[Feature] allow tensorclass to be customized

[vmoens]

Stack from ghstack (oldest at bottom):

• -> [Feature] allow tensorclass to be customized #1080

[github-actions]

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

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

Expand to view detailed results

Name	Max	Mean	Ops	Ops on Repo HEAD	Change
test_plain_set_nested	50.9850μs	17.5032μs	57.1323 KOps/s	55.2817 KOps/s	$\color{#35bf28}+3.35\%$
test_plain_set_stack_nested	43.2900μs	17.8190μs	56.1200 KOps/s	54.9033 KOps/s	$\color{#35bf28}+2.22\%$
test_plain_set_nested_inplace	80.8100μs	19.3263μs	51.7430 KOps/s	50.0216 KOps/s	$\color{#35bf28}+3.44\%$
test_plain_set_stack_nested_inplace	0.1707ms	19.4080μs	51.5250 KOps/s	50.4927 KOps/s	$\color{#35bf28}+2.04\%$
test_items	48.5220μs	4.1474μs	241.1162 KOps/s	240.2752 KOps/s	$\color{#35bf28}+0.35\%$
test_items_nested	0.5822ms	0.3445ms	2.9031 KOps/s	2.9316 KOps/s	$\color{#d91a1a}-0.97\%$
test_items_nested_locked	0.5242ms	0.3448ms	2.9004 KOps/s	2.9514 KOps/s	$\color{#d91a1a}-1.73\%$
test_items_nested_leaf	0.1478ms	71.2852μs	14.0282 KOps/s	14.1811 KOps/s	$\color{#d91a1a}-1.08\%$
test_items_stack_nested	0.5619ms	0.3483ms	2.8709 KOps/s	2.9263 KOps/s	$\color{#d91a1a}-1.89\%$
test_items_stack_nested_leaf	0.1553ms	74.7213μs	13.3831 KOps/s	13.8487 KOps/s	$\color{#d91a1a}-3.36\%$
test_items_stack_nested_locked	0.6392ms	0.3478ms	2.8753 KOps/s	2.9598 KOps/s	$\color{#d91a1a}-2.86\%$
test_keys	30.3760μs	3.5013μs	285.6067 KOps/s	285.6426 KOps/s	$\color{#d91a1a}-0.01\%$
test_keys_nested	0.2211ms	0.1376ms	7.2686 KOps/s	7.4323 KOps/s	$\color{#d91a1a}-2.20\%$
test_keys_nested_locked	1.9424ms	0.1419ms	7.0472 KOps/s	7.1715 KOps/s	$\color{#d91a1a}-1.73\%$
test_keys_nested_leaf	0.2494ms	0.1150ms	8.6964 KOps/s	8.6117 KOps/s	$\color{#35bf28}+0.98\%$
test_keys_stack_nested	0.2578ms	0.1354ms	7.3840 KOps/s	7.3749 KOps/s	$\color{#35bf28}+0.12\%$
test_keys_stack_nested_leaf	0.2066ms	0.1158ms	8.6332 KOps/s	8.6987 KOps/s	$\color{#d91a1a}-0.75\%$
test_keys_stack_nested_locked	0.4424ms	0.1452ms	6.8881 KOps/s	7.1164 KOps/s	$\color{#d91a1a}-3.21\%$
test_values	9.4696μs	1.0387μs	962.7773 KOps/s	891.1586 KOps/s	$\textbf{\color{#35bf28}+8.04\%}$
test_values_nested	0.1120ms	55.3620μs	18.0629 KOps/s	18.2139 KOps/s	$\color{#d91a1a}-0.83\%$
test_values_nested_locked	0.1032ms	55.6893μs	17.9568 KOps/s	18.2651 KOps/s	$\color{#d91a1a}-1.69\%$
test_values_nested_leaf	0.3163ms	59.8717μs	16.7024 KOps/s	16.8452 KOps/s	$\color{#d91a1a}-0.85\%$
test_values_stack_nested	0.1179ms	56.9341μs	17.5642 KOps/s	17.8085 KOps/s	$\color{#d91a1a}-1.37\%$
test_values_stack_nested_leaf	0.1204ms	59.8490μs	16.7087 KOps/s	16.8668 KOps/s	$\color{#d91a1a}-0.94\%$
test_values_stack_nested_locked	0.1199ms	56.4820μs	17.7048 KOps/s	17.8228 KOps/s	$\color{#d91a1a}-0.66\%$
test_membership	40.2260μs	0.8708μs	1.1483 MOps/s	1.1667 MOps/s	$\color{#d91a1a}-1.58\%$
test_membership_nested	30.6770μs	2.7258μs	366.8632 KOps/s	362.7692 KOps/s	$\color{#35bf28}+1.13\%$
test_membership_nested_leaf	33.1810μs	2.7525μs	363.3115 KOps/s	360.7172 KOps/s	$\color{#35bf28}+0.72\%$
test_membership_stacked_nested	20.5680μs	2.7095μs	369.0761 KOps/s	366.7860 KOps/s	$\color{#35bf28}+0.62\%$
test_membership_stacked_nested_leaf	23.2130μs	2.7352μs	365.6014 KOps/s	348.0464 KOps/s	$\textbf{\color{#35bf28}+5.04\%}$
test_membership_nested_last	45.7450μs	3.9333μs	254.2418 KOps/s	245.6591 KOps/s	$\color{#35bf28}+3.49\%$
test_membership_nested_leaf_last	51.1450μs	3.9730μs	251.6962 KOps/s	243.9172 KOps/s	$\color{#35bf28}+3.19\%$
test_membership_stacked_nested_last	25.6970μs	4.6493μs	215.0872 KOps/s	76.7336 KOps/s	$\textbf{\color{#35bf28}+180.30\%}$
test_membership_stacked_nested_leaf_last	57.0460μs	4.6078μs	217.0228 KOps/s	76.2671 KOps/s	$\textbf{\color{#35bf28}+184.56\%}$
test_nested_getleaf	39.8940μs	10.5476μs	94.8079 KOps/s	94.6177 KOps/s	$\color{#35bf28}+0.20\%$
test_nested_get	57.6770μs	9.9090μs	100.9183 KOps/s	99.9943 KOps/s	$\color{#35bf28}+0.92\%$
test_stacked_getleaf	56.6550μs	10.4121μs	96.0423 KOps/s	94.3861 KOps/s	$\color{#35bf28}+1.75\%$
test_stacked_get	47.0070μs	10.0419μs	99.5825 KOps/s	100.8635 KOps/s	$\color{#d91a1a}-1.27\%$
test_nested_getitemleaf	59.1500μs	11.0907μs	90.1657 KOps/s	91.1051 KOps/s	$\color{#d91a1a}-1.03\%$
test_nested_getitem	62.7970μs	10.3589μs	96.5350 KOps/s	97.9766 KOps/s	$\color{#d91a1a}-1.47\%$
test_stacked_getitemleaf	36.9780μs	10.9846μs	91.0369 KOps/s	91.2714 KOps/s	$\color{#d91a1a}-0.26\%$
test_stacked_getitem	65.9830μs	10.2225μs	97.8230 KOps/s	98.1420 KOps/s	$\color{#d91a1a}-0.33\%$
test_lock_nested	3.0901ms	0.4374ms	2.2862 KOps/s	2.2355 KOps/s	$\color{#35bf28}+2.27\%$
test_lock_stack_nested	0.4980ms	0.4031ms	2.4807 KOps/s	2.4922 KOps/s	$\color{#d91a1a}-0.46\%$
test_unlock_nested	0.8967ms	0.3540ms	2.8248 KOps/s	2.7498 KOps/s	$\color{#35bf28}+2.73\%$
test_unlock_stack_nested	0.5978ms	0.3221ms	3.1048 KOps/s	3.1140 KOps/s	$\color{#d91a1a}-0.30\%$
test_flatten_speed	0.1866ms	90.9702μs	10.9926 KOps/s	10.9403 KOps/s	$\color{#35bf28}+0.48\%$
test_unflatten_speed	0.5352ms	0.4561ms	2.1925 KOps/s	2.0869 KOps/s	$\textbf{\color{#35bf28}+5.06\%}$
test_common_ops	3.8484ms	0.7705ms	1.2979 KOps/s	1.2099 KOps/s	$\textbf{\color{#35bf28}+7.27\%}$
test_creation	57.7280μs	2.0320μs	492.1333 KOps/s	490.2415 KOps/s	$\color{#35bf28}+0.39\%$
test_creation_empty	42.7400μs	10.9658μs	91.1929 KOps/s	84.2569 KOps/s	$\textbf{\color{#35bf28}+8.23\%}$
test_creation_nested_1	45.6540μs	13.4874μs	74.1430 KOps/s	67.2356 KOps/s	$\textbf{\color{#35bf28}+10.27\%}$
test_creation_nested_2	51.5260μs	18.0533μs	55.3916 KOps/s	53.1864 KOps/s	$\color{#35bf28}+4.15\%$
test_clone	1.4621ms	13.2966μs	75.2074 KOps/s	75.3176 KOps/s	$\color{#d91a1a}-0.15\%$
test_getitem[int]	0.8368ms	12.6105μs	79.2990 KOps/s	77.6600 KOps/s	$\color{#35bf28}+2.11\%$
