[Feature] TensorDict.<reduction>(dim='feature')

[vmoens]

Stack from ghstack (oldest at bottom):

• [Doc,BugFix,Feature] Better doc for reductions #1122
• -> [Feature] TensorDict.<reduction>(dim='feature') #1121

[github-actions]

$\color{#D29922}\textsf{\Large&amp;#x26A0;\kern{0.2cm}\normalsize Warning}$ Result of CPU Benchmark Tests

Total Benchmarks: 217. Improved: $\large\color{#35bf28}12$. Worsened: $\large\color{#d91a1a}13$.

Expand to view detailed results

Name	Max	Mean	Ops	Ops on Repo HEAD	Change
test_plain_set_nested	46.5770μs	18.8717μs	52.9893 KOps/s	55.4831 KOps/s	$\color{#d91a1a}-4.49\%$
test_plain_set_stack_nested	52.1370μs	19.2934μs	51.8311 KOps/s	54.5593 KOps/s	$\textbf{\color{#d91a1a}-5.00\%}$
test_plain_set_nested_inplace	60.5530μs	21.4758μs	46.5641 KOps/s	50.4287 KOps/s	$\textbf{\color{#d91a1a}-7.66\%}$
test_plain_set_stack_nested_inplace	79.0390μs	20.9721μs	47.6824 KOps/s	50.2772 KOps/s	$\textbf{\color{#d91a1a}-5.16\%}$
test_items	23.1130μs	4.0780μs	245.2192 KOps/s	245.2626 KOps/s	$\color{#d91a1a}-0.02\%$
test_items_nested	0.7400ms	0.4011ms	2.4930 KOps/s	2.5574 KOps/s	$\color{#d91a1a}-2.52\%$
test_items_nested_locked	0.8960ms	0.3975ms	2.5157 KOps/s	2.5676 KOps/s	$\color{#d91a1a}-2.02\%$
test_items_nested_leaf	0.1298ms	71.4034μs	14.0049 KOps/s	13.7712 KOps/s	$\color{#35bf28}+1.70\%$
test_items_stack_nested	0.7316ms	0.4001ms	2.4993 KOps/s	2.5478 KOps/s	$\color{#d91a1a}-1.90\%$
test_items_stack_nested_leaf	0.1411ms	72.3471μs	13.8223 KOps/s	13.6395 KOps/s	$\color{#35bf28}+1.34\%$
test_items_stack_nested_locked	0.7787ms	0.4012ms	2.4924 KOps/s	2.5444 KOps/s	$\color{#d91a1a}-2.04\%$
test_keys	28.3830μs	3.4969μs	285.9693 KOps/s	287.9936 KOps/s	$\color{#d91a1a}-0.70\%$
test_keys_nested	0.2226ms	0.1377ms	7.2613 KOps/s	7.3402 KOps/s	$\color{#d91a1a}-1.07\%$
test_keys_nested_locked	0.7500ms	0.1442ms	6.9358 KOps/s	6.9468 KOps/s	$\color{#d91a1a}-0.16\%$
test_keys_nested_leaf	1.7828ms	0.1192ms	8.3871 KOps/s	8.4782 KOps/s	$\color{#d91a1a}-1.07\%$
test_keys_stack_nested	0.2195ms	0.1369ms	7.3055 KOps/s	7.2193 KOps/s	$\color{#35bf28}+1.19\%$
test_keys_stack_nested_leaf	0.2283ms	0.1204ms	8.3078 KOps/s	8.4767 KOps/s	$\color{#d91a1a}-1.99\%$
test_keys_stack_nested_locked	0.2467ms	0.1439ms	6.9490 KOps/s	7.0241 KOps/s	$\color{#d91a1a}-1.07\%$
test_values	36.3656μs	1.0729μs	932.0488 KOps/s	968.4179 KOps/s	$\color{#d91a1a}-3.76\%$
test_values_nested	0.1023ms	55.5018μs	18.0174 KOps/s	18.1351 KOps/s	$\color{#d91a1a}-0.65\%$
test_values_nested_locked	0.1386ms	55.3207μs	18.0764 KOps/s	18.1391 KOps/s	$\color{#d91a1a}-0.35\%$
test_values_nested_leaf	0.1220ms	60.4682μs	16.5376 KOps/s	15.8978 KOps/s	$\color{#35bf28}+4.02\%$
test_values_stack_nested	0.2137ms	55.7536μs	17.9360 KOps/s	17.3363 KOps/s	$\color{#35bf28}+3.46\%$
test_values_stack_nested_leaf	0.1165ms	61.0149μs	16.3894 KOps/s	16.4520 KOps/s	$\color{#d91a1a}-0.38\%$
test_values_stack_nested_locked	0.1071ms	56.0927μs	17.8276 KOps/s	17.8695 KOps/s	$\color{#d91a1a}-0.23\%$
test_membership	14.1770μs	0.8578μs	1.1657 MOps/s	1.4183 MOps/s	$\textbf{\color{#d91a1a}-17.81\%}$
test_membership_nested	44.2720μs	2.9913μs	334.3067 KOps/s	334.7078 KOps/s	$\color{#d91a1a}-0.12\%$
test_membership_nested_leaf	25.9080μs	2.9975μs	333.6122 KOps/s	337.9482 KOps/s	$\color{#d91a1a}-1.28\%$
test_membership_stacked_nested	31.6690μs	2.9780μs	335.7959 KOps/s	341.9644 KOps/s	$\color{#d91a1a}-1.80\%$
test_membership_stacked_nested_leaf	25.2970μs	2.9997μs	333.3668 KOps/s	339.9705 KOps/s	$\color{#d91a1a}-1.94\%$
test_membership_nested_last	30.5870μs	4.3773μs	228.4519 KOps/s	235.2443 KOps/s	$\color{#d91a1a}-2.89\%$
test_membership_nested_leaf_last	25.5670μs	4.3148μs	231.7625 KOps/s	234.2868 KOps/s	$\color{#d91a1a}-1.08\%$
test_membership_stacked_nested_last	23.1730μs	4.3528μs	229.7396 KOps/s	195.5843 KOps/s	$\textbf{\color{#35bf28}+17.46\%}$
test_membership_stacked_nested_leaf_last	36.4980μs	4.3139μs	231.8067 KOps/s	203.9664 KOps/s	$\textbf{\color{#35bf28}+13.65\%}$
test_nested_getleaf	54.7640μs	11.1621μs	89.5889 KOps/s	93.1920 KOps/s	$\color{#d91a1a}-3.87\%$
test_nested_get	33.3920μs	10.4509μs	95.6857 KOps/s	95.8663 KOps/s	$\color{#d91a1a}-0.19\%$
test_stacked_getleaf	45.2720μs	10.9743μs	91.1217 KOps/s	92.0891 KOps/s	$\color{#d91a1a}-1.05\%$
test_stacked_get	52.4800μs	10.6097μs	94.2529 KOps/s	96.7738 KOps/s	$\color{#d91a1a}-2.60\%$
test_nested_getitemleaf	47.7390μs	11.4535μs	87.3094 KOps/s	88.1215 KOps/s	$\color{#d91a1a}-0.92\%$
test_nested_getitem	56.8660μs	11.1778μs	89.4628 KOps/s	95.2570 KOps/s	$\textbf{\color{#d91a1a}-6.08\%}$
test_stacked_getitemleaf	48.3300μs	11.7080μs	85.4115 KOps/s	89.9798 KOps/s	$\textbf{\color{#d91a1a}-5.08\%}$
test_stacked_getitem	45.9560μs	10.9997μs	90.9115 KOps/s	93.8048 KOps/s	$\color{#d91a1a}-3.08\%$
test_lock_nested	1.8608ms	0.4417ms	2.2639 KOps/s	2.2189 KOps/s	$\color{#35bf28}+2.03\%$
test_lock_stack_nested	0.5222ms	0.4123ms	2.4255 KOps/s	2.3964 KOps/s	$\color{#35bf28}+1.21\%$
test_unlock_nested	1.0197ms	0.3677ms	2.7193 KOps/s	2.7125 KOps/s	$\color{#35bf28}+0.25\%$
test_unlock_stack_nested	0.8294ms	0.3355ms	2.9805 KOps/s	2.9916 KOps/s	$\color{#d91a1a}-0.37\%$
test_flatten_speed	0.1792ms	95.7300μs	10.4460 KOps/s	10.4451 KOps/s	$+0.01\%$
test_unflatten_speed	0.7077ms	0.4979ms	2.0086 KOps/s	2.0391 KOps/s	$\color{#d91a1a}-1.50\%$
test_common_ops	6.6208ms	0.8091ms	1.2360 KOps/s	1.2442 KOps/s	$\color{#d91a1a}-0.66\%$
test_creation	27.2810μs	2.1053μs	474.9810 KOps/s	473.0011 KOps/s	$\color{#35bf28}+0.42\%$
test_creation_empty	41.8470μs	11.7958μs	84.7760 KOps/s	90.2984 KOps/s	$\textbf{\color{#d91a1a}-6.12\%}$
test_creation_nested_1	40.9260μs	14.4013μs	69.4381 KOps/s	71.6645 KOps/s	$\color{#d91a1a}-3.11\%$
test_creation_nested_2	65.9930μs	19.3309μs	51.7307 KOps/s	54.8798 KOps/s	$\textbf{\color{#d91a1a}-5.74\%}$
