#35bf28}+0.51\%$
test_func_call_cm_runtime[False-compile-overhead]	0.8900ms	0.4318ms	2.3159 KOps/s	2.3481 KOps/s	$\color{#d91a1a}-1.37\%$
test_func_call_cm_runtime[True-eager]	1.0329ms	0.9019ms	1.1088 KOps/s	1.1196 KOps/s	$\color{#d91a1a}-0.97\%$
test_func_call_cm_runtime[True-compile]	0.7186ms	0.4938ms	2.0253 KOps/s	2.0113 KOps/s	$\color{#35bf28}+0.70\%$
test_func_call_cm_runtime[True-compile-overhead]	0.8794ms	0.4954ms	2.0186 KOps/s	2.0054 KOps/s	$\color{#35bf28}+0.66\%$
test_vmap_func_call_cm_runtime[eager]	2.4579ms	1.8983ms	526.7883 Ops/s	518.1470 Ops/s	$\color{#35bf28}+1.67\%$
test_vmap_func_call_cm_runtime[compile]	0.6540ms	0.5191ms	1.9264 KOps/s	1.8981 KOps/s	$\color{#35bf28}+1.49\%$
test_vmap_func_call_cm_runtime[compile-overhead]	0.7202ms	0.5247ms	1.9058 KOps/s	1.8866 KOps/s	$\color{#35bf28}+1.02\%$
test_distributed	0.2403ms	0.1255ms	7.9709 KOps/s	7.8263 KOps/s	$\color{#35bf28}+1.85\%$
test_tdmodule	99.7860μs	27.6518μs	36.1641 KOps/s	36.0669 KOps/s	$\color{#35bf28}+0.27\%$
test_tdmodule_dispatch	79.6990μs	48.5244μs	20.6082 KOps/s	19.4319 KOps/s	$\textbf{\color{#35bf28}+6.05\%}$
test_tdseq	49.0820μs	27.3153μs	36.6095 KOps/s	35.2808 KOps/s	$\color{#35bf28}+3.77\%$
test_tdseq_dispatch	86.8930μs	53.1970μs	18.7981 KOps/s	18.3843 KOps/s	$\color{#35bf28}+2.25\%$
test_instantiation_functorch	2.0437ms	1.5493ms	645.4560 Ops/s	648.3995 Ops/s	$\color{#d91a1a}-0.45\%$
test_exec_functorch	0.2659ms	0.1811ms	5.5213 KOps/s	5.3988 KOps/s	$\color{#35bf28}+2.27\%$
test_exec_functional_call	0.2722ms	0.1723ms	5.8036 KOps/s	5.6431 KOps/s	$\color{#35bf28}+2.84\%$
test_exec_td_decorator	0.5221ms	0.2301ms	4.3461 KOps/s	4.2733 KOps/s	$\color{#35bf28}+1.70\%$
test_vmap_mlp_speed_decorator[True-True]	1.1949ms	0.6681ms	1.4968 KOps/s	1.4873 KOps/s	$\color{#35bf28}+0.64\%$
test_vmap_mlp_speed_decorator[True-False]	1.0177ms	0.6627ms	1.5089 KOps/s	1.4954 KOps/s	$\color{#35bf28}+0.90\%$
test_vmap_mlp_speed_decorator[False-True]	0.8287ms	0.5301ms	1.8864 KOps/s	1.8639 KOps/s	$\color{#35bf28}+1.21\%$
test_vmap_mlp_speed_decorator[False-False]	0.7041ms	0.5258ms	1.9017 KOps/s	1.8521 KOps/s	$\color{#35bf28}+2.68\%$
test_to_module_speed[True]	1.4183ms	1.2951ms	772.1513 Ops/s	774.9341 Ops/s	$\color{#d91a1a}-0.36\%$
test_to_module_speed[False]	1.3495ms	1.2542ms	797.3329 Ops/s	787.9784 Ops/s	$\color{#35bf28}+1.19\%$
test_tc_init	89.1370μs	46.9950μs	21.2789 KOps/s	20.5525 KOps/s	$\color{#35bf28}+3.53\%$
test_tc_init_nested	0.1549ms	92.5622μs	10.8035 KOps/s	10.4041 KOps/s	$\color{#35bf28}+3.84\%$
test_tc_first_layer_tensor	32.6520μs	1.5551μs	643.0276 KOps/s	657.7241 KOps/s	$\color{#d91a1a}-2.23\%$
test_tc_first_layer_nontensor	44.5730μs	4.8001μs	208.3281 KOps/s	207.6254 KOps/s	$\color{#35bf28}+0.34\%$
test_tc_second_layer_tensor	24.6160μs	2.8755μs	347.7624 KOps/s	348.5899 KOps/s	$\color{#d91a1a}-0.24\%$
test_tc_second_layer_nontensor	49.6430μs	6.1211μs	163.3685 KOps/s	163.9168 KOps/s	$\color{#d91a1a}-0.33\%$
test_unbind	0.2109s	13.9836ms	71.5124 Ops/s	81.4954 Ops/s	$\textbf{\color{#d91a1a}-12.25\%}$
test_full_like	8.1356ms	7.2293ms	138.3251 Ops/s	76.9472 Ops/s	$\textbf{\color{#35bf28}+79.77\%}$
test_zeros_like	3.2371ms	2.7219ms	367.3911 Ops/s	133.0216 Ops/s	$\textbf{\color{#35bf28}+176.19\%}$
test_ones_like	11.9239ms	6.2486ms	160.0349 Ops/s	125.9021 Ops/s	$\textbf{\color{#35bf28}+27.11\%}$
test_clone	12.5400ms	8.0934ms	123.5579 Ops/s	103.1864 Ops/s	$\textbf{\color{#35bf28}+19.74\%}$
test_squeeze	65.4120μs	11.6145μs	86.0995 KOps/s	85.1183 KOps/s	$\color{#35bf28}+1.15\%$
test_unsqueeze	0.2971ms	89.6676μs	11.1523 KOps/s	11.0582 KOps/s	$\color{#35bf28}+0.85\%$
test_split	0.4025ms	0.1938ms	5.1607 KOps/s	5.2258 KOps/s	$\color{#d91a1a}-1.24\%$
test_permute	0.4298ms	0.2211ms	4.5222 KOps/s	4.6250 KOps/s	$\color{#d91a1a}-2.22\%$
test_stack	33.0455ms	26.0787ms	38.3455 Ops/s	39.7281 Ops/s	$\color{#d91a1a}-3.48\%$
test_cat	30.4396ms	26.5291ms	37.6944 Ops/s	38.2788 Ops/s	$\color{#d91a1a}-1.53\%$

[github-actions]

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

Total Benchmarks: 229. Improved: $\large\color{#35bf28}12$. Worsened: $\large\color{#d91a1a}23$.

