[BugFix] Cloning empty tensordicts

[vmoens]

Stack from ghstack (oldest at bottom):

• [Feature] TensorDict.separates #1120
• -> [BugFix] Cloning empty tensordicts #1119

[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: $\large\color{#d91a1a}11$.

Expand to view detailed results

Name	Max	Mean	Ops	Ops on Repo HEAD	Change
test_plain_set_nested	38.1820μs	18.5676μs	53.8572 KOps/s	53.0714 KOps/s	$\color{#35bf28}+1.48\%$
test_plain_set_stack_nested	45.8660μs	18.9302μs	52.8256 KOps/s	52.8906 KOps/s	$\color{#d91a1a}-0.12\%$
test_plain_set_nested_inplace	53.2690μs	20.5534μs	48.6538 KOps/s	48.6819 KOps/s	$\color{#d91a1a}-0.06\%$
test_plain_set_stack_nested_inplace	58.8910μs	20.6207μs	48.4949 KOps/s	48.8316 KOps/s	$\color{#d91a1a}-0.69\%$
test_items	39.2330μs	4.2112μs	237.4627 KOps/s	236.4612 KOps/s	$\color{#35bf28}+0.42\%$
test_items_nested	0.4921ms	0.4033ms	2.4793 KOps/s	2.5029 KOps/s	$\color{#d91a1a}-0.94\%$
test_items_nested_locked	0.7328ms	0.4048ms	2.4702 KOps/s	2.4315 KOps/s	$\color{#35bf28}+1.59\%$
test_items_nested_leaf	0.1388ms	71.4923μs	13.9875 KOps/s	14.2042 KOps/s	$\color{#d91a1a}-1.53\%$
test_items_stack_nested	0.7491ms	0.4087ms	2.4465 KOps/s	2.4981 KOps/s	$\color{#d91a1a}-2.06\%$
test_items_stack_nested_leaf	0.1421ms	74.1757μs	13.4815 KOps/s	13.4330 KOps/s	$\color{#35bf28}+0.36\%$
test_items_stack_nested_locked	0.5945ms	0.4086ms	2.4471 KOps/s	2.4828 KOps/s	$\color{#d91a1a}-1.44\%$
test_keys	22.5020μs	3.8147μs	262.1464 KOps/s	275.4219 KOps/s	$\color{#d91a1a}-4.82\%$
test_keys_nested	0.2192ms	0.1340ms	7.4626 KOps/s	7.2202 KOps/s	$\color{#35bf28}+3.36\%$
test_keys_nested_locked	0.6685ms	0.1396ms	7.1616 KOps/s	6.9236 KOps/s	$\color{#35bf28}+3.44\%$
test_keys_nested_leaf	0.2045ms	0.1156ms	8.6474 KOps/s	8.5161 KOps/s	$\color{#35bf28}+1.54\%$
test_keys_stack_nested	0.2818ms	0.1351ms	7.4014 KOps/s	7.2036 KOps/s	$\color{#35bf28}+2.75\%$
test_keys_stack_nested_leaf	0.2259ms	0.1184ms	8.4480 KOps/s	8.4788 KOps/s	$\color{#d91a1a}-0.36\%$
test_keys_stack_nested_locked	0.2317ms	0.1419ms	7.0483 KOps/s	6.9743 KOps/s	$\color{#35bf28}+1.06\%$
test_values	5.0574μs	1.0238μs	976.7415 KOps/s	928.3699 KOps/s	$\textbf{\color{#35bf28}+5.21\%}$
test_values_nested	0.1079ms	55.2228μs	18.1085 KOps/s	18.3984 KOps/s	$\color{#d91a1a}-1.58\%$
test_values_nested_locked	0.1120ms	54.5641μs	18.3271 KOps/s	18.3010 KOps/s	$\color{#35bf28}+0.14\%$
test_values_nested_leaf	0.1124ms	59.4488μs	16.8212 KOps/s	16.6194 KOps/s	$\color{#35bf28}+1.21\%$
test_values_stack_nested	0.1028ms	55.2947μs	18.0849 KOps/s	17.8524 KOps/s	$\color{#35bf28}+1.30\%$
test_values_stack_nested_leaf	0.1156ms	60.0678μs	16.6478 KOps/s	15.6909 KOps/s	$\textbf{\color{#35bf28}+6.10\%}$
test_values_stack_nested_locked	98.8940μs	55.2748μs	18.0914 KOps/s	18.1795 KOps/s	$\color{#d91a1a}-0.48\%$
test_membership	14.9480μs	0.9092μs	1.0999 MOps/s	1.1101 MOps/s	$\color{#d91a1a}-0.92\%$
test_membership_nested	25.2370μs	3.0021μs	333.0980 KOps/s	349.6728 KOps/s	$\color{#d91a1a}-4.74\%$
test_membership_nested_leaf	27.5820μs	3.0587μs	326.9326 KOps/s	346.4622 KOps/s	$\textbf{\color{#d91a1a}-5.64\%}$
test_membership_stacked_nested	21.7410μs	2.9803μs	335.5403 KOps/s	348.8006 KOps/s	$\color{#d91a1a}-3.80\%$
test_membership_stacked_nested_leaf	27.8730μs	3.0058μs	332.6916 KOps/s	345.9757 KOps/s	$\color{#d91a1a}-3.84\%$
test_membership_nested_last	64.8930μs	4.2878μs	233.2203 KOps/s	238.8893 KOps/s	$\color{#d91a1a}-2.37\%$
test_membership_nested_leaf_last	28.5830μs	4.3418μs	230.3213 KOps/s	241.0591 KOps/s	$\color{#d91a1a}-4.45\%$
test_membership_stacked_nested_last	26.4490μs	4.3096μs	232.0406 KOps/s	242.4345 KOps/s	$\color{#d91a1a}-4.29\%$
test_membership_stacked_nested_leaf_last	20.0970μs	4.2989μs	232.6183 KOps/s	240.3058 KOps/s	$\color{#d91a1a}-3.20\%$
test_nested_getleaf	31.4390μs	10.7457μs	93.0606 KOps/s	92.8988 KOps/s	$\color{#35bf28}+0.17\%$
test_nested_get	32.6910μs	10.1696μs	98.3324 KOps/s	98.4548 KOps/s	$\color{#d91a1a}-0.12\%$
test_stacked_getleaf	55.4630μs	10.6707μs	93.7145 KOps/s	93.8030 KOps/s	$\color{#d91a1a}-0.09\%$
test_stacked_get	30.5570μs	10.1357μs	98.6615 KOps/s	98.0436 KOps/s	$\color{#35bf28}+0.63\%$
test_nested_getitemleaf	52.9080μs	10.9960μs	90.9419 KOps/s	89.2325 KOps/s	$\color{#35bf28}+1.92\%$
test_nested_getitem	31.7700μs	10.3681μs	96.4497 KOps/s	96.5732 KOps/s	$\color{#d91a1a}-0.13\%$
test_stacked_getitemleaf	42.0090μs	11.0395μs	90.5836 KOps/s	89.7876 KOps/s	$\color{#35bf28}+0.89\%$
test_stacked_getitem	33.9140μs	10.3048μs	97.0424 KOps/s	96.6273 KOps/s	$\color{#35bf28}+0.43\%$
test_lock_nested	0.9697ms	0.4402ms	2.2716 KOps/s	2.2404 KOps/s	$\color{#35bf28}+1.39\%$
test_lock_stack_nested	0.7875ms	0.4149ms	2.4100 KOps/s	2.4067 KOps/s	$\color{#35bf28}+0.14\%$
test_unlock_nested	0.7364ms	0.3573ms	2.7991 KOps/s	2.7421 KOps/s	$\color{#35bf28}+2.08\%$
test_unlock_stack_nested	0.4914ms	0.3330ms	3.0028 KOps/s	2.9977 KOps/s	$\color{#35bf28}+0.17\%$
test_flatten_speed	0.1713ms	94.5244μs	10.5793 KOps/s	10.7040 KOps/s	$\color{#d91a1a}-1.16\%$
test_unflatten_speed	0.8154ms	0.4932ms	2.0278 KOps/s	2.0400 KOps/s	$\color{#d91a1a}-0.60\%$
test_common_ops	4.2126ms	0.8430ms	1.1862 KOps/s	1.2018 KOps/s	$\color{#d91a1a}-1.30\%$
test_creation	31.0880μs	2.0817μs	480.3669 KOps/s	491.7968 KOps/s	$\color{#d91a1a}-2.32\%$
test_creation_empty	39.6340μs	13.1458μs	76.0698 KOps/s	77.0361 KOps/s	$\color{#d91a1a}-1.25\%$
test_creation_nested_1	64.7410μs	15.8421μs	63.1230 KOps/s	63.4558 KOps/s	$\color{#d91a1a}-0.52\%$
test_creation_nested_2	44.8340μs	20.2261μs	49.4411 KOps/s	49.5253 KOps/s	$\color{#d91a1a}-0.17\%$
test_clone	0.1462ms	13.1239μs	76.1967 KOps/s	75.6453 KOps/s	$\color{#35bf28}+0.73\%$
test_getitem[int]	1.4639ms	12.6615μs	78.9798 KOps/s	77.9589 KOps/s	$\color{#35bf28}+1.31\%$
