[Quality] Avoid torch.distributed imports at root

[vmoens]

Stack from ghstack (oldest at bottom):

• [Refactor] Avoid lambda functions in core functionality #1136
• [FBCode] Deactivate vmap monkey-patching in FBCode #1135
• -> [Quality] Avoid torch.distributed imports at root #1134

[github-actions]

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

Total Benchmarks: 217. Improved: $\large\color{#35bf28}26$. Worsened: $\large\color{#d91a1a}9$.

Expand to view detailed results

Name	Max	Mean	Ops	Ops on Repo HEAD	Change
test_plain_set_nested	43.5410μs	18.4301μs	54.2590 KOps/s	56.7546 KOps/s	$\color{#d91a1a}-4.40\%$
test_plain_set_stack_nested	47.3590μs	18.3522μs	54.4893 KOps/s	54.9862 KOps/s	$\color{#d91a1a}-0.90\%$
test_plain_set_nested_inplace	74.1980μs	19.9350μs	50.1630 KOps/s	50.6672 KOps/s	$\color{#d91a1a}-1.00\%$
test_plain_set_stack_nested_inplace	58.9790μs	19.7323μs	50.6782 KOps/s	51.1127 KOps/s	$\color{#d91a1a}-0.85\%$
test_items	28.0520μs	4.1968μs	238.2742 KOps/s	238.8849 KOps/s	$\color{#d91a1a}-0.26\%$
test_items_nested	0.5753ms	0.3994ms	2.5041 KOps/s	2.5149 KOps/s	$\color{#d91a1a}-0.43\%$
test_items_nested_locked	0.7397ms	0.4025ms	2.4844 KOps/s	2.5396 KOps/s	$\color{#d91a1a}-2.17\%$
test_items_nested_leaf	0.1155ms	70.2419μs	14.2365 KOps/s	14.1354 KOps/s	$\color{#35bf28}+0.72\%$
test_items_stack_nested	0.7601ms	0.4051ms	2.4687 KOps/s	2.4859 KOps/s	$\color{#d91a1a}-0.69\%$
test_items_stack_nested_leaf	0.1323ms	72.5296μs	13.7875 KOps/s	13.5671 KOps/s	$\color{#35bf28}+1.62\%$
test_items_stack_nested_locked	0.7702ms	0.4077ms	2.4530 KOps/s	2.4961 KOps/s	$\color{#d91a1a}-1.73\%$
test_keys	52.0470μs	3.5159μs	284.4207 KOps/s	278.7372 KOps/s	$\color{#35bf28}+2.04\%$
test_keys_nested	0.2294ms	0.1365ms	7.3248 KOps/s	7.3771 KOps/s	$\color{#d91a1a}-0.71\%$
test_keys_nested_locked	0.7393ms	0.1424ms	7.0210 KOps/s	6.8795 KOps/s	$\color{#35bf28}+2.06\%$
test_keys_nested_leaf	1.9661ms	0.1165ms	8.5817 KOps/s	8.5114 KOps/s	$\color{#35bf28}+0.83\%$
test_keys_stack_nested	0.2467ms	0.1374ms	7.2795 KOps/s	7.3577 KOps/s	$\color{#d91a1a}-1.06\%$
test_keys_stack_nested_leaf	0.2299ms	0.1170ms	8.5473 KOps/s	8.5687 KOps/s	$\color{#d91a1a}-0.25\%$
test_keys_stack_nested_locked	0.3003ms	0.1403ms	7.1270 KOps/s	7.0549 KOps/s	$\color{#35bf28}+1.02\%$
test_values	9.3234μs	1.0835μs	922.8998 KOps/s	949.6284 KOps/s	$\color{#d91a1a}-2.81\%$
test_values_nested	0.1090ms	54.7677μs	18.2589 KOps/s	17.7571 KOps/s	$\color{#35bf28}+2.83\%$
test_values_nested_locked	0.3581ms	55.7024μs	17.9525 KOps/s	17.8835 KOps/s	$\color{#35bf28}+0.39\%$
test_values_nested_leaf	0.1168ms	59.8711μs	16.7026 KOps/s	16.3318 KOps/s	$\color{#35bf28}+2.27\%$
test_values_stack_nested	0.1003ms	55.3606μs	18.0634 KOps/s	16.6886 KOps/s	$\textbf{\color{#35bf28}+8.24\%}$
test_values_stack_nested_leaf	0.1172ms	59.8440μs	16.7101 KOps/s	15.1867 KOps/s	$\textbf{\color{#35bf28}+10.03\%}$
test_values_stack_nested_locked	0.1398ms	55.5277μs	18.0090 KOps/s	17.3509 KOps/s	$\color{#35bf28}+3.79\%$
test_membership	21.7900μs	0.9029μs	1.1076 MOps/s	1.1592 MOps/s	$\color{#d91a1a}-4.45\%$
test_membership_nested	0.1191ms	2.9584μs	338.0217 KOps/s	342.5148 KOps/s	$\color{#d91a1a}-1.31\%$
test_membership_nested_leaf	44.5930μs	2.9506μs	338.9104 KOps/s	336.3967 KOps/s	$\color{#35bf28}+0.75\%$
test_membership_stacked_nested	29.9760μs	2.9137μs	343.2021 KOps/s	338.2489 KOps/s	$\color{#35bf28}+1.46\%$
test_membership_stacked_nested_leaf	24.5060μs	2.8939μs	345.5552 KOps/s	342.7936 KOps/s	$\color{#35bf28}+0.81\%$
test_membership_nested_last	46.4460μs	4.2244μs	236.7212 KOps/s	238.3727 KOps/s	$\color{#d91a1a}-0.69\%$
test_membership_nested_leaf_last	20.5880μs	4.2703μs	234.1770 KOps/s	234.3640 KOps/s	$\color{#d91a1a}-0.08\%$
test_membership_stacked_nested_last	52.7780μs	4.1826μs	239.0866 KOps/s	203.1504 KOps/s	$\textbf{\color{#35bf28}+17.69\%}$
test_membership_stacked_nested_leaf_last	39.2230μs	4.2463μs	235.4972 KOps/s	194.8626 KOps/s	$\textbf{\color{#35bf28}+20.85\%}$
test_nested_getleaf	60.2320μs	10.9574μs	91.2623 KOps/s	91.7585 KOps/s	$\color{#d91a1a}-0.54\%$
test_nested_get	59.3410μs	10.4423μs	95.7645 KOps/s	91.9844 KOps/s	$\color{#35bf28}+4.11\%$
test_stacked_getleaf	32.4110μs	10.8726μs	91.9745 KOps/s	90.8673 KOps/s	$\color{#35bf28}+1.22\%$
test_stacked_get	61.4650μs	10.3342μs	96.7659 KOps/s	95.9426 KOps/s	$\color{#35bf28}+0.86\%$
test_nested_getitemleaf	65.7830μs	11.4193μs	87.5707 KOps/s	88.2937 KOps/s	$\color{#d91a1a}-0.82\%$
test_nested_getitem	34.4440μs	10.6038μs	94.3062 KOps/s	94.0639 KOps/s	$\color{#35bf28}+0.26\%$
test_stacked_getitemleaf	63.2270μs	11.1014μs	90.0787 KOps/s	87.1568 KOps/s	$\color{#35bf28}+3.35\%$
test_stacked_getitem	33.2930μs	10.5954μs	94.3802 KOps/s	94.1520 KOps/s	$\color{#35bf28}+0.24\%$
test_lock_nested	1.7478ms	0.4442ms	2.2511 KOps/s	2.2554 KOps/s	$\color{#d91a1a}-0.19\%$
test_lock_stack_nested	0.7699ms	0.4155ms	2.4066 KOps/s	2.4173 KOps/s	$\color{#d91a1a}-0.44\%$
test_unlock_nested	0.8586ms	0.3649ms	2.7408 KOps/s	2.7333 KOps/s	$\color{#35bf28}+0.27\%$
test_unlock_stack_nested	0.4930ms	0.3334ms	2.9998 KOps/s	3.0274 KOps/s	$\color{#d91a1a}-0.91\%$
test_flatten_speed	0.1874ms	94.8687μs	10.5409 KOps/s	10.6572 KOps/s	$\color{#d91a1a}-1.09\%$
test_unflatten_speed	1.0057ms	0.5018ms	1.9927 KOps/s	2.0299 KOps/s	$\color{#d91a1a}-1.84\%$
test_common_ops	1.8207ms	0.7941ms	1.2594 KOps/s	1.2770 KOps/s	$\color{#d91a1a}-1.38\%$
test_creation	23.5530μs	2.1708μs	460.6635 KOps/s	483.5192 KOps/s	$\color{#d91a1a}-4.73\%$
test_creation_empty	53.9610μs	11.7334μs	85.2266 KOps/s	92.1510 KOps/s	$\textbf{\color{#d91a1a}-7.51\%}$
test_creation_nested_1	63.7390μs	14.5043μs	68.9449 KOps/s	73.7634 KOps/s	$\textbf{\color{#d91a1a}-6.53\%}$
test_creation_nested_2	54.1110μs	18.8346μs	53.0938 KOps/s	56.0682 KOps/s	$\textbf{\color{#d91a1a}-5.30\%}$