Expand to view detailed results

Name	Max	Mean	Ops	Ops on Repo HEAD	Change
test_plain_set_nested	33.1910μs	11.1467μs	89.7130 KOps/s	98.7330 KOps/s	$\textbf{\color{#d91a1a}-9.14\%}$
test_plain_set_stack_nested	50.4210μs	11.2150μs	89.1667 KOps/s	99.0278 KOps/s	$\textbf{\color{#d91a1a}-9.96\%}$
test_plain_set_nested_inplace	55.4610μs	11.9775μs	83.4898 KOps/s	90.5493 KOps/s	$\textbf{\color{#d91a1a}-7.80\%}$
test_plain_set_stack_nested_inplace	38.7200μs	12.0119μs	83.2507 KOps/s	91.0921 KOps/s	$\textbf{\color{#d91a1a}-8.61\%}$
test_items	31.4410μs	2.9051μs	344.2199 KOps/s	341.3686 KOps/s	$\color{#35bf28}+0.84\%$
test_items_nested	0.4198ms	0.3505ms	2.8529 KOps/s	2.8562 KOps/s	$\color{#d91a1a}-0.12\%$
test_items_nested_locked	0.4220ms	0.3511ms	2.8484 KOps/s	2.8246 KOps/s	$\color{#35bf28}+0.84\%$
test_items_nested_leaf	81.1810μs	58.0229μs	17.2346 KOps/s	17.2810 KOps/s	$\color{#d91a1a}-0.27\%$
test_items_stack_nested	0.3934ms	0.3515ms	2.8448 KOps/s	2.8356 KOps/s	$\color{#35bf28}+0.32\%$
test_items_stack_nested_leaf	91.4410μs	58.1611μs	17.1936 KOps/s	16.8205 KOps/s	$\color{#35bf28}+2.22\%$
test_items_stack_nested_locked	0.3932ms	0.3533ms	2.8304 KOps/s	2.8493 KOps/s	$\color{#d91a1a}-0.66\%$
test_keys	26.5610μs	3.4505μs	289.8089 KOps/s	284.4356 KOps/s	$\color{#35bf28}+1.89\%$
test_keys_nested	0.1206ms	69.7502μs	14.3369 KOps/s	14.3233 KOps/s	$\color{#35bf28}+0.09\%$
test_keys_nested_locked	0.7609ms	75.2960μs	13.2809 KOps/s	13.0873 KOps/s	$\color{#35bf28}+1.48\%$
test_keys_nested_leaf	96.7310μs	61.1264μs	16.3596 KOps/s	16.2843 KOps/s	$\color{#35bf28}+0.46\%$
test_keys_stack_nested	0.1015ms	70.2266μs	14.2396 KOps/s	14.2895 KOps/s	$\color{#d91a1a}-0.35\%$
test_keys_stack_nested_leaf	91.3910μs	61.6095μs	16.2313 KOps/s	16.0821 KOps/s	$\color{#35bf28}+0.93\%$
test_keys_stack_nested_locked	0.1126ms	75.2680μs	13.2859 KOps/s	13.1579 KOps/s	$\color{#35bf28}+0.97\%$
test_values	5.3217μs	0.8511μs	1.1749 MOps/s	1.1734 MOps/s	$\color{#35bf28}+0.13\%$
test_values_nested	66.4500μs	31.0616μs	32.1941 KOps/s	32.1424 KOps/s	$\color{#35bf28}+0.16\%$
test_values_nested_locked	96.2910μs	32.5385μs	30.7328 KOps/s	30.7212 KOps/s	$\color{#35bf28}+0.04\%$
test_values_nested_leaf	86.7210μs	33.4553μs	29.8907 KOps/s	29.8005 KOps/s	$\color{#35bf28}+0.30\%$
test_values_stack_nested	62.6010μs	31.3183μs	31.9302 KOps/s	31.5761 KOps/s	$\color{#35bf28}+1.12\%$
test_values_stack_nested_leaf	68.9510μs	33.5217μs	29.8314 KOps/s	29.3982 KOps/s	$\color{#35bf28}+1.47\%$
test_values_stack_nested_locked	63.2700μs	32.9127μs	30.3834 KOps/s	30.3179 KOps/s	$\color{#35bf28}+0.22\%$
test_membership	3.2721μs	0.5062μs	1.9753 MOps/s	1.9565 MOps/s	$\color{#35bf28}+0.96\%$
test_membership_nested	24.1100μs	2.0118μs	497.0776 KOps/s	509.2486 KOps/s	$\color{#d91a1a}-2.39\%$
test_membership_nested_leaf	15.6350μs	2.0193μs	495.2302 KOps/s	498.1728 KOps/s	$\color{#d91a1a}-0.59\%$
test_membership_stacked_nested	44.0300μs	2.0827μs	480.1489 KOps/s	490.9721 KOps/s	$\color{#d91a1a}-2.20\%$
test_membership_stacked_nested_leaf	53.5810μs	2.0685μs	483.4305 KOps/s	489.5933 KOps/s	$\color{#d91a1a}-1.26\%$
test_membership_nested_last	34.3310μs	2.9252μs	341.8560 KOps/s	344.5376 KOps/s	$\color{#d91a1a}-0.78\%$
test_membership_nested_leaf_last	27.5700μs	2.9378μs	340.3882 KOps/s	341.7900 KOps/s	$\color{#d91a1a}-0.41\%$
test_membership_stacked_nested_last	26.5000μs	2.9297μs	341.3377 KOps/s	281.2714 KOps/s	$\textbf{\color{#35bf28}+21.36\%}$