test_getitem[slice_int]	0.1409ms	24.6275μs	40.6051 KOps/s	41.3588 KOps/s	$\color{#d91a1a}-1.82\%$
test_getitem[range]	0.2867ms	48.0760μs	20.8004 KOps/s	20.7166 KOps/s	$\color{#35bf28}+0.40\%$
test_getitem[tuple]	0.1757ms	20.4785μs	48.8317 KOps/s	49.4649 KOps/s	$\color{#d91a1a}-1.28\%$
test_getitem[list]	0.1680ms	43.8652μs	22.7971 KOps/s	23.1040 KOps/s	$\color{#d91a1a}-1.33\%$
test_setitem_dim[int]	52.7090μs	25.0512μs	39.9183 KOps/s	38.1693 KOps/s	$\color{#35bf28}+4.58\%$
test_setitem_dim[slice_int]	94.2460μs	52.4413μs	19.0689 KOps/s	18.6012 KOps/s	$\color{#35bf28}+2.51\%$
test_setitem_dim[range]	0.1227ms	73.3957μs	13.6248 KOps/s	13.3105 KOps/s	$\color{#35bf28}+2.36\%$
test_setitem_dim[tuple]	71.2830μs	41.9835μs	23.8189 KOps/s	23.7023 KOps/s	$\color{#35bf28}+0.49\%$
test_setitem	95.1880μs	20.8350μs	47.9963 KOps/s	46.2752 KOps/s	$\color{#35bf28}+3.72\%$
test_set	0.1009ms	20.8626μs	47.9326 KOps/s	47.8082 KOps/s	$\color{#35bf28}+0.26\%$
test_set_shared	1.2617ms	0.1678ms	5.9588 KOps/s	5.8211 KOps/s	$\color{#35bf28}+2.37\%$
test_update	0.1990ms	24.5335μs	40.7606 KOps/s	39.1360 KOps/s	$\color{#35bf28}+4.15\%$
test_update_nested	0.2132ms	34.6501μs	28.8599 KOps/s	28.5842 KOps/s	$\color{#35bf28}+0.96\%$
test_update__nested	0.9665ms	32.6428μs	30.6346 KOps/s	30.6230 KOps/s	$\color{#35bf28}+0.04\%$
test_set_nested	0.1001ms	22.6547μs	44.1410 KOps/s	42.7838 KOps/s	$\color{#35bf28}+3.17\%$
test_set_nested_new	81.5030μs	26.9077μs	37.1640 KOps/s	35.5283 KOps/s	$\color{#35bf28}+4.60\%$
test_select	0.1302ms	42.5122μs	23.5227 KOps/s	22.6654 KOps/s	$\color{#35bf28}+3.78\%$
test_select_nested	0.1254ms	59.8659μs	16.7040 KOps/s	16.6056 KOps/s	$\color{#35bf28}+0.59\%$
test_exclude_nested	0.1667ms	78.8373μs	12.6843 KOps/s	12.8176 KOps/s	$\color{#d91a1a}-1.04\%$
test_empty[True]	0.5932ms	0.3823ms	2.6156 KOps/s	2.6293 KOps/s	$\color{#d91a1a}-0.52\%$
test_empty[False]	8.1372μs	1.1880μs	841.7223 KOps/s	829.7162 KOps/s	$\color{#35bf28}+1.45\%$
test_unbind_speed	0.4157ms	0.2637ms	3.7929 KOps/s	3.7153 KOps/s	$\color{#35bf28}+2.09\%$
test_unbind_speed_stack0	0.4363ms	0.2617ms	3.8208 KOps/s	3.8409 KOps/s	$\color{#d91a1a}-0.52\%$
test_unbind_speed_stack1	0.1035s	0.7717ms	1.2958 KOps/s	1.4219 KOps/s	$\textbf{\color{#d91a1a}-8.87\%}$
test_split	2.0609ms	1.5635ms	639.5841 Ops/s	575.7878 Ops/s	$\textbf{\color{#35bf28}+11.08\%}$
test_chunk	0.1031s	1.8731ms	533.8848 Ops/s	573.3668 Ops/s	$\textbf{\color{#d91a1a}-6.89\%}$
test_consolidate_njt[False-None]	8.7398ms	8.3403ms	119.8995 Ops/s	121.9188 Ops/s	$\color{#d91a1a}-1.66\%$
test_creation[device0]	0.2191ms	91.2311μs	10.9612 KOps/s	10.8094 KOps/s	$\color{#35bf28}+1.40\%$
test_creation_from_tensor	0.2914ms	94.8635μs	10.5415 KOps/s	10.4047 KOps/s	$\color{#35bf28}+1.31\%$
test_add_one[memmap_tensor0]	0.2261ms	4.9579μs	201.6985 KOps/s	202.0934 KOps/s	$\color{#d91a1a}-0.20\%$
test_contiguous[memmap_tensor0]	23.6050μs	0.5256μs	1.9027 MOps/s	1.9745 MOps/s	$\color{#d91a1a}-3.64\%$
test_stack[memmap_tensor0]	60.1830μs	3.5635μs	280.6262 KOps/s	293.0185 KOps/s	$\color{#d91a1a}-4.23\%$
test_memmaptd_index	1.0029ms	0.2377ms	4.2070 KOps/s	4.2590 KOps/s	$\color{#d91a1a}-1.22\%$
test_memmaptd_index_astensor	0.6584ms	0.3151ms	3.1731 KOps/s	3.2004 KOps/s	$\color{#d91a1a}-0.85\%$
test_memmaptd_index_op	1.4488ms	0.6133ms	1.6305 KOps/s	1.6445 KOps/s	$\color{#d91a1a}-0.85\%$
test_serialize_model	0.1304s	0.1159s	8.6313 Ops/s	7.3737 Ops/s	$\textbf{\color{#35bf28}+17.06\%}$
test_serialize_model_pickle	0.4424s	0.3889s	2.5717 Ops/s	2.5561 Ops/s	$\color{#35bf28}+0.61\%$
test_serialize_weights	0.2206s	0.1307s	7.6512 Ops/s	8.6465 Ops/s	$\textbf{\color{#d91a1a}-11.51\%}$
test_serialize_weights_returnearly	0.1796s	0.1588s	6.2968 Ops/s	6.2367 Ops/s	$\color{#35bf28}+0.96\%$
test_serialize_weights_pickle	0.8928s	0.7430s	1.3458 Ops/s	2.5246 Ops/s	$\textbf{\color{#d91a1a}-46.69\%}$
test_serialize_weights_filesystem	0.1436s	0.1395s	7.1678 Ops/s	6.2810 Ops/s	$\textbf{\color{#35bf28}+14.12\%}$
test_serialize_model_filesystem	0.2426s	0.1553s	6.4409 Ops/s	6.6981 Ops/s	$\color{#d91a1a}-3.84\%$
test_reshape_pytree	72.0050μs	26.8925μs	37.1851 KOps/s	37.2260 KOps/s	$\color{#d91a1a}-0.11\%$
test_reshape_td	68.8790μs	32.8094μs	30.4791 KOps/s	30.7122 KOps/s	$\color{#d91a1a}-0.76\%$
test_view_pytree	0.1357ms	28.2564μs	35.3902 KOps/s	37.2540 KOps/s	$\textbf{\color{#d91a1a}-5.00\%}$
test_view_td	83.3560μs	37.1419μs	26.9238 KOps/s	26.0911 KOps/s	$\color{#35bf28}+3.19\%$
test_unbind_pytree	66.7650μs	30.5795μs	32.7017 KOps/s	33.0897 KOps/s	$\color{#d91a1a}-1.17\%$
test_unbind_td	0.3462ms	38.4920μs	25.9794 KOps/s	25.9146 KOps/s	$\color{#35bf28}+0.25\%$
test_split_pytree	65.6930μs	30.2861μs	33.0184 KOps/s	33.5124 KOps/s	$\color{#d91a1a}-1.47\%$
test_split_td	0.5300ms	43.2250μs	23.1347 KOps/s	22.6682 KOps/s	$\color{#35bf28}+2.06\%$
test_add_pytree	0.1049ms	35.7724μs	27.9545 KOps/s	27.8678 KOps/s	$\color{#35bf28}+0.31\%$
test_add_td	0.1323ms	56.3518μs	17.7457 KOps/s	15.5968 KOps/s	$\textbf{\color{#35bf28}+13.78\%}$
test_compile_add_one_nested[tensordict-compile]	0.2458ms	62.6928μs	15.9508 KOps/s	16.1178 KOps/s	$\color{#d91a1a}-1.04\%$
test_compile_add_one_nested[tensordict-eager]	1.3862ms	0.1613ms	6.1977 KOps/s	6.1820 KOps/s	$\color{#35bf28}+0.25\%$
test_compile_add_one_nested[pytree-compile]	0.1135ms	46.3722μs	21.5647 KOps/s	22.5037 KOps/s	$\color{#d91a1a}-4.17\%$
test_compile_add_one_nested[pytree-eager]	0.2847ms	0.1188ms	8.4194 KOps/s	8.4175 KOps/s	$\color{#35bf28}+0.02\%$
test_compile_copy_nested[tensordict-compile]	97.7430μs	25.6263μs	39.0225 KOps/s	38.0485 KOps/s	$\color{#35bf28}+2.56\%$
test_compile_copy_nested[tensordict-eager]	0.1292ms	54.1218μs	18.4768 KOps/s	18.4728 KOps/s	$\color{#35bf28}+0.02\%$