test_clone	0.2292ms	13.2350μs	75.5571 KOps/s	75.3051 KOps/s	$\color{#35bf28}+0.33\%$
test_getitem[int]	1.2972ms	12.7915μs	78.1772 KOps/s	80.2240 KOps/s	$\color{#d91a1a}-2.55\%$
test_getitem[slice_int]	0.2574ms	25.4468μs	39.2976 KOps/s	39.6218 KOps/s	$\color{#d91a1a}-0.82\%$
test_getitem[range]	0.1652ms	48.2100μs	20.7426 KOps/s	19.3949 KOps/s	$\textbf{\color{#35bf28}+6.95\%}$
test_getitem[tuple]	0.1308ms	20.7358μs	48.2259 KOps/s	48.5897 KOps/s	$\color{#d91a1a}-0.75\%$
test_getitem[list]	0.4936ms	43.4214μs	23.0301 KOps/s	22.9274 KOps/s	$\color{#35bf28}+0.45\%$
test_setitem_dim[int]	49.7520μs	25.4261μs	39.3296 KOps/s	38.6585 KOps/s	$\color{#35bf28}+1.74\%$
test_setitem_dim[slice_int]	0.1184ms	54.1522μs	18.4665 KOps/s	18.5033 KOps/s	$\color{#d91a1a}-0.20\%$
test_setitem_dim[range]	0.1461ms	74.1165μs	13.4923 KOps/s	13.5418 KOps/s	$\color{#d91a1a}-0.37\%$
test_setitem_dim[tuple]	71.0420μs	40.8805μs	24.4615 KOps/s	23.5014 KOps/s	$\color{#35bf28}+4.09\%$
test_setitem	0.3099ms	20.7021μs	48.3042 KOps/s	48.8766 KOps/s	$\color{#d91a1a}-1.17\%$
test_set	0.3106ms	20.1488μs	49.6307 KOps/s	50.2381 KOps/s	$\color{#d91a1a}-1.21\%$
test_set_shared	1.2644ms	0.1729ms	5.7821 KOps/s	5.9311 KOps/s	$\color{#d91a1a}-2.51\%$
test_update	0.3098ms	23.5185μs	42.5197 KOps/s	44.2643 KOps/s	$\color{#d91a1a}-3.94\%$
test_update_nested	0.2762ms	34.9260μs	28.6320 KOps/s	30.0645 KOps/s	$\color{#d91a1a}-4.76\%$
test_update__nested	0.4041ms	33.3976μs	29.9423 KOps/s	30.6325 KOps/s	$\color{#d91a1a}-2.25\%$
test_set_nested	0.2981ms	22.8887μs	43.6898 KOps/s	45.1828 KOps/s	$\color{#d91a1a}-3.30\%$
test_set_nested_new	0.3209ms	27.9125μs	35.8263 KOps/s	37.4077 KOps/s	$\color{#d91a1a}-4.23\%$
test_select	0.3535ms	43.5669μs	22.9532 KOps/s	23.1593 KOps/s	$\color{#d91a1a}-0.89\%$
test_select_nested	0.1219ms	60.1683μs	16.6201 KOps/s	16.8466 KOps/s	$\color{#d91a1a}-1.34\%$
test_exclude_nested	0.1469ms	78.6500μs	12.7146 KOps/s	13.0180 KOps/s	$\color{#d91a1a}-2.33\%$
test_empty[True]	0.5210ms	0.3885ms	2.5743 KOps/s	2.6319 KOps/s	$\color{#d91a1a}-2.19\%$
test_empty[False]	7.3978μs	1.2045μs	830.2408 KOps/s	820.4894 KOps/s	$\color{#35bf28}+1.19\%$
test_unbind_speed	0.3794ms	0.2648ms	3.7764 KOps/s	3.7544 KOps/s	$\color{#35bf28}+0.59\%$
test_unbind_speed_stack0	0.4376ms	0.2636ms	3.7942 KOps/s	3.8382 KOps/s	$\color{#d91a1a}-1.15\%$
test_unbind_speed_stack1	0.1104s	0.7737ms	1.2925 KOps/s	1.4027 KOps/s	$\textbf{\color{#d91a1a}-7.85\%}$
test_split	0.1177s	1.7599ms	568.2162 Ops/s	574.8821 Ops/s	$\color{#d91a1a}-1.16\%$
test_chunk	0.1024s	1.7034ms	587.0598 Ops/s	576.9642 Ops/s	$\color{#35bf28}+1.75\%$
test_consolidate_njt[False-None]	11.3033ms	8.1724ms	122.3633 Ops/s	122.9608 Ops/s	$\color{#d91a1a}-0.49\%$
test_creation[device0]	0.2235ms	91.2835μs	10.9549 KOps/s	10.7076 KOps/s	$\color{#35bf28}+2.31\%$
test_creation_from_tensor	3.3529ms	95.6241μs	10.4576 KOps/s	10.3732 KOps/s	$\color{#35bf28}+0.81\%$
test_add_one[memmap_tensor0]	0.2851ms	4.9199μs	203.2578 KOps/s	210.5007 KOps/s	$\color{#d91a1a}-3.44\%$
test_contiguous[memmap_tensor0]	12.9240μs	0.5168μs	1.9349 MOps/s	1.9788 MOps/s	$\color{#d91a1a}-2.22\%$
test_stack[memmap_tensor0]	63.3080μs	3.4474μs	290.0722 KOps/s	297.1332 KOps/s	$\color{#d91a1a}-2.38\%$
test_memmaptd_index	1.0759ms	0.2430ms	4.1153 KOps/s	4.2785 KOps/s	$\color{#d91a1a}-3.81\%$
test_memmaptd_index_astensor	0.6596ms	0.3192ms	3.1329 KOps/s	3.1299 KOps/s	$\color{#35bf28}+0.10\%$
test_memmaptd_index_op	1.3883ms	0.5912ms	1.6916 KOps/s	1.7302 KOps/s	$\color{#d91a1a}-2.23\%$
test_serialize_model	0.1197s	0.1155s	8.6597 Ops/s	7.3021 Ops/s	$\textbf{\color{#35bf28}+18.59\%}$
test_serialize_model_pickle	0.4541s	0.3963s	2.5235 Ops/s	2.5353 Ops/s	$\color{#d91a1a}-0.47\%$
test_serialize_weights	0.1272s	0.1171s	8.5386 Ops/s	8.8687 Ops/s	$\color{#d91a1a}-3.72\%$
test_serialize_weights_returnearly	0.1744s	0.1622s	6.1647 Ops/s	6.4959 Ops/s	$\textbf{\color{#d91a1a}-5.10\%}$
test_serialize_weights_pickle	0.4934s	0.4029s	2.4819 Ops/s	2.4326 Ops/s	$\color{#35bf28}+2.03\%$
test_serialize_weights_filesystem	0.2523s	0.1594s	6.2739 Ops/s	6.8807 Ops/s	$\textbf{\color{#d91a1a}-8.82\%}$
test_serialize_model_filesystem	0.1758s	0.1525s	6.5574 Ops/s	5.9816 Ops/s	$\textbf{\color{#35bf28}+9.63\%}$
test_reshape_pytree	75.0100μs	27.1766μs	36.7964 KOps/s	38.0091 KOps/s	$\color{#d91a1a}-3.19\%$
test_reshape_td	74.2980μs	33.4946μs	29.8556 KOps/s	30.5400 KOps/s	$\color{#d91a1a}-2.24\%$
test_view_pytree	67.9270μs	27.1367μs	36.8505 KOps/s	37.8727 KOps/s	$\color{#d91a1a}-2.70\%$
test_view_td	0.1005ms	38.5984μs	25.9078 KOps/s	25.9238 KOps/s	$\color{#d91a1a}-0.06\%$
test_unbind_pytree	77.8750μs	30.1855μs	33.1284 KOps/s	33.0601 KOps/s	$\color{#35bf28}+0.21\%$
test_unbind_td	0.3423ms	38.1062μs	26.2424 KOps/s	25.9360 KOps/s	$\color{#35bf28}+1.18\%$
test_split_pytree	82.1230μs	29.6265μs	33.7536 KOps/s	33.9104 KOps/s	$\color{#d91a1a}-0.46\%$
test_split_td	0.5156ms	45.5446μs	21.9565 KOps/s	22.3368 KOps/s	$\color{#d91a1a}-1.70\%$
test_add_pytree	87.2830μs	35.5934μs	28.0951 KOps/s	28.5995 KOps/s	$\color{#d91a1a}-1.76\%$
test_add_td	0.2920ms	61.2999μs	16.3133 KOps/s	18.9561 KOps/s	$\textbf{\color{#d91a1a}-13.94\%}$
test_compile_add_one_nested[tensordict-compile]	0.1369ms	62.7730μs	15.9304 KOps/s	16.0731 KOps/s	$\color{#d91a1a}-0.89\%$
test_compile_add_one_nested[tensordict-eager]	0.8180ms	0.1613ms	6.2014 KOps/s	6.2287 KOps/s	$\color{#d91a1a}-0.44\%$
test_compile_add_one_nested[pytree-compile]	0.1292ms	46.0911μs	21.6961 KOps/s	21.6929 KOps/s	$\color{#35bf28}+0.02\%$
test_compile_add_one_nested[pytree-eager]	0.3203ms	0.1186ms	8.4334 KOps/s	8.5667 KOps/s	$\color{#d91a1a}-1.56\%$
test_compile_copy_nested[tensordict-compile]	0.1414ms	26.3945μs	37.8867 KOps/s	37.4193 KOps/s	$\color{#35bf28}+1.25\%$