test_clone	1.3663ms	13.1169μs	76.2377 KOps/s	75.0308 KOps/s	$\color{#35bf28}+1.61\%$
test_getitem[int]	0.8502ms	12.6266μs	79.1976 KOps/s	75.5386 KOps/s	$\color{#35bf28}+4.84\%$
test_getitem[slice_int]	0.1403ms	24.5394μs	40.7507 KOps/s	38.1230 KOps/s	$\textbf{\color{#35bf28}+6.89\%}$
test_getitem[range]	0.1739ms	45.9735μs	21.7517 KOps/s	20.5428 KOps/s	$\textbf{\color{#35bf28}+5.88\%}$
test_getitem[tuple]	0.1486ms	20.5018μs	48.7762 KOps/s	47.4913 KOps/s	$\color{#35bf28}+2.71\%$
test_getitem[list]	0.6778ms	42.1813μs	23.7072 KOps/s	22.4317 KOps/s	$\textbf{\color{#35bf28}+5.69\%}$
test_setitem_dim[int]	65.5220μs	25.0777μs	39.8760 KOps/s	37.7132 KOps/s	$\textbf{\color{#35bf28}+5.73\%}$
test_setitem_dim[slice_int]	0.1005ms	52.6702μs	18.9861 KOps/s	18.5506 KOps/s	$\color{#35bf28}+2.35\%$
test_setitem_dim[range]	0.1139ms	73.7888μs	13.5522 KOps/s	13.6839 KOps/s	$\color{#d91a1a}-0.96\%$
test_setitem_dim[tuple]	79.7390μs	40.9690μs	24.4087 KOps/s	24.2754 KOps/s	$\color{#35bf28}+0.55\%$
test_setitem	0.2992ms	20.4102μs	48.9952 KOps/s	47.7192 KOps/s	$\color{#35bf28}+2.67\%$
test_set	61.4450μs	20.1889μs	49.5321 KOps/s	49.8311 KOps/s	$\color{#d91a1a}-0.60\%$
test_set_shared	4.9586ms	0.1661ms	6.0219 KOps/s	5.8375 KOps/s	$\color{#35bf28}+3.16\%$
test_update	0.2910ms	23.2034μs	43.0971 KOps/s	44.7617 KOps/s	$\color{#d91a1a}-3.72\%$
test_update_nested	0.3412ms	33.5034μs	29.8478 KOps/s	30.1476 KOps/s	$\color{#d91a1a}-0.99\%$
test_update__nested	0.9509ms	32.3877μs	30.8760 KOps/s	29.9881 KOps/s	$\color{#35bf28}+2.96\%$
test_set_nested	0.3029ms	22.5650μs	44.3164 KOps/s	44.3194 KOps/s	$-0.01\%$
test_set_nested_new	0.3371ms	26.9096μs	37.1615 KOps/s	36.8485 KOps/s	$\color{#35bf28}+0.85\%$
test_select	0.3385ms	43.3327μs	23.0772 KOps/s	23.1463 KOps/s	$\color{#d91a1a}-0.30\%$
test_select_nested	0.1184ms	59.9520μs	16.6800 KOps/s	16.5281 KOps/s	$\color{#35bf28}+0.92\%$
test_exclude_nested	0.1432ms	78.1573μs	12.7947 KOps/s	12.7634 KOps/s	$\color{#35bf28}+0.25\%$
test_empty[True]	0.5205ms	0.3794ms	2.6356 KOps/s	2.5583 KOps/s	$\color{#35bf28}+3.02\%$
test_empty[False]	7.7093μs	1.2675μs	788.9724 KOps/s	825.2033 KOps/s	$\color{#d91a1a}-4.39\%$
test_unbind_speed	0.3452ms	0.2582ms	3.8724 KOps/s	3.7684 KOps/s	$\color{#35bf28}+2.76\%$
test_unbind_speed_stack0	0.4563ms	0.2565ms	3.8991 KOps/s	3.9185 KOps/s	$\color{#d91a1a}-0.49\%$
test_unbind_speed_stack1	0.1041s	0.7663ms	1.3049 KOps/s	1.4626 KOps/s	$\textbf{\color{#d91a1a}-10.78\%}$
test_split	0.1061s	1.7285ms	578.5347 Ops/s	562.7803 Ops/s	$\color{#35bf28}+2.80\%$
test_chunk	0.1141s	1.7505ms	571.2594 Ops/s	560.8005 Ops/s	$\color{#35bf28}+1.87\%$
test_consolidate_njt[False-None]	9.2629ms	8.0162ms	124.7475 Ops/s	123.8444 Ops/s	$\color{#35bf28}+0.73\%$
test_creation[device0]	0.2330ms	91.1631μs	10.9694 KOps/s	10.9759 KOps/s	$\color{#d91a1a}-0.06\%$
test_creation_from_tensor	3.6791ms	94.8671μs	10.5411 KOps/s	10.5344 KOps/s	$\color{#35bf28}+0.06\%$
test_add_one[memmap_tensor0]	0.3205ms	4.7236μs	211.7036 KOps/s	194.2059 KOps/s	$\textbf{\color{#35bf28}+9.01\%}$
test_contiguous[memmap_tensor0]	29.4750μs	0.4954μs	2.0184 MOps/s	1.9379 MOps/s	$\color{#35bf28}+4.15\%$
test_stack[memmap_tensor0]	79.0370μs	3.2429μs	308.3706 KOps/s	270.8130 KOps/s	$\textbf{\color{#35bf28}+13.87\%}$
test_memmaptd_index	0.9531ms	0.2378ms	4.2045 KOps/s	4.1402 KOps/s	$\color{#35bf28}+1.55\%$
test_memmaptd_index_astensor	0.5961ms	0.3157ms	3.1678 KOps/s	3.1195 KOps/s	$\color{#35bf28}+1.55\%$
test_memmaptd_index_op	0.9671ms	0.5912ms	1.6915 KOps/s	1.7218 KOps/s	$\color{#d91a1a}-1.76\%$
test_serialize_model	0.1234s	0.1162s	8.6029 Ops/s	7.5223 Ops/s	$\textbf{\color{#35bf28}+14.37\%}$
test_serialize_model_pickle	0.4909s	0.4009s	2.4941 Ops/s	2.5490 Ops/s	$\color{#d91a1a}-2.15\%$
test_serialize_weights	0.1221s	0.1139s	8.7791 Ops/s	8.6279 Ops/s	$\color{#35bf28}+1.75\%$
test_serialize_weights_returnearly	0.1770s	0.1613s	6.2010 Ops/s	6.2926 Ops/s	$\color{#d91a1a}-1.46\%$
test_serialize_weights_pickle	1.1089s	0.7066s	1.4153 Ops/s	2.4664 Ops/s	$\textbf{\color{#d91a1a}-42.62\%}$
test_serialize_weights_filesystem	0.1461s	0.1411s	7.0861 Ops/s	7.0263 Ops/s	$\color{#35bf28}+0.85\%$
test_serialize_model_filesystem	0.1523s	0.1424s	7.0219 Ops/s	6.1495 Ops/s	$\textbf{\color{#35bf28}+14.19\%}$
test_reshape_pytree	76.2020μs	27.1957μs	36.7706 KOps/s	35.6946 KOps/s	$\color{#35bf28}+3.01\%$
test_reshape_td	82.5340μs	32.7572μs	30.5276 KOps/s	30.1733 KOps/s	$\color{#35bf28}+1.17\%$
test_view_pytree	62.0760μs	27.0348μs	36.9894 KOps/s	36.1786 KOps/s	$\color{#35bf28}+2.24\%$
test_view_td	0.2194ms	37.9093μs	26.3788 KOps/s	25.8734 KOps/s	$\color{#35bf28}+1.95\%$
test_unbind_pytree	72.1740μs	29.9216μs	33.4207 KOps/s	32.7467 KOps/s	$\color{#35bf28}+2.06\%$
test_unbind_td	0.3744ms	38.8667μs	25.7290 KOps/s	25.3864 KOps/s	$\color{#35bf28}+1.35\%$
test_split_pytree	72.4440μs	29.4333μs	33.9751 KOps/s	32.6842 KOps/s	$\color{#35bf28}+3.95\%$
test_split_td	0.1081s	54.1327μs	18.4731 KOps/s	21.3996 KOps/s	$\textbf{\color{#d91a1a}-13.68\%}$
test_add_pytree	0.1042ms	35.2590μs	28.3615 KOps/s	27.6137 KOps/s	$\color{#35bf28}+2.71\%$
test_add_td	0.1213ms	57.9860μs	17.2455 KOps/s	17.8674 KOps/s	$\color{#d91a1a}-3.48\%$
test_compile_add_one_nested[tensordict-compile]	0.1561ms	61.2064μs	16.3382 KOps/s	15.7102 KOps/s	$\color{#35bf28}+4.00\%$
test_compile_add_one_nested[tensordict-eager]	0.4035ms	0.1598ms	6.2587 KOps/s	6.2132 KOps/s	$\color{#35bf28}+0.73\%$
test_compile_add_one_nested[pytree-compile]	97.5720μs	44.9326μs	22.2555 KOps/s	21.9296 KOps/s	$\color{#35bf28}+1.49\%$
test_compile_add_one_nested[pytree-eager]	0.2882ms	0.1200ms	8.3302 KOps/s	8.3683 KOps/s	$\color{#d91a1a}-0.46\%$
test_compile_copy_nested[tensordict-compile]	84.5680μs	26.0079μs	38.4499 KOps/s	35.0506 KOps/s	$\textbf{\color{#35bf28}+9.70\%}$