test_compile_copy_nested[pytree-compile]	0.1697ms	78.7529μs	12.6979 KOps/s	12.4621 KOps/s	$\color{#35bf28}+1.89\%$
test_compile_copy_nested[pytree-eager]	0.1236ms	68.5477μs	14.5884 KOps/s	14.4654 KOps/s	$\color{#35bf28}+0.85\%$
test_compile_add_one_flat[tensordict-compile]	0.2227ms	0.1050ms	9.5262 KOps/s	9.5469 KOps/s	$\color{#d91a1a}-0.22\%$
test_compile_add_one_flat[tensordict-eager]	0.4421ms	0.2010ms	4.9762 KOps/s	5.0054 KOps/s	$\color{#d91a1a}-0.58\%$
test_compile_add_one_flat[tensorclass-compile]	0.1199ms	45.0126μs	22.2160 KOps/s	22.7373 KOps/s	$\color{#d91a1a}-2.29\%$
test_compile_add_one_flat[tensorclass-eager]	0.4785ms	62.9560μs	15.8841 KOps/s	16.0494 KOps/s	$\color{#d91a1a}-1.03\%$
test_compile_add_one_flat[pytree-compile]	0.3387ms	0.1048ms	9.5418 KOps/s	9.7625 KOps/s	$\color{#d91a1a}-2.26\%$
test_compile_add_one_flat[pytree-eager]	0.3565ms	0.2015ms	4.9629 KOps/s	4.9919 KOps/s	$\color{#d91a1a}-0.58\%$
test_compile_add_self_flat[tensordict-eager]	0.4871ms	0.2108ms	4.7449 KOps/s	4.7677 KOps/s	$\color{#d91a1a}-0.48\%$
test_compile_add_self_flat[tensordict-compile]	0.2056ms	0.1057ms	9.4569 KOps/s	9.5029 KOps/s	$\color{#d91a1a}-0.48\%$
test_compile_add_self_flat[tensorclass-eager]	0.1848ms	54.7716μs	18.2576 KOps/s	18.3231 KOps/s	$\color{#d91a1a}-0.36\%$
test_compile_add_self_flat[tensorclass-compile]	0.1053ms	46.1782μs	21.6553 KOps/s	21.8380 KOps/s	$\color{#d91a1a}-0.84\%$
test_compile_add_self_flat[pytree-eager]	1.6173ms	0.1606ms	6.2267 KOps/s	6.2786 KOps/s	$\color{#d91a1a}-0.83\%$
test_compile_add_self_flat[pytree-compile]	0.2207ms	0.1024ms	9.7662 KOps/s	9.7275 KOps/s	$\color{#35bf28}+0.40\%$
test_compile_copy_flat[tensordict-compile]	54.2610μs	20.5554μs	48.6489 KOps/s	48.5948 KOps/s	$\color{#35bf28}+0.11\%$
test_compile_copy_flat[tensordict-eager]	0.1450ms	57.9870μs	17.2452 KOps/s	17.0083 KOps/s	$\color{#35bf28}+1.39\%$
test_compile_copy_flat[pytree-compile]	0.1699ms	81.4333μs	12.2800 KOps/s	12.1130 KOps/s	$\color{#35bf28}+1.38\%$
test_compile_copy_flat[pytree-eager]	0.1434ms	69.1853μs	14.4539 KOps/s	14.2501 KOps/s	$\color{#35bf28}+1.43\%$
test_compile_assign_and_add[tensordict-compile]	1.2512ms	0.2097ms	4.7677 KOps/s	4.8644 KOps/s	$\color{#d91a1a}-1.99\%$
test_compile_assign_and_add[tensordict-eager]	1.4104ms	1.2737ms	785.1186 Ops/s	795.1480 Ops/s	$\color{#d91a1a}-1.26\%$
test_compile_assign_and_add[pytree-compile]	0.4093ms	0.2053ms	4.8704 KOps/s	4.9139 KOps/s	$\color{#d91a1a}-0.88\%$
test_compile_assign_and_add[pytree-eager]	0.9003ms	0.7772ms	1.2866 KOps/s	1.2890 KOps/s	$\color{#d91a1a}-0.19\%$
test_compile_assign_and_add_stack[compile]	0.6338ms	0.4545ms	2.2004 KOps/s	2.1809 KOps/s	$\color{#35bf28}+0.90\%$
test_compile_assign_and_add_stack[eager]	0.2583s	3.4325ms	291.3327 Ops/s	367.2628 Ops/s	$\textbf{\color{#d91a1a}-20.67\%}$
test_compile_indexing[tensor-tensordict-compile]	0.1074ms	37.2117μs	26.8732 KOps/s	27.7367 KOps/s	$\color{#d91a1a}-3.11\%$
test_compile_indexing[tensor-tensordict-eager]	0.4170ms	32.3764μs	30.8867 KOps/s	30.6183 KOps/s	$\color{#35bf28}+0.88\%$
test_compile_indexing[tensor-tensorclass-compile]	83.0460μs	29.5380μs	33.8547 KOps/s	33.9370 KOps/s	$\color{#d91a1a}-0.24\%$
test_compile_indexing[tensor-tensorclass-eager]	76.4130μs	23.3776μs	42.7760 KOps/s	42.2173 KOps/s	$\color{#35bf28}+1.32\%$
test_compile_indexing[tensor-pytree-compile]	0.1007ms	30.2485μs	33.0595 KOps/s	32.6172 KOps/s	$\color{#35bf28}+1.36\%$
test_compile_indexing[tensor-pytree-eager]	76.1520μs	23.6151μs	42.3458 KOps/s	41.8760 KOps/s	$\color{#35bf28}+1.12\%$
test_compile_indexing[slice-tensordict-compile]	0.1381ms	51.5391μs	19.4028 KOps/s	19.5052 KOps/s	$\color{#d91a1a}-0.52\%$
test_compile_indexing[slice-tensordict-eager]	0.5269ms	19.3577μs	51.6590 KOps/s	49.0211 KOps/s	$\textbf{\color{#35bf28}+5.38\%}$
test_compile_indexing[slice-tensorclass-compile]	96.6910μs	43.5396μs	22.9676 KOps/s	22.8585 KOps/s	$\color{#35bf28}+0.48\%$
test_compile_indexing[slice-tensorclass-eager]	53.8610μs	18.9315μs	52.8219 KOps/s	52.5172 KOps/s	$\color{#35bf28}+0.58\%$
test_compile_indexing[slice-pytree-compile]	0.1546ms	44.3650μs	22.5403 KOps/s	22.1095 KOps/s	$\color{#35bf28}+1.95\%$
test_compile_indexing[slice-pytree-eager]	64.2400μs	18.7521μs	53.3272 KOps/s	52.2877 KOps/s	$\color{#35bf28}+1.99\%$
test_compile_indexing[int-tensordict-compile]	0.1480ms	52.4649μs	19.0603 KOps/s	19.0442 KOps/s	$\color{#35bf28}+0.08\%$
test_compile_indexing[int-tensordict-eager]	0.8865ms	19.2788μs	51.8704 KOps/s	49.9228 KOps/s	$\color{#35bf28}+3.90\%$
test_compile_indexing[int-tensorclass-compile]	0.1008ms	44.8479μs	22.2976 KOps/s	22.2369 KOps/s	$\color{#35bf28}+0.27\%$
test_compile_indexing[int-tensorclass-eager]	0.2602ms	18.7366μs	53.3716 KOps/s	52.8206 KOps/s	$\color{#35bf28}+1.04\%$
test_compile_indexing[int-pytree-compile]	0.1182ms	44.6923μs	22.3752 KOps/s	22.4096 KOps/s	$\color{#d91a1a}-0.15\%$
test_compile_indexing[int-pytree-eager]	74.9100μs	18.7794μs	53.2500 KOps/s	51.8839 KOps/s	$\color{#35bf28}+2.63\%$
test_mod_add[eager]	0.1707ms	35.3779μs	28.2662 KOps/s	28.5094 KOps/s	$\color{#d91a1a}-0.85\%$
test_mod_add[compile]	0.1598ms	48.3768μs	20.6711 KOps/s	20.6338 KOps/s	$\color{#35bf28}+0.18\%$
test_mod_add[compile-overhead]	0.1100ms	47.9607μs	20.8504 KOps/s	20.7147 KOps/s	$\color{#35bf28}+0.66\%$
test_mod_wrap[eager]	0.4379ms	0.2264ms	4.4166 KOps/s	4.4247 KOps/s	$\color{#d91a1a}-0.18\%$
test_mod_wrap[compile]	0.3293ms	0.2088ms	4.7893 KOps/s	4.7577 KOps/s	$\color{#35bf28}+0.67\%$
test_mod_wrap[compile-overhead]	0.4174ms	0.2067ms	4.8376 KOps/s	4.8100 KOps/s	$\color{#35bf28}+0.57\%$
test_mod_wrap_and_backward[eager]	13.9390ms	11.5665ms	86.4565 Ops/s	91.4447 Ops/s	$\textbf{\color{#d91a1a}-5.45\%}$
test_mod_wrap_and_backward[compile]	14.6951ms	12.3398ms	81.0388 Ops/s	89.4004 Ops/s	$\textbf{\color{#d91a1a}-9.35\%}$
