[Feature] UnbatchedTensor

[vmoens]

Stack from ghstack (oldest at bottom):

• -> [Feature] UnbatchedTensor #1170
• [Feature] Add missing __torch_function__ #1169

[github-actions]

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

Total Benchmarks: 217. Improved: $\large\color{#35bf28}20$. Worsened: $\large\color{#d91a1a}8$.

Expand to view detailed results

Name	Max	Mean	Ops	Ops on Repo HEAD	Change
test_plain_set_nested	53.3500μs	20.5679μs	48.6195 KOps/s	48.6590 KOps/s	$\color{#d91a1a}-0.08\%$
test_plain_set_stack_nested	49.3620μs	20.9773μs	47.6706 KOps/s	48.7385 KOps/s	$\color{#d91a1a}-2.19\%$
test_plain_set_nested_inplace	66.1530μs	22.5404μs	44.3649 KOps/s	44.8090 KOps/s	$\color{#d91a1a}-0.99\%$
test_plain_set_stack_nested_inplace	66.2230μs	22.3847μs	44.6734 KOps/s	44.3933 KOps/s	$\color{#35bf28}+0.63\%$
test_items	57.0660μs	4.1750μs	239.5237 KOps/s	244.6243 KOps/s	$\color{#d91a1a}-2.09\%$
test_items_nested	0.6230ms	0.3936ms	2.5410 KOps/s	2.5421 KOps/s	$\color{#d91a1a}-0.05\%$
test_items_nested_locked	0.6748ms	0.3912ms	2.5565 KOps/s	2.5507 KOps/s	$\color{#35bf28}+0.23\%$
test_items_nested_leaf	0.1382ms	76.9874μs	12.9891 KOps/s	13.0276 KOps/s	$\color{#d91a1a}-0.30\%$
test_items_stack_nested	0.5805ms	0.3944ms	2.5355 KOps/s	2.5205 KOps/s	$\color{#35bf28}+0.60\%$
test_items_stack_nested_leaf	0.1562ms	76.8054μs	13.0199 KOps/s	12.6675 KOps/s	$\color{#35bf28}+2.78\%$
test_items_stack_nested_locked	0.7431ms	0.3955ms	2.5285 KOps/s	2.5083 KOps/s	$\color{#35bf28}+0.80\%$
test_keys	43.9620μs	3.5328μs	283.0584 KOps/s	280.6814 KOps/s	$\color{#35bf28}+0.85\%$
test_keys_nested	0.2866ms	0.1642ms	6.0916 KOps/s	6.2066 KOps/s	$\color{#d91a1a}-1.85\%$
test_keys_nested_locked	0.6924ms	0.1699ms	5.8851 KOps/s	5.9376 KOps/s	$\color{#d91a1a}-0.88\%$
test_keys_nested_leaf	0.2551ms	0.1434ms	6.9742 KOps/s	7.0904 KOps/s	$\color{#d91a1a}-1.64\%$
test_keys_stack_nested	0.2648ms	0.1638ms	6.1062 KOps/s	5.9645 KOps/s	$\color{#35bf28}+2.38\%$
test_keys_stack_nested_leaf	0.2252ms	0.1436ms	6.9626 KOps/s	7.1610 KOps/s	$\color{#d91a1a}-2.77\%$
test_keys_stack_nested_locked	0.2604ms	0.1702ms	5.8759 KOps/s	5.8804 KOps/s	$\color{#d91a1a}-0.08\%$
test_values	7.2536μs	1.0332μs	967.8897 KOps/s	867.8004 KOps/s	$\textbf{\color{#35bf28}+11.53\%}$
test_values_nested	0.1206ms	62.3655μs	16.0345 KOps/s	16.3538 KOps/s	$\color{#d91a1a}-1.95\%$
test_values_nested_locked	0.1363ms	62.2160μs	16.0730 KOps/s	16.4133 KOps/s	$\color{#d91a1a}-2.07\%$
test_values_nested_leaf	0.1253ms	70.8958μs	14.1052 KOps/s	13.9567 KOps/s	$\color{#35bf28}+1.06\%$
test_values_stack_nested	0.1399ms	61.7020μs	16.2069 KOps/s	15.9368 KOps/s	$\color{#35bf28}+1.70\%$
test_values_stack_nested_leaf	0.1708ms	70.6482μs	14.1546 KOps/s	14.0110 KOps/s	$\color{#35bf28}+1.03\%$
test_values_stack_nested_locked	0.1296ms	61.8814μs	16.1599 KOps/s	15.8828 KOps/s	$\color{#35bf28}+1.74\%$
test_membership	61.6250μs	0.8934μs	1.1193 MOps/s	1.1797 MOps/s	$\textbf{\color{#d91a1a}-5.12\%}$
test_membership_nested	34.8750μs	2.9366μs	340.5331 KOps/s	352.8445 KOps/s	$\color{#d91a1a}-3.49\%$
test_membership_nested_leaf	28.4130μs	2.9306μs	341.2244 KOps/s	351.5709 KOps/s	$\color{#d91a1a}-2.94\%$
test_membership_stacked_nested	47.3480μs	2.9280μs	341.5308 KOps/s	351.6117 KOps/s	$\color{#d91a1a}-2.87\%$
test_membership_stacked_nested_leaf	28.1930μs	2.9121μs	343.3942 KOps/s	342.5905 KOps/s	$\color{#35bf28}+0.23\%$
test_membership_nested_last	31.0180μs	4.4214μs	226.1727 KOps/s	235.8693 KOps/s	$\color{#d91a1a}-4.11\%$
test_membership_nested_leaf_last	45.1450μs	4.4435μs	225.0501 KOps/s	235.9800 KOps/s	$\color{#d91a1a}-4.63\%$
test_membership_stacked_nested_last	28.0430μs	4.3594μs	229.3895 KOps/s	228.9228 KOps/s	$\color{#35bf28}+0.20\%$
test_membership_stacked_nested_leaf_last	51.3560μs	4.4027μs	227.1329 KOps/s	233.4638 KOps/s	$\color{#d91a1a}-2.71\%$
test_nested_getleaf	52.8290μs	10.3560μs	96.5624 KOps/s	98.7977 KOps/s	$\color{#d91a1a}-2.26\%$
test_nested_get	52.0870μs	9.9248μs	100.7575 KOps/s	101.4317 KOps/s	$\color{#d91a1a}-0.66\%$
test_stacked_getleaf	50.0130μs	10.4114μs	96.0490 KOps/s	95.6232 KOps/s	$\color{#35bf28}+0.45\%$
test_stacked_get	39.3230μs	9.8581μs	101.4392 KOps/s	100.3051 KOps/s	$\color{#35bf28}+1.13\%$
test_nested_getitemleaf	45.7550μs	11.0635μs	90.3872 KOps/s	90.9636 KOps/s	$\color{#d91a1a}-0.63\%$
test_nested_getitem	41.4370μs	10.5195μs	95.0611 KOps/s	97.7230 KOps/s	$\color{#d91a1a}-2.72\%$
test_stacked_getitemleaf	44.5130μs	11.0863μs	90.2012 KOps/s	90.8823 KOps/s	$\color{#d91a1a}-0.75\%$
test_stacked_getitem	49.5530μs	10.5590μs	94.7055 KOps/s	98.6709 KOps/s	$\color{#d91a1a}-4.02\%$
test_lock_nested	1.2966ms	0.4756ms	2.1025 KOps/s	1.6123 KOps/s	$\textbf{\color{#35bf28}+30.41\%}$
test_lock_stack_nested	0.8264ms	0.4491ms	2.2266 KOps/s	2.0790 KOps/s	$\textbf{\color{#35bf28}+7.10\%}$
test_unlock_nested	1.1007ms	0.3931ms	2.5439 KOps/s	2.3286 KOps/s	$\textbf{\color{#35bf28}+9.25\%}$
test_unlock_stack_nested	0.7082ms	0.3634ms	2.7518 KOps/s	2.5794 KOps/s	$\textbf{\color{#35bf28}+6.68\%}$
test_flatten_speed	0.1776ms	99.1364μs	10.0871 KOps/s	9.9111 KOps/s	$\color{#35bf28}+1.78\%$
test_unflatten_speed	0.9477ms	0.5158ms	1.9387 KOps/s	1.9237 KOps/s	$\color{#35bf28}+0.78\%$
test_common_ops	5.1514ms	0.8284ms	1.2072 KOps/s	1.2146 KOps/s	$\color{#d91a1a}-0.61\%$
test_creation	25.8780μs	2.4875μs	402.0110 KOps/s	407.4576 KOps/s	$\color{#d91a1a}-1.34\%$
test_creation_empty	59.3610μs	12.1824μs	82.0857 KOps/s	85.3025 KOps/s	$\color{#d91a1a}-3.77\%$
test_creation_nested_1	59.5610μs	15.3004μs	65.3580 KOps/s	68.0419 KOps/s	$\color{#d91a1a}-3.94\%$
test_creation_nested_2	74.2280μs	19.6982μs	50.7661 KOps/s	51.6774 KOps/s	$\color{#d91a1a}-1.76\%$
test_clone	49.9430μs	13.5858μs	73.6064 KOps/s	72.8769 KOps/s	$\color{#35bf28}+1.00\%$