test_getitem[slice_int]	0.1792ms	24.1021μs	41.4901 KOps/s	39.4217 KOps/s	$\textbf{\color{#35bf28}+5.25\%}$
test_getitem[range]	0.2604ms	49.6157μs	20.1549 KOps/s	19.3264 KOps/s	$\color{#35bf28}+4.29\%$
test_getitem[tuple]	0.1377ms	19.7584μs	50.6115 KOps/s	48.3280 KOps/s	$\color{#35bf28}+4.72\%$
test_getitem[list]	0.3663ms	44.4689μs	22.4876 KOps/s	21.1484 KOps/s	$\textbf{\color{#35bf28}+6.33\%}$
test_setitem_dim[int]	53.3390μs	24.5543μs	40.7261 KOps/s	39.6389 KOps/s	$\color{#35bf28}+2.74\%$
test_setitem_dim[slice_int]	81.8520μs	50.3101μs	19.8767 KOps/s	18.7450 KOps/s	$\textbf{\color{#35bf28}+6.04\%}$
test_setitem_dim[range]	0.1247ms	72.6828μs	13.7584 KOps/s	13.0344 KOps/s	$\textbf{\color{#35bf28}+5.55\%}$
test_setitem_dim[tuple]	85.9100μs	39.6057μs	25.2489 KOps/s	23.8056 KOps/s	$\textbf{\color{#35bf28}+6.06\%}$
test_setitem	0.2051ms	20.4519μs	48.8952 KOps/s	47.6164 KOps/s	$\color{#35bf28}+2.69\%$
test_set	0.1605ms	19.9871μs	50.0324 KOps/s	48.4982 KOps/s	$\color{#35bf28}+3.16\%$
test_set_shared	1.2490ms	0.1685ms	5.9337 KOps/s	5.6490 KOps/s	$\textbf{\color{#35bf28}+5.04\%}$
test_update	0.2024ms	22.5405μs	44.3646 KOps/s	42.5637 KOps/s	$\color{#35bf28}+4.23\%$
test_update_nested	0.2284ms	32.2918μs	30.9676 KOps/s	28.8027 KOps/s	$\textbf{\color{#35bf28}+7.52\%}$
test_update__nested	1.1385ms	32.5211μs	30.7493 KOps/s	30.0194 KOps/s	$\color{#35bf28}+2.43\%$
test_set_nested	0.1786ms	21.7243μs	46.0314 KOps/s	43.8074 KOps/s	$\textbf{\color{#35bf28}+5.08\%}$
test_set_nested_new	0.1645ms	26.4662μs	37.7841 KOps/s	36.0417 KOps/s	$\color{#35bf28}+4.83\%$
test_select	0.1612ms	41.5459μs	24.0698 KOps/s	22.6621 KOps/s	$\textbf{\color{#35bf28}+6.21\%}$
test_select_nested	0.1249ms	58.7617μs	17.0179 KOps/s	16.2105 KOps/s	$\color{#35bf28}+4.98\%$
test_exclude_nested	0.1460ms	73.4331μs	13.6178 KOps/s	13.0072 KOps/s	$\color{#35bf28}+4.69\%$
test_empty[True]	0.5817ms	0.3460ms	2.8899 KOps/s	2.8502 KOps/s	$\color{#35bf28}+1.39\%$
test_empty[False]	12.3505μs	1.2037μs	830.8051 KOps/s	823.4668 KOps/s	$\color{#35bf28}+0.89\%$
test_unbind_speed	0.4085ms	0.2553ms	3.9172 KOps/s	3.8359 KOps/s	$\color{#35bf28}+2.12\%$
test_unbind_speed_stack0	0.4150ms	0.2537ms	3.9414 KOps/s	3.9831 KOps/s	$\color{#d91a1a}-1.05\%$
test_unbind_speed_stack1	0.1164s	0.7763ms	1.2881 KOps/s	1.4442 KOps/s	$\textbf{\color{#d91a1a}-10.81\%}$
test_split	1.8042ms	1.5475ms	646.2188 Ops/s	558.0143 Ops/s	$\textbf{\color{#35bf28}+15.81\%}$
test_chunk	0.1127s	1.8883ms	529.5629 Ops/s	560.3730 Ops/s	$\textbf{\color{#d91a1a}-5.50\%}$
test_consolidate_njt[False-None]	11.1333ms	8.0382ms	124.4060 Ops/s	120.5901 Ops/s	$\color{#35bf28}+3.16\%$
test_creation[device0]	0.2419ms	89.9273μs	11.1201 KOps/s	10.3498 KOps/s	$\textbf{\color{#35bf28}+7.44\%}$
test_creation_from_tensor	3.7894ms	94.5355μs	10.5780 KOps/s	10.3983 KOps/s	$\color{#35bf28}+1.73\%$
test_add_one[memmap_tensor0]	0.1929ms	4.7544μs	210.3318 KOps/s	205.0665 KOps/s	$\color{#35bf28}+2.57\%$
test_contiguous[memmap_tensor0]	39.4040μs	0.5077μs	1.9695 MOps/s	1.9635 MOps/s	$\color{#35bf28}+0.31\%$
test_stack[memmap_tensor0]	54.9120μs	3.4160μs	292.7430 KOps/s	279.0151 KOps/s	$\color{#35bf28}+4.92\%$
test_memmaptd_index	1.0297ms	0.2318ms	4.3134 KOps/s	4.2300 KOps/s	$\color{#35bf28}+1.97\%$
test_memmaptd_index_astensor	0.8424ms	0.3077ms	3.2500 KOps/s	3.1607 KOps/s	$\color{#35bf28}+2.82\%$
test_memmaptd_index_op	0.9346ms	0.5647ms	1.7708 KOps/s	1.6186 KOps/s	$\textbf{\color{#35bf28}+9.40\%}$
test_serialize_model	0.1361s	0.1206s	8.2917 Ops/s	8.5808 Ops/s	$\color{#d91a1a}-3.37\%$
test_serialize_model_pickle	0.4489s	0.3920s	2.5512 Ops/s	2.5594 Ops/s	$\color{#d91a1a}-0.32\%$
test_serialize_weights	0.2289s	0.1346s	7.4306 Ops/s	8.6655 Ops/s	$\textbf{\color{#d91a1a}-14.25\%}$
test_serialize_weights_returnearly	0.1660s	0.1588s	6.2955 Ops/s	6.2687 Ops/s	$\color{#35bf28}+0.43\%$
test_serialize_weights_pickle	1.1110s	0.6983s	1.4321 Ops/s	2.2816 Ops/s	$\textbf{\color{#d91a1a}-37.23\%}$
test_serialize_weights_filesystem	0.1532s	0.1448s	6.9053 Ops/s	7.0174 Ops/s	$\color{#d91a1a}-1.60\%$
test_serialize_model_filesystem	0.2566s	0.1619s	6.1756 Ops/s	5.8358 Ops/s	$\textbf{\color{#35bf28}+5.82\%}$
test_reshape_pytree	58.9700μs	26.9622μs	37.0889 KOps/s	37.7935 KOps/s	$\color{#d91a1a}-1.86\%$
test_reshape_td	0.2225ms	34.6983μs	28.8199 KOps/s	30.9375 KOps/s	$\textbf{\color{#d91a1a}-6.84\%}$
test_view_pytree	0.1356ms	27.4967μs	36.3680 KOps/s	37.6193 KOps/s	$\color{#d91a1a}-3.33\%$
test_view_td	76.7220μs	37.6322μs	26.5730 KOps/s	25.9530 KOps/s	$\color{#35bf28}+2.39\%$
test_unbind_pytree	98.6130μs	29.5671μs	33.8214 KOps/s	33.5996 KOps/s	$\color{#35bf28}+0.66\%$
test_unbind_td	0.3300ms	38.3741μs	26.0592 KOps/s	26.1963 KOps/s	$\color{#d91a1a}-0.52\%$
test_split_pytree	65.8020μs	29.8225μs	33.5317 KOps/s	34.0420 KOps/s	$\color{#d91a1a}-1.50\%$
test_split_td	0.4808ms	44.3916μs	22.5268 KOps/s	21.9331 KOps/s	$\color{#35bf28}+2.71\%$
test_add_pytree	90.4080μs	36.4013μs	27.4716 KOps/s	28.2362 KOps/s	$\color{#d91a1a}-2.71\%$
test_add_td	0.1434ms	54.5873μs	18.3193 KOps/s	16.9281 KOps/s	$\textbf{\color{#35bf28}+8.22\%}$
test_compile_add_one_nested[tensordict-compile]	0.1156ms	63.2820μs	15.8023 KOps/s	16.0400 KOps/s	$\color{#d91a1a}-1.48\%$
test_compile_add_one_nested[tensordict-eager]	0.3848ms	0.1606ms	6.2262 KOps/s	6.2840 KOps/s	$\color{#d91a1a}-0.92\%$
test_compile_add_one_nested[pytree-compile]	0.1227ms	45.8266μs	21.8214 KOps/s	22.1030 KOps/s	$\color{#d91a1a}-1.27\%$
test_compile_add_one_nested[pytree-eager]	0.2464ms	0.1193ms	8.3831 KOps/s	8.4043 KOps/s	$\color{#d91a1a}-0.25\%$
test_compile_copy_nested[tensordict-compile]	0.1044ms	26.4928μs	37.7461 KOps/s	39.3624 KOps/s	$\color{#d91a1a}-4.11\%$
test_compile_copy_nested[tensordict-eager]	0.1562ms	53.4490μs	18.7094 KOps/s	18.5705 KOps/s	$\color{#35bf28}+0.75\%$