test_mod_wrap_and_backward[compile-overhead]	16.6685ms	12.6446ms	79.0853 Ops/s	89.4451 Ops/s	$\textbf{\color{#d91a1a}-11.58\%}$
test_seq_add[eager]	0.2074ms	0.1156ms	8.6513 KOps/s	8.7195 KOps/s	$\color{#d91a1a}-0.78\%$
test_seq_add[compile]	0.1553ms	63.2726μs	15.8046 KOps/s	16.3084 KOps/s	$\color{#d91a1a}-3.09\%$
test_seq_add[compile-overhead]	0.1312ms	60.0448μs	16.6542 KOps/s	16.7576 KOps/s	$\color{#d91a1a}-0.62\%$
test_seq_wrap[eager]	0.6331ms	0.4531ms	2.2069 KOps/s	2.2359 KOps/s	$\color{#d91a1a}-1.29\%$
test_seq_wrap[compile]	0.3268ms	0.2286ms	4.3745 KOps/s	4.2987 KOps/s	$\color{#35bf28}+1.76\%$
test_seq_wrap[compile-overhead]	0.4288ms	0.2278ms	4.3894 KOps/s	4.3210 KOps/s	$\color{#35bf28}+1.58\%$
test_func_call_runtime[False-eager]	0.9520ms	0.5533ms	1.8074 KOps/s	1.8250 KOps/s	$\color{#d91a1a}-0.97\%$
test_func_call_runtime[False-compile]	0.8050ms	0.4319ms	2.3155 KOps/s	2.2673 KOps/s	$\color{#35bf28}+2.12\%$
test_func_call_runtime[False-compile-overhead]	0.7535ms	0.4305ms	2.3231 KOps/s	2.2748 KOps/s	$\color{#35bf28}+2.12\%$
test_func_call_runtime[True-eager]	1.2819ms	0.7693ms	1.2998 KOps/s	1.3159 KOps/s	$\color{#d91a1a}-1.22\%$
test_func_call_runtime[True-compile]	0.6119ms	0.4691ms	2.1317 KOps/s	2.1088 KOps/s	$\color{#35bf28}+1.08\%$
test_func_call_runtime[True-compile-overhead]	0.8980ms	0.4697ms	2.1292 KOps/s	2.1037 KOps/s	$\color{#35bf28}+1.21\%$
test_func_call_cm_runtime[False-eager]	1.1554ms	0.5542ms	1.8045 KOps/s	1.8245 KOps/s	$\color{#d91a1a}-1.10\%$
test_func_call_cm_runtime[False-compile]	0.9024ms	0.4337ms	2.3057 KOps/s	2.2800 KOps/s	$\color{#35bf28}+1.13\%$
test_func_call_cm_runtime[False-compile-overhead]	0.5739ms	0.4313ms	2.3186 KOps/s	2.2933 KOps/s	$\color{#35bf28}+1.10\%$
test_func_call_cm_runtime[True-eager]	1.0932ms	0.9147ms	1.0932 KOps/s	1.1086 KOps/s	$\color{#d91a1a}-1.38\%$
test_func_call_cm_runtime[True-compile]	0.7914ms	0.4972ms	2.0114 KOps/s	1.9870 KOps/s	$\color{#35bf28}+1.23\%$
test_func_call_cm_runtime[True-compile-overhead]	0.5891ms	0.4916ms	2.0343 KOps/s	1.9813 KOps/s	$\color{#35bf28}+2.68\%$
test_vmap_func_call_cm_runtime[eager]	2.7651ms	1.9049ms	524.9609 Ops/s	525.6960 Ops/s	$\color{#d91a1a}-0.14\%$
test_vmap_func_call_cm_runtime[compile]	0.7014ms	0.5200ms	1.9231 KOps/s	1.9023 KOps/s	$\color{#35bf28}+1.09\%$
test_vmap_func_call_cm_runtime[compile-overhead]	0.9948ms	0.5219ms	1.9162 KOps/s	1.8246 KOps/s	$\textbf{\color{#35bf28}+5.02\%}$
test_distributed	0.2747ms	0.1267ms	7.8951 KOps/s	7.8084 KOps/s	$\color{#35bf28}+1.11\%$
test_tdmodule	65.8430μs	27.9712μs	35.7510 KOps/s	35.9776 KOps/s	$\color{#d91a1a}-0.63\%$
test_tdmodule_dispatch	96.0300μs	51.0671μs	19.5821 KOps/s	19.5583 KOps/s	$\color{#35bf28}+0.12\%$
test_tdseq	45.8750μs	27.0832μs	36.9232 KOps/s	35.8001 KOps/s	$\color{#35bf28}+3.14\%$
test_tdseq_dispatch	85.3200μs	53.5806μs	18.6635 KOps/s	18.4375 KOps/s	$\color{#35bf28}+1.23\%$
test_instantiation_functorch	2.7961ms	1.6416ms	609.1711 Ops/s	623.8423 Ops/s	$\color{#d91a1a}-2.35\%$
test_exec_functorch	0.3243ms	0.1788ms	5.5940 KOps/s	5.4197 KOps/s	$\color{#35bf28}+3.22\%$
test_exec_functional_call	0.3432ms	0.1738ms	5.7553 KOps/s	5.7095 KOps/s	$\color{#35bf28}+0.80\%$
test_exec_td_decorator	0.5242ms	0.2302ms	4.3439 KOps/s	4.3168 KOps/s	$\color{#35bf28}+0.63\%$
test_vmap_mlp_speed_decorator[True-True]	1.2370ms	0.6892ms	1.4510 KOps/s	1.4839 KOps/s	$\color{#d91a1a}-2.22\%$
test_vmap_mlp_speed_decorator[True-False]	1.0010ms	0.6509ms	1.5363 KOps/s	1.5120 KOps/s	$\color{#35bf28}+1.60\%$
test_vmap_mlp_speed_decorator[False-True]	0.7234ms	0.5190ms	1.9269 KOps/s	1.8723 KOps/s	$\color{#35bf28}+2.91\%$
test_vmap_mlp_speed_decorator[False-False]	0.8549ms	0.5208ms	1.9201 KOps/s	1.8760 KOps/s	$\color{#35bf28}+2.35\%$
test_to_module_speed[True]	1.6258ms	1.2806ms	780.8685 Ops/s	774.4413 Ops/s	$\color{#35bf28}+0.83\%$
test_to_module_speed[False]	2.5337ms	1.2693ms	787.8194 Ops/s	786.3981 Ops/s	$\color{#35bf28}+0.18\%$
test_tc_init	86.3910μs	47.5338μs	21.0377 KOps/s	20.1913 KOps/s	$\color{#35bf28}+4.19\%$
test_tc_init_nested	0.1763ms	93.4589μs	10.6999 KOps/s	10.1360 KOps/s	$\textbf{\color{#35bf28}+5.56\%}$
test_tc_first_layer_tensor	50.1870μs	1.5096μs	662.4226 KOps/s	657.6907 KOps/s	$\color{#35bf28}+0.72\%$
test_tc_first_layer_nontensor	18.5350μs	4.7744μs	209.4484 KOps/s	211.9351 KOps/s	$\color{#d91a1a}-1.17\%$
test_tc_second_layer_tensor	43.2510μs	2.8259μs	353.8697 KOps/s	358.3271 KOps/s	$\color{#d91a1a}-1.24\%$
test_tc_second_layer_nontensor	34.0530μs	6.1210μs	163.3713 KOps/s	167.5387 KOps/s	$\color{#d91a1a}-2.49\%$
test_unbind	0.2237s	12.6507ms	79.0470 Ops/s	80.0149 Ops/s	$\color{#d91a1a}-1.21\%$
test_full_like	10.0217ms	7.9263ms	126.1624 Ops/s	131.6882 Ops/s	$\color{#d91a1a}-4.20\%$
test_zeros_like	3.8806ms	3.1003ms	322.5503 Ops/s	362.6984 Ops/s	$\textbf{\color{#d91a1a}-11.07\%}$
test_ones_like	4.2119ms	3.5703ms	280.0849 Ops/s	153.1859 Ops/s	$\textbf{\color{#35bf28}+82.84\%}$
test_clone	6.9673ms	5.5191ms	181.1903 Ops/s	121.2558 Ops/s	$\textbf{\color{#35bf28}+49.43\%}$
test_squeeze	64.6500μs	11.5688μs	86.4393 KOps/s	84.0607 KOps/s	$\color{#35bf28}+2.83\%$
test_unsqueeze	0.1663ms	89.7796μs	11.1384 KOps/s	10.8330 KOps/s	$\color{#35bf28}+2.82\%$
test_split	0.5345ms	0.1928ms	5.1875 KOps/s	5.0525 KOps/s	$\color{#35bf28}+2.67\%$
test_permute	0.3325ms	0.2236ms	4.4730 KOps/s	4.5449 KOps/s	$\color{#d91a1a}-1.58\%$
test_stack	29.1634ms	26.3862ms	37.8986 Ops/s	38.0057 Ops/s	$\color{#d91a1a}-0.28\%$
test_cat	32.7882ms	25.5175ms	39.1888 Ops/s	39.0490 Ops/s	$\color{#35bf28}+0.36\%$

[github-actions]

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

Total Benchmarks: 229. Improved: $\large\color{#35bf28}10$. Worsened: $\large\color{#d91a1a}33$.