test_compile_copy_nested[tensordict-eager]	0.1224ms	53.5662μs	18.6685 KOps/s	18.6058 KOps/s	$\color{#35bf28}+0.34\%$
test_compile_copy_nested[pytree-compile]	0.1737ms	79.3515μs	12.6022 KOps/s	12.5224 KOps/s	$\color{#35bf28}+0.64\%$
test_compile_copy_nested[pytree-eager]	2.7558ms	68.0082μs	14.7041 KOps/s	14.6993 KOps/s	$\color{#35bf28}+0.03\%$
test_compile_add_one_flat[tensordict-compile]	0.2273ms	0.1046ms	9.5647 KOps/s	9.2300 KOps/s	$\color{#35bf28}+3.63\%$
test_compile_add_one_flat[tensordict-eager]	0.3131ms	0.1972ms	5.0703 KOps/s	4.9115 KOps/s	$\color{#35bf28}+3.23\%$
test_compile_add_one_flat[tensorclass-compile]	0.1064ms	44.1751μs	22.6372 KOps/s	21.2262 KOps/s	$\textbf{\color{#35bf28}+6.65\%}$
test_compile_add_one_flat[tensorclass-eager]	0.5513ms	60.9492μs	16.4071 KOps/s	16.2142 KOps/s	$\color{#35bf28}+1.19\%$
test_compile_add_one_flat[pytree-compile]	0.2164ms	0.1027ms	9.7379 KOps/s	9.5369 KOps/s	$\color{#35bf28}+2.11\%$
test_compile_add_one_flat[pytree-eager]	0.2883ms	0.2002ms	4.9938 KOps/s	4.9633 KOps/s	$\color{#35bf28}+0.62\%$
test_compile_add_self_flat[tensordict-eager]	0.3792ms	0.2069ms	4.8338 KOps/s	4.6942 KOps/s	$\color{#35bf28}+2.97\%$
test_compile_add_self_flat[tensordict-compile]	0.2214ms	0.1039ms	9.6223 KOps/s	9.5364 KOps/s	$\color{#35bf28}+0.90\%$
test_compile_add_self_flat[tensorclass-eager]	0.1693ms	54.9825μs	18.1876 KOps/s	18.2439 KOps/s	$\color{#d91a1a}-0.31\%$
test_compile_add_self_flat[tensorclass-compile]	96.8810μs	45.5802μs	21.9394 KOps/s	21.1829 KOps/s	$\color{#35bf28}+3.57\%$
test_compile_add_self_flat[pytree-eager]	0.5782ms	0.1585ms	6.3102 KOps/s	6.2881 KOps/s	$\color{#35bf28}+0.35\%$
test_compile_add_self_flat[pytree-compile]	0.2026ms	0.1025ms	9.7531 KOps/s	9.6708 KOps/s	$\color{#35bf28}+0.85\%$
test_compile_copy_flat[tensordict-compile]	70.6820μs	20.7126μs	48.2797 KOps/s	45.1753 KOps/s	$\textbf{\color{#35bf28}+6.87\%}$
test_compile_copy_flat[tensordict-eager]	0.1589ms	60.5905μs	16.5042 KOps/s	16.7925 KOps/s	$\color{#d91a1a}-1.72\%$
test_compile_copy_flat[pytree-compile]	0.1584ms	81.0871μs	12.3324 KOps/s	12.2722 KOps/s	$\color{#35bf28}+0.49\%$
test_compile_copy_flat[pytree-eager]	0.1556ms	68.5937μs	14.5786 KOps/s	14.3457 KOps/s	$\color{#35bf28}+1.62\%$
test_compile_assign_and_add[tensordict-compile]	0.3019ms	0.2042ms	4.8981 KOps/s	4.8379 KOps/s	$\color{#35bf28}+1.24\%$
test_compile_assign_and_add[tensordict-eager]	2.5015ms	1.3159ms	759.9149 Ops/s	768.9831 Ops/s	$\color{#d91a1a}-1.18\%$
test_compile_assign_and_add[pytree-compile]	0.3220ms	0.1971ms	5.0731 KOps/s	4.9070 KOps/s	$\color{#35bf28}+3.39\%$
test_compile_assign_and_add[pytree-eager]	1.3495ms	0.7736ms	1.2927 KOps/s	1.2690 KOps/s	$\color{#35bf28}+1.87\%$
test_compile_assign_and_add_stack[compile]	0.5380ms	0.4423ms	2.2607 KOps/s	2.1732 KOps/s	$\color{#35bf28}+4.02\%$
test_compile_assign_and_add_stack[eager]	3.4488ms	2.5818ms	387.3200 Ops/s	385.6056 Ops/s	$\color{#35bf28}+0.44\%$
test_compile_indexing[tensor-tensordict-compile]	0.1250ms	34.7766μs	28.7550 KOps/s	26.9807 KOps/s	$\textbf{\color{#35bf28}+6.58\%}$
test_compile_indexing[tensor-tensordict-eager]	0.5529ms	33.1057μs	30.2062 KOps/s	30.1061 KOps/s	$\color{#35bf28}+0.33\%$
test_compile_indexing[tensor-tensorclass-compile]	69.0890μs	28.2200μs	35.4358 KOps/s	33.5975 KOps/s	$\textbf{\color{#35bf28}+5.47\%}$
test_compile_indexing[tensor-tensorclass-eager]	71.0320μs	23.1401μs	43.2150 KOps/s	41.4173 KOps/s	$\color{#35bf28}+4.34\%$
test_compile_indexing[tensor-pytree-compile]	73.9280μs	29.2936μs	34.1371 KOps/s	33.0434 KOps/s	$\color{#35bf28}+3.31\%$
test_compile_indexing[tensor-pytree-eager]	66.0330μs	23.3956μs	42.7431 KOps/s	42.3181 KOps/s	$\color{#35bf28}+1.00\%$
test_compile_indexing[slice-tensordict-compile]	0.1423ms	49.8322μs	20.0673 KOps/s	19.0591 KOps/s	$\textbf{\color{#35bf28}+5.29\%}$
test_compile_indexing[slice-tensordict-eager]	0.5434ms	20.0387μs	49.9033 KOps/s	47.5964 KOps/s	$\color{#35bf28}+4.85\%$
test_compile_indexing[slice-tensorclass-compile]	0.1051ms	43.2209μs	23.1369 KOps/s	22.0161 KOps/s	$\textbf{\color{#35bf28}+5.09\%}$
test_compile_indexing[slice-tensorclass-eager]	58.6600μs	19.3908μs	51.5709 KOps/s	51.7176 KOps/s	$\color{#d91a1a}-0.28\%$
test_compile_indexing[slice-pytree-compile]	0.1307ms	44.4189μs	22.5130 KOps/s	21.7948 KOps/s	$\color{#35bf28}+3.29\%$
test_compile_indexing[slice-pytree-eager]	65.5530μs	19.3740μs	51.6156 KOps/s	51.8418 KOps/s	$\color{#d91a1a}-0.44\%$
test_compile_indexing[int-tensordict-compile]	0.1040ms	51.2658μs	19.5062 KOps/s	18.7086 KOps/s	$\color{#35bf28}+4.26\%$
test_compile_indexing[int-tensordict-eager]	0.8729ms	20.1648μs	49.5913 KOps/s	48.2298 KOps/s	$\color{#35bf28}+2.82\%$
test_compile_indexing[int-tensorclass-compile]	0.1162ms	44.5127μs	22.4655 KOps/s	21.9905 KOps/s	$\color{#35bf28}+2.16\%$
test_compile_indexing[int-tensorclass-eager]	0.2824ms	19.3727μs	51.6191 KOps/s	51.7972 KOps/s	$\color{#d91a1a}-0.34\%$
test_compile_indexing[int-pytree-compile]	0.1528ms	44.2944μs	22.5762 KOps/s	22.3082 KOps/s	$\color{#35bf28}+1.20\%$
test_compile_indexing[int-pytree-eager]	71.8650μs	19.4005μs	51.5451 KOps/s	52.4106 KOps/s	$\color{#d91a1a}-1.65\%$
test_mod_add[eager]	0.3562ms	34.9879μs	28.5813 KOps/s	29.3376 KOps/s	$\color{#d91a1a}-2.58\%$
test_mod_add[compile]	0.1247ms	46.7111μs	21.4082 KOps/s	20.6576 KOps/s	$\color{#35bf28}+3.63\%$
test_mod_add[compile-overhead]	0.1060ms	45.8890μs	21.7917 KOps/s	20.6281 KOps/s	$\textbf{\color{#35bf28}+5.64\%}$
test_mod_wrap[eager]	0.4403ms	0.2232ms	4.4797 KOps/s	4.4111 KOps/s	$\color{#35bf28}+1.56\%$
test_mod_wrap[compile]	0.3136ms	0.2018ms	4.9551 KOps/s	4.7368 KOps/s	$\color{#35bf28}+4.61\%$
test_mod_wrap[compile-overhead]	0.3433ms	0.2001ms	4.9978 KOps/s	4.7775 KOps/s	$\color{#35bf28}+4.61\%$
test_mod_wrap_and_backward[eager]	13.4312ms	11.5987ms	86.2163 Ops/s	84.7359 Ops/s	$\color{#35bf28}+1.75\%$