test_compile_copy_nested[pytree-compile]	0.1793ms	79.4138μs	12.5923 KOps/s	12.8273 KOps/s	$\color{#d91a1a}-1.83\%$
test_compile_copy_nested[pytree-eager]	0.1251ms	67.3308μs	14.8520 KOps/s	14.7881 KOps/s	$\color{#35bf28}+0.43\%$
test_compile_add_one_flat[tensordict-compile]	0.2100ms	0.1035ms	9.6577 KOps/s	9.6788 KOps/s	$\color{#d91a1a}-0.22\%$
test_compile_add_one_flat[tensordict-eager]	0.3339ms	0.1977ms	5.0592 KOps/s	5.0972 KOps/s	$\color{#d91a1a}-0.75\%$
test_compile_add_one_flat[tensorclass-compile]	0.1158ms	44.2935μs	22.5767 KOps/s	22.5124 KOps/s	$\color{#35bf28}+0.29\%$
test_compile_add_one_flat[tensorclass-eager]	0.5218ms	62.9309μs	15.8905 KOps/s	15.9429 KOps/s	$\color{#d91a1a}-0.33\%$
test_compile_add_one_flat[pytree-compile]	0.2136ms	0.1029ms	9.7143 KOps/s	9.8194 KOps/s	$\color{#d91a1a}-1.07\%$
test_compile_add_one_flat[pytree-eager]	0.3077ms	0.2005ms	4.9880 KOps/s	4.7789 KOps/s	$\color{#35bf28}+4.37\%$
test_compile_add_self_flat[tensordict-eager]	0.4466ms	0.2097ms	4.7686 KOps/s	4.8023 KOps/s	$\color{#d91a1a}-0.70\%$
test_compile_add_self_flat[tensordict-compile]	0.2271ms	0.1072ms	9.3272 KOps/s	9.5034 KOps/s	$\color{#d91a1a}-1.85\%$
test_compile_add_self_flat[tensorclass-eager]	0.2640ms	56.3087μs	17.7592 KOps/s	18.2452 KOps/s	$\color{#d91a1a}-2.66\%$
test_compile_add_self_flat[tensorclass-compile]	0.1397ms	45.6547μs	21.9035 KOps/s	22.4508 KOps/s	$\color{#d91a1a}-2.44\%$
test_compile_add_self_flat[pytree-eager]	0.6299ms	0.1588ms	6.2958 KOps/s	6.1796 KOps/s	$\color{#35bf28}+1.88\%$
test_compile_add_self_flat[pytree-compile]	0.2029ms	0.1029ms	9.7165 KOps/s	9.7642 KOps/s	$\color{#d91a1a}-0.49\%$
test_compile_copy_flat[tensordict-compile]	0.1033ms	21.3799μs	46.7729 KOps/s	47.4131 KOps/s	$\color{#d91a1a}-1.35\%$
test_compile_copy_flat[tensordict-eager]	0.1263ms	59.1238μs	16.9137 KOps/s	17.2318 KOps/s	$\color{#d91a1a}-1.85\%$
test_compile_copy_flat[pytree-compile]	0.1635ms	81.6158μs	12.2525 KOps/s	12.5182 KOps/s	$\color{#d91a1a}-2.12\%$
test_compile_copy_flat[pytree-eager]	0.1449ms	69.0438μs	14.4835 KOps/s	13.7806 KOps/s	$\textbf{\color{#35bf28}+5.10\%}$
test_compile_assign_and_add[tensordict-compile]	0.3054ms	0.2105ms	4.7511 KOps/s	4.7951 KOps/s	$\color{#d91a1a}-0.92\%$
test_compile_assign_and_add[tensordict-eager]	1.3981ms	1.2529ms	798.1516 Ops/s	775.6477 Ops/s	$\color{#35bf28}+2.90\%$
test_compile_assign_and_add[pytree-compile]	0.3589ms	0.2114ms	4.7301 KOps/s	4.9229 KOps/s	$\color{#d91a1a}-3.92\%$
test_compile_assign_and_add[pytree-eager]	1.0614ms	0.7688ms	1.3007 KOps/s	1.2837 KOps/s	$\color{#35bf28}+1.33\%$
test_compile_assign_and_add_stack[compile]	0.5761ms	0.4617ms	2.1659 KOps/s	2.2363 KOps/s	$\color{#d91a1a}-3.15\%$
test_compile_assign_and_add_stack[eager]	5.8609ms	2.5897ms	386.1409 Ops/s	367.4728 Ops/s	$\textbf{\color{#35bf28}+5.08\%}$
test_compile_indexing[tensor-tensordict-compile]	0.1022ms	35.7853μs	27.9445 KOps/s	28.5357 KOps/s	$\color{#d91a1a}-2.07\%$
test_compile_indexing[tensor-tensordict-eager]	0.7882ms	34.4907μs	28.9933 KOps/s	29.9899 KOps/s	$\color{#d91a1a}-3.32\%$
test_compile_indexing[tensor-tensorclass-compile]	0.1719ms	29.2060μs	34.2396 KOps/s	34.9533 KOps/s	$\color{#d91a1a}-2.04\%$
test_compile_indexing[tensor-tensorclass-eager]	83.0750μs	23.3096μs	42.9007 KOps/s	44.0134 KOps/s	$\color{#d91a1a}-2.53\%$
test_compile_indexing[tensor-pytree-compile]	77.1440μs	29.8631μs	33.4861 KOps/s	32.8245 KOps/s	$\color{#35bf28}+2.02\%$
test_compile_indexing[tensor-pytree-eager]	68.2470μs	22.8469μs	43.7696 KOps/s	44.0266 KOps/s	$\color{#d91a1a}-0.58\%$
test_compile_indexing[slice-tensordict-compile]	0.1240ms	52.4014μs	19.0835 KOps/s	19.3633 KOps/s	$\color{#d91a1a}-1.45\%$
test_compile_indexing[slice-tensordict-eager]	0.5811ms	19.8605μs	50.3513 KOps/s	49.6803 KOps/s	$\color{#35bf28}+1.35\%$
test_compile_indexing[slice-tensorclass-compile]	0.1089ms	43.7320μs	22.8666 KOps/s	22.1298 KOps/s	$\color{#35bf28}+3.33\%$
test_compile_indexing[slice-tensorclass-eager]	0.1459ms	19.0639μs	52.4552 KOps/s	53.6701 KOps/s	$\color{#d91a1a}-2.26\%$
test_compile_indexing[slice-pytree-compile]	0.1395ms	44.9433μs	22.2503 KOps/s	21.9181 KOps/s	$\color{#35bf28}+1.52\%$
test_compile_indexing[slice-pytree-eager]	0.1189ms	19.0755μs	52.4232 KOps/s	53.7187 KOps/s	$\color{#d91a1a}-2.41\%$
test_compile_indexing[int-tensordict-compile]	0.2112ms	54.6795μs	18.2884 KOps/s	18.9059 KOps/s	$\color{#d91a1a}-3.27\%$
test_compile_indexing[int-tensordict-eager]	5.1502ms	20.1279μs	49.6823 KOps/s	51.4345 KOps/s	$\color{#d91a1a}-3.41\%$
test_compile_indexing[int-tensorclass-compile]	92.3620μs	44.3050μs	22.5708 KOps/s	21.8333 KOps/s	$\color{#35bf28}+3.38\%$
test_compile_indexing[int-tensorclass-eager]	75.5610μs	18.9692μs	52.7169 KOps/s	53.3008 KOps/s	$\color{#d91a1a}-1.10\%$
test_compile_indexing[int-pytree-compile]	0.1356ms	44.4587μs	22.4928 KOps/s	21.9228 KOps/s	$\color{#35bf28}+2.60\%$
test_compile_indexing[int-pytree-eager]	50.7950μs	18.7645μs	53.2921 KOps/s	53.8953 KOps/s	$\color{#d91a1a}-1.12\%$
test_mod_add[eager]	69.0990μs	25.7486μs	38.8371 KOps/s	36.1039 KOps/s	$\textbf{\color{#35bf28}+7.57\%}$
test_mod_add[compile]	0.1121ms	44.6154μs	22.4138 KOps/s	23.2907 KOps/s	$\color{#d91a1a}-3.77\%$
test_mod_add[compile-overhead]	0.1220ms	44.7882μs	22.3273 KOps/s	22.8674 KOps/s	$\color{#d91a1a}-2.36\%$
test_mod_wrap[eager]	0.3430ms	0.2148ms	4.6554 KOps/s	4.6247 KOps/s	$\color{#35bf28}+0.66\%$
test_mod_wrap[compile]	2.1776ms	0.2034ms	4.9167 KOps/s	4.8615 KOps/s	$\color{#35bf28}+1.13\%$
test_mod_wrap[compile-overhead]	1.8340ms	0.2030ms	4.9262 KOps/s	4.8688 KOps/s	$\color{#35bf28}+1.18\%$
test_mod_wrap_and_backward[eager]	14.2195ms	11.8695ms	84.2498 Ops/s	77.1237 Ops/s	$\textbf{\color{#35bf28}+9.24\%}$
test_mod_wrap_and_backward[compile]	15.1752ms	12.9415ms	77.2707 Ops/s	77.1599 Ops/s	$\color{#35bf28}+0.14\%$