test_membership_stacked_nested_leaf_last	39.3710μs	2.9710μs	336.5889 KOps/s	279.9409 KOps/s	$\textbf{\color{#35bf28}+20.24\%}$
test_nested_getleaf	34.9400μs	6.1009μs	163.9115 KOps/s	163.7321 KOps/s	$\color{#35bf28}+0.11\%$
test_nested_get	39.5800μs	5.8402μs	171.2281 KOps/s	172.0528 KOps/s	$\color{#d91a1a}-0.48\%$
test_stacked_getleaf	39.4800μs	6.1113μs	163.6326 KOps/s	164.2807 KOps/s	$\color{#d91a1a}-0.39\%$
test_stacked_get	41.7600μs	5.8290μs	171.5573 KOps/s	171.2403 KOps/s	$\color{#35bf28}+0.19\%$
test_nested_getitemleaf	35.1200μs	6.2165μs	160.8631 KOps/s	161.3817 KOps/s	$\color{#d91a1a}-0.32\%$
test_nested_getitem	37.9900μs	5.8564μs	170.7532 KOps/s	169.5891 KOps/s	$\color{#35bf28}+0.69\%$
test_stacked_getitemleaf	40.5400μs	6.1637μs	162.2401 KOps/s	160.9330 KOps/s	$\color{#35bf28}+0.81\%$
test_stacked_getitem	36.1000μs	5.8769μs	170.1570 KOps/s	170.2291 KOps/s	$\color{#d91a1a}-0.04\%$
test_lock_nested	0.9253ms	0.3590ms	2.7858 KOps/s	2.6923 KOps/s	$\color{#35bf28}+3.47\%$
test_lock_stack_nested	0.3953ms	0.3345ms	2.9894 KOps/s	2.9782 KOps/s	$\color{#35bf28}+0.37\%$
test_unlock_nested	0.6504ms	0.3031ms	3.2989 KOps/s	3.2788 KOps/s	$\color{#35bf28}+0.61\%$
test_unlock_stack_nested	0.3089ms	0.2739ms	3.6515 KOps/s	3.6734 KOps/s	$\color{#d91a1a}-0.60\%$
test_flatten_speed	0.1047ms	74.6648μs	13.3932 KOps/s	13.5151 KOps/s	$\color{#d91a1a}-0.90\%$
test_unflatten_speed	0.3611ms	0.3054ms	3.2742 KOps/s	3.2775 KOps/s	$\color{#d91a1a}-0.10\%$
test_common_ops	1.5965ms	0.6079ms	1.6450 KOps/s	1.7195 KOps/s	$\color{#d91a1a}-4.33\%$
test_creation	0.1074ms	1.4642μs	682.9852 KOps/s	672.6815 KOps/s	$\color{#35bf28}+1.53\%$
test_creation_empty	31.2200μs	8.5971μs	116.3188 KOps/s	153.4251 KOps/s	$\textbf{\color{#d91a1a}-24.19\%}$
test_creation_nested_1	32.3010μs	10.2020μs	98.0197 KOps/s	124.5378 KOps/s	$\textbf{\color{#d91a1a}-21.29\%}$
test_creation_nested_2	44.8700μs	12.6432μs	79.0939 KOps/s	94.0992 KOps/s	$\textbf{\color{#d91a1a}-15.95\%}$
test_clone	0.1159ms	10.3660μs	96.4689 KOps/s	93.6182 KOps/s	$\color{#35bf28}+3.05\%$
test_getitem[int]	1.1478ms	10.3665μs	96.4645 KOps/s	94.5748 KOps/s	$\color{#35bf28}+2.00\%$
test_getitem[slice_int]	0.1258ms	20.5254μs	48.7201 KOps/s	47.6808 KOps/s	$\color{#35bf28}+2.18\%$
test_getitem[range]	0.1387ms	37.6767μs	26.5416 KOps/s	26.3357 KOps/s	$\color{#35bf28}+0.78\%$
test_getitem[tuple]	0.1068ms	17.8308μs	56.0826 KOps/s	55.0087 KOps/s	$\color{#35bf28}+1.95\%$
test_getitem[list]	0.1318ms	32.4783μs	30.7898 KOps/s	30.4360 KOps/s	$\color{#35bf28}+1.16\%$
test_setitem_dim[int]	92.5810μs	18.6076μs	53.7416 KOps/s	54.6371 KOps/s	$\color{#d91a1a}-1.64\%$
test_setitem_dim[slice_int]	62.1700μs	38.1717μs	26.1974 KOps/s	26.6299 KOps/s	$\color{#d91a1a}-1.62\%$
test_setitem_dim[range]	94.1210μs	53.3894μs	18.7303 KOps/s	18.9544 KOps/s	$\color{#d91a1a}-1.18\%$
test_setitem_dim[tuple]	53.6110μs	31.2072μs	32.0439 KOps/s	31.9594 KOps/s	$\color{#35bf28}+0.26\%$
test_setitem	48.0600μs	15.2914μs	65.3961 KOps/s	70.2949 KOps/s	$\textbf{\color{#d91a1a}-6.97\%}$
test_set	0.1244ms	14.8938μs	67.1421 KOps/s	72.5388 KOps/s	$\textbf{\color{#d91a1a}-7.44\%}$
test_set_shared	1.8354ms	0.1455ms	6.8731 KOps/s	6.7878 KOps/s	$\color{#35bf28}+1.26\%$
test_update	0.5192ms	17.9132μs	55.8248 KOps/s	62.2484 KOps/s	$\textbf{\color{#d91a1a}-10.32\%}$
test_update_nested	65.9600μs	22.8441μs	43.7749 KOps/s	47.9582 KOps/s	$\textbf{\color{#d91a1a}-8.72\%}$