test_getitem[int]	1.4638ms	13.0965μs	76.3562 KOps/s	77.4859 KOps/s	$\color{#d91a1a}-1.46\%$
test_getitem[slice_int]	0.1658ms	25.0657μs	39.8951 KOps/s	40.2189 KOps/s	$\color{#d91a1a}-0.81\%$
test_getitem[range]	0.2215ms	50.4593μs	19.8179 KOps/s	20.3555 KOps/s	$\color{#d91a1a}-2.64\%$
test_getitem[tuple]	0.1693ms	20.2805μs	49.3084 KOps/s	48.6413 KOps/s	$\color{#35bf28}+1.37\%$
test_getitem[list]	0.2109ms	45.6506μs	21.9055 KOps/s	22.4210 KOps/s	$\color{#d91a1a}-2.30\%$
test_setitem_dim[int]	67.7960μs	26.3557μs	37.9425 KOps/s	40.9427 KOps/s	$\textbf{\color{#d91a1a}-7.33\%}$
test_setitem_dim[slice_int]	0.1187ms	52.4323μs	19.0722 KOps/s	19.3270 KOps/s	$\color{#d91a1a}-1.32\%$
test_setitem_dim[range]	0.1200ms	74.6818μs	13.3901 KOps/s	13.8317 KOps/s	$\color{#d91a1a}-3.19\%$
test_setitem_dim[tuple]	85.3290μs	41.4222μs	24.1416 KOps/s	25.1835 KOps/s	$\color{#d91a1a}-4.14\%$
test_setitem	77.9140μs	21.2929μs	46.9641 KOps/s	47.0667 KOps/s	$\color{#d91a1a}-0.22\%$
test_set	68.4680μs	20.6032μs	48.5361 KOps/s	47.7877 KOps/s	$\color{#35bf28}+1.57\%$
test_set_shared	4.9751ms	0.1829ms	5.4689 KOps/s	5.4115 KOps/s	$\color{#35bf28}+1.06\%$
test_update	0.4708ms	23.4809μs	42.5878 KOps/s	42.9975 KOps/s	$\color{#d91a1a}-0.95\%$
test_update_nested	0.3800ms	33.4083μs	29.9326 KOps/s	29.5808 KOps/s	$\color{#35bf28}+1.19\%$
test_update__nested	0.5539ms	33.8538μs	29.5388 KOps/s	28.9596 KOps/s	$\color{#35bf28}+2.00\%$
test_set_nested	0.4360ms	22.5578μs	44.3306 KOps/s	43.4893 KOps/s	$\color{#35bf28}+1.93\%$
test_set_nested_new	91.4000μs	27.0916μs	36.9119 KOps/s	35.8379 KOps/s	$\color{#35bf28}+3.00\%$
test_select	0.1015ms	43.0460μs	23.2309 KOps/s	22.4102 KOps/s	$\color{#35bf28}+3.66\%$
test_select_nested	0.1301ms	61.9627μs	16.1387 KOps/s	15.8082 KOps/s	$\color{#35bf28}+2.09\%$
test_exclude_nested	0.1500ms	79.5998μs	12.5628 KOps/s	12.2784 KOps/s	$\color{#35bf28}+2.32\%$
test_empty[True]	0.5412ms	0.4034ms	2.4789 KOps/s	2.4479 KOps/s	$\color{#35bf28}+1.27\%$
test_empty[False]	45.0540μs	1.4953μs	668.7813 KOps/s	723.4456 KOps/s	$\textbf{\color{#d91a1a}-7.56\%}$
test_unbind_speed	0.4540ms	0.2686ms	3.7235 KOps/s	3.7044 KOps/s	$\color{#35bf28}+0.51\%$
test_unbind_speed_stack0	0.4705ms	0.2672ms	3.7432 KOps/s	3.7333 KOps/s	$\color{#35bf28}+0.26\%$
test_unbind_speed_stack1	0.1253s	0.8557ms	1.1687 KOps/s	1.4989 KOps/s	$\textbf{\color{#d91a1a}-22.03\%}$
test_split	0.1270s	1.8147ms	551.0476 Ops/s	499.2232 Ops/s	$\textbf{\color{#35bf28}+10.38\%}$
test_chunk	1.9985ms	1.6077ms	621.9900 Ops/s	499.0749 Ops/s	$\textbf{\color{#35bf28}+24.63\%}$
test_consolidate_njt[False-None]	0.1331s	9.7196ms	102.8846 Ops/s	108.8334 Ops/s	$\textbf{\color{#d91a1a}-5.47\%}$
test_creation[device0]	0.3445ms	92.3694μs	10.8261 KOps/s	10.4468 KOps/s	$\color{#35bf28}+3.63\%$
test_creation_from_tensor	5.0527ms	95.6403μs	10.4558 KOps/s	10.3654 KOps/s	$\color{#35bf28}+0.87\%$
test_add_one[memmap_tensor0]	0.8522ms	5.0786μs	196.9065 KOps/s	197.5912 KOps/s	$\color{#d91a1a}-0.35\%$
test_contiguous[memmap_tensor0]	36.1970μs	0.5136μs	1.9469 MOps/s	1.9449 MOps/s	$\color{#35bf28}+0.10\%$
test_stack[memmap_tensor0]	0.2133ms	3.6089μs	277.0915 KOps/s	286.0283 KOps/s	$\color{#d91a1a}-3.12\%$
test_memmaptd_index	1.1609ms	0.2417ms	4.1382 KOps/s	3.9760 KOps/s	$\color{#35bf28}+4.08\%$
test_memmaptd_index_astensor	0.6655ms	0.3318ms	3.0141 KOps/s	2.9116 KOps/s	$\color{#35bf28}+3.52\%$
test_memmaptd_index_op	1.3498ms	0.6169ms	1.6211 KOps/s	1.5944 KOps/s	$\color{#35bf28}+1.68\%$
test_serialize_model	0.1360s	0.1278s	7.8251 Ops/s	7.8968 Ops/s	$\color{#d91a1a}-0.91\%$
test_serialize_model_pickle	0.5097s	0.4042s	2.4738 Ops/s	2.4422 Ops/s	$\color{#35bf28}+1.30\%$
test_serialize_weights	0.1302s	0.1225s	8.1616 Ops/s	6.7889 Ops/s	$\textbf{\color{#35bf28}+20.22\%}$
test_serialize_weights_returnearly	0.2831s	0.1818s	5.4995 Ops/s	6.0130 Ops/s	$\textbf{\color{#d91a1a}-8.54\%}$
test_serialize_weights_pickle	0.4837s	0.4070s	2.4572 Ops/s	2.5235 Ops/s	$\color{#d91a1a}-2.63\%$
test_serialize_weights_filesystem	0.1599s	0.1530s	6.5366 Ops/s	6.2744 Ops/s	$\color{#35bf28}+4.18\%$
test_serialize_model_filesystem	0.1687s	0.1578s	6.3371 Ops/s	5.7230 Ops/s	$\textbf{\color{#35bf28}+10.73\%}$
test_reshape_pytree	0.1283ms	26.5931μs	37.6037 KOps/s	37.5755 KOps/s	$\color{#35bf28}+0.08\%$
test_reshape_td	84.8080μs	32.5948μs	30.6798 KOps/s	29.8670 KOps/s	$\color{#35bf28}+2.72\%$
test_view_pytree	0.1138ms	26.7303μs	37.4107 KOps/s	37.5031 KOps/s	$\color{#d91a1a}-0.25\%$
test_view_td	0.1136ms	37.5066μs	26.6619 KOps/s	25.7395 KOps/s	$\color{#35bf28}+3.58\%$
test_unbind_pytree	0.1174ms	29.4598μs	33.9445 KOps/s	33.8915 KOps/s	$\color{#35bf28}+0.16\%$
test_unbind_td	0.3818ms	39.4114μs	25.3733 KOps/s	25.1346 KOps/s	$\color{#35bf28}+0.95\%$
test_split_pytree	85.9500μs	29.7720μs	33.5886 KOps/s	33.5579 KOps/s	$\color{#35bf28}+0.09\%$
test_split_td	0.5631ms	45.6722μs	21.8952 KOps/s	21.8816 KOps/s	$\color{#35bf28}+0.06\%$
test_add_pytree	92.5220μs	34.8481μs	28.6960 KOps/s	28.0266 KOps/s	$\color{#35bf28}+2.39\%$
test_add_td	0.3266ms	57.9810μs	17.2470 KOps/s	17.7635 KOps/s	$\color{#d91a1a}-2.91\%$
test_compile_add_one_nested[tensordict-compile]	0.1645ms	63.5835μs	15.7273 KOps/s	15.5877 KOps/s	$\color{#35bf28}+0.90\%$
test_compile_add_one_nested[tensordict-eager]	0.4413ms	0.1740ms	5.7477 KOps/s	5.6206 KOps/s	$\color{#35bf28}+2.26\%$
test_compile_add_one_nested[pytree-compile]	0.1416ms	45.1231μs	22.1616 KOps/s	21.4329 KOps/s	$\color{#35bf28}+3.40\%$
test_compile_add_one_nested[pytree-eager]	0.2394ms	0.1183ms	8.4560 KOps/s	8.3681 KOps/s	$\color{#35bf28}+1.05\%$
test_compile_copy_nested[tensordict-compile]	0.1085ms	26.1726μs	38.2079 KOps/s	38.8174 KOps/s	$\color{#d91a1a}-1.57\%$
test_compile_copy_nested[tensordict-eager]	0.1367ms	57.9499μs	17.2563 KOps/s	16.7387 KOps/s	$\color{#35bf28}+3.09\%$