test_compile_copy_nested[tensordict-eager]	0.1169ms	54.6961μs	18.2828 KOps/s	18.6063 KOps/s	$\color{#d91a1a}-1.74\%$
test_compile_copy_nested[pytree-compile]	0.1951ms	79.1964μs	12.6268 KOps/s	12.6780 KOps/s	$\color{#d91a1a}-0.40\%$
test_compile_copy_nested[pytree-eager]	0.1326ms	68.4893μs	14.6008 KOps/s	14.8018 KOps/s	$\color{#d91a1a}-1.36\%$
test_compile_add_one_flat[tensordict-compile]	0.2185ms	0.1067ms	9.3752 KOps/s	9.5616 KOps/s	$\color{#d91a1a}-1.95\%$
test_compile_add_one_flat[tensordict-eager]	0.3339ms	0.2010ms	4.9758 KOps/s	4.9685 KOps/s	$\color{#35bf28}+0.15\%$
test_compile_add_one_flat[tensorclass-compile]	0.2402ms	46.3198μs	21.5891 KOps/s	20.5725 KOps/s	$\color{#35bf28}+4.94\%$
test_compile_add_one_flat[tensorclass-eager]	1.3488ms	62.8243μs	15.9174 KOps/s	16.3813 KOps/s	$\color{#d91a1a}-2.83\%$
test_compile_add_one_flat[pytree-compile]	0.1897ms	0.1056ms	9.4667 KOps/s	9.5776 KOps/s	$\color{#d91a1a}-1.16\%$
test_compile_add_one_flat[pytree-eager]	0.4255ms	0.2034ms	4.9171 KOps/s	4.9418 KOps/s	$\color{#d91a1a}-0.50\%$
test_compile_add_self_flat[tensordict-eager]	0.3452ms	0.2135ms	4.6847 KOps/s	4.6932 KOps/s	$\color{#d91a1a}-0.18\%$
test_compile_add_self_flat[tensordict-compile]	0.2370ms	0.1071ms	9.3388 KOps/s	9.5538 KOps/s	$\color{#d91a1a}-2.25\%$
test_compile_add_self_flat[tensorclass-eager]	0.3378ms	55.4747μs	18.0262 KOps/s	18.3785 KOps/s	$\color{#d91a1a}-1.92\%$
test_compile_add_self_flat[tensorclass-compile]	0.2087ms	47.3735μs	21.1089 KOps/s	22.0484 KOps/s	$\color{#d91a1a}-4.26\%$
test_compile_add_self_flat[pytree-eager]	0.6124ms	0.1587ms	6.3019 KOps/s	6.2167 KOps/s	$\color{#35bf28}+1.37\%$
test_compile_add_self_flat[pytree-compile]	0.2141ms	0.1037ms	9.6420 KOps/s	9.6784 KOps/s	$\color{#d91a1a}-0.38\%$
test_compile_copy_flat[tensordict-compile]	75.9810μs	21.1844μs	47.2044 KOps/s	45.6878 KOps/s	$\color{#35bf28}+3.32\%$
test_compile_copy_flat[tensordict-eager]	0.1171ms	60.8060μs	16.4458 KOps/s	16.7311 KOps/s	$\color{#d91a1a}-1.71\%$
test_compile_copy_flat[pytree-compile]	0.1554ms	80.1601μs	12.4750 KOps/s	12.4550 KOps/s	$\color{#35bf28}+0.16\%$
test_compile_copy_flat[pytree-eager]	0.1693ms	68.1438μs	14.6748 KOps/s	14.7021 KOps/s	$\color{#d91a1a}-0.19\%$
test_compile_assign_and_add[tensordict-compile]	0.4569ms	0.2138ms	4.6775 KOps/s	4.8844 KOps/s	$\color{#d91a1a}-4.24\%$
test_compile_assign_and_add[tensordict-eager]	1.4963ms	1.3164ms	759.6627 Ops/s	777.4561 Ops/s	$\color{#d91a1a}-2.29\%$
test_compile_assign_and_add[pytree-compile]	0.2790ms	0.2088ms	4.7884 KOps/s	4.9412 KOps/s	$\color{#d91a1a}-3.09\%$
test_compile_assign_and_add[pytree-eager]	0.9125ms	0.7881ms	1.2688 KOps/s	1.3006 KOps/s	$\color{#d91a1a}-2.45\%$
test_compile_assign_and_add_stack[compile]	0.7801ms	0.4688ms	2.1330 KOps/s	2.2308 KOps/s	$\color{#d91a1a}-4.38\%$
test_compile_assign_and_add_stack[eager]	2.9527ms	2.6469ms	377.7947 Ops/s	376.2672 Ops/s	$\color{#35bf28}+0.41\%$
test_compile_indexing[tensor-tensordict-compile]	0.1454ms	36.4230μs	27.4552 KOps/s	28.3907 KOps/s	$\color{#d91a1a}-3.30\%$
test_compile_indexing[tensor-tensordict-eager]	0.6471ms	32.9684μs	30.3321 KOps/s	30.2804 KOps/s	$\color{#35bf28}+0.17\%$
test_compile_indexing[tensor-tensorclass-compile]	0.1023ms	30.1810μs	33.1334 KOps/s	34.1557 KOps/s	$\color{#d91a1a}-2.99\%$
test_compile_indexing[tensor-tensorclass-eager]	68.1270μs	23.4121μs	42.7129 KOps/s	42.8806 KOps/s	$\color{#d91a1a}-0.39\%$
test_compile_indexing[tensor-pytree-compile]	0.1074ms	31.0303μs	32.2266 KOps/s	33.7327 KOps/s	$\color{#d91a1a}-4.46\%$
test_compile_indexing[tensor-pytree-eager]	92.3820μs	23.1553μs	43.1867 KOps/s	43.5209 KOps/s	$\color{#d91a1a}-0.77\%$
test_compile_indexing[slice-tensordict-compile]	0.1286ms	52.0418μs	19.2153 KOps/s	19.3468 KOps/s	$\color{#d91a1a}-0.68\%$
test_compile_indexing[slice-tensordict-eager]	0.4216ms	20.3126μs	49.2306 KOps/s	50.7910 KOps/s	$\color{#d91a1a}-3.07\%$
test_compile_indexing[slice-tensorclass-compile]	0.1032ms	44.2754μs	22.5859 KOps/s	22.7526 KOps/s	$\color{#d91a1a}-0.73\%$
test_compile_indexing[slice-tensorclass-eager]	85.4390μs	18.7584μs	53.3094 KOps/s	53.0458 KOps/s	$\color{#35bf28}+0.50\%$
test_compile_indexing[slice-pytree-compile]	0.1251ms	44.8164μs	22.3133 KOps/s	22.3317 KOps/s	$\color{#d91a1a}-0.08\%$
test_compile_indexing[slice-pytree-eager]	54.1110μs	18.7828μs	53.2403 KOps/s	53.2193 KOps/s	$\color{#35bf28}+0.04\%$
test_compile_indexing[int-tensordict-compile]	0.1164ms	52.8907μs	18.9069 KOps/s	18.9524 KOps/s	$\color{#d91a1a}-0.24\%$
test_compile_indexing[int-tensordict-eager]	1.0024ms	19.9591μs	50.1024 KOps/s	51.1753 KOps/s	$\color{#d91a1a}-2.10\%$
test_compile_indexing[int-tensorclass-compile]	0.1048ms	44.7402μs	22.3512 KOps/s	22.1817 KOps/s	$\color{#35bf28}+0.76\%$
test_compile_indexing[int-tensorclass-eager]	81.2010μs	18.6524μs	53.6124 KOps/s	53.5070 KOps/s	$\color{#35bf28}+0.20\%$
test_compile_indexing[int-pytree-compile]	0.1140ms	44.7737μs	22.3345 KOps/s	22.2798 KOps/s	$\color{#35bf28}+0.25\%$
test_compile_indexing[int-pytree-eager]	65.6230μs	18.9043μs	52.8981 KOps/s	53.0472 KOps/s	$\color{#d91a1a}-0.28\%$
test_mod_add[eager]	0.1059ms	35.1202μs	28.4737 KOps/s	28.4267 KOps/s	$\color{#35bf28}+0.17\%$
test_mod_add[compile]	0.1324ms	46.4675μs	21.5204 KOps/s	20.9646 KOps/s	$\color{#35bf28}+2.65\%$
test_mod_add[compile-overhead]	0.1136ms	46.1229μs	21.6812 KOps/s	21.1657 KOps/s	$\color{#35bf28}+2.44\%$
test_mod_wrap[eager]	0.3914ms	0.2237ms	4.4708 KOps/s	4.4673 KOps/s	$\color{#35bf28}+0.08\%$
test_mod_wrap[compile]	0.3249ms	0.2082ms	4.8039 KOps/s	4.8985 KOps/s	$\color{#d91a1a}-1.93\%$
test_mod_wrap[compile-overhead]	0.4527ms	0.2077ms	4.8149 KOps/s	4.9390 KOps/s	$\color{#d91a1a}-2.51\%$
test_mod_wrap_and_backward[eager]	13.7092ms	11.7953ms	84.7798 Ops/s	85.4702 Ops/s	$\color{#d91a1a}-0.81\%$
test_mod_wrap_and_backward[compile]	11.9642ms	11.1120ms	89.9925 Ops/s	78.4122 Ops/s	$\textbf{\color{#35bf28}+14.77\%}$