test_update__nested	0.4733ms	24.5228μs	40.7784 KOps/s	40.3362 KOps/s	$\color{#35bf28}+1.10\%$
test_set_nested	0.1290ms	16.0782μs	62.1961 KOps/s	66.5039 KOps/s	$\textbf{\color{#d91a1a}-6.48\%}$
test_set_nested_new	0.1415ms	18.2562μs	54.7760 KOps/s	57.5473 KOps/s	$\color{#d91a1a}-4.82\%$
test_select	0.2248ms	30.3088μs	32.9937 KOps/s	34.3869 KOps/s	$\color{#d91a1a}-4.05\%$
test_select_nested	73.2610μs	42.3668μs	23.6034 KOps/s	23.9156 KOps/s	$\color{#d91a1a}-1.31\%$
test_exclude_nested	0.1129ms	62.1018μs	16.1026 KOps/s	16.4909 KOps/s	$\color{#d91a1a}-2.35\%$
test_empty[True]	0.3402ms	0.2721ms	3.6746 KOps/s	3.6261 KOps/s	$\color{#35bf28}+1.34\%$
test_empty[False]	3.8280μs	0.7477μs	1.3375 MOps/s	1.3327 MOps/s	$\color{#35bf28}+0.36\%$
test_to	86.6810μs	54.9607μs	18.1948 KOps/s	17.7672 KOps/s	$\color{#35bf28}+2.41\%$
test_to_nonblocking	76.4710μs	45.9196μs	21.7772 KOps/s	21.5986 KOps/s	$\color{#35bf28}+0.83\%$
test_unbind_speed	0.3121ms	0.2244ms	4.4569 KOps/s	4.4885 KOps/s	$\color{#d91a1a}-0.70\%$
test_unbind_speed_stack0	0.3085ms	0.2286ms	4.3743 KOps/s	4.3448 KOps/s	$\color{#35bf28}+0.68\%$
test_unbind_speed_stack1	99.9329ms	0.6493ms	1.5402 KOps/s	1.5605 KOps/s	$\color{#d91a1a}-1.30\%$
test_split	98.4185ms	1.7174ms	582.2831 Ops/s	586.1055 Ops/s	$\color{#d91a1a}-0.65\%$
test_chunk	0.1010s	1.5883ms	629.5948 Ops/s	699.1171 Ops/s	$\textbf{\color{#d91a1a}-9.94\%}$
test_consolidate[False-None]	2.7042ms	2.6172ms	382.0821 Ops/s	338.4213 Ops/s	$\textbf{\color{#35bf28}+12.90\%}$
test_consolidate[default-None]	2.2738ms	1.6991ms	588.5380 Ops/s	580.3087 Ops/s	$\color{#35bf28}+1.42\%$
test_consolidate[reduce-overhead-None]	1.8776ms	1.7247ms	579.8267 Ops/s	565.8293 Ops/s	$\color{#35bf28}+2.47\%$
test_consolidate_njt[False-None]	6.6927ms	6.4878ms	154.1365 Ops/s	152.2057 Ops/s	$\color{#35bf28}+1.27\%$
test_to[False-False-None]	1.7934ms	1.6737ms	597.4822 Ops/s	593.5639 Ops/s	$\color{#35bf28}+0.66\%$
test_to[True-False-None]	1.5830ms	1.2829ms	779.4767 Ops/s	796.1403 Ops/s	$\color{#d91a1a}-2.09\%$
test_to[within-False-None]	4.3024ms	4.0348ms	247.8444 Ops/s	250.0095 Ops/s	$\color{#d91a1a}-0.87\%$
test_to[True-default-None]	5.3168ms	5.0127ms	199.4932 Ops/s	194.1921 Ops/s	$\color{#35bf28}+2.73\%$
test_to_njt[False-False-None]	6.9762ms	6.8557ms	145.8647 Ops/s	143.8748 Ops/s	$\color{#35bf28}+1.38\%$
test_to_njt[True-False-None]	6.4178ms	5.4077ms	184.9206 Ops/s	182.9370 Ops/s	$\color{#35bf28}+1.08\%$
test_to_njt[within-False-None]	12.0288ms	11.8752ms	84.2094 Ops/s	82.8409 Ops/s	$\color{#35bf28}+1.65\%$
test_creation[device0]	0.4707ms	78.2463μs	12.7802 KOps/s	12.3740 KOps/s	$\color{#35bf28}+3.28\%$
test_creation_from_tensor	0.5274ms	82.2823μs	12.1533 KOps/s	11.9020 KOps/s	$\color{#35bf28}+2.11\%$
test_add_one[memmap_tensor0]	0.2233ms	6.7328μs	148.5260 KOps/s	142.9803 KOps/s	$\color{#35bf28}+3.88\%$
test_contiguous[memmap_tensor0]	2.0150μs	0.4146μs	2.4117 MOps/s	2.2539 MOps/s	$\textbf{\color{#35bf28}+7.00\%}$
test_stack[memmap_tensor0]	43.4910μs	4.3282μs	231.0435 KOps/s	222.8131 KOps/s	$\color{#35bf28}+3.69\%$
test_memmaptd_index	1.8960ms	0.2426ms	4.1223 KOps/s	3.9922 KOps/s	$\color{#35bf28}+3.26\%$
test_memmaptd_index_astensor	0.5773ms	0.3000ms	3.3337 KOps/s	3.2170 KOps/s	$\color{#35bf28}+3.63\%$
test_memmaptd_index_op	1.0666ms	0.5816ms	1.7195 KOps/s	1.7631 KOps/s	$\color{#d91a1a}-2.47\%$
test_serialize_model	0.1309s	0.1301s	7.6874 Ops/s	7.6704 Ops/s	$\color{#35bf28}+0.22\%$