test_mod_wrap_and_backward[compile-overhead]	15.0021ms	12.9863ms	77.0043 Ops/s	75.4662 Ops/s	$\color{#35bf28}+2.04\%$
test_seq_add[eager]	0.1578ms	89.9240μs	11.1205 KOps/s	10.4878 KOps/s	$\textbf{\color{#35bf28}+6.03\%}$
test_seq_add[compile]	0.1301ms	59.7788μs	16.7283 KOps/s	16.6746 KOps/s	$\color{#35bf28}+0.32\%$
test_seq_add[compile-overhead]	0.1248ms	58.3501μs	17.1379 KOps/s	17.1891 KOps/s	$\color{#d91a1a}-0.30\%$
test_seq_wrap[eager]	0.5909ms	0.3829ms	2.6116 KOps/s	2.4273 KOps/s	$\textbf{\color{#35bf28}+7.59\%}$
test_seq_wrap[compile]	0.4228ms	0.2278ms	4.3904 KOps/s	4.3962 KOps/s	$\color{#d91a1a}-0.13\%$
test_seq_wrap[compile-overhead]	0.4337ms	0.2254ms	4.4371 KOps/s	4.4283 KOps/s	$\color{#35bf28}+0.20\%$
test_func_call_runtime[False-eager]	0.7717ms	0.5534ms	1.8071 KOps/s	1.7582 KOps/s	$\color{#35bf28}+2.78\%$
test_func_call_runtime[False-compile]	0.6395ms	0.4209ms	2.3759 KOps/s	2.3466 KOps/s	$\color{#35bf28}+1.25\%$
test_func_call_runtime[False-compile-overhead]	0.5829ms	0.4234ms	2.3617 KOps/s	2.3428 KOps/s	$\color{#35bf28}+0.81\%$
test_func_call_runtime[True-eager]	1.1218ms	0.7609ms	1.3143 KOps/s	1.3210 KOps/s	$\color{#d91a1a}-0.51\%$
test_func_call_runtime[True-compile]	0.8148ms	0.4609ms	2.1695 KOps/s	2.1602 KOps/s	$\color{#35bf28}+0.43\%$
test_func_call_runtime[True-compile-overhead]	0.6190ms	0.4630ms	2.1600 KOps/s	2.1530 KOps/s	$\color{#35bf28}+0.33\%$
test_func_call_cm_runtime[False-eager]	1.2039ms	0.5624ms	1.7780 KOps/s	1.8268 KOps/s	$\color{#d91a1a}-2.67\%$
test_func_call_cm_runtime[False-compile]	0.5780ms	0.4244ms	2.3565 KOps/s	2.3550 KOps/s	$\color{#35bf28}+0.07\%$
test_func_call_cm_runtime[False-compile-overhead]	0.9299ms	0.4218ms	2.3705 KOps/s	2.3489 KOps/s	$\color{#35bf28}+0.92\%$
test_func_call_cm_runtime[True-eager]	1.0570ms	0.9040ms	1.1062 KOps/s	1.1059 KOps/s	$\color{#35bf28}+0.03\%$
test_func_call_cm_runtime[True-compile]	1.2313ms	0.4875ms	2.0512 KOps/s	2.0306 KOps/s	$\color{#35bf28}+1.02\%$
test_func_call_cm_runtime[True-compile-overhead]	0.7261ms	0.4886ms	2.0467 KOps/s	2.0284 KOps/s	$\color{#35bf28}+0.90\%$
test_vmap_func_call_cm_runtime[eager]	2.4282ms	1.8823ms	531.2646 Ops/s	517.3762 Ops/s	$\color{#35bf28}+2.68\%$
test_vmap_func_call_cm_runtime[compile]	0.6974ms	0.5088ms	1.9654 KOps/s	1.9329 KOps/s	$\color{#35bf28}+1.68\%$
test_vmap_func_call_cm_runtime[compile-overhead]	0.6138ms	0.5056ms	1.9777 KOps/s	1.9429 KOps/s	$\color{#35bf28}+1.79\%$
test_distributed	0.3392ms	0.1268ms	7.8861 KOps/s	7.7617 KOps/s	$\color{#35bf28}+1.60\%$
test_tdmodule	44.6740μs	18.2764μs	54.7153 KOps/s	50.3696 KOps/s	$\textbf{\color{#35bf28}+8.63\%}$
test_tdmodule_dispatch	66.5140μs	35.7200μs	27.9955 KOps/s	26.3081 KOps/s	$\textbf{\color{#35bf28}+6.41\%}$
test_tdseq	46.1960μs	21.0099μs	47.5965 KOps/s	44.1430 KOps/s	$\textbf{\color{#35bf28}+7.82\%}$
test_tdseq_dispatch	76.5120μs	41.6109μs	24.0321 KOps/s	23.6039 KOps/s	$\color{#35bf28}+1.81\%$
test_instantiation_functorch	2.1208ms	1.5368ms	650.6954 Ops/s	657.6047 Ops/s	$\color{#d91a1a}-1.05\%$
test_exec_functorch	0.3456ms	0.1835ms	5.4505 KOps/s	5.4662 KOps/s	$\color{#d91a1a}-0.29\%$
test_exec_functional_call	0.3357ms	0.1765ms	5.6663 KOps/s	5.6692 KOps/s	$\color{#d91a1a}-0.05\%$
test_exec_td_decorator	0.5408ms	0.2305ms	4.3380 KOps/s	4.3457 KOps/s	$\color{#d91a1a}-0.18\%$
test_vmap_mlp_speed_decorator[True-True]	0.7989ms	0.6307ms	1.5855 KOps/s	1.5431 KOps/s	$\color{#35bf28}+2.75\%$
test_vmap_mlp_speed_decorator[True-False]	0.9712ms	0.6317ms	1.5830 KOps/s	1.5380 KOps/s	$\color{#35bf28}+2.93\%$
test_vmap_mlp_speed_decorator[False-True]	0.7638ms	0.5172ms	1.9337 KOps/s	1.8927 KOps/s	$\color{#35bf28}+2.16\%$
test_vmap_mlp_speed_decorator[False-False]	0.7328ms	0.5175ms	1.9322 KOps/s	1.8466 KOps/s	$\color{#35bf28}+4.63\%$
test_to_module_speed[True]	1.5510ms	1.2911ms	774.5616 Ops/s	784.8408 Ops/s	$\color{#d91a1a}-1.31\%$
test_to_module_speed[False]	1.6625ms	1.2625ms	792.0681 Ops/s	797.8408 Ops/s	$\color{#d91a1a}-0.72\%$
test_tc_init	89.1050μs	45.0352μs	22.2049 KOps/s	21.3839 KOps/s	$\color{#35bf28}+3.84\%$
test_tc_init_nested	0.1702ms	91.0522μs	10.9827 KOps/s	10.7106 KOps/s	$\color{#35bf28}+2.54\%$
test_tc_first_layer_tensor	40.6960μs	1.5284μs	654.2715 KOps/s	661.2306 KOps/s	$\color{#d91a1a}-1.05\%$
test_tc_first_layer_nontensor	28.7130μs	4.7284μs	211.4890 KOps/s	215.6634 KOps/s	$\color{#d91a1a}-1.94\%$
test_tc_second_layer_tensor	42.6590μs	2.8206μs	354.5295 KOps/s	356.5423 KOps/s	$\color{#d91a1a}-0.56\%$
test_tc_second_layer_nontensor	31.2480μs	6.0308μs	165.8142 KOps/s	169.6419 KOps/s	$\color{#d91a1a}-2.26\%$
test_unbind	0.2403s	14.1888ms	70.4782 Ops/s	77.4163 Ops/s	$\textbf{\color{#d91a1a}-8.96\%}$
test_full_like	9.9245ms	8.2730ms	120.8746 Ops/s	78.9645 Ops/s	$\textbf{\color{#35bf28}+53.07\%}$
test_zeros_like	3.6937ms	3.1546ms	316.9993 Ops/s	122.2463 Ops/s	$\textbf{\color{#35bf28}+159.31\%}$
test_ones_like	4.1287ms	3.5831ms	279.0913 Ops/s	123.3096 Ops/s	$\textbf{\color{#35bf28}+126.33\%}$
test_clone	6.6814ms	5.9049ms	169.3519 Ops/s	101.2904 Ops/s	$\textbf{\color{#35bf28}+67.19\%}$
test_squeeze	63.3380μs	11.9947μs	83.3699 KOps/s	82.3051 KOps/s	$\color{#35bf28}+1.29\%$
test_unsqueeze	0.1656ms	88.9454μs	11.2429 KOps/s	11.1270 KOps/s	$\color{#35bf28}+1.04\%$
test_split	0.4837ms	0.1846ms	5.4158 KOps/s	5.2027 KOps/s	$\color{#35bf28}+4.10\%$
test_permute	0.3142ms	0.2172ms	4.6036 KOps/s	4.5547 KOps/s	$\color{#35bf28}+1.07\%$
test_stack	30.3634ms	26.1641ms	38.2202 Ops/s	38.2684 Ops/s	$\color{#d91a1a}-0.13\%$
test_cat	33.6727ms	26.0917ms	38.3264 Ops/s	38.4214 Ops/s	$\color{#d91a1a}-0.25\%$

[github-actions]

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

Total Benchmarks: 229. Improved: $\large\color{#35bf28}22$. Worsened: $\large\color{#d91a1a}10$.