test_mod_wrap_and_backward[compile-overhead]	13.2555ms	11.5945ms	86.2475 Ops/s	73.5592 Ops/s	$\textbf{\color{#35bf28}+17.25\%}$
test_seq_add[eager]	0.1953ms	0.1130ms	8.8530 KOps/s	8.8537 KOps/s	$-0.01\%$
test_seq_add[compile]	0.1734ms	61.9775μs	16.1349 KOps/s	15.9631 KOps/s	$\color{#35bf28}+1.08\%$
test_seq_add[compile-overhead]	0.1295ms	59.6622μs	16.7610 KOps/s	16.2060 KOps/s	$\color{#35bf28}+3.42\%$
test_seq_wrap[eager]	0.8031ms	0.4370ms	2.2882 KOps/s	2.2646 KOps/s	$\color{#35bf28}+1.04\%$
test_seq_wrap[compile]	0.3212ms	0.2287ms	4.3728 KOps/s	4.3851 KOps/s	$\color{#d91a1a}-0.28\%$
test_seq_wrap[compile-overhead]	0.5237ms	0.2320ms	4.3096 KOps/s	4.4284 KOps/s	$\color{#d91a1a}-2.68\%$
test_func_call_runtime[False-eager]	1.0512ms	0.5350ms	1.8692 KOps/s	1.8242 KOps/s	$\color{#35bf28}+2.47\%$
test_func_call_runtime[False-compile]	0.8098ms	0.4279ms	2.3373 KOps/s	2.3887 KOps/s	$\color{#d91a1a}-2.15\%$
test_func_call_runtime[False-compile-overhead]	0.8473ms	0.4245ms	2.3557 KOps/s	2.3713 KOps/s	$\color{#d91a1a}-0.66\%$
test_func_call_runtime[True-eager]	0.8564ms	0.7373ms	1.3562 KOps/s	1.3202 KOps/s	$\color{#35bf28}+2.73\%$
test_func_call_runtime[True-compile]	0.5847ms	0.4614ms	2.1672 KOps/s	2.1545 KOps/s	$\color{#35bf28}+0.59\%$
test_func_call_runtime[True-compile-overhead]	0.9104ms	0.4666ms	2.1431 KOps/s	2.1790 KOps/s	$\color{#d91a1a}-1.65\%$
test_func_call_cm_runtime[False-eager]	0.9636ms	0.5310ms	1.8833 KOps/s	1.8111 KOps/s	$\color{#35bf28}+3.98\%$
test_func_call_cm_runtime[False-compile]	0.7807ms	0.4264ms	2.3454 KOps/s	2.3745 KOps/s	$\color{#d91a1a}-1.23\%$
test_func_call_cm_runtime[False-compile-overhead]	0.5335ms	0.4264ms	2.3454 KOps/s	2.3654 KOps/s	$\color{#d91a1a}-0.85\%$
test_func_call_cm_runtime[True-eager]	1.6538ms	0.8866ms	1.1280 KOps/s	1.1089 KOps/s	$\color{#35bf28}+1.72\%$
test_func_call_cm_runtime[True-compile]	0.6164ms	0.4899ms	2.0413 KOps/s	2.0383 KOps/s	$\color{#35bf28}+0.15\%$
test_func_call_cm_runtime[True-compile-overhead]	1.0113ms	0.4952ms	2.0194 KOps/s	2.0413 KOps/s	$\color{#d91a1a}-1.07\%$
test_vmap_func_call_cm_runtime[eager]	2.4942ms	1.8657ms	535.9866 Ops/s	533.9930 Ops/s	$\color{#35bf28}+0.37\%$
test_vmap_func_call_cm_runtime[compile]	0.6879ms	0.5231ms	1.9117 KOps/s	1.9065 KOps/s	$\color{#35bf28}+0.27\%$
test_vmap_func_call_cm_runtime[compile-overhead]	0.7150ms	0.5224ms	1.9141 KOps/s	1.9055 KOps/s	$\color{#35bf28}+0.45\%$
test_distributed	0.2829ms	0.1251ms	7.9943 KOps/s	7.7502 KOps/s	$\color{#35bf28}+3.15\%$
test_tdmodule	45.4240μs	26.2654μs	38.0729 KOps/s	36.8850 KOps/s	$\color{#35bf28}+3.22\%$
test_tdmodule_dispatch	83.0140μs	48.6217μs	20.5669 KOps/s	20.1671 KOps/s	$\color{#35bf28}+1.98\%$
test_tdseq	58.4590μs	26.3074μs	38.0121 KOps/s	36.9568 KOps/s	$\color{#35bf28}+2.86\%$
test_tdseq_dispatch	93.6840μs	51.4193μs	19.4479 KOps/s	19.5989 KOps/s	$\color{#d91a1a}-0.77\%$
test_instantiation_functorch	2.1840ms	1.5592ms	641.3375 Ops/s	640.7101 Ops/s	$\color{#35bf28}+0.10\%$
test_exec_functorch	0.4056ms	0.1771ms	5.6478 KOps/s	5.4886 KOps/s	$\color{#35bf28}+2.90\%$
test_exec_functional_call	0.3889ms	0.1729ms	5.7853 KOps/s	5.7884 KOps/s	$\color{#d91a1a}-0.05\%$
test_exec_td_decorator	0.6060ms	0.2273ms	4.3992 KOps/s	4.2427 KOps/s	$\color{#35bf28}+3.69\%$
test_vmap_mlp_speed_decorator[True-True]	0.9227ms	0.6500ms	1.5384 KOps/s	1.5386 KOps/s	$\color{#d91a1a}-0.02\%$
test_vmap_mlp_speed_decorator[True-False]	0.9091ms	0.6489ms	1.5412 KOps/s	1.5074 KOps/s	$\color{#35bf28}+2.24\%$
test_vmap_mlp_speed_decorator[False-True]	0.8015ms	0.5225ms	1.9140 KOps/s	1.9077 KOps/s	$\color{#35bf28}+0.33\%$
test_vmap_mlp_speed_decorator[False-False]	0.6931ms	0.5193ms	1.9256 KOps/s	1.8906 KOps/s	$\color{#35bf28}+1.85\%$
test_to_module_speed[True]	1.9570ms	1.2975ms	770.7319 Ops/s	767.2519 Ops/s	$\color{#35bf28}+0.45\%$
test_to_module_speed[False]	1.9838ms	1.2634ms	791.4843 Ops/s	789.8247 Ops/s	$\color{#35bf28}+0.21\%$
test_tc_init	87.9140μs	45.8432μs	21.8135 KOps/s	21.0349 KOps/s	$\color{#35bf28}+3.70\%$
test_tc_init_nested	0.1559ms	90.1290μs	11.0952 KOps/s	10.8736 KOps/s	$\color{#35bf28}+2.04\%$
test_tc_first_layer_tensor	24.2950μs	1.5333μs	652.2051 KOps/s	658.3637 KOps/s	$\color{#d91a1a}-0.94\%$
test_tc_first_layer_nontensor	48.2000μs	4.8114μs	207.8377 KOps/s	205.4476 KOps/s	$\color{#35bf28}+1.16\%$
test_tc_second_layer_tensor	40.3950μs	2.8294μs	353.4345 KOps/s	336.4271 KOps/s	$\textbf{\color{#35bf28}+5.06\%}$
test_tc_second_layer_nontensor	27.7320μs	6.1441μs	162.7576 KOps/s	163.4452 KOps/s	$\color{#d91a1a}-0.42\%$
test_unbind	0.2337s	12.8671ms	77.7173 Ops/s	80.7436 Ops/s	$\color{#d91a1a}-3.75\%$
test_full_like	11.5778ms	8.1609ms	122.5362 Ops/s	80.2472 Ops/s	$\textbf{\color{#35bf28}+52.70\%}$
test_zeros_like	5.5324ms	3.4075ms	293.4716 Ops/s	124.9594 Ops/s	$\textbf{\color{#35bf28}+134.85\%}$
test_ones_like	6.2906ms	3.7043ms	269.9560 Ops/s	124.2350 Ops/s	$\textbf{\color{#35bf28}+117.29\%}$
test_clone	7.1438ms	5.5349ms	180.6729 Ops/s	102.5211 Ops/s	$\textbf{\color{#35bf28}+76.23\%}$
test_squeeze	62.1250μs	11.5839μs	86.3267 KOps/s	83.9207 KOps/s	$\color{#35bf28}+2.87\%$
test_unsqueeze	0.2966ms	89.2230μs	11.2079 KOps/s	11.2675 KOps/s	$\color{#d91a1a}-0.53\%$
test_split	0.3673ms	0.1928ms	5.1861 KOps/s	5.1641 KOps/s	$\color{#35bf28}+0.43\%$
test_permute	0.3545ms	0.2177ms	4.5932 KOps/s	4.6053 KOps/s	$\color{#d91a1a}-0.26\%$
test_stack	30.1093ms	26.6396ms	37.5382 Ops/s	38.9251 Ops/s	$\color{#d91a1a}-3.56\%$
test_cat	31.8705ms	27.2319ms	36.7217 Ops/s	39.7687 Ops/s	$\textbf{\color{#d91a1a}-7.66\%}$

[github-actions]

$\color{#D29922}\textsf{\Large&amp;#x26A0;\kern{0.2cm}\normalsize Warning}$ Result of GPU Benchmark Tests

Total Benchmarks: 229. Improved: $\large\color{#35bf28}15$. Worsened: $\large\color{#d91a1a}31$.