test_serialize_model_pickle	1.3822s	1.2192s	0.8202 Ops/s	0.8199 Ops/s	$\color{#35bf28}+0.04\%$
test_serialize_weights	0.1307s	0.1300s	7.6919 Ops/s	7.7268 Ops/s	$\color{#d91a1a}-0.45\%$
test_serialize_weights_returnearly	0.5245s	72.8446ms	13.7279 Ops/s	15.3146 Ops/s	$\textbf{\color{#d91a1a}-10.36\%}$
test_serialize_weights_pickle	1.3826s	1.2200s	0.8196 Ops/s	0.8207 Ops/s	$\color{#d91a1a}-0.13\%$
test_reshape_pytree	54.0500μs	22.1331μs	45.1812 KOps/s	43.7270 KOps/s	$\color{#35bf28}+3.33\%$
test_reshape_td	61.1500μs	26.0000μs	38.4616 KOps/s	36.7920 KOps/s	$\color{#35bf28}+4.54\%$
test_view_pytree	46.5110μs	22.1915μs	45.0623 KOps/s	44.9582 KOps/s	$\color{#35bf28}+0.23\%$
test_view_td	64.4810μs	28.9856μs	34.4999 KOps/s	34.0358 KOps/s	$\color{#35bf28}+1.36\%$
test_unbind_pytree	0.1570ms	27.5893μs	36.2460 KOps/s	35.2087 KOps/s	$\color{#35bf28}+2.95\%$
test_unbind_td	0.7968ms	34.4746μs	29.0069 KOps/s	28.5286 KOps/s	$\color{#35bf28}+1.68\%$
test_split_pytree	61.3300μs	29.3730μs	34.0449 KOps/s	32.8563 KOps/s	$\color{#35bf28}+3.62\%$
test_split_td	1.0103ms	37.5055μs	26.6628 KOps/s	25.8675 KOps/s	$\color{#35bf28}+3.07\%$
test_add_pytree	75.3800μs	34.0507μs	29.3679 KOps/s	29.1994 KOps/s	$\color{#35bf28}+0.58\%$
test_add_td	71.5410μs	44.2670μs	22.5902 KOps/s	23.2661 KOps/s	$\color{#d91a1a}-2.91\%$
test_compile_add_one_nested[tensordict-compile]	0.1755ms	0.1185ms	8.4424 KOps/s	8.0261 KOps/s	$\textbf{\color{#35bf28}+5.19\%}$
test_compile_add_one_nested[tensordict-eager]	0.2441ms	0.1208ms	8.2780 KOps/s	7.8651 KOps/s	$\textbf{\color{#35bf28}+5.25\%}$
test_compile_add_one_nested[pytree-compile]	0.1338ms	93.5971μs	10.6841 KOps/s	10.4183 KOps/s	$\color{#35bf28}+2.55\%$
test_compile_add_one_nested[pytree-eager]	0.2060ms	0.1457ms	6.8615 KOps/s	6.6547 KOps/s	$\color{#35bf28}+3.11\%$
test_compile_copy_nested[tensordict-compile]	55.2900μs	21.3776μs	46.7780 KOps/s	43.7365 KOps/s	$\textbf{\color{#35bf28}+6.95\%}$
test_compile_copy_nested[tensordict-eager]	55.1810μs	26.6561μs	37.5148 KOps/s	37.0051 KOps/s	$\color{#35bf28}+1.38\%$
test_compile_copy_nested[pytree-compile]	0.1021ms	63.8072μs	15.6722 KOps/s	15.2538 KOps/s	$\color{#35bf28}+2.74\%$
test_compile_copy_nested[pytree-eager]	0.1768ms	49.0705μs	20.3788 KOps/s	19.8420 KOps/s	$\color{#35bf28}+2.71\%$
test_compile_add_one_flat[tensordict-compile]	0.2146ms	0.1406ms	7.1100 KOps/s	6.8546 KOps/s	$\color{#35bf28}+3.73\%$
test_compile_add_one_flat[tensordict-eager]	0.3429ms	0.2064ms	4.8447 KOps/s	4.8116 KOps/s	$\color{#35bf28}+0.69\%$
test_compile_add_one_flat[tensorclass-compile]	0.1406ms	96.1748μs	10.3977 KOps/s	10.2392 KOps/s	$\color{#35bf28}+1.55\%$
test_compile_add_one_flat[tensorclass-eager]	0.1364ms	51.4987μs	19.4180 KOps/s	19.0413 KOps/s	$\color{#35bf28}+1.98\%$
test_compile_add_one_flat[pytree-compile]	0.1796ms	0.1341ms	7.4543 KOps/s	7.2964 KOps/s	$\color{#35bf28}+2.17\%$
test_compile_add_one_flat[pytree-eager]	0.6392ms	0.4723ms	2.1171 KOps/s	2.0627 KOps/s	$\color{#35bf28}+2.64\%$
test_compile_add_self_flat[tensordict-eager]	0.3593ms	0.2455ms	4.0729 KOps/s	4.0295 KOps/s	$\color{#35bf28}+1.08\%$
test_compile_add_self_flat[tensordict-compile]	0.1853ms	0.1430ms	6.9928 KOps/s	7.0185 KOps/s	$\color{#d91a1a}-0.37\%$
test_compile_add_self_flat[tensorclass-eager]	0.1584ms	62.6732μs	15.9558 KOps/s	15.9463 KOps/s	$\color{#35bf28}+0.06\%$
test_compile_add_self_flat[tensorclass-compile]	0.1451ms	99.2794μs	10.0726 KOps/s	10.1076 KOps/s	$\color{#d91a1a}-0.35\%$