test_compile_copy_nested[pytree-compile]	0.1831ms	77.1306μs	12.9650 KOps/s	12.7638 KOps/s	$\color{#35bf28}+1.58\%$
test_compile_copy_nested[pytree-eager]	0.1215ms	66.9230μs	14.9425 KOps/s	14.8448 KOps/s	$\color{#35bf28}+0.66\%$
test_compile_add_one_flat[tensordict-compile]	0.2218ms	0.1053ms	9.4927 KOps/s	9.3134 KOps/s	$\color{#35bf28}+1.93\%$
test_compile_add_one_flat[tensordict-eager]	0.4940ms	0.2180ms	4.5877 KOps/s	4.6372 KOps/s	$\color{#d91a1a}-1.07\%$
test_compile_add_one_flat[tensorclass-compile]	0.1384ms	46.1396μs	21.6734 KOps/s	21.1178 KOps/s	$\color{#35bf28}+2.63\%$
test_compile_add_one_flat[tensorclass-eager]	0.6635ms	66.4241μs	15.0548 KOps/s	14.4377 KOps/s	$\color{#35bf28}+4.27\%$
test_compile_add_one_flat[pytree-compile]	0.1981ms	0.1024ms	9.7657 KOps/s	9.6768 KOps/s	$\color{#35bf28}+0.92\%$
test_compile_add_one_flat[pytree-eager]	0.4147ms	0.2061ms	4.8525 KOps/s	4.9699 KOps/s	$\color{#d91a1a}-2.36\%$
test_compile_add_self_flat[tensordict-eager]	0.3971ms	0.2339ms	4.2752 KOps/s	4.2538 KOps/s	$\color{#35bf28}+0.50\%$
test_compile_add_self_flat[tensordict-compile]	0.2760ms	0.1082ms	9.2407 KOps/s	9.2989 KOps/s	$\color{#d91a1a}-0.63\%$
test_compile_add_self_flat[tensorclass-eager]	0.2897ms	62.2503μs	16.0642 KOps/s	16.0095 KOps/s	$\color{#35bf28}+0.34\%$
test_compile_add_self_flat[tensorclass-compile]	0.1523ms	47.4329μs	21.0824 KOps/s	21.1063 KOps/s	$\color{#d91a1a}-0.11\%$
test_compile_add_self_flat[pytree-eager]	0.3305ms	0.1673ms	5.9790 KOps/s	6.3688 KOps/s	$\textbf{\color{#d91a1a}-6.12\%}$
test_compile_add_self_flat[pytree-compile]	0.2175ms	0.1023ms	9.7791 KOps/s	9.5918 KOps/s	$\color{#35bf28}+1.95\%$
test_compile_copy_flat[tensordict-compile]	89.1460μs	21.1665μs	47.2446 KOps/s	46.7870 KOps/s	$\color{#35bf28}+0.98\%$
test_compile_copy_flat[tensordict-eager]	0.1544ms	66.6294μs	15.0084 KOps/s	15.1751 KOps/s	$\color{#d91a1a}-1.10\%$
test_compile_copy_flat[pytree-compile]	0.1730ms	78.7057μs	12.7056 KOps/s	12.6411 KOps/s	$\color{#35bf28}+0.51\%$
test_compile_copy_flat[pytree-eager]	0.1423ms	67.9216μs	14.7229 KOps/s	14.7415 KOps/s	$\color{#d91a1a}-0.13\%$
test_compile_assign_and_add[tensordict-compile]	0.4395ms	0.2098ms	4.7666 KOps/s	4.7734 KOps/s	$\color{#d91a1a}-0.14\%$
test_compile_assign_and_add[tensordict-eager]	2.2891ms	1.3541ms	738.5184 Ops/s	761.3921 Ops/s	$\color{#d91a1a}-3.00\%$
test_compile_assign_and_add[pytree-compile]	0.3135ms	0.2038ms	4.9067 KOps/s	4.8570 KOps/s	$\color{#35bf28}+1.02\%$
test_compile_assign_and_add[pytree-eager]	0.9670ms	0.7796ms	1.2828 KOps/s	1.2818 KOps/s	$\color{#35bf28}+0.08\%$
test_compile_assign_and_add_stack[compile]	0.5957ms	0.4504ms	2.2202 KOps/s	2.1330 KOps/s	$\color{#35bf28}+4.09\%$
test_compile_assign_and_add_stack[eager]	4.7833ms	2.7771ms	360.0829 Ops/s	345.8552 Ops/s	$\color{#35bf28}+4.11\%$
test_compile_indexing[tensor-tensordict-compile]	0.1264ms	37.3369μs	26.7832 KOps/s	27.0557 KOps/s	$\color{#d91a1a}-1.01\%$
test_compile_indexing[tensor-tensordict-eager]	0.7948ms	34.2026μs	29.2376 KOps/s	28.9715 KOps/s	$\color{#35bf28}+0.92\%$
test_compile_indexing[tensor-tensorclass-compile]	0.1134ms	29.5477μs	33.8436 KOps/s	33.9424 KOps/s	$\color{#d91a1a}-0.29\%$
test_compile_indexing[tensor-tensorclass-eager]	0.1196ms	23.0669μs	43.3522 KOps/s	43.4029 KOps/s	$\color{#d91a1a}-0.12\%$
test_compile_indexing[tensor-pytree-compile]	0.1426ms	30.2352μs	33.0740 KOps/s	32.6290 KOps/s	$\color{#35bf28}+1.36\%$
test_compile_indexing[tensor-pytree-eager]	99.4250μs	23.2460μs	43.0182 KOps/s	42.8631 KOps/s	$\color{#35bf28}+0.36\%$
test_compile_indexing[slice-tensordict-compile]	0.1338ms	52.5933μs	19.0138 KOps/s	18.8201 KOps/s	$\color{#35bf28}+1.03\%$
test_compile_indexing[slice-tensordict-eager]	0.9647ms	20.8930μs	47.8630 KOps/s	48.1646 KOps/s	$\color{#d91a1a}-0.63\%$
test_compile_indexing[slice-tensorclass-compile]	0.1626ms	44.0632μs	22.6947 KOps/s	22.1041 KOps/s	$\color{#35bf28}+2.67\%$
test_compile_indexing[slice-tensorclass-eager]	0.1364ms	19.2010μs	52.0806 KOps/s	51.2345 KOps/s	$\color{#35bf28}+1.65\%$
test_compile_indexing[slice-pytree-compile]	0.1152ms	44.4662μs	22.4890 KOps/s	21.4747 KOps/s	$\color{#35bf28}+4.72\%$
test_compile_indexing[slice-pytree-eager]	72.4350μs	18.6984μs	53.4806 KOps/s	52.2586 KOps/s	$\color{#35bf28}+2.34\%$
test_compile_indexing[int-tensordict-compile]	0.1217ms	54.1201μs	18.4774 KOps/s	18.6430 KOps/s	$\color{#d91a1a}-0.89\%$
test_compile_indexing[int-tensordict-eager]	1.3089ms	20.6669μs	48.3866 KOps/s	48.5937 KOps/s	$\color{#d91a1a}-0.43\%$
test_compile_indexing[int-tensorclass-compile]	0.1160ms	45.0306μs	22.2071 KOps/s	21.5411 KOps/s	$\color{#35bf28}+3.09\%$
test_compile_indexing[int-tensorclass-eager]	65.4720μs	18.8078μs	53.1695 KOps/s	52.0524 KOps/s	$\color{#35bf28}+2.15\%$
test_compile_indexing[int-pytree-compile]	0.1277ms	44.9413μs	22.2512 KOps/s	21.5218 KOps/s	$\color{#35bf28}+3.39\%$
test_compile_indexing[int-pytree-eager]	72.5150μs	18.8581μs	53.0275 KOps/s	52.3619 KOps/s	$\color{#35bf28}+1.27\%$
test_mod_add[eager]	0.1334ms	35.4892μs	28.1776 KOps/s	26.7831 KOps/s	$\textbf{\color{#35bf28}+5.21\%}$
test_mod_add[compile]	0.1236ms	48.1798μs	20.7556 KOps/s	20.0485 KOps/s	$\color{#35bf28}+3.53\%$
test_mod_add[compile-overhead]	0.1171ms	48.1102μs	20.7856 KOps/s	19.7216 KOps/s	$\textbf{\color{#35bf28}+5.40\%}$
test_mod_wrap[eager]	0.4154ms	0.2325ms	4.3011 KOps/s	4.2344 KOps/s	$\color{#35bf28}+1.57\%$
test_mod_wrap[compile]	0.3272ms	0.2012ms	4.9694 KOps/s	4.6883 KOps/s	$\textbf{\color{#35bf28}+6.00\%}$
test_mod_wrap[compile-overhead]	0.4673ms	0.2049ms	4.8805 KOps/s	4.7393 KOps/s	$\color{#35bf28}+2.98\%$
test_mod_wrap_and_backward[eager]	20.6115ms	12.7180ms	78.6290 Ops/s	63.4358 Ops/s	$\textbf{\color{#35bf28}+23.95\%}$