Expand to view detailed results

Name	Max	Mean	Ops	Ops on Repo HEAD	Change
test_plain_set_nested	44.6910μs	11.4245μs	87.5310 KOps/s	95.1466 KOps/s	$\textbf{\color{#d91a1a}-8.00\%}$
test_plain_set_stack_nested	37.0910μs	11.5304μs	86.7270 KOps/s	94.9236 KOps/s	$\textbf{\color{#d91a1a}-8.63\%}$
test_plain_set_nested_inplace	43.6510μs	12.4288μs	80.4581 KOps/s	86.6852 KOps/s	$\textbf{\color{#d91a1a}-7.18\%}$
test_plain_set_stack_nested_inplace	51.0010μs	12.3759μs	80.8025 KOps/s	87.1947 KOps/s	$\textbf{\color{#d91a1a}-7.33\%}$
test_items	30.3700μs	2.8941μs	345.5356 KOps/s	342.1358 KOps/s	$\color{#35bf28}+0.99\%$
test_items_nested	0.3902ms	0.3505ms	2.8528 KOps/s	2.8181 KOps/s	$\color{#35bf28}+1.23\%$
test_items_nested_locked	0.4084ms	0.3500ms	2.8568 KOps/s	2.8229 KOps/s	$\color{#35bf28}+1.20\%$
test_items_nested_leaf	81.2810μs	58.6516μs	17.0498 KOps/s	17.3046 KOps/s	$\color{#d91a1a}-1.47\%$
test_items_stack_nested	0.3912ms	0.3498ms	2.8592 KOps/s	2.7969 KOps/s	$\color{#35bf28}+2.23\%$
test_items_stack_nested_leaf	82.4210μs	57.9901μs	17.2443 KOps/s	17.2933 KOps/s	$\color{#d91a1a}-0.28\%$
test_items_stack_nested_locked	0.4048ms	0.3481ms	2.8730 KOps/s	2.8034 KOps/s	$\color{#35bf28}+2.48\%$
test_keys	28.9100μs	3.4640μs	288.6799 KOps/s	289.8759 KOps/s	$\color{#d91a1a}-0.41\%$
test_keys_nested	0.1047ms	69.8461μs	14.3172 KOps/s	14.0840 KOps/s	$\color{#35bf28}+1.66\%$
test_keys_nested_locked	0.8141ms	75.4046μs	13.2618 KOps/s	12.9437 KOps/s	$\color{#35bf28}+2.46\%$
test_keys_nested_leaf	95.9810μs	61.4198μs	16.2814 KOps/s	16.0835 KOps/s	$\color{#35bf28}+1.23\%$
test_keys_stack_nested	0.1065ms	69.8202μs	14.3225 KOps/s	14.1109 KOps/s	$\color{#35bf28}+1.50\%$
test_keys_stack_nested_leaf	88.9910μs	61.2761μs	16.3196 KOps/s	16.1695 KOps/s	$\color{#35bf28}+0.93\%$
test_keys_stack_nested_locked	0.1086ms	75.3160μs	13.2774 KOps/s	13.0591 KOps/s	$\color{#35bf28}+1.67\%$
test_values	6.1100μs	0.8600μs	1.1629 MOps/s	1.1769 MOps/s	$\color{#d91a1a}-1.19\%$
test_values_nested	57.2100μs	31.3587μs	31.8891 KOps/s	32.0315 KOps/s	$\color{#d91a1a}-0.44\%$
test_values_nested_locked	60.2210μs	32.8966μs	30.3983 KOps/s	30.3734 KOps/s	$\color{#35bf28}+0.08\%$
test_values_nested_leaf	57.5010μs	33.7651μs	29.6164 KOps/s	29.4430 KOps/s	$\color{#35bf28}+0.59\%$
test_values_stack_nested	69.7800μs	31.4438μs	31.8028 KOps/s	31.9287 KOps/s	$\color{#d91a1a}-0.39\%$
test_values_stack_nested_leaf	59.1700μs	33.7020μs	29.6718 KOps/s	29.6901 KOps/s	$\color{#d91a1a}-0.06\%$
test_values_stack_nested_locked	81.2910μs	32.3980μs	30.8661 KOps/s	30.4088 KOps/s	$\color{#35bf28}+1.50\%$
test_membership	1.4515μs	0.5089μs	1.9651 MOps/s	1.9650 MOps/s	$+0.01\%$
test_membership_nested	14.6950μs	2.0321μs	492.0939 KOps/s	486.8578 KOps/s	$\color{#35bf28}+1.08\%$
test_membership_nested_leaf	30.1800μs	2.0308μs	492.4138 KOps/s	487.7232 KOps/s	$\color{#35bf28}+0.96\%$
test_membership_stacked_nested	30.3900μs	2.1197μs	471.7632 KOps/s	485.0897 KOps/s	$\color{#d91a1a}-2.75\%$
test_membership_stacked_nested_leaf	27.8710μs	2.1106μs	473.8022 KOps/s	480.2792 KOps/s	$\color{#d91a1a}-1.35\%$
test_membership_nested_last	29.5110μs	2.9954μs	333.8405 KOps/s	337.8990 KOps/s	$\color{#d91a1a}-1.20\%$
test_membership_nested_leaf_last	33.2300μs	2.9936μs	334.0434 KOps/s	333.3471 KOps/s	$\color{#35bf28}+0.21\%$
test_membership_stacked_nested_last	32.5100μs	2.9674μs	336.9994 KOps/s	338.3797 KOps/s	$\color{#d91a1a}-0.41\%$
test_membership_stacked_nested_leaf_last	24.3000μs	2.9514μs	338.8235 KOps/s	335.1988 KOps/s	$\color{#35bf28}+1.08\%$
test_nested_getleaf	33.5800μs	6.1803μs	161.8042 KOps/s	161.5012 KOps/s	$\color{#35bf28}+0.19\%$
test_nested_get	59.0110μs	5.8660μs	170.4733 KOps/s	170.5100 KOps/s	$\color{#d91a1a}-0.02\%$
test_stacked_getleaf	40.3800μs	6.1675μs	162.1414 KOps/s	164.0029 KOps/s	$\color{#d91a1a}-1.14\%$
test_stacked_get	0.8383ms	5.8563μs	170.7555 KOps/s	171.5536 KOps/s	$\color{#d91a1a}-0.47\%$
test_nested_getitemleaf	35.7210μs	6.1747μs	161.9521 KOps/s	160.5980 KOps/s	$\color{#35bf28}+0.84\%$
test_nested_getitem	33.1310μs	5.8892μs	169.8030 KOps/s	168.4001 KOps/s	$\color{#35bf28}+0.83\%$
test_stacked_getitemleaf	42.6300μs	6.1998μs	161.2958 KOps/s	160.3834 KOps/s	$\color{#35bf28}+0.57\%$
test_stacked_getitem	38.6500μs	5.8776μs	170.1371 KOps/s	167.7456 KOps/s	$\color{#35bf28}+1.43\%$
test_lock_nested	0.7929ms	0.3686ms	2.7131 KOps/s	2.6742 KOps/s	$\color{#35bf28}+1.46\%$
test_lock_stack_nested	0.3922ms	0.3407ms	2.9354 KOps/s	2.9466 KOps/s	$\color{#d91a1a}-0.38\%$
test_unlock_nested	0.6546ms	0.3094ms	3.2323 KOps/s	3.2191 KOps/s	$\color{#35bf28}+0.41\%$
test_unlock_stack_nested	0.3465ms	0.2806ms	3.5644 KOps/s	3.5957 KOps/s	$\color{#d91a1a}-0.87\%$
test_flatten_speed	0.1166ms	74.6597μs	13.3941 KOps/s	13.5005 KOps/s	$\color{#d91a1a}-0.79\%$
test_unflatten_speed	0.3680ms	0.3077ms	3.2503 KOps/s	3.2841 KOps/s	$\color{#d91a1a}-1.03\%$
test_common_ops	1.5390ms	0.6172ms	1.6203 KOps/s	1.7456 KOps/s	$\textbf{\color{#d91a1a}-7.18\%}$
test_creation	0.1001ms	1.4865μs	672.7253 KOps/s	665.1705 KOps/s	$\color{#35bf28}+1.14\%$
test_creation_empty	39.3600μs	9.2537μs	108.0651 KOps/s	144.7944 KOps/s	$\textbf{\color{#d91a1a}-25.37\%}$
test_creation_nested_1	34.4100μs	10.7401μs	93.1093 KOps/s	117.3025 KOps/s	$\textbf{\color{#d91a1a}-20.62\%}$
test_creation_nested_2	38.2100μs	13.1883μs	75.8247 KOps/s	89.9152 KOps/s	$\textbf{\color{#d91a1a}-15.67\%}$
test_clone	0.1205ms	10.8320μs	92.3194 KOps/s	98.8229 KOps/s	$\textbf{\color{#d91a1a}-6.58\%}$
test_getitem[int]	1.7304ms	10.7306μs	93.1916 KOps/s	93.2721 KOps/s	$\color{#d91a1a}-0.09\%$
test_getitem[slice_int]	0.1093ms	21.0421μs	47.5237 KOps/s	48.5398 KOps/s	$\color{#d91a1a}-2.09\%$
test_getitem[range]	0.1347ms	37.4917μs	26.6726 KOps/s	27.9295 KOps/s	$\color{#d91a1a}-4.50\%$
test_getitem[tuple]	0.1075ms	18.3152μs	54.5993 KOps/s	55.1826 KOps/s	$\color{#d91a1a}-1.06\%$
test_getitem[list]	0.2968ms	33.0211μs	30.2837 KOps/s	31.1394 KOps/s	$\color{#d91a1a}-2.75\%$
test_setitem_dim[int]	39.0900μs	19.3836μs	51.5901 KOps/s	54.2637 KOps/s	$\color{#d91a1a}-4.93\%$