test_compile_add_self_flat[pytree-eager]	0.5485ms	0.4042ms	2.4741 KOps/s	2.4646 KOps/s	$\color{#35bf28}+0.38\%$
test_compile_add_self_flat[pytree-compile]	0.2575ms	0.1402ms	7.1303 KOps/s	7.3957 KOps/s	$\color{#d91a1a}-3.59\%$
test_compile_copy_flat[tensordict-compile]	56.4710μs	17.6231μs	56.7436 KOps/s	55.9402 KOps/s	$\color{#35bf28}+1.44\%$
test_compile_copy_flat[tensordict-eager]	0.1039ms	26.9964μs	37.0420 KOps/s	37.7281 KOps/s	$\color{#d91a1a}-1.82\%$
test_compile_copy_flat[pytree-compile]	0.2035ms	69.8711μs	14.3121 KOps/s	14.2676 KOps/s	$\color{#35bf28}+0.31\%$
test_compile_copy_flat[pytree-eager]	88.1710μs	51.3935μs	19.4577 KOps/s	19.4944 KOps/s	$\color{#d91a1a}-0.19\%$
test_compile_assign_and_add[tensordict-compile]	1.6560ms	0.3960ms	2.5252 KOps/s	2.0945 KOps/s	$\textbf{\color{#35bf28}+20.56\%}$
test_compile_assign_and_add[tensordict-eager]	2.8714ms	2.5665ms	389.6327 Ops/s	376.3000 Ops/s	$\color{#35bf28}+3.54\%$
test_compile_assign_and_add[pytree-compile]	1.6259ms	0.4450ms	2.2473 KOps/s	2.2082 KOps/s	$\color{#35bf28}+1.77\%$
test_compile_assign_and_add[pytree-eager]	2.8439ms	2.5976ms	384.9757 Ops/s	383.6552 Ops/s	$\color{#35bf28}+0.34\%$
test_compile_indexing[tensor-tensordict-compile]	0.5408ms	0.1179ms	8.4796 KOps/s	8.6859 KOps/s	$\color{#d91a1a}-2.37\%$
test_compile_indexing[tensor-tensordict-eager]	0.5762ms	79.2890μs	12.6121 KOps/s	12.5522 KOps/s	$\color{#35bf28}+0.48\%$
test_compile_indexing[tensor-tensorclass-compile]	0.6012ms	0.1087ms	9.1962 KOps/s	9.0866 KOps/s	$\color{#35bf28}+1.21\%$
test_compile_indexing[tensor-tensorclass-eager]	0.2162ms	67.9074μs	14.7259 KOps/s	14.1059 KOps/s	$\color{#35bf28}+4.40\%$
test_compile_indexing[tensor-pytree-compile]	0.2611ms	0.1139ms	8.7833 KOps/s	8.9285 KOps/s	$\color{#d91a1a}-1.63\%$
test_compile_indexing[tensor-pytree-eager]	0.2222ms	70.6783μs	14.1486 KOps/s	14.1032 KOps/s	$\color{#35bf28}+0.32\%$
test_compile_indexing[slice-tensordict-compile]	0.1460ms	0.1004ms	9.9642 KOps/s	9.9176 KOps/s	$\color{#35bf28}+0.47\%$
test_compile_indexing[slice-tensordict-eager]	0.1402ms	16.9702μs	58.9269 KOps/s	51.0413 KOps/s	$\textbf{\color{#35bf28}+15.45\%}$
test_compile_indexing[slice-tensorclass-compile]	0.1492ms	99.8473μs	10.0153 KOps/s	9.8681 KOps/s	$\color{#35bf28}+1.49\%$
test_compile_indexing[slice-tensorclass-eager]	51.4300μs	17.2116μs	58.1004 KOps/s	60.8027 KOps/s	$\color{#d91a1a}-4.44\%$
test_compile_indexing[slice-pytree-compile]	0.1414ms	0.1016ms	9.8415 KOps/s	9.7360 KOps/s	$\color{#35bf28}+1.08\%$
test_compile_indexing[slice-pytree-eager]	0.1532ms	16.7822μs	59.5871 KOps/s	61.7090 KOps/s	$\color{#d91a1a}-3.44\%$
test_compile_indexing[int-tensordict-compile]	0.2597ms	0.1007ms	9.9313 KOps/s	9.3112 KOps/s	$\textbf{\color{#35bf28}+6.66\%}$
test_compile_indexing[int-tensordict-eager]	0.5774ms	16.8592μs	59.3147 KOps/s	57.0920 KOps/s	$\color{#35bf28}+3.89\%$
test_compile_indexing[int-tensorclass-compile]	0.1350ms	96.2489μs	10.3897 KOps/s	10.2477 KOps/s	$\color{#35bf28}+1.39\%$
test_compile_indexing[int-tensorclass-eager]	42.8700μs	15.9928μs	62.5283 KOps/s	63.3775 KOps/s	$\color{#d91a1a}-1.34\%$
test_compile_indexing[int-pytree-compile]	0.2453ms	0.1010ms	9.9014 KOps/s	9.7325 KOps/s	$\color{#35bf28}+1.74\%$
test_compile_indexing[int-pytree-eager]	0.4979ms	16.7455μs	59.7176 KOps/s	62.0440 KOps/s	$\color{#d91a1a}-3.75\%$