Expand to view detailed results

Name	Max	Mean	Ops	Ops on Repo HEAD	Change
test_plain_set_nested	30.6400μs	11.0298μs	90.6637 KOps/s	85.2034 KOps/s	$\textbf{\color{#35bf28}+6.41\%}$
test_plain_set_stack_nested	39.4110μs	11.1360μs	89.7990 KOps/s	86.1697 KOps/s	$\color{#35bf28}+4.21\%$
test_plain_set_nested_inplace	0.4038ms	12.0047μs	83.3008 KOps/s	80.0803 KOps/s	$\color{#35bf28}+4.02\%$
test_plain_set_stack_nested_inplace	0.3903ms	12.0142μs	83.2347 KOps/s	79.7488 KOps/s	$\color{#35bf28}+4.37\%$
test_items	21.1010μs	2.8571μs	350.0048 KOps/s	340.0693 KOps/s	$\color{#35bf28}+2.92\%$
test_items_nested	0.6935ms	0.3206ms	3.1193 KOps/s	3.1486 KOps/s	$\color{#d91a1a}-0.93\%$
test_items_nested_locked	0.6949ms	0.3220ms	3.1055 KOps/s	3.1466 KOps/s	$\color{#d91a1a}-1.31\%$
test_items_nested_leaf	0.4340ms	58.3381μs	17.1415 KOps/s	17.1229 KOps/s	$\color{#35bf28}+0.11\%$
test_items_stack_nested	0.7060ms	0.3201ms	3.1237 KOps/s	3.1197 KOps/s	$\color{#35bf28}+0.13\%$
test_items_stack_nested_leaf	0.4302ms	59.0322μs	16.9399 KOps/s	16.9439 KOps/s	$\color{#d91a1a}-0.02\%$
test_items_stack_nested_locked	0.7169ms	0.3233ms	3.0928 KOps/s	3.1258 KOps/s	$\color{#d91a1a}-1.06\%$
test_keys	32.0200μs	3.4614μs	288.8991 KOps/s	288.6467 KOps/s	$\color{#35bf28}+0.09\%$
test_keys_nested	0.4583ms	69.3884μs	14.4116 KOps/s	14.3150 KOps/s	$\color{#35bf28}+0.67\%$
test_keys_nested_locked	0.7627ms	74.9163μs	13.3482 KOps/s	13.2147 KOps/s	$\color{#35bf28}+1.01\%$
test_keys_nested_leaf	0.4324ms	62.0190μs	16.1241 KOps/s	16.3482 KOps/s	$\color{#d91a1a}-1.37\%$
test_keys_stack_nested	0.1257ms	70.3367μs	14.2173 KOps/s	13.9700 KOps/s	$\color{#35bf28}+1.77\%$
test_keys_stack_nested_leaf	0.4362ms	61.6028μs	16.2330 KOps/s	16.1622 KOps/s	$\color{#35bf28}+0.44\%$
test_keys_stack_nested_locked	0.4499ms	76.0355μs	13.1518 KOps/s	13.0968 KOps/s	$\color{#35bf28}+0.42\%$
test_values	65.4060μs	0.8420μs	1.1876 MOps/s	1.1829 MOps/s	$\color{#35bf28}+0.40\%$
test_values_nested	66.4110μs	31.3467μs	31.9013 KOps/s	32.0224 KOps/s	$\color{#d91a1a}-0.38\%$
test_values_nested_locked	60.5500μs	32.9403μs	30.3580 KOps/s	30.4742 KOps/s	$\color{#d91a1a}-0.38\%$
test_values_nested_leaf	54.5510μs	33.7122μs	29.6628 KOps/s	29.7106 KOps/s	$\color{#d91a1a}-0.16\%$
test_values_stack_nested	59.6710μs	31.7395μs	31.5064 KOps/s	31.6531 KOps/s	$\color{#d91a1a}-0.46\%$
test_values_stack_nested_leaf	75.1910μs	34.2246μs	29.2187 KOps/s	29.0778 KOps/s	$\color{#35bf28}+0.48\%$
test_values_stack_nested_locked	59.7510μs	33.5109μs	29.8411 KOps/s	30.2148 KOps/s	$\color{#d91a1a}-1.24\%$
test_membership	1.6166μs	0.5197μs	1.9241 MOps/s	1.9263 MOps/s	$\color{#d91a1a}-0.12\%$
test_membership_nested	13.3800μs	1.8628μs	536.8168 KOps/s	543.1905 KOps/s	$\color{#d91a1a}-1.17\%$
test_membership_nested_leaf	11.8770μs	1.8172μs	550.3097 KOps/s	536.3013 KOps/s	$\color{#35bf28}+2.61\%$
test_membership_stacked_nested	35.3210μs	1.9490μs	513.0874 KOps/s	514.0016 KOps/s	$\color{#d91a1a}-0.18\%$
test_membership_stacked_nested_leaf	30.4600μs	1.9520μs	512.2882 KOps/s	522.6111 KOps/s	$\color{#d91a1a}-1.98\%$
test_membership_nested_last	38.0800μs	2.7633μs	361.8868 KOps/s	364.8927 KOps/s	$\color{#d91a1a}-0.82\%$
test_membership_nested_leaf_last	25.3900μs	2.7917μs	358.2104 KOps/s	361.2259 KOps/s	$\color{#d91a1a}-0.83\%$
test_membership_stacked_nested_last	42.4010μs	3.3472μs	298.7567 KOps/s	268.1016 KOps/s	$\textbf{\color{#35bf28}+11.43\%}$
test_membership_stacked_nested_leaf_last	0.7785ms	3.1759μs	314.8701 KOps/s	269.8069 KOps/s	$\textbf{\color{#35bf28}+16.70\%}$
test_nested_getleaf	34.2910μs	5.9435μs	168.2521 KOps/s	167.4980 KOps/s	$\color{#35bf28}+0.45\%$
test_nested_get	28.4500μs	5.6376μs	177.3818 KOps/s	176.4344 KOps/s	$\color{#35bf28}+0.54\%$
test_stacked_getleaf	30.8100μs	5.9563μs	167.8902 KOps/s	167.0318 KOps/s	$\color{#35bf28}+0.51\%$
test_stacked_get	28.2210μs	5.6145μs	178.1097 KOps/s	176.6182 KOps/s	$\color{#35bf28}+0.84\%$
test_nested_getitemleaf	32.8210μs	6.0417μs	165.5162 KOps/s	164.4359 KOps/s	$\color{#35bf28}+0.66\%$
test_nested_getitem	30.0700μs	5.7267μs	174.6207 KOps/s	174.4036 KOps/s	$\color{#35bf28}+0.12\%$
test_stacked_getitemleaf	31.3500μs	6.0357μs	165.6811 KOps/s	164.5472 KOps/s	$\color{#35bf28}+0.69\%$
test_stacked_getitem	26.3410μs	5.7269μs	174.6135 KOps/s	174.9187 KOps/s	$\color{#d91a1a}-0.17\%$
test_lock_nested	9.1584ms	0.3672ms	2.7234 KOps/s	2.7080 KOps/s	$\color{#35bf28}+0.57\%$
test_lock_stack_nested	0.3764ms	0.3302ms	3.0289 KOps/s	2.9893 KOps/s	$\color{#35bf28}+1.32\%$
test_unlock_nested	0.6611ms	0.3024ms	3.3073 KOps/s	3.3597 KOps/s	$\color{#d91a1a}-1.56\%$
test_unlock_stack_nested	0.3154ms	0.2714ms	3.6848 KOps/s	3.6983 KOps/s	$\color{#d91a1a}-0.37\%$
test_flatten_speed	0.1046ms	73.3030μs	13.6420 KOps/s	13.7461 KOps/s	$\color{#d91a1a}-0.76\%$
test_unflatten_speed	0.3377ms	0.2873ms	3.4812 KOps/s	3.4272 KOps/s	$\color{#35bf28}+1.58\%$
test_common_ops	1.7584ms	0.5892ms	1.6971 KOps/s	1.6643 KOps/s	$\color{#35bf28}+1.98\%$
test_creation	0.1204ms	1.4929μs	669.8547 KOps/s	676.6796 KOps/s	$\color{#d91a1a}-1.01\%$
test_creation_empty	39.2410μs	8.2591μs	121.0787 KOps/s	104.4622 KOps/s	$\textbf{\color{#35bf28}+15.91\%}$
test_creation_nested_1	1.6593ms	9.6878μs	103.2221 KOps/s	90.6371 KOps/s	$\textbf{\color{#35bf28}+13.89\%}$
test_creation_nested_2	48.8510μs	12.1293μs	82.4450 KOps/s	75.2270 KOps/s	$\textbf{\color{#35bf28}+9.59\%}$
test_clone	58.4610μs	10.5414μs	94.8642 KOps/s	99.2059 KOps/s	$\color{#d91a1a}-4.38\%$
test_getitem[int]	92.7821ms	15.2685μs	65.4943 KOps/s	97.5402 KOps/s	$\textbf{\color{#d91a1a}-32.85\%}$
test_getitem[slice_int]	0.1103ms	20.2091μs	49.4827 KOps/s	48.9503 KOps/s	$\color{#35bf28}+1.09\%$
test_getitem[range]	0.1321ms	35.9340μs	27.8288 KOps/s	28.1575 KOps/s	$\color{#d91a1a}-1.17\%$
test_getitem[tuple]	0.1045ms	17.6389μs	56.6928 KOps/s	56.3888 KOps/s	$\color{#35bf28}+0.54\%$
test_getitem[list]	0.2336ms	31.8923μs	31.3555 KOps/s	31.6371 KOps/s	$\color{#d91a1a}-0.89\%$