test_setitem_dim[slice_int]	74.7110μs	38.4229μs	26.0261 KOps/s	26.6220 KOps/s	$\color{#d91a1a}-2.24\%$
test_setitem_dim[range]	75.5810μs	52.9053μs	18.9017 KOps/s	19.0992 KOps/s	$\color{#d91a1a}-1.03\%$
test_setitem_dim[tuple]	63.9100μs	32.5819μs	30.6918 KOps/s	31.5164 KOps/s	$\color{#d91a1a}-2.62\%$
test_setitem	52.8000μs	16.2965μs	61.3628 KOps/s	71.4521 KOps/s	$\textbf{\color{#d91a1a}-14.12\%}$
test_set	0.1120ms	15.7163μs	63.6280 KOps/s	73.0642 KOps/s	$\textbf{\color{#d91a1a}-12.91\%}$
test_set_shared	1.8671ms	0.1481ms	6.7507 KOps/s	6.9283 KOps/s	$\color{#d91a1a}-2.56\%$
test_update	0.3189ms	19.0079μs	52.6096 KOps/s	63.3627 KOps/s	$\textbf{\color{#d91a1a}-16.97\%}$
test_update_nested	0.1200ms	24.7229μs	40.4483 KOps/s	47.4850 KOps/s	$\textbf{\color{#d91a1a}-14.82\%}$
test_update__nested	0.6471ms	24.7222μs	40.4495 KOps/s	42.4468 KOps/s	$\color{#d91a1a}-4.71\%$
test_set_nested	0.1175ms	17.3884μs	57.5096 KOps/s	68.0068 KOps/s	$\textbf{\color{#d91a1a}-15.44\%}$
test_set_nested_new	0.1151ms	19.5029μs	51.2744 KOps/s	58.0325 KOps/s	$\textbf{\color{#d91a1a}-11.65\%}$
test_select	0.1280ms	31.3008μs	31.9481 KOps/s	33.2827 KOps/s	$\color{#d91a1a}-4.01\%$
test_select_nested	76.5300μs	42.4114μs	23.5786 KOps/s	23.8207 KOps/s	$\color{#d91a1a}-1.02\%$
test_exclude_nested	87.8110μs	64.2663μs	15.5603 KOps/s	15.9580 KOps/s	$\color{#d91a1a}-2.49\%$
test_empty[True]	0.3226ms	0.2856ms	3.5009 KOps/s	3.5800 KOps/s	$\color{#d91a1a}-2.21\%$
test_empty[False]	3.4961μs	0.7633μs	1.3102 MOps/s	1.3287 MOps/s	$\color{#d91a1a}-1.39\%$
test_to	87.7410μs	60.8579μs	16.4317 KOps/s	18.1991 KOps/s	$\textbf{\color{#d91a1a}-9.71\%}$
test_to_nonblocking	81.4710μs	46.0582μs	21.7117 KOps/s	22.1821 KOps/s	$\color{#d91a1a}-2.12\%$
test_unbind_speed	0.3014ms	0.2327ms	4.2973 KOps/s	4.2800 KOps/s	$\color{#35bf28}+0.40\%$
test_unbind_speed_stack0	0.2895ms	0.2348ms	4.2591 KOps/s	4.3031 KOps/s	$\color{#d91a1a}-1.02\%$
test_unbind_speed_stack1	93.9709ms	0.6523ms	1.5331 KOps/s	1.5265 KOps/s	$\color{#35bf28}+0.43\%$
test_split	97.6146ms	1.7393ms	574.9382 Ops/s	636.9599 Ops/s	$\textbf{\color{#d91a1a}-9.74\%}$
test_chunk	99.2093ms	1.6092ms	621.4165 Ops/s	585.4764 Ops/s	$\textbf{\color{#35bf28}+6.14\%}$
test_consolidate[False-None]	2.6466ms	2.5963ms	385.1586 Ops/s	382.6090 Ops/s	$\color{#35bf28}+0.67\%$
test_consolidate[default-None]	1.7311ms	1.6541ms	604.5650 Ops/s	615.7457 Ops/s	$\color{#d91a1a}-1.82\%$
test_consolidate[reduce-overhead-None]	1.7783ms	1.6906ms	591.5232 Ops/s	598.2741 Ops/s	$\color{#d91a1a}-1.13\%$
test_consolidate_njt[False-None]	6.6490ms	6.5182ms	153.4169 Ops/s	154.6799 Ops/s	$\color{#d91a1a}-0.82\%$
test_to[False-False-None]	1.7697ms	1.6846ms	593.6196 Ops/s	613.4932 Ops/s	$\color{#d91a1a}-3.24\%$
test_to[True-False-None]	1.5552ms	1.3140ms	761.0420 Ops/s	756.6162 Ops/s	$\color{#35bf28}+0.58\%$
test_to[within-False-None]	4.0965ms	4.0371ms	247.7000 Ops/s	251.7431 Ops/s	$\color{#d91a1a}-1.61\%$
test_to[True-default-None]	5.6817ms	5.3078ms	188.4020 Ops/s	197.5637 Ops/s	$\color{#d91a1a}-4.64\%$
test_to_njt[False-False-None]	7.1071ms	6.9831ms	143.2031 Ops/s	143.2857 Ops/s	$\color{#d91a1a}-0.06\%$
test_to_njt[True-False-None]	5.7271ms	5.4877ms	182.2269 Ops/s	181.7592 Ops/s	$\color{#35bf28}+0.26\%$
test_to_njt[within-False-None]	12.2067ms	12.1005ms	82.6411 Ops/s	82.9670 Ops/s	$\color{#d91a1a}-0.39\%$
test_creation[device0]	0.5409ms	79.2187μs	12.6233 KOps/s	12.4028 KOps/s	$\color{#35bf28}+1.78\%$
test_creation_from_tensor	0.5351ms	85.7129μs	11.6669 KOps/s	12.1861 KOps/s	$\color{#d91a1a}-4.26\%$
test_add_one[memmap_tensor0]	0.3128ms	6.8509μs	145.9673 KOps/s	154.3136 KOps/s	$\textbf{\color{#d91a1a}-5.41\%}$
test_contiguous[memmap_tensor0]	1.8130μs	0.3992μs	2.5050 MOps/s	2.5276 MOps/s	$\color{#d91a1a}-0.89\%$
test_stack[memmap_tensor0]	33.8710μs	4.4710μs	223.6650 KOps/s	232.4311 KOps/s	$\color{#d91a1a}-3.77\%$
test_memmaptd_index	1.7410ms	0.2495ms	4.0076 KOps/s	3.9958 KOps/s	$\color{#35bf28}+0.30\%$
test_memmaptd_index_astensor	0.5931ms	0.3104ms	3.2220 KOps/s	3.2566 KOps/s	$\color{#d91a1a}-1.06\%$
test_memmaptd_index_op	1.0121ms	0.5995ms	1.6680 KOps/s	1.7766 KOps/s	$\textbf{\color{#d91a1a}-6.11\%}$
test_serialize_model	0.1320s	0.1304s	7.6665 Ops/s	7.6765 Ops/s	$\color{#d91a1a}-0.13\%$
test_serialize_model_pickle	1.3511s	1.2154s	0.8228 Ops/s	0.8252 Ops/s	$\color{#d91a1a}-0.30\%$
test_serialize_weights	0.4228s	0.1716s	5.8263 Ops/s	7.7113 Ops/s	$\textbf{\color{#d91a1a}-24.45\%}$
test_serialize_weights_returnearly	0.2975s	55.5337ms	18.0071 Ops/s	15.1471 Ops/s	$\textbf{\color{#35bf28}+18.88\%}$
test_serialize_weights_pickle	1.3738s	1.2219s	0.8184 Ops/s	0.8187 Ops/s	$\color{#d91a1a}-0.03\%$
test_reshape_pytree	51.3000μs	22.1355μs	45.1764 KOps/s	44.7398 KOps/s	$\color{#35bf28}+0.98\%$
test_reshape_td	0.1080ms	25.3805μs	39.4004 KOps/s	37.1029 KOps/s	$\textbf{\color{#35bf28}+6.19\%}$
test_view_pytree	0.1048ms	21.9151μs	45.6306 KOps/s	45.2987 KOps/s	$\color{#35bf28}+0.73\%$
test_view_td	65.9500μs	30.7390μs	32.5319 KOps/s	33.9173 KOps/s	$\color{#d91a1a}-4.08\%$
test_unbind_pytree	64.6610μs	28.1452μs	35.5301 KOps/s	35.8412 KOps/s	$\color{#d91a1a}-0.87\%$
test_unbind_td	0.8748ms	35.9921μs	27.7839 KOps/s	27.8086 KOps/s	$\color{#d91a1a}-0.09\%$
test_split_pytree	58.4400μs	30.1964μs	33.1166 KOps/s	33.3920 KOps/s	$\color{#d91a1a}-0.82\%$
test_split_td	0.9906ms	38.3672μs	26.0639 KOps/s	25.6520 KOps/s	$\color{#35bf28}+1.61\%$
test_add_pytree	71.5510μs	35.1447μs	28.4538 KOps/s	29.4774 KOps/s	$\color{#d91a1a}-3.47\%$
test_add_td	0.1465ms	48.7934μs	20.4946 KOps/s	21.7093 KOps/s	$\textbf{\color{#d91a1a}-5.60\%}$
test_compile_add_one_nested[tensordict-compile]	0.1754ms	0.1194ms	8.3743 KOps/s	7.9996 KOps/s	$\color{#35bf28}+4.68\%$
test_compile_add_one_nested[tensordict-eager]	0.2205ms	0.1249ms	8.0082 KOps/s	7.9139 KOps/s	$\color{#35bf28}+1.19\%$
test_compile_add_one_nested[pytree-compile]	0.2001ms	95.1586μs	10.5088 KOps/s	10.3370 KOps/s	$\color{#35bf28}+1.66\%$