test_mod_add[eager]	0.1022ms	37.8617μs	26.4119 KOps/s	27.3349 KOps/s	$\color{#d91a1a}-3.38\%$
test_mod_add[compile]	0.2677ms	84.6622μs	11.8116 KOps/s	12.0766 KOps/s	$\color{#d91a1a}-2.19\%$
test_mod_add[compile-overhead]	0.3387ms	0.1694ms	5.9045 KOps/s	5.3858 KOps/s	$\textbf{\color{#35bf28}+9.63\%}$
test_mod_wrap[eager]	0.4238ms	0.2638ms	3.7909 KOps/s	3.9754 KOps/s	$\color{#d91a1a}-4.64\%$
test_mod_wrap[compile]	0.4219ms	0.2790ms	3.5840 KOps/s	3.2655 KOps/s	$\textbf{\color{#35bf28}+9.75\%}$
test_mod_wrap[compile-overhead]	7.1184ms	3.8114ms	262.3735 Ops/s	263.0406 Ops/s	$\color{#d91a1a}-0.25\%$
test_mod_wrap_and_backward[eager]	1.6413ms	1.4474ms	690.9147 Ops/s	725.4516 Ops/s	$\color{#d91a1a}-4.76\%$
test_mod_wrap_and_backward[compile]	1.4486ms	1.3549ms	738.0695 Ops/s	791.7996 Ops/s	$\textbf{\color{#d91a1a}-6.79\%}$
test_mod_wrap_and_backward[compile-overhead]	1.5987ms	1.0447ms	957.1963 Ops/s	1.0635 KOps/s	$\textbf{\color{#d91a1a}-9.99\%}$
test_seq_add[eager]	0.2616ms	0.1104ms	9.0584 KOps/s	8.8978 KOps/s	$\color{#35bf28}+1.81\%$
test_seq_add[compile]	0.1361ms	86.2874μs	11.5892 KOps/s	11.1929 KOps/s	$\color{#35bf28}+3.54\%$
test_seq_add[compile-overhead]	0.3227ms	0.1272ms	7.8631 KOps/s	7.5861 KOps/s	$\color{#35bf28}+3.65\%$
test_seq_wrap[eager]	0.4783ms	0.4139ms	2.4161 KOps/s	2.3568 KOps/s	$\color{#35bf28}+2.51\%$
test_seq_wrap[compile]	0.4492ms	0.2954ms	3.3848 KOps/s	3.2551 KOps/s	$\color{#35bf28}+3.98\%$
test_seq_wrap[compile-overhead]	0.2710ms	0.2205ms	4.5351 KOps/s	4.4559 KOps/s	$\color{#35bf28}+1.78\%$
test_func_call_runtime[False-eager]	0.8509ms	0.7296ms	1.3706 KOps/s	1.3615 KOps/s	$\color{#35bf28}+0.67\%$
test_func_call_runtime[False-compile]	0.9344ms	0.7489ms	1.3352 KOps/s	1.3589 KOps/s	$\color{#d91a1a}-1.74\%$
test_func_call_runtime[False-compile-overhead]	0.4769ms	0.3610ms	2.7702 KOps/s	2.7306 KOps/s	$\color{#35bf28}+1.45\%$
test_func_call_runtime[True-eager]	1.0303ms	0.8880ms	1.1261 KOps/s	1.1084 KOps/s	$\color{#35bf28}+1.60\%$
test_func_call_runtime[True-compile]	0.9212ms	0.7894ms	1.2668 KOps/s	1.3179 KOps/s	$\color{#d91a1a}-3.88\%$
test_func_call_runtime[True-compile-overhead]	0.4349ms	0.3836ms	2.6068 KOps/s	2.5788 KOps/s	$\color{#35bf28}+1.08\%$
test_func_call_cm_runtime[False-eager]	0.8039ms	0.7230ms	1.3831 KOps/s	1.3495 KOps/s	$\color{#35bf28}+2.50\%$
test_func_call_cm_runtime[False-compile]	0.8433ms	0.7533ms	1.3275 KOps/s	1.3544 KOps/s	$\color{#d91a1a}-1.98\%$
test_func_call_cm_runtime[False-compile-overhead]	0.4156ms	0.3630ms	2.7546 KOps/s	2.7279 KOps/s	$\color{#35bf28}+0.98\%$
test_func_call_cm_runtime[True-eager]	1.1390ms	0.9863ms	1.0138 KOps/s	994.3820 Ops/s	$\color{#35bf28}+1.96\%$
test_func_call_cm_runtime[True-compile]	0.8449ms	0.7898ms	1.2661 KOps/s	1.2681 KOps/s	$\color{#d91a1a}-0.16\%$
test_func_call_cm_runtime[True-compile-overhead]	0.6321ms	0.4123ms	2.4253 KOps/s	2.4059 KOps/s	$\color{#35bf28}+0.81\%$
test_vmap_func_call_cm_runtime[eager]	2.5479ms	2.0599ms	485.4598 Ops/s	478.7605 Ops/s	$\color{#35bf28}+1.40\%$
test_vmap_func_call_cm_runtime[compile]	0.9410ms	0.8035ms	1.2446 KOps/s	1.2191 KOps/s	$\color{#35bf28}+2.10\%$
test_vmap_func_call_cm_runtime[compile-overhead]	0.4583ms	0.4116ms	2.4297 KOps/s	2.3838 KOps/s	$\color{#35bf28}+1.92\%$
test_distributed	2.7510ms	0.2365ms	4.2284 KOps/s	8.7912 KOps/s	$\textbf{\color{#d91a1a}-51.90\%}$
test_tdmodule	32.0010μs	19.0046μs	52.6187 KOps/s	56.5904 KOps/s	$\textbf{\color{#d91a1a}-7.02\%}$