test_mod_wrap_and_backward[compile]	15.4701ms	11.9163ms	83.9188 Ops/s	67.4303 Ops/s	$\textbf{\color{#35bf28}+24.45\%}$
test_mod_wrap_and_backward[compile-overhead]	13.1598ms	11.4591ms	87.2671 Ops/s	62.5524 Ops/s	$\textbf{\color{#35bf28}+39.51\%}$
test_seq_add[eager]	0.2544ms	0.1186ms	8.4283 KOps/s	8.2312 KOps/s	$\color{#35bf28}+2.39\%$
test_seq_add[compile]	0.1479ms	63.2920μs	15.7998 KOps/s	15.4217 KOps/s	$\color{#35bf28}+2.45\%$
test_seq_add[compile-overhead]	0.1335ms	61.5294μs	16.2524 KOps/s	15.9062 KOps/s	$\color{#35bf28}+2.18\%$
test_seq_wrap[eager]	0.8460ms	0.4583ms	2.1818 KOps/s	2.1494 KOps/s	$\color{#35bf28}+1.50\%$
test_seq_wrap[compile]	0.4603ms	0.2270ms	4.4054 KOps/s	4.1829 KOps/s	$\textbf{\color{#35bf28}+5.32\%}$
test_seq_wrap[compile-overhead]	0.4018ms	0.2274ms	4.3966 KOps/s	4.3484 KOps/s	$\color{#35bf28}+1.11\%$
test_func_call_runtime[False-eager]	0.8139ms	0.5693ms	1.7566 KOps/s	1.8101 KOps/s	$\color{#d91a1a}-2.96\%$
test_func_call_runtime[False-compile]	0.8203ms	0.4255ms	2.3504 KOps/s	2.3133 KOps/s	$\color{#35bf28}+1.60\%$
test_func_call_runtime[False-compile-overhead]	0.6118ms	0.4264ms	2.3451 KOps/s	2.3104 KOps/s	$\color{#35bf28}+1.50\%$
test_func_call_runtime[True-eager]	1.1107ms	0.7806ms	1.2811 KOps/s	1.2660 KOps/s	$\color{#35bf28}+1.19\%$
test_func_call_runtime[True-compile]	0.9095ms	0.4702ms	2.1266 KOps/s	2.1134 KOps/s	$\color{#35bf28}+0.62\%$
test_func_call_runtime[True-compile-overhead]	0.6376ms	0.4702ms	2.1266 KOps/s	2.1241 KOps/s	$\color{#35bf28}+0.12\%$
test_func_call_cm_runtime[False-eager]	0.8438ms	0.5581ms	1.7919 KOps/s	1.7912 KOps/s	$\color{#35bf28}+0.04\%$
test_func_call_cm_runtime[False-compile]	0.8682ms	0.4277ms	2.3379 KOps/s	2.3192 KOps/s	$\color{#35bf28}+0.81\%$
test_func_call_cm_runtime[False-compile-overhead]	0.5539ms	0.4233ms	2.3624 KOps/s	2.3227 KOps/s	$\color{#35bf28}+1.71\%$
test_func_call_cm_runtime[True-eager]	1.9394ms	0.9408ms	1.0629 KOps/s	1.0765 KOps/s	$\color{#d91a1a}-1.27\%$
test_func_call_cm_runtime[True-compile]	0.9015ms	0.4903ms	2.0394 KOps/s	2.0200 KOps/s	$\color{#35bf28}+0.96\%$
test_func_call_cm_runtime[True-compile-overhead]	0.6566ms	0.4927ms	2.0297 KOps/s	2.0144 KOps/s	$\color{#35bf28}+0.76\%$
test_vmap_func_call_cm_runtime[eager]	2.8424ms	2.0582ms	485.8600 Ops/s	467.2149 Ops/s	$\color{#35bf28}+3.99\%$
test_vmap_func_call_cm_runtime[compile]	0.7828ms	0.5289ms	1.8907 KOps/s	1.9225 KOps/s	$\color{#d91a1a}-1.65\%$
test_vmap_func_call_cm_runtime[compile-overhead]	1.1823ms	0.5455ms	1.8331 KOps/s	1.9087 KOps/s	$\color{#d91a1a}-3.96\%$
test_distributed	0.4005ms	0.1256ms	7.9644 KOps/s	7.6594 KOps/s	$\color{#35bf28}+3.98\%$
test_tdmodule	0.1260ms	27.2754μs	36.6630 KOps/s	37.6900 KOps/s	$\color{#d91a1a}-2.72\%$
test_tdmodule_dispatch	0.1084ms	50.4772μs	19.8109 KOps/s	20.2123 KOps/s	$\color{#d91a1a}-1.99\%$
test_tdseq	86.1800μs	30.3149μs	32.9871 KOps/s	33.5381 KOps/s	$\color{#d91a1a}-1.64\%$
test_tdseq_dispatch	0.1033ms	55.6276μs	17.9767 KOps/s	17.8085 KOps/s	$\color{#35bf28}+0.94\%$
test_instantiation_functorch	2.2982ms	1.5886ms	629.4734 Ops/s	642.5866 Ops/s	$\color{#d91a1a}-2.04\%$
test_exec_functorch	0.3045ms	0.1816ms	5.5062 KOps/s	5.4410 KOps/s	$\color{#35bf28}+1.20\%$
test_exec_functional_call	0.3313ms	0.1769ms	5.6534 KOps/s	5.6079 KOps/s	$\color{#35bf28}+0.81\%$
test_exec_td_decorator	0.5975ms	0.2361ms	4.2361 KOps/s	4.2452 KOps/s	$\color{#d91a1a}-0.21\%$
test_vmap_mlp_speed_decorator[True-True]	1.1847ms	0.6767ms	1.4778 KOps/s	1.4899 KOps/s	$\color{#d91a1a}-0.81\%$
test_vmap_mlp_speed_decorator[True-False]	1.2866ms	0.6860ms	1.4577 KOps/s	1.4911 KOps/s	$\color{#d91a1a}-2.24\%$
test_vmap_mlp_speed_decorator[False-True]	0.9396ms	0.5402ms	1.8511 KOps/s	1.8656 KOps/s	$\color{#d91a1a}-0.78\%$
test_vmap_mlp_speed_decorator[False-False]	0.8631ms	0.5427ms	1.8426 KOps/s	1.8673 KOps/s	$\color{#d91a1a}-1.32\%$
test_to_module_speed[True]	2.4548ms	1.3548ms	738.1003 Ops/s	729.8391 Ops/s	$\color{#35bf28}+1.13\%$
test_to_module_speed[False]	2.1937ms	1.3543ms	738.3897 Ops/s	762.0032 Ops/s	$\color{#d91a1a}-3.10\%$
test_tc_init	88.9750μs	46.4037μs	21.5500 KOps/s	21.5807 KOps/s	$\color{#d91a1a}-0.14\%$
test_tc_init_nested	0.1699ms	93.9529μs	10.6436 KOps/s	10.7354 KOps/s	$\color{#d91a1a}-0.86\%$
test_tc_first_layer_tensor	31.1380μs	1.6020μs	624.2298 KOps/s	625.6424 KOps/s	$\color{#d91a1a}-0.23\%$
test_tc_first_layer_nontensor	32.9610μs	4.6057μs	217.1206 KOps/s	211.5221 KOps/s	$\color{#35bf28}+2.65\%$
test_tc_second_layer_tensor	23.8240μs	2.8577μs	349.9261 KOps/s	348.5311 KOps/s	$\color{#35bf28}+0.40\%$
test_tc_second_layer_nontensor	55.2530μs	5.9729μs	167.4230 KOps/s	164.2193 KOps/s	$\color{#35bf28}+1.95\%$
test_unbind	0.2749s	16.2281ms	61.6216 Ops/s	50.8657 Ops/s	$\textbf{\color{#35bf28}+21.15\%}$
test_full_like	16.8880ms	11.5881ms	86.2951 Ops/s	71.2935 Ops/s	$\textbf{\color{#35bf28}+21.04\%}$
test_zeros_like	7.8381ms	4.4552ms	224.4584 Ops/s	244.7646 Ops/s	$\textbf{\color{#d91a1a}-8.30\%}$
test_ones_like	7.2662ms	5.0027ms	199.8922 Ops/s	129.4478 Ops/s	$\textbf{\color{#35bf28}+54.42\%}$
test_clone	12.1308ms	8.1177ms	123.1879 Ops/s	93.0662 Ops/s	$\textbf{\color{#35bf28}+32.37\%}$
test_squeeze	76.7230μs	12.2337μs	81.7413 KOps/s	81.3912 KOps/s	$\color{#35bf28}+0.43\%$
test_unsqueeze	0.2018ms	93.6463μs	10.6785 KOps/s	10.7109 KOps/s	$\color{#d91a1a}-0.30\%$
test_split	0.5395ms	0.1980ms	5.0517 KOps/s	5.0024 KOps/s	$\color{#35bf28}+0.99\%$
test_permute	0.4062ms	0.2022ms	4.9460 KOps/s	4.9657 KOps/s	$\color{#d91a1a}-0.40\%$
test_stack	39.6256ms	31.7117ms	31.5341 Ops/s	30.6699 Ops/s	$\color{#35bf28}+2.82\%$
test_cat	45.3636ms	32.3152ms	30.9452 Ops/s	32.2363 Ops/s	$\color{#d91a1a}-4.01\%$

[github-actions]

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

Total Benchmarks: 229. Improved: $\large\color{#35bf28}35$. Worsened: $\large\color{#d91a1a}17$.