test_compile_add_one_nested[pytree-eager]	1.4759ms	0.1514ms	6.6059 KOps/s	6.7625 KOps/s	$\color{#d91a1a}-2.32\%$
test_compile_copy_nested[tensordict-compile]	54.6810μs	22.1753μs	45.0953 KOps/s	39.8373 KOps/s	$\textbf{\color{#35bf28}+13.20\%}$
test_compile_copy_nested[tensordict-eager]	60.1900μs	26.2752μs	38.0587 KOps/s	37.2705 KOps/s	$\color{#35bf28}+2.11\%$
test_compile_copy_nested[pytree-compile]	0.3679ms	64.8455μs	15.4213 KOps/s	15.3625 KOps/s	$\color{#35bf28}+0.38\%$
test_compile_copy_nested[pytree-eager]	0.1042ms	49.2631μs	20.2992 KOps/s	19.7739 KOps/s	$\color{#35bf28}+2.66\%$
test_compile_add_one_flat[tensordict-compile]	0.1918ms	0.1452ms	6.8862 KOps/s	6.9166 KOps/s	$\color{#d91a1a}-0.44\%$
test_compile_add_one_flat[tensordict-eager]	0.3202ms	0.2070ms	4.8315 KOps/s	4.8586 KOps/s	$\color{#d91a1a}-0.56\%$
test_compile_add_one_flat[tensorclass-compile]	0.1360ms	97.6320μs	10.2425 KOps/s	10.2139 KOps/s	$\color{#35bf28}+0.28\%$
test_compile_add_one_flat[tensorclass-eager]	0.1088ms	52.4306μs	19.0728 KOps/s	19.4920 KOps/s	$\color{#d91a1a}-2.15\%$
test_compile_add_one_flat[pytree-compile]	0.2842ms	0.1374ms	7.2760 KOps/s	7.3526 KOps/s	$\color{#d91a1a}-1.04\%$
test_compile_add_one_flat[pytree-eager]	0.5296ms	0.4867ms	2.0548 KOps/s	2.0980 KOps/s	$\color{#d91a1a}-2.06\%$
test_compile_add_self_flat[tensordict-eager]	0.3941ms	0.2471ms	4.0471 KOps/s	4.0259 KOps/s	$\color{#35bf28}+0.53\%$
test_compile_add_self_flat[tensordict-compile]	0.1943ms	0.1451ms	6.8909 KOps/s	6.9717 KOps/s	$\color{#d91a1a}-1.16\%$
test_compile_add_self_flat[tensorclass-eager]	0.1530ms	65.9331μs	15.1669 KOps/s	15.5665 KOps/s	$\color{#d91a1a}-2.57\%$
test_compile_add_self_flat[tensorclass-compile]	0.1547ms	0.1032ms	9.6914 KOps/s	9.8538 KOps/s	$\color{#d91a1a}-1.65\%$
test_compile_add_self_flat[pytree-eager]	0.4826ms	0.4119ms	2.4278 KOps/s	2.5008 KOps/s	$\color{#d91a1a}-2.92\%$
test_compile_add_self_flat[pytree-compile]	0.1751ms	0.1368ms	7.3094 KOps/s	7.4405 KOps/s	$\color{#d91a1a}-1.76\%$
test_compile_copy_flat[tensordict-compile]	62.1210μs	19.5671μs	51.1062 KOps/s	51.4626 KOps/s	$\color{#d91a1a}-0.69\%$
test_compile_copy_flat[tensordict-eager]	63.3710μs	26.9287μs	37.1352 KOps/s	36.4406 KOps/s	$\color{#35bf28}+1.91\%$
test_compile_copy_flat[pytree-compile]	0.1701ms	69.4226μs	14.4045 KOps/s	14.3316 KOps/s	$\color{#35bf28}+0.51\%$
test_compile_copy_flat[pytree-eager]	79.7510μs	50.9418μs	19.6302 KOps/s	19.4143 KOps/s	$\color{#35bf28}+1.11\%$
test_compile_assign_and_add[tensordict-compile]	1.6356ms	0.3954ms	2.5290 KOps/s	2.2325 KOps/s	$\textbf{\color{#35bf28}+13.28\%}$
test_compile_assign_and_add[tensordict-eager]	2.8375ms	2.6277ms	380.5548 Ops/s	387.0478 Ops/s	$\color{#d91a1a}-1.68\%$
test_compile_assign_and_add[pytree-compile]	1.5800ms	0.4291ms	2.3303 KOps/s	2.2653 KOps/s	$\color{#35bf28}+2.87\%$
test_compile_assign_and_add[pytree-eager]	2.7238ms	2.6176ms	382.0237 Ops/s	384.4001 Ops/s	$\color{#d91a1a}-0.62\%$
test_compile_indexing[tensor-tensordict-compile]	0.4599ms	0.1176ms	8.5029 KOps/s	8.6468 KOps/s	$\color{#d91a1a}-1.66\%$
test_compile_indexing[tensor-tensordict-eager]	0.5572ms	82.3779μs	12.1392 KOps/s	12.3127 KOps/s	$\color{#d91a1a}-1.41\%$
test_compile_indexing[tensor-tensorclass-compile]	0.5167ms	0.1094ms	9.1409 KOps/s	9.1060 KOps/s	$\color{#35bf28}+0.38\%$
test_compile_indexing[tensor-tensorclass-eager]	0.1244ms	68.2942μs	14.6425 KOps/s	13.8937 KOps/s	$\textbf{\color{#35bf28}+5.39\%}$
test_compile_indexing[tensor-pytree-compile]	0.2788ms	0.1060ms	9.4350 KOps/s	9.4145 KOps/s	$\color{#35bf28}+0.22\%$
test_compile_indexing[tensor-pytree-eager]	0.2249ms	69.8071μs	14.3252 KOps/s	14.1189 KOps/s	$\color{#35bf28}+1.46\%$
test_compile_indexing[slice-tensordict-compile]	0.1417ms	99.9616μs	10.0038 KOps/s	9.9473 KOps/s	$\color{#35bf28}+0.57\%$
test_compile_indexing[slice-tensordict-eager]	0.1588ms	17.7177μs	56.4408 KOps/s	58.3028 KOps/s	$\color{#d91a1a}-3.19\%$
test_compile_indexing[slice-tensorclass-compile]	0.2148ms	95.4068μs	10.4814 KOps/s	10.4998 KOps/s	$\color{#d91a1a}-0.17\%$
test_compile_indexing[slice-tensorclass-eager]	63.5700μs	16.0408μs	62.3411 KOps/s	63.0356 KOps/s	$\color{#d91a1a}-1.10\%$
test_compile_indexing[slice-pytree-compile]	0.1987ms	98.4387μs	10.1586 KOps/s	10.3902 KOps/s	$\color{#d91a1a}-2.23\%$
test_compile_indexing[slice-pytree-eager]	47.7900μs	16.0945μs	62.1329 KOps/s	63.9217 KOps/s	$\color{#d91a1a}-2.80\%$
test_compile_indexing[int-tensordict-compile]	0.1612ms	0.1010ms	9.9000 KOps/s	9.9585 KOps/s	$\color{#d91a1a}-0.59\%$
test_compile_indexing[int-tensordict-eager]	0.5703ms	17.5177μs	57.0851 KOps/s	58.0082 KOps/s	$\color{#d91a1a}-1.59\%$
test_compile_indexing[int-tensorclass-compile]	0.1448ms	0.1011ms	9.8954 KOps/s	10.4893 KOps/s	$\textbf{\color{#d91a1a}-5.66\%}$
test_compile_indexing[int-tensorclass-eager]	50.3800μs	16.3560μs	61.1395 KOps/s	63.7286 KOps/s	$\color{#d91a1a}-4.06\%$
test_compile_indexing[int-pytree-compile]	0.1432ms	99.8346μs	10.0166 KOps/s	10.4491 KOps/s	$\color{#d91a1a}-4.14\%$
test_compile_indexing[int-pytree-eager]	0.1290ms	19.1345μs	52.2615 KOps/s	64.1114 KOps/s	$\textbf{\color{#d91a1a}-18.48\%}$
test_mod_add[eager]	90.0710μs	38.2920μs	26.1151 KOps/s	27.6225 KOps/s	$\textbf{\color{#d91a1a}-5.46\%}$
test_mod_add[compile]	0.2860ms	84.9127μs	11.7768 KOps/s	11.8828 KOps/s	$\color{#d91a1a}-0.89\%$
test_mod_add[compile-overhead]	0.3305ms	0.1678ms	5.9588 KOps/s	5.7340 KOps/s	$\color{#35bf28}+3.92\%$
test_mod_wrap[eager]	0.3858ms	0.2650ms	3.7738 KOps/s	3.9271 KOps/s	$\color{#d91a1a}-3.90\%$
test_mod_wrap[compile]	0.8204ms	0.3003ms	3.3301 KOps/s	3.4652 KOps/s	$\color{#d91a1a}-3.90\%$
test_mod_wrap[compile-overhead]	7.1170ms	3.8133ms	262.2382 Ops/s	265.6966 Ops/s	$\color{#d91a1a}-1.30\%$
test_mod_wrap_and_backward[eager]	1.5074ms	1.3604ms	735.0559 Ops/s	692.9063 Ops/s	$\textbf{\color{#35bf28}+6.08\%}$
test_mod_wrap_and_backward[compile]	1.4123ms	1.2684ms	788.4061 Ops/s	729.0813 Ops/s	$\textbf{\color{#35bf28}+8.14\%}$