test_tdmodule_dispatch	83.9110μs	34.3159μs	29.1410 KOps/s	31.2856 KOps/s	$\textbf{\color{#d91a1a}-6.85\%}$
test_tdseq	40.0900μs	19.0248μs	52.5629 KOps/s	54.9556 KOps/s	$\color{#d91a1a}-4.35\%$
test_tdseq_dispatch	64.7510μs	37.0104μs	27.0194 KOps/s	28.8575 KOps/s	$\textbf{\color{#d91a1a}-6.37\%}$
test_instantiation_functorch	1.6932ms	1.5634ms	639.6281 Ops/s	636.7230 Ops/s	$\color{#35bf28}+0.46\%$
test_exec_functorch	0.1983ms	0.1447ms	6.9085 KOps/s	6.9167 KOps/s	$\color{#d91a1a}-0.12\%$
test_exec_functional_call	0.2057ms	0.1400ms	7.1448 KOps/s	7.1737 KOps/s	$\color{#d91a1a}-0.40\%$
test_exec_td_decorator	0.3742ms	0.1871ms	5.3447 KOps/s	5.3490 KOps/s	$\color{#d91a1a}-0.08\%$
test_vmap_mlp_speed_decorator[True-True]	0.8963ms	0.6882ms	1.4531 KOps/s	1.4539 KOps/s	$\color{#d91a1a}-0.05\%$
test_vmap_mlp_speed_decorator[True-False]	0.9235ms	0.6861ms	1.4575 KOps/s	1.4504 KOps/s	$\color{#35bf28}+0.49\%$
test_vmap_mlp_speed_decorator[False-True]	0.7216ms	0.5962ms	1.6772 KOps/s	1.6484 KOps/s	$\color{#35bf28}+1.75\%$
test_vmap_mlp_speed_decorator[False-False]	0.8156ms	0.5976ms	1.6734 KOps/s	1.6786 KOps/s	$\color{#d91a1a}-0.31\%$
test_vmap_transformer_speed_decorator[True-True]	20.3638ms	19.4625ms	51.3808 Ops/s	52.3957 Ops/s	$\color{#d91a1a}-1.94\%$
test_vmap_transformer_speed_decorator[True-False]	20.5579ms	19.5417ms	51.1726 Ops/s	52.3197 Ops/s	$\color{#d91a1a}-2.19\%$
test_vmap_transformer_speed_decorator[False-True]	19.8694ms	19.3031ms	51.8050 Ops/s	52.8125 Ops/s	$\color{#d91a1a}-1.91\%$
test_vmap_transformer_speed_decorator[False-False]	19.9713ms	19.2663ms	51.9040 Ops/s	52.6550 Ops/s	$\color{#d91a1a}-1.43\%$
test_to_module_speed[True]	1.2743ms	0.9352ms	1.0693 KOps/s	1.0616 KOps/s	$\color{#35bf28}+0.73\%$
test_to_module_speed[False]	1.4133ms	0.9190ms	1.0881 KOps/s	1.0777 KOps/s	$\color{#35bf28}+0.97\%$
test_tc_init	64.7010μs	35.0830μs	28.5038 KOps/s	30.0350 KOps/s	$\textbf{\color{#d91a1a}-5.10\%}$
test_tc_init_nested	0.1061ms	71.2113μs	14.0427 KOps/s	14.5636 KOps/s	$\color{#d91a1a}-3.58\%$
test_tc_first_layer_tensor	4.7987μs	0.7013μs	1.4260 MOps/s	1.4024 MOps/s	$\color{#35bf28}+1.68\%$
test_tc_first_layer_nontensor	22.9010μs	2.3288μs	429.3968 KOps/s	427.7521 KOps/s	$\color{#35bf28}+0.38\%$
test_tc_second_layer_tensor	22.5778μs	1.4128μs	707.8046 KOps/s	702.4825 KOps/s	$\color{#35bf28}+0.76\%$
test_tc_second_layer_nontensor	0.2796ms	3.0501μs	327.8617 KOps/s	323.6898 KOps/s	$\color{#35bf28}+1.29\%$
test_unbind	0.2301s	10.0434ms	99.5677 Ops/s	151.2010 Ops/s	$\textbf{\color{#d91a1a}-34.15\%}$
test_full_like	13.2578ms	9.7277ms	102.7996 Ops/s	101.3283 Ops/s	$\color{#35bf28}+1.45\%$
test_zeros_like	9.4929ms	7.2851ms	137.2672 Ops/s	227.4447 Ops/s	$\textbf{\color{#d91a1a}-39.65\%}$
test_ones_like	5.3032ms	4.4257ms	225.9514 Ops/s	231.6334 Ops/s	$\color{#d91a1a}-2.45\%$
test_clone	7.7813ms	6.8354ms	146.2981 Ops/s	145.3506 Ops/s	$\color{#35bf28}+0.65\%$
test_squeeze	0.1541ms	9.0631μs	110.3377 KOps/s	108.7411 KOps/s	$\color{#35bf28}+1.47\%$
test_unsqueeze	0.1136ms	66.5784μs	15.0199 KOps/s	14.7476 KOps/s	$\color{#35bf28}+1.85\%$
test_split	0.3972ms	0.1512ms	6.6146 KOps/s	6.4233 KOps/s	$\color{#35bf28}+2.98\%$
test_permute	0.2760ms	0.1710ms	5.8493 KOps/s	5.5920 KOps/s	$\color{#35bf28}+4.60\%$
test_stack	54.2521ms	52.5314ms	19.0362 Ops/s	19.3588 Ops/s	$\color{#d91a1a}-1.67\%$
test_cat	52.2857ms	51.4988ms	19.4179 Ops/s	19.3274 Ops/s	$\color{#35bf28}+0.47\%$