Expand to view detailed results

Name	Max	Mean	Ops	Ops on Repo HEAD	Change
test_plain_set_nested	0.3967ms	11.1975μs	89.3057 KOps/s	78.2651 KOps/s	$\textbf{\color{#35bf28}+14.11\%}$
test_plain_set_stack_nested	37.7610μs	11.5410μs	86.6476 KOps/s	77.0228 KOps/s	$\textbf{\color{#35bf28}+12.50\%}$
test_plain_set_nested_inplace	0.3989ms	12.4017μs	80.6340 KOps/s	72.2581 KOps/s	$\textbf{\color{#35bf28}+11.59\%}$
test_plain_set_stack_nested_inplace	55.0220μs	12.4905μs	80.0607 KOps/s	71.9302 KOps/s	$\textbf{\color{#35bf28}+11.30\%}$
test_items	31.5310μs	2.9032μs	344.4514 KOps/s	336.7721 KOps/s	$\color{#35bf28}+2.28\%$
test_items_nested	0.7522ms	0.3612ms	2.7687 KOps/s	2.7802 KOps/s	$\color{#d91a1a}-0.41\%$
test_items_nested_locked	0.7613ms	0.3644ms	2.7443 KOps/s	2.7529 KOps/s	$\color{#d91a1a}-0.31\%$
test_items_nested_leaf	88.6340μs	57.8702μs	17.2800 KOps/s	17.1509 KOps/s	$\color{#35bf28}+0.75\%$
test_items_stack_nested	0.7623ms	0.3643ms	2.7448 KOps/s	2.8023 KOps/s	$\color{#d91a1a}-2.05\%$
test_items_stack_nested_leaf	0.4678ms	58.9768μs	16.9558 KOps/s	16.5219 KOps/s	$\color{#35bf28}+2.63\%$
test_items_stack_nested_locked	0.4991ms	0.3625ms	2.7589 KOps/s	2.7769 KOps/s	$\color{#d91a1a}-0.65\%$
test_keys	35.4410μs	3.4949μs	286.1332 KOps/s	291.8173 KOps/s	$\color{#d91a1a}-1.95\%$
test_keys_nested	0.1333ms	88.2676μs	11.3292 KOps/s	12.1831 KOps/s	$\textbf{\color{#d91a1a}-7.01\%}$
test_keys_nested_locked	0.8146ms	93.6598μs	10.6769 KOps/s	11.3990 KOps/s	$\textbf{\color{#d91a1a}-6.33\%}$
test_keys_nested_leaf	0.1272ms	78.3781μs	12.7587 KOps/s	13.7230 KOps/s	$\textbf{\color{#d91a1a}-7.03\%}$
test_keys_stack_nested	0.1395ms	89.3664μs	11.1899 KOps/s	12.0634 KOps/s	$\textbf{\color{#d91a1a}-7.24\%}$
test_keys_stack_nested_leaf	0.1295ms	79.9151μs	12.5133 KOps/s	13.5795 KOps/s	$\textbf{\color{#d91a1a}-7.85\%}$
test_keys_stack_nested_locked	0.1611ms	96.0333μs	10.4131 KOps/s	11.3289 KOps/s	$\textbf{\color{#d91a1a}-8.08\%}$
test_values	5.3103μs	0.8535μs	1.1716 MOps/s	1.1776 MOps/s	$\color{#d91a1a}-0.51\%$
test_values_nested	67.6220μs	37.2980μs	26.8111 KOps/s	29.0247 KOps/s	$\textbf{\color{#d91a1a}-7.63\%}$
test_values_nested_locked	86.6930μs	38.6941μs	25.8438 KOps/s	27.6203 KOps/s	$\textbf{\color{#d91a1a}-6.43\%}$
test_values_nested_leaf	69.8530μs	41.4700μs	24.1138 KOps/s	25.3669 KOps/s	$\color{#d91a1a}-4.94\%$
test_values_stack_nested	79.5630μs	38.0046μs	26.3126 KOps/s	28.5240 KOps/s	$\textbf{\color{#d91a1a}-7.75\%}$
test_values_stack_nested_leaf	0.1143ms	41.8545μs	23.8923 KOps/s	25.4482 KOps/s	$\textbf{\color{#d91a1a}-6.11\%}$
test_values_stack_nested_locked	74.3130μs	39.7158μs	25.1789 KOps/s	27.4248 KOps/s	$\textbf{\color{#d91a1a}-8.19\%}$
test_membership	1.9226μs	0.5082μs	1.9677 MOps/s	1.9418 MOps/s	$\color{#35bf28}+1.34\%$
test_membership_nested	18.2560μs	1.9526μs	512.1270 KOps/s	512.6203 KOps/s	$\color{#d91a1a}-0.10\%$
test_membership_nested_leaf	18.8210μs	1.9836μs	504.1352 KOps/s	505.3926 KOps/s	$\color{#d91a1a}-0.25\%$
test_membership_stacked_nested	32.2420μs	2.0388μs	490.4956 KOps/s	484.6061 KOps/s	$\color{#35bf28}+1.22\%$
test_membership_stacked_nested_leaf	23.5200μs	2.0179μs	495.5566 KOps/s	483.9393 KOps/s	$\color{#35bf28}+2.40\%$
test_membership_nested_last	39.5510μs	3.0685μs	325.8872 KOps/s	327.2809 KOps/s	$\color{#d91a1a}-0.43\%$
test_membership_nested_leaf_last	37.5320μs	3.0846μs	324.1913 KOps/s	321.2412 KOps/s	$\color{#35bf28}+0.92\%$
test_membership_stacked_nested_last	35.7920μs	3.0639μs	326.3836 KOps/s	120.9242 KOps/s	$\textbf{\color{#35bf28}+169.91\%}$
test_membership_stacked_nested_leaf_last	42.5220μs	3.0847μs	324.1833 KOps/s	121.9812 KOps/s	$\textbf{\color{#35bf28}+165.76\%}$
test_nested_getleaf	36.7910μs	6.0891μs	164.2267 KOps/s	162.3881 KOps/s	$\color{#35bf28}+1.13\%$
test_nested_get	30.7620μs	5.7869μs	172.8046 KOps/s	171.9933 KOps/s	$\color{#35bf28}+0.47\%$
test_stacked_getleaf	42.2510μs	6.1172μs	163.4733 KOps/s	163.7006 KOps/s	$\color{#d91a1a}-0.14\%$
test_stacked_get	27.3010μs	5.7772μs	173.0931 KOps/s	172.8298 KOps/s	$\color{#35bf28}+0.15\%$
test_nested_getitemleaf	32.2810μs	6.3617μs	157.1896 KOps/s	161.7074 KOps/s	$\color{#d91a1a}-2.79\%$
test_nested_getitem	31.7010μs	6.0338μs	165.7322 KOps/s	166.7548 KOps/s	$\color{#d91a1a}-0.61\%$
test_stacked_getitemleaf	40.6910μs	6.3833μs	156.6583 KOps/s	160.0527 KOps/s	$\color{#d91a1a}-2.12\%$
test_stacked_getitem	31.2310μs	6.0743μs	164.6270 KOps/s	168.2540 KOps/s	$\color{#d91a1a}-2.16\%$
test_lock_nested	0.6993ms	0.3733ms	2.6791 KOps/s	2.6434 KOps/s	$\color{#35bf28}+1.35\%$
test_lock_stack_nested	0.3967ms	0.3454ms	2.8951 KOps/s	2.9426 KOps/s	$\color{#d91a1a}-1.61\%$
test_unlock_nested	0.6251ms	0.3154ms	3.1709 KOps/s	3.1577 KOps/s	$\color{#35bf28}+0.42\%$
test_unlock_stack_nested	0.3375ms	0.2862ms	3.4939 KOps/s	3.5999 KOps/s	$\color{#d91a1a}-2.94\%$
test_flatten_speed	0.1423ms	73.0494μs	13.6894 KOps/s	13.4292 KOps/s	$\color{#35bf28}+1.94\%$
test_unflatten_speed	0.4388ms	0.3158ms	3.1663 KOps/s	3.0967 KOps/s	$\color{#35bf28}+2.25\%$
test_common_ops	1.5616ms	0.5806ms	1.7223 KOps/s	1.6022 KOps/s	$\textbf{\color{#35bf28}+7.50\%}$
test_creation	0.1929ms	1.7389μs	575.0670 KOps/s	569.5233 KOps/s	$\color{#35bf28}+0.97\%$
test_creation_empty	36.8420μs	6.9528μs	143.8261 KOps/s	108.5713 KOps/s	$\textbf{\color{#35bf28}+32.47\%}$
test_creation_nested_1	33.1910μs	8.5789μs	116.5650 KOps/s	90.8440 KOps/s	$\textbf{\color{#35bf28}+28.31\%}$
test_creation_nested_2	44.0520μs	11.3194μs	88.3438 KOps/s	72.5512 KOps/s	$\textbf{\color{#35bf28}+21.77\%}$
test_clone	45.8620μs	10.9446μs	91.3691 KOps/s	97.7012 KOps/s	$\textbf{\color{#d91a1a}-6.48\%}$
test_getitem[int]	1.8032ms	11.3495μs	88.1096 KOps/s	91.1449 KOps/s	$\color{#d91a1a}-3.33\%$
test_getitem[slice_int]	0.1132ms	21.3129μs	46.9200 KOps/s	47.7795 KOps/s	$\color{#d91a1a}-1.80\%$
test_getitem[range]	0.1368ms	37.2299μs	26.8601 KOps/s	25.8509 KOps/s	$\color{#35bf28}+3.90\%$
test_getitem[tuple]	0.1078ms	18.7957μs	53.2035 KOps/s	54.0529 KOps/s	$\color{#d91a1a}-1.57\%$