test_mod_wrap_and_backward[compile]	19.0724ms	12.4183ms	80.5264 Ops/s	79.1348 Ops/s	$\color{#35bf28}+1.76\%$
test_mod_wrap_and_backward[compile-overhead]	12.4300ms	11.1296ms	89.8507 Ops/s	75.9496 Ops/s	$\textbf{\color{#35bf28}+18.30\%}$
test_seq_add[eager]	0.2379ms	0.1124ms	8.8975 KOps/s	9.0822 KOps/s	$\color{#d91a1a}-2.03\%$
test_seq_add[compile]	0.1066ms	60.9707μs	16.4013 KOps/s	15.8282 KOps/s	$\color{#35bf28}+3.62\%$
test_seq_add[compile-overhead]	0.1741ms	58.8631μs	16.9886 KOps/s	16.4253 KOps/s	$\color{#35bf28}+3.43\%$
test_seq_wrap[eager]	0.8666ms	0.4397ms	2.2741 KOps/s	2.2677 KOps/s	$\color{#35bf28}+0.28\%$
test_seq_wrap[compile]	0.4110ms	0.2232ms	4.4813 KOps/s	4.2090 KOps/s	$\textbf{\color{#35bf28}+6.47\%}$
test_seq_wrap[compile-overhead]	0.3910ms	0.2222ms	4.4999 KOps/s	4.2957 KOps/s	$\color{#35bf28}+4.75\%$
test_func_call_runtime[False-eager]	0.9239ms	0.5465ms	1.8298 KOps/s	1.8050 KOps/s	$\color{#35bf28}+1.38\%$
test_func_call_runtime[False-compile]	0.8073ms	0.4213ms	2.3737 KOps/s	2.2959 KOps/s	$\color{#35bf28}+3.39\%$
test_func_call_runtime[False-compile-overhead]	0.5212ms	0.4213ms	2.3737 KOps/s	2.2975 KOps/s	$\color{#35bf28}+3.32\%$
test_func_call_runtime[True-eager]	1.1824ms	0.7575ms	1.3202 KOps/s	1.2961 KOps/s	$\color{#35bf28}+1.86\%$
test_func_call_runtime[True-compile]	0.8238ms	0.4658ms	2.1469 KOps/s	2.0853 KOps/s	$\color{#35bf28}+2.95\%$
test_func_call_runtime[True-compile-overhead]	0.5748ms	0.4642ms	2.1541 KOps/s	2.0925 KOps/s	$\color{#35bf28}+2.94\%$
test_func_call_cm_runtime[False-eager]	0.7458ms	0.5439ms	1.8384 KOps/s	1.8062 KOps/s	$\color{#35bf28}+1.78\%$
test_func_call_cm_runtime[False-compile]	0.9013ms	0.4209ms	2.3759 KOps/s	2.3107 KOps/s	$\color{#35bf28}+2.82\%$
test_func_call_cm_runtime[False-compile-overhead]	0.7860ms	0.4226ms	2.3663 KOps/s	2.3164 KOps/s	$\color{#35bf28}+2.15\%$
test_func_call_cm_runtime[True-eager]	1.0535ms	0.8935ms	1.1193 KOps/s	1.1071 KOps/s	$\color{#35bf28}+1.10\%$
test_func_call_cm_runtime[True-compile]	0.7966ms	0.4904ms	2.0392 KOps/s	2.0122 KOps/s	$\color{#35bf28}+1.34\%$
test_func_call_cm_runtime[True-compile-overhead]	0.6045ms	0.4834ms	2.0685 KOps/s	2.0128 KOps/s	$\color{#35bf28}+2.77\%$
test_vmap_func_call_cm_runtime[eager]	3.2219ms	1.8666ms	535.7409 Ops/s	527.4590 Ops/s	$\color{#35bf28}+1.57\%$
test_vmap_func_call_cm_runtime[compile]	0.7114ms	0.5141ms	1.9450 KOps/s	1.8843 KOps/s	$\color{#35bf28}+3.22\%$
test_vmap_func_call_cm_runtime[compile-overhead]	0.6328ms	0.5133ms	1.9483 KOps/s	1.8868 KOps/s	$\color{#35bf28}+3.26\%$
test_distributed	0.2224ms	0.1254ms	7.9776 KOps/s	7.6644 KOps/s	$\color{#35bf28}+4.09\%$
test_tdmodule	57.6480μs	26.2231μs	38.1343 KOps/s	38.3117 KOps/s	$\color{#d91a1a}-0.46\%$
test_tdmodule_dispatch	81.9330μs	47.8501μs	20.8986 KOps/s	20.5457 KOps/s	$\color{#35bf28}+1.72\%$
test_tdseq	42.3280μs	26.5696μs	37.6370 KOps/s	38.0765 KOps/s	$\color{#d91a1a}-1.15\%$
test_tdseq_dispatch	92.1620μs	51.1636μs	19.5452 KOps/s	20.0698 KOps/s	$\color{#d91a1a}-2.61\%$
test_instantiation_functorch	2.7351ms	1.5309ms	653.1977 Ops/s	646.4587 Ops/s	$\color{#35bf28}+1.04\%$
test_exec_functorch	0.2933ms	0.1807ms	5.5334 KOps/s	5.3813 KOps/s	$\color{#35bf28}+2.83\%$
test_exec_functional_call	0.4015ms	0.1743ms	5.7373 KOps/s	5.8295 KOps/s	$\color{#d91a1a}-1.58\%$
test_exec_td_decorator	0.5770ms	0.2286ms	4.3753 KOps/s	4.1729 KOps/s	$\color{#35bf28}+4.85\%$
test_vmap_mlp_speed_decorator[True-True]	0.9696ms	0.6652ms	1.5033 KOps/s	1.5188 KOps/s	$\color{#d91a1a}-1.02\%$
test_vmap_mlp_speed_decorator[True-False]	0.8629ms	0.6450ms	1.5503 KOps/s	1.5496 KOps/s	$\color{#35bf28}+0.05\%$
test_vmap_mlp_speed_decorator[False-True]	0.8145ms	0.5188ms	1.9277 KOps/s	1.8944 KOps/s	$\color{#35bf28}+1.76\%$
test_vmap_mlp_speed_decorator[False-False]	0.9514ms	0.5275ms	1.8956 KOps/s	1.8888 KOps/s	$\color{#35bf28}+0.36\%$
test_to_module_speed[True]	1.8791ms	1.2769ms	783.1539 Ops/s	763.2110 Ops/s	$\color{#35bf28}+2.61\%$
test_to_module_speed[False]	1.9707ms	1.2611ms	792.9718 Ops/s	798.5748 Ops/s	$\color{#d91a1a}-0.70\%$
test_tc_init	89.0960μs	48.1496μs	20.7686 KOps/s	22.4475 KOps/s	$\textbf{\color{#d91a1a}-7.48\%}$
test_tc_init_nested	0.1869ms	93.7676μs	10.6647 KOps/s	10.9262 KOps/s	$\color{#d91a1a}-2.39\%$
test_tc_first_layer_tensor	41.5000μs	1.4867μs	672.6243 KOps/s	659.7877 KOps/s	$\color{#35bf28}+1.95\%$
test_tc_first_layer_nontensor	29.8560μs	4.7152μs	212.0795 KOps/s	204.6342 KOps/s	$\color{#35bf28}+3.64\%$
test_tc_second_layer_tensor	34.3440μs	2.7573μs	362.6695 KOps/s	362.5930 KOps/s	$\color{#35bf28}+0.02\%$
test_tc_second_layer_nontensor	53.9810μs	6.0627μs	164.9421 KOps/s	167.2721 KOps/s	$\color{#d91a1a}-1.39\%$
test_unbind	0.2240s	13.0261ms	76.7692 Ops/s	75.9648 Ops/s	$\color{#35bf28}+1.06\%$
test_full_like	8.9543ms	7.2131ms	138.6362 Ops/s	130.3522 Ops/s	$\textbf{\color{#35bf28}+6.36\%}$
test_zeros_like	3.6484ms	2.7963ms	357.6154 Ops/s	327.9688 Ops/s	$\textbf{\color{#35bf28}+9.04\%}$
test_ones_like	11.9553ms	6.3870ms	156.5668 Ops/s	292.1165 Ops/s	$\textbf{\color{#d91a1a}-46.40\%}$
test_clone	10.7872ms	8.2736ms	120.8660 Ops/s	156.6369 Ops/s	$\textbf{\color{#d91a1a}-22.84\%}$
test_squeeze	79.7890μs	12.0364μs	83.0813 KOps/s	81.9017 KOps/s	$\color{#35bf28}+1.44\%$
test_unsqueeze	0.1518ms	89.5803μs	11.1632 KOps/s	10.9897 KOps/s	$\color{#35bf28}+1.58\%$
test_split	0.5494ms	0.1950ms	5.1276 KOps/s	5.0201 KOps/s	$\color{#35bf28}+2.14\%$
test_permute	0.3865ms	0.2041ms	4.8997 KOps/s	4.8554 KOps/s	$\color{#35bf28}+0.91\%$
test_stack	28.6310ms	24.6912ms	40.5002 Ops/s	37.6961 Ops/s	$\textbf{\color{#35bf28}+7.44\%}$
test_cat	27.4176ms	24.5151ms	40.7912 Ops/s	37.8607 Ops/s	$\textbf{\color{#35bf28}+7.74\%}$

[github-actions]

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

Total Benchmarks: 229. Improved: $\large\color{#35bf28}20$. Worsened: $\large\color{#d91a1a}32$.