Expand to view detailed results

Name	Max	Mean	Ops	Ops on Repo HEAD	Change
test_plain_set_nested	31.5310μs	11.4966μs	86.9824 KOps/s	93.4053 KOps/s	$\textbf{\color{#d91a1a}-6.88\%}$
test_plain_set_stack_nested	34.6700μs	11.4790μs	87.1155 KOps/s	93.5376 KOps/s	$\textbf{\color{#d91a1a}-6.87\%}$
test_plain_set_nested_inplace	37.3010μs	12.3116μs	81.2241 KOps/s	86.0871 KOps/s	$\textbf{\color{#d91a1a}-5.65\%}$
test_plain_set_stack_nested_inplace	45.4010μs	12.4193μs	80.5196 KOps/s	86.7603 KOps/s	$\textbf{\color{#d91a1a}-7.19\%}$
test_items	24.9810μs	2.9012μs	344.6835 KOps/s	341.9352 KOps/s	$\color{#35bf28}+0.80\%$
test_items_nested	0.3789ms	0.3482ms	2.8722 KOps/s	2.8318 KOps/s	$\color{#35bf28}+1.43\%$
test_items_nested_locked	0.4127ms	0.3491ms	2.8643 KOps/s	2.8331 KOps/s	$\color{#35bf28}+1.10\%$
test_items_nested_leaf	80.2310μs	57.7618μs	17.3125 KOps/s	17.2681 KOps/s	$\color{#35bf28}+0.26\%$
test_items_stack_nested	0.4079ms	0.3520ms	2.8408 KOps/s	2.8259 KOps/s	$\color{#35bf28}+0.53\%$
test_items_stack_nested_leaf	86.5000μs	58.6471μs	17.0511 KOps/s	16.9442 KOps/s	$\color{#35bf28}+0.63\%$
test_items_stack_nested_locked	0.3746ms	0.3515ms	2.8449 KOps/s	2.8170 KOps/s	$\color{#35bf28}+0.99\%$
test_keys	33.1800μs	3.7021μs	270.1191 KOps/s	292.1499 KOps/s	$\textbf{\color{#d91a1a}-7.54\%}$
test_keys_nested	99.1310μs	70.0521μs	14.2751 KOps/s	14.2313 KOps/s	$\color{#35bf28}+0.31\%$
test_keys_nested_locked	0.7041ms	76.1489μs	13.1322 KOps/s	13.0255 KOps/s	$\color{#35bf28}+0.82\%$
test_keys_nested_leaf	94.8410μs	60.9257μs	16.4134 KOps/s	16.2554 KOps/s	$\color{#35bf28}+0.97\%$
test_keys_stack_nested	0.1044ms	70.8573μs	14.1129 KOps/s	14.0932 KOps/s	$\color{#35bf28}+0.14\%$
test_keys_stack_nested_leaf	98.6610μs	61.4114μs	16.2836 KOps/s	15.9430 KOps/s	$\color{#35bf28}+2.14\%$
test_keys_stack_nested_locked	0.1062ms	76.3142μs	13.1037 KOps/s	13.0925 KOps/s	$\color{#35bf28}+0.09\%$
test_values	8.7802μs	0.8474μs	1.1801 MOps/s	1.1727 MOps/s	$\color{#35bf28}+0.62\%$
test_values_nested	55.5710μs	31.1672μs	32.0850 KOps/s	32.3531 KOps/s	$\color{#d91a1a}-0.83\%$
test_values_nested_locked	59.5710μs	32.5991μs	30.6757 KOps/s	30.7078 KOps/s	$\color{#d91a1a}-0.10\%$
test_values_nested_leaf	57.2810μs	33.3587μs	29.9772 KOps/s	29.8760 KOps/s	$\color{#35bf28}+0.34\%$
test_values_stack_nested	52.4110μs	31.3702μs	31.8773 KOps/s	31.6551 KOps/s	$\color{#35bf28}+0.70\%$
test_values_stack_nested_leaf	59.9010μs	33.6816μs	29.6898 KOps/s	29.4339 KOps/s	$\color{#35bf28}+0.87\%$
test_values_stack_nested_locked	0.1236ms	33.3930μs	29.9464 KOps/s	30.3629 KOps/s	$\color{#d91a1a}-1.37\%$
test_membership	1.4760μs	0.5168μs	1.9351 MOps/s	1.9693 MOps/s	$\color{#d91a1a}-1.73\%$
test_membership_nested	29.7910μs	2.0460μs	488.7680 KOps/s	498.9046 KOps/s	$\color{#d91a1a}-2.03\%$
test_membership_nested_leaf	16.6505μs	1.9956μs	501.1013 KOps/s	504.1315 KOps/s	$\color{#d91a1a}-0.60\%$
test_membership_stacked_nested	33.5200μs	2.0808μs	480.5782 KOps/s	480.8136 KOps/s	$\color{#d91a1a}-0.05\%$
test_membership_stacked_nested_leaf	28.7100μs	2.0529μs	487.1119 KOps/s	482.1761 KOps/s	$\color{#35bf28}+1.02\%$
test_membership_nested_last	30.0410μs	2.8939μs	345.5504 KOps/s	343.9843 KOps/s	$\color{#35bf28}+0.46\%$
test_membership_nested_leaf_last	29.1600μs	2.9231μs	342.1040 KOps/s	338.5766 KOps/s	$\color{#35bf28}+1.04\%$
test_membership_stacked_nested_last	25.3700μs	3.3408μs	299.3337 KOps/s	347.2531 KOps/s	$\textbf{\color{#d91a1a}-13.80\%}$
test_membership_stacked_nested_leaf_last	42.4500μs	3.3096μs	302.1516 KOps/s	342.5276 KOps/s	$\textbf{\color{#d91a1a}-11.79\%}$
test_nested_getleaf	30.4100μs	6.1511μs	162.5713 KOps/s	162.5516 KOps/s	$\color{#35bf28}+0.01\%$
test_nested_get	31.9000μs	5.8160μs	171.9391 KOps/s	171.3314 KOps/s	$\color{#35bf28}+0.35\%$
test_stacked_getleaf	38.6210μs	6.1392μs	162.8890 KOps/s	163.5920 KOps/s	$\color{#d91a1a}-0.43\%$
test_stacked_get	26.6900μs	5.8005μs	172.3993 KOps/s	172.7816 KOps/s	$\color{#d91a1a}-0.22\%$
test_nested_getitemleaf	33.8000μs	6.2288μs	160.5440 KOps/s	159.8520 KOps/s	$\color{#35bf28}+0.43\%$
test_nested_getitem	35.5610μs	5.9004μs	169.4792 KOps/s	168.5306 KOps/s	$\color{#35bf28}+0.56\%$
test_stacked_getitemleaf	44.9400μs	6.2219μs	160.7227 KOps/s	159.8207 KOps/s	$\color{#35bf28}+0.56\%$
test_stacked_getitem	32.6900μs	5.8824μs	169.9978 KOps/s	170.1987 KOps/s	$\color{#d91a1a}-0.12\%$
test_lock_nested	1.1588ms	0.3596ms	2.7807 KOps/s	2.6915 KOps/s	$\color{#35bf28}+3.31\%$
test_lock_stack_nested	0.3861ms	0.3305ms	3.0258 KOps/s	3.0126 KOps/s	$\color{#35bf28}+0.44\%$
test_unlock_nested	0.6233ms	0.2990ms	3.3448 KOps/s	3.2833 KOps/s	$\color{#35bf28}+1.87\%$
test_unlock_stack_nested	0.3275ms	0.2710ms	3.6906 KOps/s	3.6751 KOps/s	$\color{#35bf28}+0.42\%$
test_flatten_speed	0.1059ms	74.1427μs	13.4875 KOps/s	13.2400 KOps/s	$\color{#35bf28}+1.87\%$
test_unflatten_speed	0.3399ms	0.3014ms	3.3176 KOps/s	3.2650 KOps/s	$\color{#35bf28}+1.61\%$
test_common_ops	1.6038ms	0.6104ms	1.6383 KOps/s	1.7269 KOps/s	$\textbf{\color{#d91a1a}-5.13\%}$
test_creation	0.1864ms	1.4689μs	680.7897 KOps/s	692.1686 KOps/s	$\color{#d91a1a}-1.64\%$
test_creation_empty	30.9610μs	9.2424μs	108.1966 KOps/s	134.3623 KOps/s	$\textbf{\color{#d91a1a}-19.47\%}$
test_creation_nested_1	51.3110μs	10.7231μs	93.2569 KOps/s	114.5620 KOps/s	$\textbf{\color{#d91a1a}-18.60\%}$
test_creation_nested_2	52.3500μs	13.1266μs	76.1809 KOps/s	87.9016 KOps/s	$\textbf{\color{#d91a1a}-13.33\%}$
test_clone	1.9153ms	10.4191μs	95.9774 KOps/s	99.2732 KOps/s	$\color{#d91a1a}-3.32\%$
test_getitem[int]	1.1452ms	10.4622μs	95.5822 KOps/s	95.9968 KOps/s	$\color{#d91a1a}-0.43\%$
test_getitem[slice_int]	0.1130ms	20.1777μs	49.5597 KOps/s	49.5495 KOps/s	$\color{#35bf28}+0.02\%$
test_getitem[range]	0.1344ms	39.1556μs	25.5391 KOps/s	27.6726 KOps/s	$\textbf{\color{#d91a1a}-7.71\%}$
test_getitem[tuple]	0.1084ms	17.7601μs	56.3060 KOps/s	55.8395 KOps/s	$\color{#35bf28}+0.84\%$
test_getitem[list]	0.1311ms	33.4153μs	29.9264 KOps/s	30.8307 KOps/s	$\color{#d91a1a}-2.93\%$