test_getitem[list]	0.1597ms	32.4661μs	30.8014 KOps/s	31.1136 KOps/s	$\color{#d91a1a}-1.00\%$
test_setitem_dim[int]	37.2310μs	19.5029μs	51.2743 KOps/s	53.9591 KOps/s	$\color{#d91a1a}-4.98\%$
test_setitem_dim[slice_int]	66.1730μs	38.4305μs	26.0210 KOps/s	25.9404 KOps/s	$\color{#35bf28}+0.31\%$
test_setitem_dim[range]	81.8830μs	52.0890μs	19.1979 KOps/s	18.8058 KOps/s	$\color{#35bf28}+2.09\%$
test_setitem_dim[tuple]	58.9920μs	32.1073μs	31.1456 KOps/s	32.5284 KOps/s	$\color{#d91a1a}-4.25\%$
test_setitem	49.9520μs	14.3795μs	69.5432 KOps/s	66.7330 KOps/s	$\color{#35bf28}+4.21\%$
test_set	49.4020μs	13.7543μs	72.7044 KOps/s	67.9367 KOps/s	$\textbf{\color{#35bf28}+7.02\%}$
test_set_shared	1.6608ms	0.1515ms	6.5992 KOps/s	6.6792 KOps/s	$\color{#d91a1a}-1.20\%$
test_update	0.4668ms	15.9222μs	62.8056 KOps/s	55.3758 KOps/s	$\textbf{\color{#35bf28}+13.42\%}$
test_update_nested	0.1184ms	21.3477μs	46.8435 KOps/s	42.4369 KOps/s	$\textbf{\color{#35bf28}+10.38\%}$
test_update__nested	1.2862ms	25.4112μs	39.3527 KOps/s	40.5874 KOps/s	$\color{#d91a1a}-3.04\%$
test_set_nested	0.1247ms	15.1475μs	66.0176 KOps/s	63.3865 KOps/s	$\color{#35bf28}+4.15\%$
test_set_nested_new	0.1203ms	17.7960μs	56.1923 KOps/s	54.5334 KOps/s	$\color{#35bf28}+3.04\%$
test_select	83.2130μs	29.3188μs	34.1078 KOps/s	33.1654 KOps/s	$\color{#35bf28}+2.84\%$
test_select_nested	77.2130μs	44.2448μs	22.6015 KOps/s	22.5653 KOps/s	$\color{#35bf28}+0.16\%$
test_exclude_nested	0.1141ms	62.9298μs	15.8907 KOps/s	15.9458 KOps/s	$\color{#d91a1a}-0.35\%$
test_empty[True]	0.6891ms	0.2951ms	3.3887 KOps/s	3.4427 KOps/s	$\color{#d91a1a}-1.57\%$
test_empty[False]	41.2455μs	0.8272μs	1.2089 MOps/s	1.2223 MOps/s	$\color{#d91a1a}-1.10\%$
test_to	88.1340μs	57.0337μs	17.5335 KOps/s	17.6578 KOps/s	$\color{#d91a1a}-0.70\%$
test_to_nonblocking	0.9618ms	48.9686μs	20.4213 KOps/s	21.1963 KOps/s	$\color{#d91a1a}-3.66\%$
test_unbind_speed	0.6614ms	0.2410ms	4.1492 KOps/s	4.2059 KOps/s	$\color{#d91a1a}-1.35\%$
test_unbind_speed_stack0	0.6464ms	0.2431ms	4.1134 KOps/s	4.2600 KOps/s	$\color{#d91a1a}-3.44\%$
test_unbind_speed_stack1	94.2779ms	0.6710ms	1.4903 KOps/s	1.5041 KOps/s	$\color{#d91a1a}-0.92\%$
test_split	95.5320ms	1.7071ms	585.7878 Ops/s	617.3459 Ops/s	$\textbf{\color{#d91a1a}-5.11\%}$
test_chunk	95.9011ms	1.7130ms	583.7676 Ops/s	613.9080 Ops/s	$\color{#d91a1a}-4.91\%$
test_consolidate[False-None]	98.0917ms	3.0344ms	329.5538 Ops/s	332.6700 Ops/s	$\color{#d91a1a}-0.94\%$
test_consolidate[default-None]	1.9227ms	1.7687ms	565.3947 Ops/s	586.5856 Ops/s	$\color{#d91a1a}-3.61\%$
test_consolidate[reduce-overhead-None]	1.8735ms	1.7783ms	562.3458 Ops/s	567.9303 Ops/s	$\color{#d91a1a}-0.98\%$
test_consolidate_njt[False-None]	0.3028s	8.2237ms	121.5991 Ops/s	113.0538 Ops/s	$\textbf{\color{#35bf28}+7.56\%}$
test_to[False-False-None]	1.8406ms	1.7265ms	579.2150 Ops/s	573.7178 Ops/s	$\color{#35bf28}+0.96\%$
test_to[True-False-None]	1.4263ms	1.3247ms	754.9050 Ops/s	768.2462 Ops/s	$\color{#d91a1a}-1.74\%$
test_to[within-False-None]	4.3801ms	4.1481ms	241.0715 Ops/s	241.4042 Ops/s	$\color{#d91a1a}-0.14\%$
test_to[True-default-None]	5.2621ms	5.1317ms	194.8685 Ops/s	189.9998 Ops/s	$\color{#35bf28}+2.56\%$
test_to_njt[False-False-None]	6.8793ms	6.7480ms	148.1914 Ops/s	144.2342 Ops/s	$\color{#35bf28}+2.74\%$
test_to_njt[True-False-None]	5.7425ms	5.3315ms	187.5659 Ops/s	179.7203 Ops/s	$\color{#35bf28}+4.37\%$
test_to_njt[within-False-None]	11.8247ms	11.7224ms	85.3067 Ops/s	81.6011 Ops/s	$\color{#35bf28}+4.54\%$
test_creation[device0]	0.4638ms	79.6123μs	12.5609 KOps/s	12.4708 KOps/s	$\color{#35bf28}+0.72\%$
test_creation_from_tensor	0.4706ms	83.0086μs	12.0469 KOps/s	11.8355 KOps/s	$\color{#35bf28}+1.79\%$
test_add_one[memmap_tensor0]	0.4678ms	6.4886μs	154.1153 KOps/s	159.0490 KOps/s	$\color{#d91a1a}-3.10\%$
test_contiguous[memmap_tensor0]	4.7337μs	0.4015μs	2.4906 MOps/s	2.5390 MOps/s	$\color{#d91a1a}-1.91\%$
test_stack[memmap_tensor0]	42.4510μs	4.8773μs	205.0323 KOps/s	213.3569 KOps/s	$\color{#d91a1a}-3.90\%$
test_memmaptd_index	2.0705ms	0.2613ms	3.8268 KOps/s	3.8734 KOps/s	$\color{#d91a1a}-1.20\%$
test_memmaptd_index_astensor	0.8222ms	0.3213ms	3.1122 KOps/s	3.1130 KOps/s	$\color{#d91a1a}-0.03\%$
test_memmaptd_index_op	1.4494ms	0.5584ms	1.7907 KOps/s	1.6831 KOps/s	$\textbf{\color{#35bf28}+6.39\%}$
test_serialize_model	0.1322s	0.1311s	7.6292 Ops/s	7.6213 Ops/s	$\color{#35bf28}+0.10\%$
test_serialize_model_pickle	1.3494s	1.2163s	0.8222 Ops/s	0.8236 Ops/s	$\color{#d91a1a}-0.17\%$
test_serialize_weights	0.1340s	0.1308s	7.6472 Ops/s	7.6515 Ops/s	$\color{#d91a1a}-0.06\%$
test_serialize_weights_returnearly	0.3443s	60.8175ms	16.4426 Ops/s	12.6522 Ops/s	$\textbf{\color{#35bf28}+29.96\%}$
test_serialize_weights_pickle	1.3564s	1.2127s	0.8246 Ops/s	0.8211 Ops/s	$\color{#35bf28}+0.42\%$
test_reshape_pytree	69.6830μs	22.1808μs	45.0841 KOps/s	45.7273 KOps/s	$\color{#d91a1a}-1.41\%$
test_reshape_td	57.8620μs	26.6657μs	37.5014 KOps/s	36.6081 KOps/s	$\color{#35bf28}+2.44\%$
test_view_pytree	53.4720μs	22.0353μs	45.3816 KOps/s	45.8446 KOps/s	$\color{#d91a1a}-1.01\%$
test_view_td	60.2030μs	29.3418μs	34.0810 KOps/s	33.2689 KOps/s	$\color{#35bf28}+2.44\%$
test_unbind_pytree	68.9120μs	28.4337μs	35.1695 KOps/s	36.0727 KOps/s	$\color{#d91a1a}-2.50\%$
test_unbind_td	0.6059ms	36.5805μs	27.3370 KOps/s	27.5051 KOps/s	$\color{#d91a1a}-0.61\%$
test_split_pytree	61.2520μs	30.5973μs	32.6826 KOps/s	32.6778 KOps/s	$\color{#35bf28}+0.01\%$
test_split_td	0.7775ms	40.7569μs	24.5357 KOps/s	25.9045 KOps/s	$\textbf{\color{#d91a1a}-5.28\%}$
test_add_pytree	63.4120μs	33.2455μs	30.0793 KOps/s	30.2455 KOps/s	$\color{#d91a1a}-0.55\%$
test_add_td	0.1925ms	47.6821μs	20.9722 KOps/s	19.7487 KOps/s	$\textbf{\color{#35bf28}+6.20\%}$
test_compile_add_one_nested[tensordict-compile]	0.1735ms	0.1179ms	8.4782 KOps/s	8.1308 KOps/s	$\color{#35bf28}+4.27\%$
test_compile_add_one_nested[tensordict-eager]	0.2301ms	0.1305ms	7.6651 KOps/s	7.5263 KOps/s	$\color{#35bf28}+1.84\%$
test_compile_add_one_nested[pytree-compile]	0.2043ms	94.3155μs	10.6027 KOps/s	10.5679 KOps/s	$\color{#35bf28}+0.33\%$