Expand to view detailed results

Name	Max	Mean	Ops	Ops on Repo HEAD	Change
test_plain_set_nested	34.8710μs	11.0041μs	90.8748 KOps/s	99.2428 KOps/s	$\textbf{\color{#d91a1a}-8.43\%}$
test_plain_set_stack_nested	34.8010μs	11.0622μs	90.3981 KOps/s	99.1158 KOps/s	$\textbf{\color{#d91a1a}-8.80\%}$
test_plain_set_nested_inplace	0.2097ms	11.9140μs	83.9351 KOps/s	91.2335 KOps/s	$\textbf{\color{#d91a1a}-8.00\%}$
test_plain_set_stack_nested_inplace	42.0310μs	11.9248μs	83.8591 KOps/s	91.2350 KOps/s	$\textbf{\color{#d91a1a}-8.08\%}$
test_items	36.4910μs	2.8894μs	346.0947 KOps/s	343.0515 KOps/s	$\color{#35bf28}+0.89\%$
test_items_nested	0.5518ms	0.3553ms	2.8148 KOps/s	2.8045 KOps/s	$\color{#35bf28}+0.37\%$
test_items_nested_locked	0.4142ms	0.3573ms	2.7984 KOps/s	2.7966 KOps/s	$\color{#35bf28}+0.06\%$
test_items_nested_leaf	82.0110μs	57.9861μs	17.2455 KOps/s	17.1991 KOps/s	$\color{#35bf28}+0.27\%$
test_items_stack_nested	0.4009ms	0.3545ms	2.8206 KOps/s	2.7645 KOps/s	$\color{#35bf28}+2.03\%$
test_items_stack_nested_leaf	86.0520μs	57.7730μs	17.3091 KOps/s	16.6326 KOps/s	$\color{#35bf28}+4.07\%$
test_items_stack_nested_locked	0.3962ms	0.3563ms	2.8062 KOps/s	2.7876 KOps/s	$\color{#35bf28}+0.67\%$
test_keys	33.8600μs	3.4879μs	286.7093 KOps/s	288.1990 KOps/s	$\color{#d91a1a}-0.52\%$
test_keys_nested	0.1117ms	70.5725μs	14.1698 KOps/s	13.9837 KOps/s	$\color{#35bf28}+1.33\%$
test_keys_nested_locked	0.7801ms	76.8735μs	13.0084 KOps/s	12.9818 KOps/s	$\color{#35bf28}+0.20\%$
test_keys_nested_leaf	97.4020μs	61.6455μs	16.2218 KOps/s	16.1851 KOps/s	$\color{#35bf28}+0.23\%$
test_keys_stack_nested	0.1277ms	70.7646μs	14.1314 KOps/s	13.9639 KOps/s	$\color{#35bf28}+1.20\%$
test_keys_stack_nested_leaf	0.1011ms	61.2538μs	16.3255 KOps/s	15.8927 KOps/s	$\color{#35bf28}+2.72\%$
test_keys_stack_nested_locked	0.1125ms	76.2963μs	13.1068 KOps/s	12.9378 KOps/s	$\color{#35bf28}+1.31\%$
test_values	6.5102μs	0.8530μs	1.1724 MOps/s	1.1348 MOps/s	$\color{#35bf28}+3.31\%$
test_values_nested	95.2620μs	31.4049μs	31.8421 KOps/s	32.1721 KOps/s	$\color{#d91a1a}-1.03\%$
test_values_nested_locked	64.3210μs	33.2048μs	30.1162 KOps/s	30.7125 KOps/s	$\color{#d91a1a}-1.94\%$
test_values_nested_leaf	62.6310μs	33.6128μs	29.7506 KOps/s	29.9822 KOps/s	$\color{#d91a1a}-0.77\%$
test_values_stack_nested	72.7710μs	31.5240μs	31.7219 KOps/s	31.3094 KOps/s	$\color{#35bf28}+1.32\%$
test_values_stack_nested_leaf	60.2410μs	33.9152μs	29.4853 KOps/s	29.1574 KOps/s	$\color{#35bf28}+1.12\%$
test_values_stack_nested_locked	64.9410μs	33.0922μs	30.2186 KOps/s	29.9292 KOps/s	$\color{#35bf28}+0.97\%$
test_membership	1.9350μs	0.5257μs	1.9023 MOps/s	1.8586 MOps/s	$\color{#35bf28}+2.35\%$
test_membership_nested	19.5150μs	2.0086μs	497.8522 KOps/s	493.8639 KOps/s	$\color{#35bf28}+0.81\%$
test_membership_nested_leaf	15.8955μs	2.0152μs	496.2238 KOps/s	492.7361 KOps/s	$\color{#35bf28}+0.71\%$
test_membership_stacked_nested	29.6500μs	2.1069μs	474.6322 KOps/s	480.3819 KOps/s	$\color{#d91a1a}-1.20\%$
test_membership_stacked_nested_leaf	24.7910μs	2.1431μs	466.6126 KOps/s	482.9710 KOps/s	$\color{#d91a1a}-3.39\%$
test_membership_nested_last	26.6410μs	2.9834μs	335.1864 KOps/s	334.4776 KOps/s	$\color{#35bf28}+0.21\%$
test_membership_nested_leaf_last	25.7000μs	2.9732μs	336.3377 KOps/s	337.0392 KOps/s	$\color{#d91a1a}-0.21\%$
test_membership_stacked_nested_last	36.0110μs	3.0141μs	331.7715 KOps/s	295.5516 KOps/s	$\textbf{\color{#35bf28}+12.25\%}$
test_membership_stacked_nested_leaf_last	41.6310μs	2.9638μs	337.4099 KOps/s	297.0103 KOps/s	$\textbf{\color{#35bf28}+13.60\%}$
test_nested_getleaf	30.5110μs	6.1781μs	161.8613 KOps/s	162.6317 KOps/s	$\color{#d91a1a}-0.47\%$
test_nested_get	48.9010μs	5.8865μs	169.8792 KOps/s	169.3999 KOps/s	$\color{#35bf28}+0.28\%$
test_stacked_getleaf	41.0110μs	6.1916μs	161.5084 KOps/s	162.5174 KOps/s	$\color{#d91a1a}-0.62\%$
test_stacked_get	32.1300μs	5.8915μs	169.7359 KOps/s	171.7732 KOps/s	$\color{#d91a1a}-1.19\%$
test_nested_getitemleaf	41.0110μs	6.2612μs	159.7140 KOps/s	160.7271 KOps/s	$\color{#d91a1a}-0.63\%$
test_nested_getitem	32.7410μs	5.9501μs	168.0655 KOps/s	168.3493 KOps/s	$\color{#d91a1a}-0.17\%$
test_stacked_getitemleaf	43.6710μs	6.2497μs	160.0090 KOps/s	160.7908 KOps/s	$\color{#d91a1a}-0.49\%$
test_stacked_getitem	27.0510μs	5.9391μs	168.3752 KOps/s	169.6052 KOps/s	$\color{#d91a1a}-0.73\%$
test_lock_nested	1.0829ms	0.3666ms	2.7274 KOps/s	2.6574 KOps/s	$\color{#35bf28}+2.64\%$
test_lock_stack_nested	0.3829ms	0.3356ms	2.9794 KOps/s	2.9374 KOps/s	$\color{#35bf28}+1.43\%$
test_unlock_nested	0.6567ms	0.3037ms	3.2923 KOps/s	3.2342 KOps/s	$\color{#35bf28}+1.80\%$
test_unlock_stack_nested	0.3777ms	0.2728ms	3.6651 KOps/s	3.6093 KOps/s	$\color{#35bf28}+1.55\%$
test_flatten_speed	0.1266ms	74.2452μs	13.4689 KOps/s	13.3398 KOps/s	$\color{#35bf28}+0.97\%$
test_unflatten_speed	0.3545ms	0.3059ms	3.2690 KOps/s	3.2359 KOps/s	$\color{#35bf28}+1.02\%$
test_common_ops	1.5069ms	0.5900ms	1.6949 KOps/s	1.8026 KOps/s	$\textbf{\color{#d91a1a}-5.97\%}$
test_creation	0.1915ms	1.4520μs	688.7262 KOps/s	672.4305 KOps/s	$\color{#35bf28}+2.42\%$
test_creation_empty	31.7600μs	7.8731μs	127.0142 KOps/s	164.9215 KOps/s	$\textbf{\color{#d91a1a}-22.99\%}$
test_creation_nested_1	1.6214ms	9.3727μs	106.6923 KOps/s	131.5830 KOps/s	$\textbf{\color{#d91a1a}-18.92\%}$
test_creation_nested_2	61.8210μs	11.9320μs	83.8085 KOps/s	98.7094 KOps/s	$\textbf{\color{#d91a1a}-15.10\%}$
test_clone	0.1421ms	9.9534μs	100.4685 KOps/s	96.5006 KOps/s	$\color{#35bf28}+4.11\%$
test_getitem[int]	1.1859ms	10.6200μs	94.1621 KOps/s	91.3714 KOps/s	$\color{#35bf28}+3.05\%$
test_getitem[slice_int]	0.1327ms	20.4255μs	48.9584 KOps/s	46.9091 KOps/s	$\color{#35bf28}+4.37\%$
test_getitem[range]	0.1821ms	39.0707μs	25.5946 KOps/s	26.9495 KOps/s	$\textbf{\color{#d91a1a}-5.03\%}$
test_getitem[tuple]	0.1057ms	18.0626μs	55.3630 KOps/s	53.9004 KOps/s	$\color{#35bf28}+2.71\%$
test_getitem[list]	0.1327ms	32.0609μs	31.1906 KOps/s	28.8805 KOps/s	$\textbf{\color{#35bf28}+8.00\%}$