test_setitem_dim[int]	38.0600μs	19.1063μs	52.3387 KOps/s	55.2942 KOps/s	$\textbf{\color{#d91a1a}-5.35\%}$
test_setitem_dim[slice_int]	86.0100μs	39.8050μs	25.1225 KOps/s	27.1617 KOps/s	$\textbf{\color{#d91a1a}-7.51\%}$
test_setitem_dim[range]	79.4210μs	55.9252μs	17.8810 KOps/s	19.3869 KOps/s	$\textbf{\color{#d91a1a}-7.77\%}$
test_setitem_dim[tuple]	54.1900μs	32.5853μs	30.6887 KOps/s	31.8249 KOps/s	$\color{#d91a1a}-3.57\%$
test_setitem	0.1347ms	15.4333μs	64.7949 KOps/s	70.4507 KOps/s	$\textbf{\color{#d91a1a}-8.03\%}$
test_set	0.1310ms	14.8811μs	67.1992 KOps/s	72.2131 KOps/s	$\textbf{\color{#d91a1a}-6.94\%}$
test_set_shared	1.7101ms	0.1442ms	6.9327 KOps/s	6.9767 KOps/s	$\color{#d91a1a}-0.63\%$
test_update	0.5253ms	18.1105μs	55.2165 KOps/s	60.2293 KOps/s	$\textbf{\color{#d91a1a}-8.32\%}$
test_update_nested	0.1246ms	23.6313μs	42.3168 KOps/s	44.7656 KOps/s	$\textbf{\color{#d91a1a}-5.47\%}$
test_update__nested	1.1115ms	24.7471μs	40.4087 KOps/s	43.0207 KOps/s	$\textbf{\color{#d91a1a}-6.07\%}$
test_set_nested	0.1215ms	15.7675μs	63.4218 KOps/s	67.1321 KOps/s	$\textbf{\color{#d91a1a}-5.53\%}$
test_set_nested_new	0.1257ms	18.2529μs	54.7858 KOps/s	57.1524 KOps/s	$\color{#d91a1a}-4.14\%$
test_select	65.0010μs	30.8990μs	32.3635 KOps/s	33.6252 KOps/s	$\color{#d91a1a}-3.75\%$
test_select_nested	75.0410μs	41.3762μs	24.1685 KOps/s	23.7582 KOps/s	$\color{#35bf28}+1.73\%$
test_exclude_nested	90.7410μs	61.3906μs	16.2891 KOps/s	16.1609 KOps/s	$\color{#35bf28}+0.79\%$
test_empty[True]	0.3158ms	0.2730ms	3.6625 KOps/s	3.6559 KOps/s	$\color{#35bf28}+0.18\%$
test_empty[False]	4.7860μs	0.7395μs	1.3523 MOps/s	1.3541 MOps/s	$\color{#d91a1a}-0.13\%$
test_to	88.1810μs	54.0930μs	18.4867 KOps/s	18.2072 KOps/s	$\color{#35bf28}+1.54\%$
test_to_nonblocking	88.1110μs	48.3879μs	20.6663 KOps/s	21.5577 KOps/s	$\color{#d91a1a}-4.13\%$
test_unbind_speed	0.8463ms	0.2264ms	4.4168 KOps/s	4.4187 KOps/s	$\color{#d91a1a}-0.04\%$
test_unbind_speed_stack0	0.2846ms	0.2253ms	4.4395 KOps/s	4.3949 KOps/s	$\color{#35bf28}+1.02\%$
test_unbind_speed_stack1	93.7262ms	0.6478ms	1.5436 KOps/s	1.5524 KOps/s	$\color{#d91a1a}-0.57\%$
test_split	94.1869ms	1.6726ms	597.8668 Ops/s	594.9992 Ops/s	$\color{#35bf28}+0.48\%$
test_chunk	95.4542ms	1.5639ms	639.4444 Ops/s	703.1940 Ops/s	$\textbf{\color{#d91a1a}-9.07\%}$
test_consolidate[False-None]	2.6436ms	2.5626ms	390.2242 Ops/s	355.7900 Ops/s	$\textbf{\color{#35bf28}+9.68\%}$
test_consolidate[default-None]	1.7668ms	1.6260ms	615.0190 Ops/s	608.0817 Ops/s	$\color{#35bf28}+1.14\%$
test_consolidate[reduce-overhead-None]	1.8675ms	1.6599ms	602.4623 Ops/s	599.1183 Ops/s	$\color{#35bf28}+0.56\%$
test_consolidate_njt[False-None]	6.6712ms	6.3601ms	157.2301 Ops/s	156.8697 Ops/s	$\color{#35bf28}+0.23\%$
test_to[False-False-None]	1.8158ms	1.6714ms	598.2873 Ops/s	588.8536 Ops/s	$\color{#35bf28}+1.60\%$
test_to[True-False-None]	1.5252ms	1.2386ms	807.3327 Ops/s	807.6642 Ops/s	$\color{#d91a1a}-0.04\%$
test_to[within-False-None]	4.1397ms	3.9070ms	255.9489 Ops/s	255.0598 Ops/s	$\color{#35bf28}+0.35\%$
test_to[True-default-None]	5.1737ms	4.9902ms	200.3923 Ops/s	191.4188 Ops/s	$\color{#35bf28}+4.69\%$
test_to_njt[False-False-None]	7.1159ms	6.8548ms	145.8829 Ops/s	145.4635 Ops/s	$\color{#35bf28}+0.29\%$
test_to_njt[True-False-None]	5.6280ms	5.3746ms	186.0618 Ops/s	184.2683 Ops/s	$\color{#35bf28}+0.97\%$
test_to_njt[within-False-None]	12.2723ms	12.0633ms	82.8958 Ops/s	82.6243 Ops/s	$\color{#35bf28}+0.33\%$
test_creation[device0]	0.5760ms	80.7403μs	12.3854 KOps/s	12.2305 KOps/s	$\color{#35bf28}+1.27\%$
test_creation_from_tensor	0.5809ms	83.9600μs	11.9104 KOps/s	11.6643 KOps/s	$\color{#35bf28}+2.11\%$
test_add_one[memmap_tensor0]	0.2023ms	6.6569μs	150.2199 KOps/s	148.1060 KOps/s	$\color{#35bf28}+1.43\%$
test_contiguous[memmap_tensor0]	1.8470μs	0.4020μs	2.4879 MOps/s	2.5048 MOps/s	$\color{#d91a1a}-0.68\%$
test_stack[memmap_tensor0]	44.1600μs	4.4543μs	224.5044 KOps/s	221.8087 KOps/s	$\color{#35bf28}+1.22\%$
test_memmaptd_index	1.8472ms	0.2495ms	4.0081 KOps/s	4.0661 KOps/s	$\color{#d91a1a}-1.43\%$
test_memmaptd_index_astensor	0.9372ms	0.3081ms	3.2457 KOps/s	3.3125 KOps/s	$\color{#d91a1a}-2.02\%$
test_memmaptd_index_op	0.9914ms	0.5958ms	1.6784 KOps/s	1.7425 KOps/s	$\color{#d91a1a}-3.68\%$
test_serialize_model	0.1314s	0.1307s	7.6536 Ops/s	7.6695 Ops/s	$\color{#d91a1a}-0.21\%$
test_serialize_model_pickle	1.3495s	1.2120s	0.8251 Ops/s	0.7766 Ops/s	$\textbf{\color{#35bf28}+6.24\%}$
test_serialize_weights	0.1320s	0.1296s	7.7165 Ops/s	7.6932 Ops/s	$\color{#35bf28}+0.30\%$
test_serialize_weights_returnearly	0.3194s	53.4081ms	18.7238 Ops/s	16.3817 Ops/s	$\textbf{\color{#35bf28}+14.30\%}$
test_serialize_weights_pickle	1.3771s	1.2208s	0.8191 Ops/s	0.8177 Ops/s	$\color{#35bf28}+0.18\%$
test_reshape_pytree	91.0310μs	21.9282μs	45.6035 KOps/s	45.0086 KOps/s	$\color{#35bf28}+1.32\%$
test_reshape_td	53.1810μs	26.3047μs	38.0161 KOps/s	37.3957 KOps/s	$\color{#35bf28}+1.66\%$
test_view_pytree	53.8700μs	21.9226μs	45.6151 KOps/s	45.1700 KOps/s	$\color{#35bf28}+0.99\%$
test_view_td	53.9010μs	28.9888μs	34.4960 KOps/s	33.7731 KOps/s	$\color{#35bf28}+2.14\%$
test_unbind_pytree	54.6610μs	27.7623μs	36.0201 KOps/s	35.9487 KOps/s	$\color{#35bf28}+0.20\%$
test_unbind_td	0.8251ms	35.1456μs	28.4531 KOps/s	28.2705 KOps/s	$\color{#35bf28}+0.65\%$
test_split_pytree	53.0510μs	29.2452μs	34.1937 KOps/s	33.5727 KOps/s	$\color{#35bf28}+1.85\%$
test_split_td	0.1741ms	37.4609μs	26.6945 KOps/s	24.9914 KOps/s	$\textbf{\color{#35bf28}+6.81\%}$
test_add_pytree	61.3600μs	33.9101μs	29.4897 KOps/s	30.0340 KOps/s	$\color{#d91a1a}-1.81\%$
test_add_td	96.1810μs	50.0709μs	19.9717 KOps/s	21.3699 KOps/s	$\textbf{\color{#d91a1a}-6.54\%}$
test_compile_add_one_nested[tensordict-compile]	0.1721ms	0.1180ms	8.4736 KOps/s	8.2548 KOps/s	$\color{#35bf28}+2.65\%$
test_compile_add_one_nested[tensordict-eager]	0.2135ms	0.1240ms	8.0661 KOps/s	7.9707 KOps/s	$\color{#35bf28}+1.20\%$
test_compile_add_one_nested[pytree-compile]	0.1642ms	96.0729μs	10.4088 KOps/s	10.5363 KOps/s	$\color{#d91a1a}-1.21\%$