test_setitem_dim[int]	39.4010μs	18.3122μs	54.6084 KOps/s	56.3104 KOps/s	$\color{#d91a1a}-3.02\%$
test_setitem_dim[slice_int]	97.3710μs	35.1315μs	28.4645 KOps/s	27.9596 KOps/s	$\color{#35bf28}+1.81\%$
test_setitem_dim[range]	82.4210μs	50.4190μs	19.8338 KOps/s	19.8001 KOps/s	$\color{#35bf28}+0.17\%$
test_setitem_dim[tuple]	51.2910μs	30.9215μs	32.3400 KOps/s	32.7637 KOps/s	$\color{#d91a1a}-1.29\%$
test_setitem	94.1020μs	14.7527μs	67.7842 KOps/s	65.1968 KOps/s	$\color{#35bf28}+3.97\%$
test_set	89.2510μs	14.3294μs	69.7868 KOps/s	68.7660 KOps/s	$\color{#35bf28}+1.48\%$
test_set_shared	1.5957ms	0.1459ms	6.8529 KOps/s	6.9214 KOps/s	$\color{#d91a1a}-0.99\%$
test_update	0.9957ms	17.3612μs	57.5997 KOps/s	53.4434 KOps/s	$\textbf{\color{#35bf28}+7.78\%}$
test_update_nested	95.2810μs	22.7634μs	43.9302 KOps/s	42.4510 KOps/s	$\color{#35bf28}+3.48\%$
test_update__nested	0.1271ms	23.9557μs	41.7438 KOps/s	42.9048 KOps/s	$\color{#d91a1a}-2.71\%$
test_set_nested	85.7610μs	15.1247μs	66.1168 KOps/s	62.9976 KOps/s	$\color{#35bf28}+4.95\%$
test_set_nested_new	94.3420μs	18.1093μs	55.2202 KOps/s	56.3168 KOps/s	$\color{#d91a1a}-1.95\%$
test_select	96.1210μs	29.8036μs	33.5530 KOps/s	34.1975 KOps/s	$\color{#d91a1a}-1.88\%$
test_select_nested	67.4910μs	42.3229μs	23.6279 KOps/s	23.6845 KOps/s	$\color{#d91a1a}-0.24\%$
test_exclude_nested	93.5420μs	57.4827μs	17.3965 KOps/s	17.1451 KOps/s	$\color{#35bf28}+1.47\%$
test_empty[True]	0.3032ms	0.2521ms	3.9671 KOps/s	3.9792 KOps/s	$\color{#d91a1a}-0.30\%$
test_empty[False]	3.4341μs	0.7678μs	1.3025 MOps/s	1.3047 MOps/s	$\color{#d91a1a}-0.17\%$
test_to	85.3120μs	55.0423μs	18.1679 KOps/s	17.8232 KOps/s	$\color{#35bf28}+1.93\%$
test_to_nonblocking	78.3310μs	45.5885μs	21.9353 KOps/s	22.0028 KOps/s	$\color{#d91a1a}-0.31\%$
test_unbind_speed	1.5721ms	0.2300ms	4.3479 KOps/s	4.3151 KOps/s	$\color{#35bf28}+0.76\%$
test_unbind_speed_stack0	0.3073ms	0.2287ms	4.3725 KOps/s	4.3083 KOps/s	$\color{#35bf28}+1.49\%$
test_unbind_speed_stack1	93.0841ms	0.6513ms	1.5354 KOps/s	1.5319 KOps/s	$\color{#35bf28}+0.23\%$
test_split	94.8592ms	1.7011ms	587.8417 Ops/s	642.0571 Ops/s	$\textbf{\color{#d91a1a}-8.44\%}$
test_chunk	1.5476ms	1.4353ms	696.7122 Ops/s	587.4610 Ops/s	$\textbf{\color{#35bf28}+18.60\%}$
test_consolidate[False-None]	97.4198ms	2.7751ms	360.3420 Ops/s	394.1560 Ops/s	$\textbf{\color{#d91a1a}-8.58\%}$
test_consolidate[default-None]	1.7908ms	1.5871ms	630.0840 Ops/s	621.8761 Ops/s	$\color{#35bf28}+1.32\%$
test_consolidate[reduce-overhead-None]	1.8009ms	1.6216ms	616.6635 Ops/s	612.2369 Ops/s	$\color{#35bf28}+0.72\%$
test_consolidate_njt[False-None]	6.8788ms	6.3364ms	157.8193 Ops/s	112.3847 Ops/s	$\textbf{\color{#35bf28}+40.43\%}$
test_to[False-False-None]	1.8577ms	1.7237ms	580.1399 Ops/s	586.5867 Ops/s	$\color{#d91a1a}-1.10\%$
test_to[True-False-None]	1.4001ms	1.2359ms	809.1050 Ops/s	823.0252 Ops/s	$\color{#d91a1a}-1.69\%$
test_to[within-False-None]	0.3381s	5.2388ms	190.8841 Ops/s	258.6629 Ops/s	$\textbf{\color{#d91a1a}-26.20\%}$
test_to[True-default-None]	5.3881ms	5.0071ms	199.7167 Ops/s	197.3797 Ops/s	$\color{#35bf28}+1.18\%$
test_to_njt[False-False-None]	7.0621ms	6.7549ms	148.0417 Ops/s	143.4625 Ops/s	$\color{#35bf28}+3.19\%$
test_to_njt[True-False-None]	5.8423ms	5.4371ms	183.9226 Ops/s	179.1999 Ops/s	$\color{#35bf28}+2.64\%$
test_to_njt[within-False-None]	12.1316ms	11.7501ms	85.1060 Ops/s	82.3208 Ops/s	$\color{#35bf28}+3.38\%$
test_creation[device0]	0.4651ms	81.2486μs	12.3079 KOps/s	12.7789 KOps/s	$\color{#d91a1a}-3.69\%$
test_creation_from_tensor	0.4926ms	84.5091μs	11.8331 KOps/s	12.2099 KOps/s	$\color{#d91a1a}-3.09\%$
test_add_one[memmap_tensor0]	0.4607ms	7.0170μs	142.5116 KOps/s	145.7487 KOps/s	$\color{#d91a1a}-2.22\%$
test_contiguous[memmap_tensor0]	2.7345μs	0.3873μs	2.5821 MOps/s	2.5404 MOps/s	$\color{#35bf28}+1.64\%$
test_stack[memmap_tensor0]	31.1410μs	4.2314μs	236.3273 KOps/s	232.9595 KOps/s	$\color{#35bf28}+1.45\%$
test_memmaptd_index	1.8664ms	0.2407ms	4.1541 KOps/s	4.1416 KOps/s	$\color{#35bf28}+0.30\%$
test_memmaptd_index_astensor	0.8253ms	0.3004ms	3.3286 KOps/s	3.3207 KOps/s	$\color{#35bf28}+0.24\%$
test_memmaptd_index_op	0.9988ms	0.5815ms	1.7198 KOps/s	1.6616 KOps/s	$\color{#35bf28}+3.50\%$
test_serialize_model	0.1317s	0.1306s	7.6574 Ops/s	7.6188 Ops/s	$\color{#35bf28}+0.51\%$
test_serialize_model_pickle	1.3479s	1.1851s	0.8438 Ops/s	0.8212 Ops/s	$\color{#35bf28}+2.75\%$
test_serialize_weights	0.1319s	0.1304s	7.6677 Ops/s	7.6153 Ops/s	$\color{#35bf28}+0.69\%$
test_serialize_weights_returnearly	0.3774s	67.0612ms	14.9117 Ops/s	10.5919 Ops/s	$\textbf{\color{#35bf28}+40.78\%}$
test_serialize_weights_pickle	1.4745s	1.2327s	0.8112 Ops/s	0.8387 Ops/s	$\color{#d91a1a}-3.28\%$
test_reshape_pytree	53.8510μs	22.3262μs	44.7905 KOps/s	45.5258 KOps/s	$\color{#d91a1a}-1.62\%$
test_reshape_td	52.6210μs	26.6260μs	37.5572 KOps/s	38.0967 KOps/s	$\color{#d91a1a}-1.42\%$
test_view_pytree	55.7210μs	22.2527μs	44.9383 KOps/s	45.7024 KOps/s	$\color{#d91a1a}-1.67\%$
test_view_td	62.1410μs	29.3372μs	34.0864 KOps/s	33.3265 KOps/s	$\color{#35bf28}+2.28\%$
test_unbind_pytree	59.2810μs	28.4804μs	35.1119 KOps/s	35.2717 KOps/s	$\color{#d91a1a}-0.45\%$
test_unbind_td	0.6355ms	35.2031μs	28.4066 KOps/s	28.0174 KOps/s	$\color{#35bf28}+1.39\%$
test_split_pytree	62.4010μs	30.0955μs	33.2276 KOps/s	33.4044 KOps/s	$\color{#d91a1a}-0.53\%$
test_split_td	0.7949ms	38.1611μs	26.2047 KOps/s	25.5353 KOps/s	$\color{#35bf28}+2.62\%$
test_add_pytree	67.4910μs	35.2577μs	28.3626 KOps/s	28.9313 KOps/s	$\color{#d91a1a}-1.97\%$
test_add_td	90.1110μs	47.9926μs	20.8365 KOps/s	20.4727 KOps/s	$\color{#35bf28}+1.78\%$
test_compile_add_one_nested[tensordict-compile]	0.1807ms	0.1177ms	8.4994 KOps/s	8.2701 KOps/s	$\color{#35bf28}+2.77\%$
test_compile_add_one_nested[tensordict-eager]	0.2182ms	0.1257ms	7.9583 KOps/s	7.8870 KOps/s	$\color{#35bf28}+0.90\%$
test_compile_add_one_nested[pytree-compile]	0.1327ms	94.6304μs	10.5674 KOps/s	9.9062 KOps/s	$\textbf{\color{#35bf28}+6.68\%}$