test_setitem_dim[int]	44.9900μs	18.7600μs	53.3048 KOps/s	55.0459 KOps/s	$\color{#d91a1a}-3.16\%$
test_setitem_dim[slice_int]	59.6710μs	38.0448μs	26.2848 KOps/s	26.8973 KOps/s	$\color{#d91a1a}-2.28\%$
test_setitem_dim[range]	78.1810μs	54.8812μs	18.2212 KOps/s	19.4689 KOps/s	$\textbf{\color{#d91a1a}-6.41\%}$
test_setitem_dim[tuple]	54.2010μs	32.5604μs	30.7121 KOps/s	32.3023 KOps/s	$\color{#d91a1a}-4.92\%$
test_setitem	0.1536ms	15.8016μs	63.2847 KOps/s	74.5713 KOps/s	$\textbf{\color{#d91a1a}-15.14\%}$
test_set	50.0710μs	14.3334μs	69.7669 KOps/s	75.5425 KOps/s	$\textbf{\color{#d91a1a}-7.65\%}$
test_set_shared	1.6386ms	0.1441ms	6.9388 KOps/s	6.8314 KOps/s	$\color{#35bf28}+1.57\%$
test_update	0.5672ms	16.6994μs	59.8823 KOps/s	66.6476 KOps/s	$\textbf{\color{#d91a1a}-10.15\%}$
test_update_nested	1.1611ms	24.2835μs	41.1802 KOps/s	49.5103 KOps/s	$\textbf{\color{#d91a1a}-16.83\%}$
test_update__nested	67.8310μs	24.5384μs	40.7524 KOps/s	42.1476 KOps/s	$\color{#d91a1a}-3.31\%$
test_set_nested	0.1329ms	14.9513μs	66.8837 KOps/s	70.1718 KOps/s	$\color{#d91a1a}-4.69\%$
test_set_nested_new	0.1330ms	17.8878μs	55.9040 KOps/s	60.5532 KOps/s	$\textbf{\color{#d91a1a}-7.68\%}$
test_select	0.1436ms	30.5214μs	32.7639 KOps/s	35.4977 KOps/s	$\textbf{\color{#d91a1a}-7.70\%}$
test_select_nested	69.6610μs	42.1584μs	23.7201 KOps/s	23.6565 KOps/s	$\color{#35bf28}+0.27\%$
test_exclude_nested	0.1012ms	61.7277μs	16.2002 KOps/s	16.2623 KOps/s	$\color{#d91a1a}-0.38\%$
test_empty[True]	0.3756ms	0.2759ms	3.6240 KOps/s	3.5650 KOps/s	$\color{#35bf28}+1.66\%$
test_empty[False]	3.3421μs	0.7787μs	1.2842 MOps/s	1.3031 MOps/s	$\color{#d91a1a}-1.45\%$
test_to	87.0120μs	54.6667μs	18.2927 KOps/s	17.2563 KOps/s	$\textbf{\color{#35bf28}+6.01\%}$
test_to_nonblocking	0.1960ms	45.4017μs	22.0256 KOps/s	21.6784 KOps/s	$\color{#35bf28}+1.60\%$
test_unbind_speed	1.6694ms	0.2339ms	4.2748 KOps/s	4.3597 KOps/s	$\color{#d91a1a}-1.95\%$
test_unbind_speed_stack0	0.2971ms	0.2284ms	4.3785 KOps/s	4.2961 KOps/s	$\color{#35bf28}+1.92\%$
test_unbind_speed_stack1	93.1758ms	0.6464ms	1.5471 KOps/s	1.6735 KOps/s	$\textbf{\color{#d91a1a}-7.55\%}$
test_split	95.5329ms	1.7094ms	585.0142 Ops/s	619.2744 Ops/s	$\textbf{\color{#d91a1a}-5.53\%}$
test_chunk	97.3832ms	1.5828ms	631.7880 Ops/s	616.9278 Ops/s	$\color{#35bf28}+2.41\%$
test_consolidate[False-None]	3.4352ms	2.6134ms	382.6479 Ops/s	344.0850 Ops/s	$\textbf{\color{#35bf28}+11.21\%}$
test_consolidate[default-None]	1.7529ms	1.6779ms	595.9765 Ops/s	583.1654 Ops/s	$\color{#35bf28}+2.20\%$
test_consolidate[reduce-overhead-None]	1.8372ms	1.7073ms	585.7086 Ops/s	569.1823 Ops/s	$\color{#35bf28}+2.90\%$
test_consolidate_njt[False-None]	6.5931ms	6.3181ms	158.2744 Ops/s	158.6550 Ops/s	$\color{#d91a1a}-0.24\%$
test_to[False-False-None]	1.7598ms	1.6320ms	612.7491 Ops/s	609.2914 Ops/s	$\color{#35bf28}+0.57\%$
test_to[True-False-None]	1.4933ms	1.2700ms	787.3749 Ops/s	774.2872 Ops/s	$\color{#35bf28}+1.69\%$
test_to[within-False-None]	4.1573ms	3.9130ms	255.5610 Ops/s	182.4377 Ops/s	$\textbf{\color{#35bf28}+40.08\%}$
test_to[True-default-None]	5.2707ms	5.0480ms	198.0974 Ops/s	191.3883 Ops/s	$\color{#35bf28}+3.51\%$
test_to_njt[False-False-None]	6.9883ms	6.8237ms	146.5479 Ops/s	140.9777 Ops/s	$\color{#35bf28}+3.95\%$
test_to_njt[True-False-None]	5.5974ms	5.3468ms	187.0291 Ops/s	179.0251 Ops/s	$\color{#35bf28}+4.47\%$
test_to_njt[within-False-None]	11.9290ms	11.8002ms	84.7447 Ops/s	80.5381 Ops/s	$\textbf{\color{#35bf28}+5.22\%}$
test_creation[device0]	0.4669ms	77.9331μs	12.8315 KOps/s	12.7164 KOps/s	$\color{#35bf28}+0.90\%$
test_creation_from_tensor	0.5989ms	81.8433μs	12.2185 KOps/s	12.3143 KOps/s	$\color{#d91a1a}-0.78\%$
test_add_one[memmap_tensor0]	0.2761ms	5.9978μs	166.7276 KOps/s	151.3426 KOps/s	$\textbf{\color{#35bf28}+10.17\%}$
test_contiguous[memmap_tensor0]	1.7561μs	0.4065μs	2.4601 MOps/s	2.4123 MOps/s	$\color{#35bf28}+1.98\%$
test_stack[memmap_tensor0]	29.3200μs	4.4055μs	226.9868 KOps/s	215.9160 KOps/s	$\textbf{\color{#35bf28}+5.13\%}$
test_memmaptd_index	1.6368ms	0.2414ms	4.1433 KOps/s	3.9541 KOps/s	$\color{#35bf28}+4.78\%$
test_memmaptd_index_astensor	0.5689ms	0.3010ms	3.3228 KOps/s	3.2359 KOps/s	$\color{#35bf28}+2.68\%$
test_memmaptd_index_op	0.9586ms	0.5438ms	1.8389 KOps/s	1.8273 KOps/s	$\color{#35bf28}+0.64\%$
test_serialize_model	0.1317s	0.1307s	7.6494 Ops/s	7.6800 Ops/s	$\color{#d91a1a}-0.40\%$
test_serialize_model_pickle	1.3524s	1.2120s	0.8251 Ops/s	0.8213 Ops/s	$\color{#35bf28}+0.46\%$
test_serialize_weights	0.1314s	0.1301s	7.6886 Ops/s	7.7170 Ops/s	$\color{#d91a1a}-0.37\%$
test_serialize_weights_returnearly	0.4289s	65.3719ms	15.2971 Ops/s	11.9756 Ops/s	$\textbf{\color{#35bf28}+27.74\%}$
test_serialize_weights_pickle	1.3520s	1.2151s	0.8230 Ops/s	0.8097 Ops/s	$\color{#35bf28}+1.64\%$
test_reshape_pytree	62.4310μs	22.2704μs	44.9026 KOps/s	43.6986 KOps/s	$\color{#35bf28}+2.76\%$
test_reshape_td	0.1307ms	26.9961μs	37.0424 KOps/s	36.6236 KOps/s	$\color{#35bf28}+1.14\%$
test_view_pytree	45.1100μs	22.1741μs	45.0977 KOps/s	44.9749 KOps/s	$\color{#35bf28}+0.27\%$
test_view_td	55.3410μs	29.9817μs	33.3536 KOps/s	31.6246 KOps/s	$\textbf{\color{#35bf28}+5.47\%}$
test_unbind_pytree	63.3410μs	27.7601μs	36.0229 KOps/s	35.4055 KOps/s	$\color{#35bf28}+1.74\%$
test_unbind_td	0.7539ms	35.2381μs	28.3784 KOps/s	27.9572 KOps/s	$\color{#35bf28}+1.51\%$
test_split_pytree	0.2229ms	29.6768μs	33.6964 KOps/s	32.9420 KOps/s	$\color{#35bf28}+2.29\%$
test_split_td	0.9425ms	38.4849μs	25.9842 KOps/s	24.7440 KOps/s	$\textbf{\color{#35bf28}+5.01\%}$
test_add_pytree	81.1810μs	33.3267μs	30.0060 KOps/s	29.1207 KOps/s	$\color{#35bf28}+3.04\%$
test_add_td	0.1010ms	45.2660μs	22.0917 KOps/s	23.5531 KOps/s	$\textbf{\color{#d91a1a}-6.21\%}$
test_compile_add_one_nested[tensordict-compile]	0.1879ms	0.1208ms	8.2770 KOps/s	8.0177 KOps/s	$\color{#35bf28}+3.23\%$
test_compile_add_one_nested[tensordict-eager]	0.2665ms	0.1242ms	8.0514 KOps/s	7.8525 KOps/s	$\color{#35bf28}+2.53\%$
test_compile_add_one_nested[pytree-compile]	0.2441ms	96.5336μs	10.3591 KOps/s	10.0702 KOps/s	$\color{#35bf28}+2.87\%$