test_compile_add_one_nested[pytree-eager]	1.2229ms	0.1485ms	6.7350 KOps/s	6.6926 KOps/s	$\color{#35bf28}+0.63\%$
test_compile_copy_nested[tensordict-compile]	53.7910μs	23.2198μs	43.0666 KOps/s	45.3002 KOps/s	$\color{#d91a1a}-4.93\%$
test_compile_copy_nested[tensordict-eager]	60.3610μs	26.3102μs	38.0081 KOps/s	34.2021 KOps/s	$\textbf{\color{#35bf28}+11.13\%}$
test_compile_copy_nested[pytree-compile]	0.2929ms	64.6555μs	15.4666 KOps/s	15.3662 KOps/s	$\color{#35bf28}+0.65\%$
test_compile_copy_nested[pytree-eager]	86.3910μs	49.2449μs	20.3067 KOps/s	20.3043 KOps/s	$\color{#35bf28}+0.01\%$
test_compile_add_one_flat[tensordict-compile]	0.1818ms	0.1421ms	7.0353 KOps/s	6.9366 KOps/s	$\color{#35bf28}+1.42\%$
test_compile_add_one_flat[tensordict-eager]	0.2955ms	0.2079ms	4.8110 KOps/s	4.8289 KOps/s	$\color{#d91a1a}-0.37\%$
test_compile_add_one_flat[tensorclass-compile]	0.1470ms	97.5724μs	10.2488 KOps/s	9.8300 KOps/s	$\color{#35bf28}+4.26\%$
test_compile_add_one_flat[tensorclass-eager]	0.1064ms	51.2880μs	19.4978 KOps/s	18.8941 KOps/s	$\color{#35bf28}+3.20\%$
test_compile_add_one_flat[pytree-compile]	0.1810ms	0.1358ms	7.3629 KOps/s	7.3757 KOps/s	$\color{#d91a1a}-0.17\%$
test_compile_add_one_flat[pytree-eager]	0.5604ms	0.4805ms	2.0810 KOps/s	2.0832 KOps/s	$\color{#d91a1a}-0.11\%$
test_compile_add_self_flat[tensordict-eager]	0.3643ms	0.2451ms	4.0808 KOps/s	4.0151 KOps/s	$\color{#35bf28}+1.63\%$
test_compile_add_self_flat[tensordict-compile]	0.1810ms	0.1428ms	7.0051 KOps/s	6.8198 KOps/s	$\color{#35bf28}+2.72\%$
test_compile_add_self_flat[tensorclass-eager]	0.1742ms	62.8632μs	15.9075 KOps/s	15.8445 KOps/s	$\color{#35bf28}+0.40\%$
test_compile_add_self_flat[tensorclass-compile]	0.1417ms	97.9750μs	10.2067 KOps/s	9.9761 KOps/s	$\color{#35bf28}+2.31\%$
test_compile_add_self_flat[pytree-eager]	0.4452ms	0.4060ms	2.4631 KOps/s	2.4628 KOps/s	$\color{#35bf28}+0.01\%$
test_compile_add_self_flat[pytree-compile]	0.1981ms	0.1349ms	7.4144 KOps/s	7.4284 KOps/s	$\color{#d91a1a}-0.19\%$
test_compile_copy_flat[tensordict-compile]	44.4910μs	19.3445μs	51.6942 KOps/s	55.9042 KOps/s	$\textbf{\color{#d91a1a}-7.53\%}$
test_compile_copy_flat[tensordict-eager]	53.2200μs	26.7268μs	37.4156 KOps/s	38.2940 KOps/s	$\color{#d91a1a}-2.29\%$
test_compile_copy_flat[pytree-compile]	0.1721ms	69.0120μs	14.4902 KOps/s	14.3281 KOps/s	$\color{#35bf28}+1.13\%$
test_compile_copy_flat[pytree-eager]	97.4810μs	51.2535μs	19.5109 KOps/s	19.3958 KOps/s	$\color{#35bf28}+0.59\%$
test_compile_assign_and_add[tensordict-compile]	1.5954ms	0.3865ms	2.5871 KOps/s	2.2332 KOps/s	$\textbf{\color{#35bf28}+15.85\%}$
test_compile_assign_and_add[tensordict-eager]	2.6518ms	2.5723ms	388.7637 Ops/s	384.3186 Ops/s	$\color{#35bf28}+1.16\%$
test_compile_assign_and_add[pytree-compile]	1.5586ms	0.3792ms	2.6368 KOps/s	2.2406 KOps/s	$\textbf{\color{#35bf28}+17.68\%}$
test_compile_assign_and_add[pytree-eager]	2.7171ms	2.6262ms	380.7791 Ops/s	376.2602 Ops/s	$\color{#35bf28}+1.20\%$
test_compile_indexing[tensor-tensordict-compile]	0.6351ms	0.1117ms	8.9505 KOps/s	8.5587 KOps/s	$\color{#35bf28}+4.58\%$
test_compile_indexing[tensor-tensordict-eager]	0.5555ms	78.5918μs	12.7240 KOps/s	11.9377 KOps/s	$\textbf{\color{#35bf28}+6.59\%}$
test_compile_indexing[tensor-tensorclass-compile]	0.4677ms	0.1050ms	9.5267 KOps/s	9.1384 KOps/s	$\color{#35bf28}+4.25\%$
test_compile_indexing[tensor-tensorclass-eager]	0.1110ms	68.1770μs	14.6677 KOps/s	14.1235 KOps/s	$\color{#35bf28}+3.85\%$
test_compile_indexing[tensor-pytree-compile]	0.1577ms	0.1056ms	9.4701 KOps/s	9.2671 KOps/s	$\color{#35bf28}+2.19\%$
test_compile_indexing[tensor-pytree-eager]	0.1398ms	68.0641μs	14.6920 KOps/s	14.4699 KOps/s	$\color{#35bf28}+1.54\%$
test_compile_indexing[slice-tensordict-compile]	0.1428ms	0.1002ms	9.9837 KOps/s	10.0778 KOps/s	$\color{#d91a1a}-0.93\%$
test_compile_indexing[slice-tensordict-eager]	0.1482ms	16.6871μs	59.9264 KOps/s	57.0380 KOps/s	$\textbf{\color{#35bf28}+5.06\%}$
test_compile_indexing[slice-tensorclass-compile]	0.1279ms	95.3931μs	10.4829 KOps/s	10.6217 KOps/s	$\color{#d91a1a}-1.31\%$
test_compile_indexing[slice-tensorclass-eager]	46.2800μs	15.6179μs	64.0290 KOps/s	64.9722 KOps/s	$\color{#d91a1a}-1.45\%$
test_compile_indexing[slice-pytree-compile]	0.1453ms	96.1855μs	10.3966 KOps/s	10.4811 KOps/s	$\color{#d91a1a}-0.81\%$
test_compile_indexing[slice-pytree-eager]	44.6300μs	15.6527μs	63.8866 KOps/s	64.2884 KOps/s	$\color{#d91a1a}-0.62\%$
test_compile_indexing[int-tensordict-compile]	0.1541ms	0.1004ms	9.9591 KOps/s	10.0081 KOps/s	$\color{#d91a1a}-0.49\%$
test_compile_indexing[int-tensordict-eager]	0.6052ms	16.3798μs	61.0509 KOps/s	57.8182 KOps/s	$\textbf{\color{#35bf28}+5.59\%}$
test_compile_indexing[int-tensorclass-compile]	0.1418ms	95.5409μs	10.4667 KOps/s	10.5195 KOps/s	$\color{#d91a1a}-0.50\%$
test_compile_indexing[int-tensorclass-eager]	48.9400μs	15.4745μs	64.6224 KOps/s	64.6503 KOps/s	$\color{#d91a1a}-0.04\%$
test_compile_indexing[int-pytree-compile]	0.1371ms	95.5687μs	10.4637 KOps/s	10.5071 KOps/s	$\color{#d91a1a}-0.41\%$
test_compile_indexing[int-pytree-eager]	91.6310μs	19.8012μs	50.5021 KOps/s	64.6714 KOps/s	$\textbf{\color{#d91a1a}-21.91\%}$
test_mod_add[eager]	0.1367ms	37.3001μs	26.8096 KOps/s	27.0701 KOps/s	$\color{#d91a1a}-0.96\%$
test_mod_add[compile]	0.1784ms	79.4768μs	12.5823 KOps/s	12.4137 KOps/s	$\color{#35bf28}+1.36\%$
test_mod_add[compile-overhead]	0.3188ms	0.1652ms	6.0526 KOps/s	5.5550 KOps/s	$\textbf{\color{#35bf28}+8.96\%}$
test_mod_wrap[eager]	0.3343ms	0.2512ms	3.9803 KOps/s	3.8585 KOps/s	$\color{#35bf28}+3.16\%$
test_mod_wrap[compile]	0.5698ms	0.2914ms	3.4313 KOps/s	3.4862 KOps/s	$\color{#d91a1a}-1.57\%$
test_mod_wrap[compile-overhead]	7.1247ms	3.8412ms	260.3367 Ops/s	262.7099 Ops/s	$\color{#d91a1a}-0.90\%$
test_mod_wrap_and_backward[eager]	1.4908ms	1.3601ms	735.2227 Ops/s	690.5795 Ops/s	$\textbf{\color{#35bf28}+6.46\%}$
test_mod_wrap_and_backward[compile]	1.3814ms	1.2601ms	793.5850 Ops/s	731.6543 Ops/s	$\textbf{\color{#35bf28}+8.46\%}$