test_mod_wrap_and_backward[compile-overhead]	1.3555ms	0.9154ms	1.0924 KOps/s	941.6609 Ops/s	$\textbf{\color{#35bf28}+16.00\%}$
test_seq_add[eager]	0.1822ms	0.1178ms	8.4870 KOps/s	9.2028 KOps/s	$\textbf{\color{#d91a1a}-7.78\%}$
test_seq_add[compile]	0.1280ms	87.6484μs	11.4092 KOps/s	11.3315 KOps/s	$\color{#35bf28}+0.69\%$
test_seq_add[compile-overhead]	0.1680ms	0.1278ms	7.8251 KOps/s	7.8448 KOps/s	$\color{#d91a1a}-0.25\%$
test_seq_wrap[eager]	0.4931ms	0.4168ms	2.3991 KOps/s	2.4183 KOps/s	$\color{#d91a1a}-0.80\%$
test_seq_wrap[compile]	0.3555ms	0.2984ms	3.3514 KOps/s	3.3073 KOps/s	$\color{#35bf28}+1.34\%$
test_seq_wrap[compile-overhead]	0.2733ms	0.2216ms	4.5116 KOps/s	4.4836 KOps/s	$\color{#35bf28}+0.62\%$
test_func_call_runtime[False-eager]	0.8359ms	0.7421ms	1.3475 KOps/s	1.3697 KOps/s	$\color{#d91a1a}-1.62\%$
test_func_call_runtime[False-compile]	0.8562ms	0.7415ms	1.3486 KOps/s	1.3409 KOps/s	$\color{#35bf28}+0.58\%$
test_func_call_runtime[False-compile-overhead]	0.4240ms	0.3592ms	2.7840 KOps/s	2.8153 KOps/s	$\color{#d91a1a}-1.11\%$
test_func_call_runtime[True-eager]	1.0160ms	0.9127ms	1.0957 KOps/s	1.1010 KOps/s	$\color{#d91a1a}-0.48\%$
test_func_call_runtime[True-compile]	0.8216ms	0.7598ms	1.3161 KOps/s	1.3067 KOps/s	$\color{#35bf28}+0.71\%$
test_func_call_runtime[True-compile-overhead]	0.4696ms	0.3789ms	2.6392 KOps/s	2.6442 KOps/s	$\color{#d91a1a}-0.19\%$
test_func_call_cm_runtime[False-eager]	0.8066ms	0.7331ms	1.3640 KOps/s	1.3832 KOps/s	$\color{#d91a1a}-1.39\%$
test_func_call_cm_runtime[False-compile]	0.8941ms	0.7442ms	1.3437 KOps/s	1.2932 KOps/s	$\color{#35bf28}+3.91\%$
test_func_call_cm_runtime[False-compile-overhead]	0.4309ms	0.3623ms	2.7599 KOps/s	2.7859 KOps/s	$\color{#d91a1a}-0.93\%$
test_func_call_cm_runtime[True-eager]	1.1153ms	1.0023ms	997.6984 Ops/s	998.6397 Ops/s	$\color{#d91a1a}-0.09\%$
test_func_call_cm_runtime[True-compile]	0.9697ms	0.8239ms	1.2137 KOps/s	1.2555 KOps/s	$\color{#d91a1a}-3.33\%$
test_func_call_cm_runtime[True-compile-overhead]	0.5028ms	0.4096ms	2.4412 KOps/s	2.4519 KOps/s	$\color{#d91a1a}-0.44\%$
test_vmap_func_call_cm_runtime[eager]	2.5223ms	2.0698ms	483.1422 Ops/s	480.9891 Ops/s	$\color{#35bf28}+0.45\%$
test_vmap_func_call_cm_runtime[compile]	0.9582ms	0.8011ms	1.2482 KOps/s	1.2369 KOps/s	$\color{#35bf28}+0.91\%$
test_vmap_func_call_cm_runtime[compile-overhead]	0.4591ms	0.4092ms	2.4435 KOps/s	2.4416 KOps/s	$\color{#35bf28}+0.08\%$
test_distributed	4.5184ms	0.2335ms	4.2826 KOps/s	8.4508 KOps/s	$\textbf{\color{#d91a1a}-49.32\%}$
test_tdmodule	59.3310μs	21.0208μs	47.5719 KOps/s	53.5360 KOps/s	$\textbf{\color{#d91a1a}-11.14\%}$
test_tdmodule_dispatch	68.8210μs	38.8416μs	25.7456 KOps/s	29.4353 KOps/s	$\textbf{\color{#d91a1a}-12.53\%}$
test_tdseq	45.1200μs	21.4692μs	46.5784 KOps/s	54.2328 KOps/s	$\textbf{\color{#d91a1a}-14.11\%}$
test_tdseq_dispatch	68.5210μs	41.2496μs	24.2427 KOps/s	28.3270 KOps/s	$\textbf{\color{#d91a1a}-14.42\%}$
test_instantiation_functorch	1.7100ms	1.5825ms	631.9089 Ops/s	639.5299 Ops/s	$\color{#d91a1a}-1.19\%$
test_exec_functorch	0.2078ms	0.1490ms	6.7135 KOps/s	7.0194 KOps/s	$\color{#d91a1a}-4.36\%$
test_exec_functional_call	0.1857ms	0.1431ms	6.9861 KOps/s	7.3470 KOps/s	$\color{#d91a1a}-4.91\%$
test_exec_td_decorator	0.3746ms	0.1890ms	5.2914 KOps/s	5.4865 KOps/s	$\color{#d91a1a}-3.56\%$
test_vmap_mlp_speed_decorator[True-True]	0.7896ms	0.6813ms	1.4678 KOps/s	1.4668 KOps/s	$\color{#35bf28}+0.07\%$
test_vmap_mlp_speed_decorator[True-False]	0.7873ms	0.6804ms	1.4698 KOps/s	1.4674 KOps/s	$\color{#35bf28}+0.17\%$
test_vmap_mlp_speed_decorator[False-True]	0.6998ms	0.5907ms	1.6930 KOps/s	1.6879 KOps/s	$\color{#35bf28}+0.30\%$
test_vmap_mlp_speed_decorator[False-False]	0.6966ms	0.5906ms	1.6933 KOps/s	1.6809 KOps/s	$\color{#35bf28}+0.74\%$
test_vmap_transformer_speed_decorator[True-True]	19.2278ms	19.1247ms	52.2883 Ops/s	52.4900 Ops/s	$\color{#d91a1a}-0.38\%$
test_vmap_transformer_speed_decorator[True-False]	19.9362ms	19.1999ms	52.0837 Ops/s	52.4896 Ops/s	$\color{#d91a1a}-0.77\%$
test_vmap_transformer_speed_decorator[False-True]	19.1595ms	19.0715ms	52.4344 Ops/s	52.8565 Ops/s	$\color{#d91a1a}-0.80\%$
test_vmap_transformer_speed_decorator[False-False]	19.1448ms	19.0157ms	52.5882 Ops/s	52.8252 Ops/s	$\color{#d91a1a}-0.45\%$
test_to_module_speed[True]	1.0597ms	0.9513ms	1.0512 KOps/s	1.0557 KOps/s	$\color{#d91a1a}-0.42\%$
test_to_module_speed[False]	1.2595ms	0.9319ms	1.0731 KOps/s	1.0772 KOps/s	$\color{#d91a1a}-0.38\%$
test_tc_init	67.8610μs	37.9221μs	26.3699 KOps/s	28.1286 KOps/s	$\textbf{\color{#d91a1a}-6.25\%}$
test_tc_init_nested	0.2409ms	76.9991μs	12.9872 KOps/s	13.6086 KOps/s	$\color{#d91a1a}-4.57\%$
test_tc_first_layer_tensor	4.0471μs	0.7002μs	1.4282 MOps/s	1.4063 MOps/s	$\color{#35bf28}+1.56\%$
test_tc_first_layer_nontensor	33.5500μs	2.3228μs	430.5208 KOps/s	431.6699 KOps/s	$\color{#d91a1a}-0.27\%$
test_tc_second_layer_tensor	13.8000μs	1.4399μs	694.4775 KOps/s	690.9010 KOps/s	$\color{#35bf28}+0.52\%$
test_tc_second_layer_nontensor	26.1110μs	3.0753μs	325.1747 KOps/s	326.5493 KOps/s	$\color{#d91a1a}-0.42\%$
test_unbind	0.2302s	10.1559ms	98.4648 Ops/s	150.1969 Ops/s	$\textbf{\color{#d91a1a}-34.44\%}$
test_full_like	10.7637ms	9.7511ms	102.5527 Ops/s	102.8823 Ops/s	$\color{#d91a1a}-0.32\%$
test_zeros_like	5.0914ms	4.4131ms	226.5968 Ops/s	227.8466 Ops/s	$\color{#d91a1a}-0.55\%$
test_ones_like	5.1380ms	4.4314ms	225.6642 Ops/s	227.1558 Ops/s	$\color{#d91a1a}-0.66\%$
test_clone	7.6786ms	6.8349ms	146.3084 Ops/s	147.1458 Ops/s	$\color{#d91a1a}-0.57\%$
test_squeeze	58.6700μs	9.1967μs	108.7352 KOps/s	108.9101 KOps/s	$\color{#d91a1a}-0.16\%$
test_unsqueeze	0.2281ms	74.5881μs	13.4070 KOps/s	14.1220 KOps/s	$\textbf{\color{#d91a1a}-5.06\%}$
test_split	0.2635ms	0.1621ms	6.1678 KOps/s	6.2794 KOps/s	$\color{#d91a1a}-1.78\%$
test_permute	0.2399ms	0.1751ms	5.7100 KOps/s	5.3652 KOps/s	$\textbf{\color{#35bf28}+6.43\%}$
test_stack	52.6180ms	51.9029ms	19.2667 Ops/s	19.3595 Ops/s	$\color{#d91a1a}-0.48\%$
test_cat	52.8605ms	51.7498ms	19.3237 Ops/s	19.2863 Ops/s	$\color{#35bf28}+0.19\%$