test_compile_add_one_nested[pytree-eager]	1.1191ms	0.1560ms	6.4087 KOps/s	6.3228 KOps/s	$\color{#35bf28}+1.36\%$
test_compile_copy_nested[tensordict-compile]	55.8010μs	22.5517μs	44.3426 KOps/s	34.3122 KOps/s	$\textbf{\color{#35bf28}+29.23\%}$
test_compile_copy_nested[tensordict-eager]	58.7410μs	26.3896μs	37.8937 KOps/s	37.6086 KOps/s	$\color{#35bf28}+0.76\%$
test_compile_copy_nested[pytree-compile]	0.2375ms	65.5916μs	15.2459 KOps/s	14.9417 KOps/s	$\color{#35bf28}+2.04\%$
test_compile_copy_nested[pytree-eager]	82.8510μs	50.0644μs	19.9743 KOps/s	19.5778 KOps/s	$\color{#35bf28}+2.03\%$
test_compile_add_one_flat[tensordict-compile]	0.2019ms	0.1421ms	7.0398 KOps/s	7.1238 KOps/s	$\color{#d91a1a}-1.18\%$
test_compile_add_one_flat[tensordict-eager]	0.2998ms	0.2084ms	4.7986 KOps/s	4.8020 KOps/s	$\color{#d91a1a}-0.07\%$
test_compile_add_one_flat[tensorclass-compile]	0.1481ms	97.5700μs	10.2491 KOps/s	10.3723 KOps/s	$\color{#d91a1a}-1.19\%$
test_compile_add_one_flat[tensorclass-eager]	0.1041ms	51.3372μs	19.4790 KOps/s	18.3600 KOps/s	$\textbf{\color{#35bf28}+6.09\%}$
test_compile_add_one_flat[pytree-compile]	0.1752ms	0.1368ms	7.3076 KOps/s	6.9662 KOps/s	$\color{#35bf28}+4.90\%$
test_compile_add_one_flat[pytree-eager]	0.5665ms	0.5049ms	1.9804 KOps/s	1.9475 KOps/s	$\color{#35bf28}+1.69\%$
test_compile_add_self_flat[tensordict-eager]	0.3596ms	0.2460ms	4.0645 KOps/s	3.9477 KOps/s	$\color{#35bf28}+2.96\%$
test_compile_add_self_flat[tensordict-compile]	0.1865ms	0.1428ms	7.0006 KOps/s	6.6627 KOps/s	$\textbf{\color{#35bf28}+5.07\%}$
test_compile_add_self_flat[tensorclass-eager]	0.1423ms	60.6259μs	16.4946 KOps/s	15.5607 KOps/s	$\textbf{\color{#35bf28}+6.00\%}$
test_compile_add_self_flat[tensorclass-compile]	0.1463ms	97.9934μs	10.2048 KOps/s	9.6264 KOps/s	$\textbf{\color{#35bf28}+6.01\%}$
test_compile_add_self_flat[pytree-eager]	0.4836ms	0.4157ms	2.4058 KOps/s	2.3045 KOps/s	$\color{#35bf28}+4.39\%$
test_compile_add_self_flat[pytree-compile]	0.1928ms	0.1392ms	7.1818 KOps/s	7.4846 KOps/s	$\color{#d91a1a}-4.05\%$
test_compile_copy_flat[tensordict-compile]	76.2610μs	18.8551μs	53.0361 KOps/s	53.9262 KOps/s	$\color{#d91a1a}-1.65\%$
test_compile_copy_flat[tensordict-eager]	65.6010μs	27.3366μs	36.5810 KOps/s	36.3626 KOps/s	$\color{#35bf28}+0.60\%$
test_compile_copy_flat[pytree-compile]	0.1025ms	68.7621μs	14.5429 KOps/s	14.5463 KOps/s	$\color{#d91a1a}-0.02\%$
test_compile_copy_flat[pytree-eager]	0.1369ms	51.2909μs	19.4966 KOps/s	19.3382 KOps/s	$\color{#35bf28}+0.82\%$
test_compile_assign_and_add[tensordict-compile]	1.5697ms	0.4333ms	2.3079 KOps/s	2.2844 KOps/s	$\color{#35bf28}+1.03\%$
test_compile_assign_and_add[tensordict-eager]	2.7484ms	2.5753ms	388.2978 Ops/s	381.2226 Ops/s	$\color{#35bf28}+1.86\%$
test_compile_assign_and_add[pytree-compile]	1.5682ms	0.4253ms	2.3512 KOps/s	2.2829 KOps/s	$\color{#35bf28}+2.99\%$
test_compile_assign_and_add[pytree-eager]	2.8358ms	2.7163ms	368.1471 Ops/s	370.7453 Ops/s	$\color{#d91a1a}-0.70\%$
test_compile_indexing[tensor-tensordict-compile]	0.3581ms	0.1154ms	8.6646 KOps/s	8.9724 KOps/s	$\color{#d91a1a}-3.43\%$
test_compile_indexing[tensor-tensordict-eager]	0.5910ms	82.0770μs	12.1837 KOps/s	11.8831 KOps/s	$\color{#35bf28}+2.53\%$
test_compile_indexing[tensor-tensorclass-compile]	0.3779ms	0.1056ms	9.4690 KOps/s	9.5660 KOps/s	$\color{#d91a1a}-1.01\%$
test_compile_indexing[tensor-tensorclass-eager]	0.1247ms	70.1039μs	14.2645 KOps/s	14.3573 KOps/s	$\color{#d91a1a}-0.65\%$
test_compile_indexing[tensor-pytree-compile]	0.1670ms	0.1105ms	9.0481 KOps/s	9.0943 KOps/s	$\color{#d91a1a}-0.51\%$
test_compile_indexing[tensor-pytree-eager]	0.1204ms	68.6596μs	14.5646 KOps/s	14.2364 KOps/s	$\color{#35bf28}+2.31\%$
test_compile_indexing[slice-tensordict-compile]	0.2587ms	0.1000ms	9.9974 KOps/s	10.0675 KOps/s	$\color{#d91a1a}-0.70\%$
test_compile_indexing[slice-tensordict-eager]	0.1411ms	16.4927μs	60.6328 KOps/s	58.7572 KOps/s	$\color{#35bf28}+3.19\%$
test_compile_indexing[slice-tensorclass-compile]	0.1627ms	96.2567μs	10.3889 KOps/s	10.5804 KOps/s	$\color{#d91a1a}-1.81\%$
test_compile_indexing[slice-tensorclass-eager]	47.4710μs	15.7766μs	63.3849 KOps/s	63.9219 KOps/s	$\color{#d91a1a}-0.84\%$
test_compile_indexing[slice-pytree-compile]	0.1497ms	95.6477μs	10.4550 KOps/s	10.4613 KOps/s	$\color{#d91a1a}-0.06\%$
test_compile_indexing[slice-pytree-eager]	53.3710μs	15.7327μs	63.5619 KOps/s	63.5371 KOps/s	$\color{#35bf28}+0.04\%$
test_compile_indexing[int-tensordict-compile]	0.1557ms	0.1010ms	9.8972 KOps/s	9.6673 KOps/s	$\color{#35bf28}+2.38\%$
test_compile_indexing[int-tensordict-eager]	0.6154ms	16.4565μs	60.7662 KOps/s	59.9283 KOps/s	$\color{#35bf28}+1.40\%$
test_compile_indexing[int-tensorclass-compile]	0.1430ms	95.4775μs	10.4737 KOps/s	10.4942 KOps/s	$\color{#d91a1a}-0.20\%$
test_compile_indexing[int-tensorclass-eager]	50.9110μs	15.8379μs	63.1398 KOps/s	63.9975 KOps/s	$\color{#d91a1a}-1.34\%$
test_compile_indexing[int-pytree-compile]	0.1560ms	95.3516μs	10.4875 KOps/s	10.5191 KOps/s	$\color{#d91a1a}-0.30\%$
test_compile_indexing[int-pytree-eager]	48.8410μs	15.7773μs	63.3820 KOps/s	63.9592 KOps/s	$\color{#d91a1a}-0.90\%$
test_mod_add[eager]	0.1497ms	31.9164μs	31.3319 KOps/s	29.8940 KOps/s	$\color{#35bf28}+4.81\%$
test_mod_add[compile]	0.3865ms	76.2567μs	13.1136 KOps/s	13.1213 KOps/s	$\color{#d91a1a}-0.06\%$
test_mod_add[compile-overhead]	0.3157ms	0.1657ms	6.0342 KOps/s	5.7490 KOps/s	$\color{#35bf28}+4.96\%$
test_mod_wrap[eager]	0.3221ms	0.2437ms	4.1038 KOps/s	4.0986 KOps/s	$\color{#35bf28}+0.13\%$
test_mod_wrap[compile]	1.5679ms	0.2789ms	3.5850 KOps/s	3.5605 KOps/s	$\color{#35bf28}+0.69\%$
test_mod_wrap[compile-overhead]	7.5216ms	4.0491ms	246.9711 Ops/s	241.8535 Ops/s	$\color{#35bf28}+2.12\%$
test_mod_wrap_and_backward[eager]	1.8083ms	1.3649ms	732.6601 Ops/s	677.1597 Ops/s	$\textbf{\color{#35bf28}+8.20\%}$
test_mod_wrap_and_backward[compile]	1.4699ms	1.2635ms	791.4513 Ops/s	731.5239 Ops/s	$\textbf{\color{#35bf28}+8.19\%}$
test_mod_wrap_and_backward[compile-overhead]	1.3699ms	0.9220ms	1.0846 KOps/s	958.0745 Ops/s	$\textbf{\color{#35bf28}+13.20\%}$