test_compile_add_one_nested[pytree-eager]	0.3108ms	0.1448ms	6.9038 KOps/s	6.6651 KOps/s	$\color{#35bf28}+3.58\%$
test_compile_copy_nested[tensordict-compile]	0.1527ms	23.3177μs	42.8859 KOps/s	37.1916 KOps/s	$\textbf{\color{#35bf28}+15.31\%}$
test_compile_copy_nested[tensordict-eager]	0.1398ms	26.9561μs	37.0974 KOps/s	36.6195 KOps/s	$\color{#35bf28}+1.31\%$
test_compile_copy_nested[pytree-compile]	0.1120ms	65.1364μs	15.3524 KOps/s	15.1034 KOps/s	$\color{#35bf28}+1.65\%$
test_compile_copy_nested[pytree-eager]	87.8920μs	49.2063μs	20.3226 KOps/s	19.9538 KOps/s	$\color{#35bf28}+1.85\%$
test_compile_add_one_flat[tensordict-compile]	0.1823ms	0.1437ms	6.9613 KOps/s	7.0175 KOps/s	$\color{#d91a1a}-0.80\%$
test_compile_add_one_flat[tensordict-eager]	0.3222ms	0.2079ms	4.8103 KOps/s	4.8165 KOps/s	$\color{#d91a1a}-0.13\%$
test_compile_add_one_flat[tensorclass-compile]	0.2527ms	0.1017ms	9.8328 KOps/s	10.1802 KOps/s	$\color{#d91a1a}-3.41\%$
test_compile_add_one_flat[tensorclass-eager]	0.2020ms	52.8713μs	18.9139 KOps/s	19.3353 KOps/s	$\color{#d91a1a}-2.18\%$
test_compile_add_one_flat[pytree-compile]	0.2315ms	0.1381ms	7.2430 KOps/s	7.3714 KOps/s	$\color{#d91a1a}-1.74\%$
test_compile_add_one_flat[pytree-eager]	0.6325ms	0.4671ms	2.1408 KOps/s	2.0730 KOps/s	$\color{#35bf28}+3.27\%$
test_compile_add_self_flat[tensordict-eager]	0.3813ms	0.2476ms	4.0383 KOps/s	4.0396 KOps/s	$\color{#d91a1a}-0.03\%$
test_compile_add_self_flat[tensordict-compile]	0.2978ms	0.1458ms	6.8581 KOps/s	6.9499 KOps/s	$\color{#d91a1a}-1.32\%$
test_compile_add_self_flat[tensorclass-eager]	0.1954ms	61.3196μs	16.3080 KOps/s	16.0530 KOps/s	$\color{#35bf28}+1.59\%$
test_compile_add_self_flat[tensorclass-compile]	0.2511ms	0.1006ms	9.9442 KOps/s	10.0914 KOps/s	$\color{#d91a1a}-1.46\%$
test_compile_add_self_flat[pytree-eager]	0.4337ms	0.3980ms	2.5124 KOps/s	2.4712 KOps/s	$\color{#35bf28}+1.67\%$
test_compile_add_self_flat[pytree-compile]	0.1774ms	0.1381ms	7.2405 KOps/s	7.4440 KOps/s	$\color{#d91a1a}-2.73\%$
test_compile_copy_flat[tensordict-compile]	55.6210μs	19.5877μs	51.0524 KOps/s	44.9562 KOps/s	$\textbf{\color{#35bf28}+13.56\%}$
test_compile_copy_flat[tensordict-eager]	0.1036ms	27.0948μs	36.9074 KOps/s	37.2064 KOps/s	$\color{#d91a1a}-0.80\%$
test_compile_copy_flat[pytree-compile]	0.1203ms	69.9991μs	14.2859 KOps/s	14.2657 KOps/s	$\color{#35bf28}+0.14\%$
test_compile_copy_flat[pytree-eager]	83.6610μs	51.3586μs	19.4709 KOps/s	19.3201 KOps/s	$\color{#35bf28}+0.78\%$
test_compile_assign_and_add[tensordict-compile]	1.6708ms	0.4047ms	2.4709 KOps/s	2.1537 KOps/s	$\textbf{\color{#35bf28}+14.73\%}$
test_compile_assign_and_add[tensordict-eager]	2.7174ms	2.5562ms	391.2004 Ops/s	392.8270 Ops/s	$\color{#d91a1a}-0.41\%$
test_compile_assign_and_add[pytree-compile]	1.6220ms	0.4382ms	2.2822 KOps/s	2.2651 KOps/s	$\color{#35bf28}+0.76\%$
test_compile_assign_and_add[pytree-eager]	2.7814ms	2.5880ms	386.4003 Ops/s	385.6575 Ops/s	$\color{#35bf28}+0.19\%$
test_compile_indexing[tensor-tensordict-compile]	0.2577ms	0.1184ms	8.4447 KOps/s	8.8490 KOps/s	$\color{#d91a1a}-4.57\%$
test_compile_indexing[tensor-tensordict-eager]	0.5532ms	81.6189μs	12.2521 KOps/s	12.7811 KOps/s	$\color{#d91a1a}-4.14\%$
test_compile_indexing[tensor-tensorclass-compile]	0.2646ms	0.1128ms	8.8616 KOps/s	9.5231 KOps/s	$\textbf{\color{#d91a1a}-6.95\%}$
test_compile_indexing[tensor-tensorclass-eager]	0.2187ms	69.8419μs	14.3181 KOps/s	15.0157 KOps/s	$\color{#d91a1a}-4.65\%$
test_compile_indexing[tensor-pytree-compile]	0.2616ms	0.1125ms	8.8889 KOps/s	9.4446 KOps/s	$\textbf{\color{#d91a1a}-5.88\%}$
test_compile_indexing[tensor-pytree-eager]	0.2209ms	70.7866μs	14.1270 KOps/s	14.9580 KOps/s	$\textbf{\color{#d91a1a}-5.56\%}$
test_compile_indexing[slice-tensordict-compile]	0.2456ms	0.1062ms	9.4120 KOps/s	9.6838 KOps/s	$\color{#d91a1a}-2.81\%$
test_compile_indexing[slice-tensordict-eager]	0.1529ms	17.2353μs	58.0205 KOps/s	56.3916 KOps/s	$\color{#35bf28}+2.89\%$
test_compile_indexing[slice-tensorclass-compile]	0.1433ms	98.7622μs	10.1253 KOps/s	10.2780 KOps/s	$\color{#d91a1a}-1.49\%$
test_compile_indexing[slice-tensorclass-eager]	56.3810μs	15.8869μs	62.9449 KOps/s	62.7754 KOps/s	$\color{#35bf28}+0.27\%$
test_compile_indexing[slice-pytree-compile]	0.1526ms	0.1019ms	9.8160 KOps/s	10.2709 KOps/s	$\color{#d91a1a}-4.43\%$
test_compile_indexing[slice-pytree-eager]	0.1539ms	15.8090μs	63.2551 KOps/s	60.3242 KOps/s	$\color{#35bf28}+4.86\%$
test_compile_indexing[int-tensordict-compile]	0.2524ms	0.1055ms	9.4764 KOps/s	9.8314 KOps/s	$\color{#d91a1a}-3.61\%$
test_compile_indexing[int-tensordict-eager]	0.5858ms	17.0207μs	58.7520 KOps/s	57.7100 KOps/s	$\color{#35bf28}+1.81\%$
test_compile_indexing[int-tensorclass-compile]	0.4868ms	0.1015ms	9.8570 KOps/s	9.8995 KOps/s	$\color{#d91a1a}-0.43\%$
test_compile_indexing[int-tensorclass-eager]	43.7310μs	15.8541μs	63.0751 KOps/s	62.7043 KOps/s	$\color{#35bf28}+0.59\%$
test_compile_indexing[int-pytree-compile]	0.4928ms	98.7879μs	10.1227 KOps/s	10.2815 KOps/s	$\color{#d91a1a}-1.54\%$
test_compile_indexing[int-pytree-eager]	45.8010μs	16.0993μs	62.1145 KOps/s	62.8196 KOps/s	$\color{#d91a1a}-1.12\%$
test_mod_add[eager]	80.7910μs	37.5291μs	26.6460 KOps/s	27.6718 KOps/s	$\color{#d91a1a}-3.71\%$
test_mod_add[compile]	0.2005ms	81.7292μs	12.2355 KOps/s	12.2280 KOps/s	$\color{#35bf28}+0.06\%$
test_mod_add[compile-overhead]	0.3241ms	0.1669ms	5.9902 KOps/s	5.6952 KOps/s	$\textbf{\color{#35bf28}+5.18\%}$
test_mod_wrap[eager]	0.3493ms	0.2604ms	3.8401 KOps/s	4.0453 KOps/s	$\textbf{\color{#d91a1a}-5.07\%}$
test_mod_wrap[compile]	0.4307ms	0.2863ms	3.4927 KOps/s	3.4573 KOps/s	$\color{#35bf28}+1.03\%$
test_mod_wrap[compile-overhead]	7.1899ms	3.7873ms	264.0407 Ops/s	265.1275 Ops/s	$\color{#d91a1a}-0.41\%$
test_mod_wrap_and_backward[eager]	1.4700ms	1.3254ms	754.4637 Ops/s	699.8996 Ops/s	$\textbf{\color{#35bf28}+7.80\%}$
test_mod_wrap_and_backward[compile]	1.3928ms	1.2535ms	797.7621 Ops/s	725.2439 Ops/s	$\textbf{\color{#35bf28}+10.00\%}$