test_compile_add_one_nested[pytree-eager]	1.0774ms	0.1509ms	6.6281 KOps/s	6.6773 KOps/s	$\color{#d91a1a}-0.74\%$
test_compile_copy_nested[tensordict-compile]	63.7320μs	22.9187μs	43.6325 KOps/s	30.1951 KOps/s	$\textbf{\color{#35bf28}+44.50\%}$
test_compile_copy_nested[tensordict-eager]	62.0520μs	29.6432μs	33.7346 KOps/s	33.6078 KOps/s	$\color{#35bf28}+0.38\%$
test_compile_copy_nested[pytree-compile]	0.4431ms	64.0379μs	15.6158 KOps/s	15.2413 KOps/s	$\color{#35bf28}+2.46\%$
test_compile_copy_nested[pytree-eager]	83.4730μs	48.7642μs	20.5069 KOps/s	20.2471 KOps/s	$\color{#35bf28}+1.28\%$
test_compile_add_one_flat[tensordict-compile]	0.1827ms	0.1414ms	7.0723 KOps/s	7.1110 KOps/s	$\color{#d91a1a}-0.54\%$
test_compile_add_one_flat[tensordict-eager]	0.3189ms	0.2159ms	4.6319 KOps/s	4.6275 KOps/s	$\color{#35bf28}+0.09\%$
test_compile_add_one_flat[tensorclass-compile]	0.1431ms	96.8478μs	10.3255 KOps/s	10.3239 KOps/s	$\color{#35bf28}+0.02\%$
test_compile_add_one_flat[tensorclass-eager]	0.1131ms	55.3411μs	18.0697 KOps/s	18.1951 KOps/s	$\color{#d91a1a}-0.69\%$
test_compile_add_one_flat[pytree-compile]	0.2427ms	0.1351ms	7.4027 KOps/s	7.4688 KOps/s	$\color{#d91a1a}-0.88\%$
test_compile_add_one_flat[pytree-eager]	0.5303ms	0.4897ms	2.0422 KOps/s	2.0611 KOps/s	$\color{#d91a1a}-0.92\%$
test_compile_add_self_flat[tensordict-eager]	0.4118ms	0.2582ms	3.8729 KOps/s	3.8136 KOps/s	$\color{#35bf28}+1.56\%$
test_compile_add_self_flat[tensordict-compile]	0.2006ms	0.1426ms	7.0104 KOps/s	7.1636 KOps/s	$\color{#d91a1a}-2.14\%$
test_compile_add_self_flat[tensorclass-eager]	0.1606ms	66.4871μs	15.0405 KOps/s	15.0314 KOps/s	$\color{#35bf28}+0.06\%$
test_compile_add_self_flat[tensorclass-compile]	0.1336ms	96.6467μs	10.3470 KOps/s	10.2373 KOps/s	$\color{#35bf28}+1.07\%$
test_compile_add_self_flat[pytree-eager]	0.4801ms	0.4247ms	2.3546 KOps/s	2.4507 KOps/s	$\color{#d91a1a}-3.92\%$
test_compile_add_self_flat[pytree-compile]	0.2045ms	0.1329ms	7.5217 KOps/s	7.4990 KOps/s	$\color{#35bf28}+0.30\%$
test_compile_copy_flat[tensordict-compile]	79.7730μs	18.5820μs	53.8154 KOps/s	28.7793 KOps/s	$\textbf{\color{#35bf28}+86.99\%}$
test_compile_copy_flat[tensordict-eager]	55.0720μs	30.7652μs	32.5043 KOps/s	31.6139 KOps/s	$\color{#35bf28}+2.82\%$
test_compile_copy_flat[pytree-compile]	0.2140ms	70.5348μs	14.1774 KOps/s	14.3632 KOps/s	$\color{#d91a1a}-1.29\%$
test_compile_copy_flat[pytree-eager]	87.0230μs	52.0480μs	19.2130 KOps/s	19.6424 KOps/s	$\color{#d91a1a}-2.19\%$
test_compile_assign_and_add[tensordict-compile]	1.6276ms	0.3904ms	2.5617 KOps/s	2.2024 KOps/s	$\textbf{\color{#35bf28}+16.31\%}$
test_compile_assign_and_add[tensordict-eager]	2.8218ms	2.6039ms	384.0373 Ops/s	389.7622 Ops/s	$\color{#d91a1a}-1.47\%$
test_compile_assign_and_add[pytree-compile]	1.6003ms	0.4358ms	2.2947 KOps/s	2.2689 KOps/s	$\color{#35bf28}+1.14\%$
test_compile_assign_and_add[pytree-eager]	2.7618ms	2.6584ms	376.1629 Ops/s	384.9034 Ops/s	$\color{#d91a1a}-2.27\%$
test_compile_indexing[tensor-tensordict-compile]	0.2187ms	0.1161ms	8.6140 KOps/s	8.9580 KOps/s	$\color{#d91a1a}-3.84\%$
test_compile_indexing[tensor-tensordict-eager]	0.6093ms	80.4755μs	12.4261 KOps/s	12.9202 KOps/s	$\color{#d91a1a}-3.82\%$
test_compile_indexing[tensor-tensorclass-compile]	0.2374ms	0.1036ms	9.6555 KOps/s	9.7005 KOps/s	$\color{#d91a1a}-0.46\%$
test_compile_indexing[tensor-tensorclass-eager]	0.1183ms	67.8409μs	14.7404 KOps/s	14.4759 KOps/s	$\color{#35bf28}+1.83\%$
test_compile_indexing[tensor-pytree-compile]	0.1989ms	0.1102ms	9.0706 KOps/s	9.4812 KOps/s	$\color{#d91a1a}-4.33\%$
test_compile_indexing[tensor-pytree-eager]	0.1602ms	69.3091μs	14.4281 KOps/s	14.3931 KOps/s	$\color{#35bf28}+0.24\%$
test_compile_indexing[slice-tensordict-compile]	0.2506ms	0.1050ms	9.5254 KOps/s	9.7458 KOps/s	$\color{#d91a1a}-2.26\%$
test_compile_indexing[slice-tensordict-eager]	0.1435ms	17.1472μs	58.3187 KOps/s	58.1011 KOps/s	$\color{#35bf28}+0.37\%$
test_compile_indexing[slice-tensorclass-compile]	0.1546ms	97.6665μs	10.2389 KOps/s	10.4085 KOps/s	$\color{#d91a1a}-1.63\%$
test_compile_indexing[slice-tensorclass-eager]	0.1043ms	15.8510μs	63.0874 KOps/s	63.3007 KOps/s	$\color{#d91a1a}-0.34\%$
test_compile_indexing[slice-pytree-compile]	0.2015ms	98.6285μs	10.1391 KOps/s	10.2586 KOps/s	$\color{#d91a1a}-1.16\%$
test_compile_indexing[slice-pytree-eager]	0.1075ms	15.9271μs	62.7862 KOps/s	64.3393 KOps/s	$\color{#d91a1a}-2.41\%$
test_compile_indexing[int-tensordict-compile]	0.1725ms	0.1031ms	9.7039 KOps/s	9.8823 KOps/s	$\color{#d91a1a}-1.81\%$
test_compile_indexing[int-tensordict-eager]	0.6068ms	17.1647μs	58.2592 KOps/s	58.6559 KOps/s	$\color{#d91a1a}-0.68\%$
test_compile_indexing[int-tensorclass-compile]	0.2012ms	98.8809μs	10.1132 KOps/s	10.3095 KOps/s	$\color{#d91a1a}-1.90\%$
test_compile_indexing[int-tensorclass-eager]	0.1049ms	15.9131μs	62.8413 KOps/s	63.8137 KOps/s	$\color{#d91a1a}-1.52\%$
test_compile_indexing[int-pytree-compile]	0.2155ms	99.9613μs	10.0039 KOps/s	10.3156 KOps/s	$\color{#d91a1a}-3.02\%$
test_compile_indexing[int-pytree-eager]	0.1031ms	15.8108μs	63.2481 KOps/s	64.3232 KOps/s	$\color{#d91a1a}-1.67\%$
test_mod_add[eager]	0.2080ms	35.4478μs	28.2105 KOps/s	24.4867 KOps/s	$\textbf{\color{#35bf28}+15.21\%}$
test_mod_add[compile]	0.1849ms	77.8906μs	12.8385 KOps/s	12.2481 KOps/s	$\color{#35bf28}+4.82\%$
test_mod_add[compile-overhead]	0.3253ms	0.1658ms	6.0324 KOps/s	5.6966 KOps/s	$\textbf{\color{#35bf28}+5.89\%}$
test_mod_wrap[eager]	0.3445ms	0.2382ms	4.1983 KOps/s	3.9099 KOps/s	$\textbf{\color{#35bf28}+7.38\%}$
test_mod_wrap[compile]	0.3810ms	0.2825ms	3.5397 KOps/s	3.5261 KOps/s	$\color{#35bf28}+0.39\%$
test_mod_wrap[compile-overhead]	6.9529ms	3.7510ms	266.5935 Ops/s	267.5136 Ops/s	$\color{#d91a1a}-0.34\%$
test_mod_wrap_and_backward[eager]	1.6992ms	1.3422ms	745.0369 Ops/s	697.5581 Ops/s	$\textbf{\color{#35bf28}+6.81\%}$
test_mod_wrap_and_backward[compile]	1.3851ms	1.2657ms	790.0522 Ops/s	728.7287 Ops/s	$\textbf{\color{#35bf28}+8.42\%}$