test_seq_add[eager]	0.5057ms	98.3289μs	10.1699 KOps/s	10.3256 KOps/s	$\color{#d91a1a}-1.51\%$
test_seq_add[compile]	0.1817ms	85.8556μs	11.6475 KOps/s	11.6985 KOps/s	$\color{#d91a1a}-0.44\%$
test_seq_add[compile-overhead]	0.5226ms	0.1267ms	7.8956 KOps/s	7.8759 KOps/s	$\color{#35bf28}+0.25\%$
test_seq_wrap[eager]	0.7856ms	0.3831ms	2.6102 KOps/s	2.5777 KOps/s	$\color{#35bf28}+1.26\%$
test_seq_wrap[compile]	0.3797ms	0.2939ms	3.4026 KOps/s	3.3284 KOps/s	$\color{#35bf28}+2.23\%$
test_seq_wrap[compile-overhead]	0.2979ms	0.2199ms	4.5468 KOps/s	4.5583 KOps/s	$\color{#d91a1a}-0.25\%$
test_func_call_runtime[False-eager]	0.8784ms	0.7543ms	1.3258 KOps/s	1.2873 KOps/s	$\color{#35bf28}+2.99\%$
test_func_call_runtime[False-compile]	0.8894ms	0.7566ms	1.3218 KOps/s	1.3438 KOps/s	$\color{#d91a1a}-1.64\%$
test_func_call_runtime[False-compile-overhead]	0.4063ms	0.3546ms	2.8201 KOps/s	2.8129 KOps/s	$\color{#35bf28}+0.25\%$
test_func_call_runtime[True-eager]	1.0317ms	0.8938ms	1.1188 KOps/s	1.0858 KOps/s	$\color{#35bf28}+3.04\%$
test_func_call_runtime[True-compile]	0.9097ms	0.7530ms	1.3280 KOps/s	1.2971 KOps/s	$\color{#35bf28}+2.38\%$
test_func_call_runtime[True-compile-overhead]	0.5312ms	0.3752ms	2.6649 KOps/s	2.6661 KOps/s	$\color{#d91a1a}-0.04\%$
test_func_call_cm_runtime[False-eager]	0.8766ms	0.7279ms	1.3738 KOps/s	1.3478 KOps/s	$\color{#35bf28}+1.93\%$
test_func_call_cm_runtime[False-compile]	0.8880ms	0.7429ms	1.3461 KOps/s	1.3410 KOps/s	$\color{#35bf28}+0.38\%$
test_func_call_cm_runtime[False-compile-overhead]	0.4480ms	0.3553ms	2.8143 KOps/s	2.8158 KOps/s	$\color{#d91a1a}-0.05\%$
test_func_call_cm_runtime[True-eager]	1.1341ms	0.9935ms	1.0066 KOps/s	980.8138 Ops/s	$\color{#35bf28}+2.63\%$
test_func_call_cm_runtime[True-compile]	0.9536ms	0.8106ms	1.2337 KOps/s	1.2540 KOps/s	$\color{#d91a1a}-1.62\%$
test_func_call_cm_runtime[True-compile-overhead]	0.6049ms	0.3977ms	2.5145 KOps/s	2.4707 KOps/s	$\color{#35bf28}+1.77\%$
test_vmap_func_call_cm_runtime[eager]	2.5487ms	2.0790ms	480.9937 Ops/s	478.8489 Ops/s	$\color{#35bf28}+0.45\%$
test_vmap_func_call_cm_runtime[compile]	1.1822ms	0.7919ms	1.2629 KOps/s	1.2453 KOps/s	$\color{#35bf28}+1.41\%$
test_vmap_func_call_cm_runtime[compile-overhead]	0.4459ms	0.4000ms	2.5003 KOps/s	2.4744 KOps/s	$\color{#35bf28}+1.05\%$
test_distributed	3.6096ms	0.1747ms	5.7255 KOps/s	8.8582 KOps/s	$\textbf{\color{#d91a1a}-35.37\%}$
test_tdmodule	32.0200μs	14.4279μs	69.3101 KOps/s	65.5955 KOps/s	$\textbf{\color{#35bf28}+5.66\%}$
test_tdmodule_dispatch	87.0720μs	28.2412μs	35.4092 KOps/s	34.0713 KOps/s	$\color{#35bf28}+3.93\%$
test_tdseq	39.8500μs	16.3133μs	61.2999 KOps/s	61.0235 KOps/s	$\color{#35bf28}+0.45\%$
test_tdseq_dispatch	58.2410μs	31.8864μs	31.3614 KOps/s	30.5972 KOps/s	$\color{#35bf28}+2.50\%$
test_instantiation_functorch	1.6228ms	1.5338ms	651.9731 Ops/s	647.0752 Ops/s	$\color{#35bf28}+0.76\%$
test_exec_functorch	0.2112ms	0.1501ms	6.6632 KOps/s	6.9853 KOps/s	$\color{#d91a1a}-4.61\%$
test_exec_functional_call	0.2580ms	0.1511ms	6.6197 KOps/s	7.1202 KOps/s	$\textbf{\color{#d91a1a}-7.03\%}$
test_exec_td_decorator	0.3946ms	0.1973ms	5.0692 KOps/s	5.4030 KOps/s	$\textbf{\color{#d91a1a}-6.18\%}$
test_vmap_mlp_speed_decorator[True-True]	0.8309ms	0.7111ms	1.4062 KOps/s	1.4848 KOps/s	$\textbf{\color{#d91a1a}-5.30\%}$
test_vmap_mlp_speed_decorator[True-False]	0.8251ms	0.7015ms	1.4256 KOps/s	1.4664 KOps/s	$\color{#d91a1a}-2.78\%$
test_vmap_mlp_speed_decorator[False-True]	0.7937ms	0.6300ms	1.5872 KOps/s	1.6217 KOps/s	$\color{#d91a1a}-2.12\%$
test_vmap_mlp_speed_decorator[False-False]	0.7685ms	0.6316ms	1.5834 KOps/s	1.6957 KOps/s	$\textbf{\color{#d91a1a}-6.62\%}$
test_vmap_transformer_speed_decorator[True-True]	20.5331ms	20.1653ms	49.5901 Ops/s	51.9841 Ops/s	$\color{#d91a1a}-4.61\%$
test_vmap_transformer_speed_decorator[True-False]	20.3941ms	20.1212ms	49.6989 Ops/s	51.8704 Ops/s	$\color{#d91a1a}-4.19\%$
test_vmap_transformer_speed_decorator[False-True]	20.4276ms	19.8827ms	50.2949 Ops/s	52.4493 Ops/s	$\color{#d91a1a}-4.11\%$
test_vmap_transformer_speed_decorator[False-False]	20.5292ms	19.2537ms	51.9382 Ops/s	52.2678 Ops/s	$\color{#d91a1a}-0.63\%$
test_to_module_speed[True]	2.1787ms	0.9448ms	1.0585 KOps/s	1.0597 KOps/s	$\color{#d91a1a}-0.12\%$
test_to_module_speed[False]	1.0213ms	0.9344ms	1.0702 KOps/s	1.0792 KOps/s	$\color{#d91a1a}-0.83\%$
test_tc_init	66.5410μs	36.6598μs	27.2778 KOps/s	25.8074 KOps/s	$\textbf{\color{#35bf28}+5.70\%}$
test_tc_init_nested	0.1117ms	74.5634μs	13.4114 KOps/s	12.6998 KOps/s	$\textbf{\color{#35bf28}+5.60\%}$
test_tc_first_layer_tensor	14.0873μs	0.7086μs	1.4113 MOps/s	1.4155 MOps/s	$\color{#d91a1a}-0.29\%$
test_tc_first_layer_nontensor	24.6000μs	2.3277μs	429.6064 KOps/s	439.3045 KOps/s	$\color{#d91a1a}-2.21\%$
test_tc_second_layer_tensor	8.6103μs	1.4369μs	695.9203 KOps/s	707.7757 KOps/s	$\color{#d91a1a}-1.68\%$
test_tc_second_layer_nontensor	25.3810μs	3.0579μs	327.0266 KOps/s	331.9561 KOps/s	$\color{#d91a1a}-1.48\%$
test_unbind	0.2276s	12.0283ms	83.1376 Ops/s	150.8182 Ops/s	$\textbf{\color{#d91a1a}-44.88\%}$
test_full_like	9.4428ms	9.1062ms	109.8149 Ops/s	108.3080 Ops/s	$\color{#35bf28}+1.39\%$
test_zeros_like	4.8492ms	4.3176ms	231.6091 Ops/s	231.0469 Ops/s	$\color{#35bf28}+0.24\%$
test_ones_like	9.1784ms	4.3359ms	230.6316 Ops/s	236.6159 Ops/s	$\color{#d91a1a}-2.53\%$
test_clone	6.6828ms	6.3717ms	156.9448 Ops/s	156.6134 Ops/s	$\color{#35bf28}+0.21\%$
test_squeeze	58.4510μs	9.2526μs	108.0773 KOps/s	108.4997 KOps/s	$\color{#d91a1a}-0.39\%$
test_unsqueeze	0.1211ms	70.9858μs	14.0873 KOps/s	14.2066 KOps/s	$\color{#d91a1a}-0.84\%$
test_split	0.4007ms	0.1556ms	6.4269 KOps/s	6.3501 KOps/s	$\color{#35bf28}+1.21\%$
test_permute	0.3269ms	0.1734ms	5.7662 KOps/s	5.7373 KOps/s	$\color{#35bf28}+0.50\%$
test_stack	51.6348ms	50.6939ms	19.7262 Ops/s	19.7124 Ops/s	$\color{#35bf28}+0.07\%$
test_cat	50.7155ms	50.4571ms	19.8188 Ops/s	19.7706 Ops/s	$\color{#35bf28}+0.24\%$