test_mod_wrap_and_backward[compile-overhead]	1.3757ms	0.9362ms	1.0681 KOps/s	964.2428 Ops/s	$\textbf{\color{#35bf28}+10.77\%}$
test_seq_add[eager]	0.1596ms	0.1122ms	8.9117 KOps/s	8.8924 KOps/s	$\color{#35bf28}+0.22\%$
test_seq_add[compile]	0.2250ms	87.1435μs	11.4753 KOps/s	11.1027 KOps/s	$\color{#35bf28}+3.36\%$
test_seq_add[compile-overhead]	0.3130ms	0.1290ms	7.7496 KOps/s	7.6057 KOps/s	$\color{#35bf28}+1.89\%$
test_seq_wrap[eager]	0.6097ms	0.4070ms	2.4573 KOps/s	2.4219 KOps/s	$\color{#35bf28}+1.46\%$
test_seq_wrap[compile]	0.4451ms	0.2962ms	3.3767 KOps/s	3.2528 KOps/s	$\color{#35bf28}+3.81\%$
test_seq_wrap[compile-overhead]	0.3601ms	0.2229ms	4.4855 KOps/s	4.3995 KOps/s	$\color{#35bf28}+1.95\%$
test_func_call_runtime[False-eager]	0.8705ms	0.7064ms	1.4157 KOps/s	1.3721 KOps/s	$\color{#35bf28}+3.17\%$
test_func_call_runtime[False-compile]	0.8989ms	0.7378ms	1.3554 KOps/s	1.3211 KOps/s	$\color{#35bf28}+2.59\%$
test_func_call_runtime[False-compile-overhead]	0.5147ms	0.3625ms	2.7587 KOps/s	2.7394 KOps/s	$\color{#35bf28}+0.70\%$
test_func_call_runtime[True-eager]	1.0039ms	0.8649ms	1.1562 KOps/s	1.1111 KOps/s	$\color{#35bf28}+4.06\%$
test_func_call_runtime[True-compile]	0.8714ms	0.7578ms	1.3195 KOps/s	1.2870 KOps/s	$\color{#35bf28}+2.52\%$
test_func_call_runtime[True-compile-overhead]	0.5176ms	0.3846ms	2.6003 KOps/s	2.5922 KOps/s	$\color{#35bf28}+0.31\%$
test_func_call_cm_runtime[False-eager]	0.8473ms	0.7046ms	1.4193 KOps/s	1.3703 KOps/s	$\color{#35bf28}+3.58\%$
test_func_call_cm_runtime[False-compile]	0.8885ms	0.7388ms	1.3536 KOps/s	1.3214 KOps/s	$\color{#35bf28}+2.43\%$
test_func_call_cm_runtime[False-compile-overhead]	0.5047ms	0.3650ms	2.7399 KOps/s	2.7348 KOps/s	$\color{#35bf28}+0.19\%$
test_func_call_cm_runtime[True-eager]	1.1246ms	0.9670ms	1.0341 KOps/s	998.2042 Ops/s	$\color{#35bf28}+3.60\%$
test_func_call_cm_runtime[True-compile]	0.9215ms	0.7848ms	1.2743 KOps/s	1.2429 KOps/s	$\color{#35bf28}+2.52\%$
test_func_call_cm_runtime[True-compile-overhead]	0.5331ms	0.4170ms	2.3981 KOps/s	2.4214 KOps/s	$\color{#d91a1a}-0.96\%$
test_vmap_func_call_cm_runtime[eager]	2.4628ms	2.0161ms	496.0027 Ops/s	492.2531 Ops/s	$\color{#35bf28}+0.76\%$
test_vmap_func_call_cm_runtime[compile]	0.9433ms	0.8147ms	1.2275 KOps/s	1.2104 KOps/s	$\color{#35bf28}+1.41\%$
test_vmap_func_call_cm_runtime[compile-overhead]	0.4668ms	0.4179ms	2.3927 KOps/s	2.3482 KOps/s	$\color{#35bf28}+1.89\%$
test_distributed	2.7818ms	0.2000ms	5.0006 KOps/s	8.6894 KOps/s	$\textbf{\color{#d91a1a}-42.45\%}$
test_tdmodule	40.7410μs	19.2123μs	52.0501 KOps/s	56.7661 KOps/s	$\textbf{\color{#d91a1a}-8.31\%}$
test_tdmodule_dispatch	0.2544ms	34.6188μs	28.8861 KOps/s	31.5588 KOps/s	$\textbf{\color{#d91a1a}-8.47\%}$
test_tdseq	30.9210μs	19.0560μs	52.4769 KOps/s	57.1716 KOps/s	$\textbf{\color{#d91a1a}-8.21\%}$
test_tdseq_dispatch	76.8810μs	36.6970μs	27.2502 KOps/s	29.6596 KOps/s	$\textbf{\color{#d91a1a}-8.12\%}$
test_instantiation_functorch	1.6753ms	1.5453ms	647.1383 Ops/s	651.8740 Ops/s	$\color{#d91a1a}-0.73\%$
test_exec_functorch	0.5304ms	0.1405ms	7.1193 KOps/s	7.1539 KOps/s	$\color{#d91a1a}-0.48\%$
test_exec_functional_call	0.5196ms	0.1297ms	7.7126 KOps/s	7.6200 KOps/s	$\color{#35bf28}+1.21\%$
test_exec_td_decorator	0.3703ms	0.1743ms	5.7357 KOps/s	5.6296 KOps/s	$\color{#35bf28}+1.88\%$
test_vmap_mlp_speed_decorator[True-True]	1.0618ms	0.6607ms	1.5135 KOps/s	1.4917 KOps/s	$\color{#35bf28}+1.46\%$
test_vmap_mlp_speed_decorator[True-False]	1.0517ms	0.6654ms	1.5029 KOps/s	1.4842 KOps/s	$\color{#35bf28}+1.26\%$
test_vmap_mlp_speed_decorator[False-True]	0.9792ms	0.5707ms	1.7522 KOps/s	1.6705 KOps/s	$\color{#35bf28}+4.89\%$
test_vmap_mlp_speed_decorator[False-False]	0.9550ms	0.5722ms	1.7477 KOps/s	1.7329 KOps/s	$\color{#35bf28}+0.85\%$
test_vmap_transformer_speed_decorator[True-True]	18.9059ms	18.5055ms	54.0380 Ops/s	53.7454 Ops/s	$\color{#35bf28}+0.54\%$
test_vmap_transformer_speed_decorator[True-False]	19.4022ms	18.6435ms	53.6380 Ops/s	53.6639 Ops/s	$\color{#d91a1a}-0.05\%$
test_vmap_transformer_speed_decorator[False-True]	19.6330ms	18.5704ms	53.8492 Ops/s	54.2341 Ops/s	$\color{#d91a1a}-0.71\%$
test_vmap_transformer_speed_decorator[False-False]	19.6033ms	18.5605ms	53.8779 Ops/s	54.1558 Ops/s	$\color{#d91a1a}-0.51\%$
test_to_module_speed[True]	1.2586ms	0.9222ms	1.0844 KOps/s	1.0731 KOps/s	$\color{#35bf28}+1.06\%$
test_to_module_speed[False]	1.5378ms	0.9077ms	1.1017 KOps/s	1.0947 KOps/s	$\color{#35bf28}+0.64\%$
test_tc_init	68.1410μs	32.8584μs	30.4336 KOps/s	33.0453 KOps/s	$\textbf{\color{#d91a1a}-7.90\%}$
test_tc_init_nested	0.1279ms	66.9432μs	14.9380 KOps/s	15.7353 KOps/s	$\textbf{\color{#d91a1a}-5.07\%}$
test_tc_first_layer_tensor	14.0187μs	0.6995μs	1.4296 MOps/s	1.4197 MOps/s	$\color{#35bf28}+0.69\%$
test_tc_first_layer_nontensor	89.9510μs	2.3059μs	433.6643 KOps/s	429.2672 KOps/s	$\color{#35bf28}+1.02\%$
test_tc_second_layer_tensor	36.4605μs	1.4325μs	698.0764 KOps/s	694.6352 KOps/s	$\color{#35bf28}+0.50\%$
test_tc_second_layer_nontensor	38.3910μs	3.0855μs	324.0953 KOps/s	327.9530 KOps/s	$\color{#d91a1a}-1.18\%$
test_unbind	0.2390s	9.9683ms	100.3182 Ops/s	149.4197 Ops/s	$\textbf{\color{#d91a1a}-32.86\%}$
test_full_like	9.9667ms	9.5087ms	105.1664 Ops/s	101.8522 Ops/s	$\color{#35bf28}+3.25\%$
test_zeros_like	9.3940ms	7.3176ms	136.6572 Ops/s	233.6041 Ops/s	$\textbf{\color{#d91a1a}-41.50\%}$
test_ones_like	4.9454ms	4.3799ms	228.3173 Ops/s	232.7041 Ops/s	$\color{#d91a1a}-1.89\%$
test_clone	7.1517ms	6.6907ms	149.4621 Ops/s	105.6851 Ops/s	$\textbf{\color{#35bf28}+41.42\%}$
test_squeeze	59.8410μs	9.1742μs	109.0008 KOps/s	109.7053 KOps/s	$\color{#d91a1a}-0.64\%$
test_unsqueeze	0.1210ms	69.0752μs	14.4770 KOps/s	14.8200 KOps/s	$\color{#d91a1a}-2.31\%$
test_split	0.3765ms	0.1546ms	6.4680 KOps/s	6.3746 KOps/s	$\color{#35bf28}+1.46\%$
test_permute	0.2464ms	0.1770ms	5.6489 KOps/s	5.6061 KOps/s	$\color{#35bf28}+0.76\%$
test_stack	52.2387ms	51.7392ms	19.3277 Ops/s	19.4173 Ops/s	$\color{#d91a1a}-0.46\%$
test_cat	52.5280ms	51.6492ms	19.3614 Ops/s	19.4230 Ops/s	$\color{#d91a1a}-0.32\%$