test_mod_wrap_and_backward[compile-overhead]	1.3551ms	0.9158ms	1.0919 KOps/s	953.7447 Ops/s	$\textbf{\color{#35bf28}+14.49\%}$
test_seq_add[eager]	0.1756ms	0.1112ms	8.9920 KOps/s	8.5530 KOps/s	$\textbf{\color{#35bf28}+5.13\%}$
test_seq_add[compile]	0.1581ms	92.0681μs	10.8615 KOps/s	11.3565 KOps/s	$\color{#d91a1a}-4.36\%$
test_seq_add[compile-overhead]	0.1773ms	0.1316ms	7.5962 KOps/s	7.8283 KOps/s	$\color{#d91a1a}-2.96\%$
test_seq_wrap[eager]	0.8242ms	0.4206ms	2.3777 KOps/s	2.3767 KOps/s	$\color{#35bf28}+0.04\%$
test_seq_wrap[compile]	0.7200ms	0.3086ms	3.2401 KOps/s	3.3353 KOps/s	$\color{#d91a1a}-2.85\%$
test_seq_wrap[compile-overhead]	0.2697ms	0.2211ms	4.5227 KOps/s	4.4563 KOps/s	$\color{#35bf28}+1.49\%$
test_func_call_runtime[False-eager]	0.8014ms	0.7068ms	1.4148 KOps/s	1.3922 KOps/s	$\color{#35bf28}+1.62\%$
test_func_call_runtime[False-compile]	1.2232ms	0.7664ms	1.3048 KOps/s	1.3492 KOps/s	$\color{#d91a1a}-3.29\%$
test_func_call_runtime[False-compile-overhead]	0.7659ms	0.3724ms	2.6856 KOps/s	2.7726 KOps/s	$\color{#d91a1a}-3.14\%$
test_func_call_runtime[True-eager]	1.3210ms	0.9102ms	1.0987 KOps/s	1.1384 KOps/s	$\color{#d91a1a}-3.49\%$
test_func_call_runtime[True-compile]	1.2279ms	0.7844ms	1.2748 KOps/s	1.3224 KOps/s	$\color{#d91a1a}-3.60\%$
test_func_call_runtime[True-compile-overhead]	0.4424ms	0.3783ms	2.6433 KOps/s	2.6313 KOps/s	$\color{#35bf28}+0.46\%$
test_func_call_cm_runtime[False-eager]	0.8108ms	0.7038ms	1.4208 KOps/s	1.3910 KOps/s	$\color{#35bf28}+2.14\%$
test_func_call_cm_runtime[False-compile]	0.8147ms	0.7363ms	1.3582 KOps/s	1.3309 KOps/s	$\color{#35bf28}+2.05\%$
test_func_call_cm_runtime[False-compile-overhead]	0.4570ms	0.3606ms	2.7730 KOps/s	2.7316 KOps/s	$\color{#35bf28}+1.52\%$
test_func_call_cm_runtime[True-eager]	1.1194ms	0.9728ms	1.0279 KOps/s	1.0098 KOps/s	$\color{#35bf28}+1.80\%$
test_func_call_cm_runtime[True-compile]	0.9412ms	0.7792ms	1.2834 KOps/s	1.2571 KOps/s	$\color{#35bf28}+2.09\%$
test_func_call_cm_runtime[True-compile-overhead]	0.8208ms	0.4059ms	2.4636 KOps/s	2.4558 KOps/s	$\color{#35bf28}+0.32\%$
test_vmap_func_call_cm_runtime[eager]	2.4641ms	2.0177ms	495.6225 Ops/s	489.2664 Ops/s	$\color{#35bf28}+1.30\%$
test_vmap_func_call_cm_runtime[compile]	1.1960ms	0.7907ms	1.2647 KOps/s	1.2364 KOps/s	$\color{#35bf28}+2.29\%$
test_vmap_func_call_cm_runtime[compile-overhead]	0.8485ms	0.4088ms	2.4459 KOps/s	2.4424 KOps/s	$\color{#35bf28}+0.14\%$
test_distributed	4.4408ms	0.1800ms	5.5556 KOps/s	8.3538 KOps/s	$\textbf{\color{#d91a1a}-33.50\%}$
test_tdmodule	0.3045ms	18.9542μs	52.7588 KOps/s	47.5214 KOps/s	$\textbf{\color{#35bf28}+11.02\%}$
test_tdmodule_dispatch	59.0930μs	33.4583μs	29.8880 KOps/s	27.0124 KOps/s	$\textbf{\color{#35bf28}+10.65\%}$
test_tdseq	44.3310μs	19.5473μs	51.1579 KOps/s	47.1525 KOps/s	$\textbf{\color{#35bf28}+8.49\%}$
test_tdseq_dispatch	60.8030μs	36.3552μs	27.5064 KOps/s	25.2633 KOps/s	$\textbf{\color{#35bf28}+8.88\%}$
test_instantiation_functorch	1.6068ms	1.5333ms	652.1730 Ops/s	652.0140 Ops/s	$\color{#35bf28}+0.02\%$
test_exec_functorch	0.1856ms	0.1429ms	6.9958 KOps/s	7.1209 KOps/s	$\color{#d91a1a}-1.76\%$
test_exec_functional_call	0.1756ms	0.1334ms	7.4960 KOps/s	7.6157 KOps/s	$\color{#d91a1a}-1.57\%$
test_exec_td_decorator	0.3836ms	0.1801ms	5.5535 KOps/s	5.6046 KOps/s	$\color{#d91a1a}-0.91\%$
test_vmap_mlp_speed_decorator[True-True]	0.7861ms	0.6671ms	1.4991 KOps/s	1.4860 KOps/s	$\color{#35bf28}+0.88\%$
test_vmap_mlp_speed_decorator[True-False]	0.7539ms	0.6609ms	1.5130 KOps/s	1.4933 KOps/s	$\color{#35bf28}+1.32\%$
test_vmap_mlp_speed_decorator[False-True]	0.7023ms	0.5791ms	1.7268 KOps/s	1.7209 KOps/s	$\color{#35bf28}+0.34\%$
test_vmap_mlp_speed_decorator[False-False]	0.7199ms	0.5848ms	1.7099 KOps/s	1.7162 KOps/s	$\color{#d91a1a}-0.37\%$
test_vmap_transformer_speed_decorator[True-True]	19.0862ms	18.7659ms	53.2880 Ops/s	53.1251 Ops/s	$\color{#35bf28}+0.31\%$
test_vmap_transformer_speed_decorator[True-False]	19.4739ms	18.7997ms	53.1925 Ops/s	53.1897 Ops/s	$+0.01\%$
test_vmap_transformer_speed_decorator[False-True]	19.7307ms	18.8049ms	53.1775 Ops/s	53.6781 Ops/s	$\color{#d91a1a}-0.93\%$
test_vmap_transformer_speed_decorator[False-False]	19.0227ms	18.6289ms	53.6800 Ops/s	53.5237 Ops/s	$\color{#35bf28}+0.29\%$
test_to_module_speed[True]	1.0787ms	0.9736ms	1.0271 KOps/s	1.0308 KOps/s	$\color{#d91a1a}-0.36\%$
test_to_module_speed[False]	1.5131ms	0.9391ms	1.0648 KOps/s	1.0372 KOps/s	$\color{#35bf28}+2.66\%$
test_tc_init	62.9830μs	34.5004μs	28.9851 KOps/s	27.4256 KOps/s	$\textbf{\color{#35bf28}+5.69\%}$
test_tc_init_nested	0.1051ms	71.9114μs	13.9060 KOps/s	13.9148 KOps/s	$\color{#d91a1a}-0.06\%$
test_tc_first_layer_tensor	32.7220μs	0.7981μs	1.2530 MOps/s	1.2304 MOps/s	$\color{#35bf28}+1.84\%$
test_tc_first_layer_nontensor	20.0810μs	2.2615μs	442.1940 KOps/s	451.0266 KOps/s	$\color{#d91a1a}-1.96\%$
test_tc_second_layer_tensor	31.2510μs	1.4944μs	669.1583 KOps/s	705.8785 KOps/s	$\textbf{\color{#d91a1a}-5.20\%}$
test_tc_second_layer_nontensor	33.6310μs	2.9997μs	333.3643 KOps/s	338.8064 KOps/s	$\color{#d91a1a}-1.61\%$
test_unbind	0.2305s	10.0063ms	99.9367 Ops/s	143.7064 Ops/s	$\textbf{\color{#d91a1a}-30.46\%}$
test_full_like	9.7356ms	9.1479ms	109.3142 Ops/s	109.1881 Ops/s	$\color{#35bf28}+0.12\%$
test_zeros_like	6.6437ms	4.3602ms	229.3470 Ops/s	140.8856 Ops/s	$\textbf{\color{#35bf28}+62.79\%}$
test_ones_like	9.2830ms	7.3040ms	136.9109 Ops/s	138.5971 Ops/s	$\color{#d91a1a}-1.22\%$
test_clone	6.6458ms	6.4393ms	155.2975 Ops/s	109.4984 Ops/s	$\textbf{\color{#35bf28}+41.83\%}$
test_squeeze	53.1620μs	9.4885μs	105.3907 KOps/s	104.6228 KOps/s	$\color{#35bf28}+0.73\%$
test_unsqueeze	0.1299ms	72.3765μs	13.8166 KOps/s	13.6171 KOps/s	$\color{#35bf28}+1.47\%$
test_split	0.3893ms	0.1590ms	6.2887 KOps/s	5.9441 KOps/s	$\textbf{\color{#35bf28}+5.80\%}$
test_permute	0.2781ms	0.1723ms	5.8039 KOps/s	5.7556 KOps/s	$\color{#35bf28}+0.84\%$
test_stack	50.7087ms	50.3082ms	19.8775 Ops/s	19.6853 Ops/s	$\color{#35bf28}+0.98\%$
test_cat	50.8993ms	50.2622ms	19.8957 Ops/s	19.8337 Ops/s	$\color{#35bf28}+0.31\%$