test_mod_wrap_and_backward[compile-overhead]	1.3534ms	0.9243ms	1.0819 KOps/s	990.1941 Ops/s	$\textbf{\color{#35bf28}+9.27\%}$
test_seq_add[eager]	0.1772ms	0.1128ms	8.8636 KOps/s	9.0353 KOps/s	$\color{#d91a1a}-1.90\%$
test_seq_add[compile]	0.2368ms	95.4114μs	10.4809 KOps/s	11.3198 KOps/s	$\textbf{\color{#d91a1a}-7.41\%}$
test_seq_add[compile-overhead]	0.1791ms	0.1272ms	7.8646 KOps/s	7.8195 KOps/s	$\color{#35bf28}+0.58\%$
test_seq_wrap[eager]	0.5504ms	0.4262ms	2.3466 KOps/s	2.3725 KOps/s	$\color{#d91a1a}-1.09\%$
test_seq_wrap[compile]	0.3737ms	0.3146ms	3.1790 KOps/s	3.3370 KOps/s	$\color{#d91a1a}-4.73\%$
test_seq_wrap[compile-overhead]	0.2651ms	0.2201ms	4.5432 KOps/s	4.3866 KOps/s	$\color{#35bf28}+3.57\%$
test_func_call_runtime[False-eager]	0.7791ms	0.7350ms	1.3605 KOps/s	1.3379 KOps/s	$\color{#35bf28}+1.69\%$
test_func_call_runtime[False-compile]	0.9294ms	0.7438ms	1.3444 KOps/s	1.3539 KOps/s	$\color{#d91a1a}-0.70\%$
test_func_call_runtime[False-compile-overhead]	0.4027ms	0.3583ms	2.7913 KOps/s	2.7978 KOps/s	$\color{#d91a1a}-0.23\%$
test_func_call_runtime[True-eager]	0.9923ms	0.8953ms	1.1170 KOps/s	1.1015 KOps/s	$\color{#35bf28}+1.41\%$
test_func_call_runtime[True-compile]	0.9250ms	0.7965ms	1.2555 KOps/s	1.3139 KOps/s	$\color{#d91a1a}-4.45\%$
test_func_call_runtime[True-compile-overhead]	0.5423ms	0.3805ms	2.6281 KOps/s	2.6345 KOps/s	$\color{#d91a1a}-0.24\%$
test_func_call_cm_runtime[False-eager]	0.8219ms	0.7284ms	1.3729 KOps/s	1.3546 KOps/s	$\color{#35bf28}+1.36\%$
test_func_call_cm_runtime[False-compile]	1.0664ms	0.7396ms	1.3521 KOps/s	1.3456 KOps/s	$\color{#35bf28}+0.49\%$
test_func_call_cm_runtime[False-compile-overhead]	0.4357ms	0.3593ms	2.7831 KOps/s	2.7952 KOps/s	$\color{#d91a1a}-0.43\%$
test_func_call_cm_runtime[True-eager]	1.0796ms	1.0002ms	999.7563 Ops/s	1.0004 KOps/s	$\color{#d91a1a}-0.07\%$
test_func_call_cm_runtime[True-compile]	1.0330ms	0.7930ms	1.2610 KOps/s	1.2574 KOps/s	$\color{#35bf28}+0.29\%$
test_func_call_cm_runtime[True-compile-overhead]	0.4880ms	0.4063ms	2.4614 KOps/s	2.4380 KOps/s	$\color{#35bf28}+0.96\%$
test_vmap_func_call_cm_runtime[eager]	2.5325ms	2.0735ms	482.2807 Ops/s	477.6691 Ops/s	$\color{#35bf28}+0.97\%$
test_vmap_func_call_cm_runtime[compile]	0.8897ms	0.8033ms	1.2449 KOps/s	1.2334 KOps/s	$\color{#35bf28}+0.94\%$
test_vmap_func_call_cm_runtime[compile-overhead]	0.4546ms	0.4056ms	2.4656 KOps/s	2.4625 KOps/s	$\color{#35bf28}+0.13\%$
test_distributed	0.6509ms	0.1526ms	6.5528 KOps/s	8.4101 KOps/s	$\textbf{\color{#d91a1a}-22.08\%}$
test_tdmodule	62.0710μs	19.9385μs	50.1542 KOps/s	52.1547 KOps/s	$\color{#d91a1a}-3.84\%$
test_tdmodule_dispatch	75.8600μs	36.2018μs	27.6229 KOps/s	28.6413 KOps/s	$\color{#d91a1a}-3.56\%$
test_tdseq	39.5610μs	19.3960μs	51.5569 KOps/s	52.7835 KOps/s	$\color{#d91a1a}-2.32\%$
test_tdseq_dispatch	57.3710μs	37.8990μs	26.3859 KOps/s	27.7619 KOps/s	$\color{#d91a1a}-4.96\%$
test_instantiation_functorch	1.6083ms	1.5203ms	657.7713 Ops/s	650.5880 Ops/s	$\color{#35bf28}+1.10\%$
test_exec_functorch	0.1861ms	0.1404ms	7.1247 KOps/s	7.0577 KOps/s	$\color{#35bf28}+0.95\%$
test_exec_functional_call	0.1866ms	0.1384ms	7.2280 KOps/s	7.4602 KOps/s	$\color{#d91a1a}-3.11\%$
test_exec_td_decorator	0.3624ms	0.1787ms	5.5956 KOps/s	5.5461 KOps/s	$\color{#35bf28}+0.89\%$
test_vmap_mlp_speed_decorator[True-True]	0.8334ms	0.6832ms	1.4637 KOps/s	1.4642 KOps/s	$\color{#d91a1a}-0.03\%$
test_vmap_mlp_speed_decorator[True-False]	0.8378ms	0.6844ms	1.4610 KOps/s	1.4532 KOps/s	$\color{#35bf28}+0.54\%$
test_vmap_mlp_speed_decorator[False-True]	0.7615ms	0.6143ms	1.6280 KOps/s	1.6620 KOps/s	$\color{#d91a1a}-2.05\%$
test_vmap_mlp_speed_decorator[False-False]	0.6983ms	0.5896ms	1.6962 KOps/s	1.6738 KOps/s	$\color{#35bf28}+1.34\%$
test_vmap_transformer_speed_decorator[True-True]	19.4675ms	19.1094ms	52.3303 Ops/s	52.1419 Ops/s	$\color{#35bf28}+0.36\%$
test_vmap_transformer_speed_decorator[True-False]	20.0625ms	19.8342ms	50.4180 Ops/s	52.0898 Ops/s	$\color{#d91a1a}-3.21\%$
test_vmap_transformer_speed_decorator[False-True]	19.6174ms	19.0913ms	52.3798 Ops/s	52.5141 Ops/s	$\color{#d91a1a}-0.26\%$
test_vmap_transformer_speed_decorator[False-False]	20.0331ms	19.4570ms	51.3954 Ops/s	52.3865 Ops/s	$\color{#d91a1a}-1.89\%$
test_to_module_speed[True]	1.0792ms	0.9315ms	1.0735 KOps/s	1.0495 KOps/s	$\color{#35bf28}+2.29\%$
test_to_module_speed[False]	1.3082ms	0.9275ms	1.0781 KOps/s	1.0884 KOps/s	$\color{#d91a1a}-0.95\%$
test_tc_init	56.2610μs	35.9321μs	27.8302 KOps/s	28.4881 KOps/s	$\color{#d91a1a}-2.31\%$
test_tc_init_nested	0.2069ms	74.4251μs	13.4363 KOps/s	14.1496 KOps/s	$\textbf{\color{#d91a1a}-5.04\%}$
test_tc_first_layer_tensor	4.8944μs	0.7022μs	1.4240 MOps/s	1.4260 MOps/s	$\color{#d91a1a}-0.14\%$
test_tc_first_layer_nontensor	18.2700μs	2.3177μs	431.4685 KOps/s	429.4303 KOps/s	$\color{#35bf28}+0.47\%$
test_tc_second_layer_tensor	11.9867μs	1.4226μs	702.9326 KOps/s	703.9102 KOps/s	$\color{#d91a1a}-0.14\%$
test_tc_second_layer_nontensor	39.8300μs	3.0466μs	328.2374 KOps/s	327.8897 KOps/s	$\color{#35bf28}+0.11\%$
test_unbind	0.2233s	9.9358ms	100.6458 Ops/s	152.4530 Ops/s	$\textbf{\color{#d91a1a}-33.98\%}$
test_full_like	10.6615ms	9.1558ms	109.2206 Ops/s	108.6675 Ops/s	$\color{#35bf28}+0.51\%$
test_zeros_like	4.8546ms	4.3008ms	232.5154 Ops/s	115.6708 Ops/s	$\textbf{\color{#35bf28}+101.01\%}$
test_ones_like	10.0966ms	7.1359ms	140.1368 Ops/s	231.5833 Ops/s	$\textbf{\color{#d91a1a}-39.49\%}$
test_clone	6.6902ms	6.3467ms	157.5620 Ops/s	158.0967 Ops/s	$\color{#d91a1a}-0.34\%$
test_squeeze	60.3500μs	9.2463μs	108.1509 KOps/s	107.8226 KOps/s	$\color{#35bf28}+0.30\%$
test_unsqueeze	0.1213ms	69.6665μs	14.3541 KOps/s	14.5829 KOps/s	$\color{#d91a1a}-1.57\%$
test_split	0.3617ms	0.1547ms	6.4651 KOps/s	6.2182 KOps/s	$\color{#35bf28}+3.97\%$
test_permute	0.2585ms	0.1746ms	5.7287 KOps/s	5.5176 KOps/s	$\color{#35bf28}+3.83\%$
test_stack	50.6203ms	50.4039ms	19.8397 Ops/s	19.8492 Ops/s	$\color{#d91a1a}-0.05\%$
test_cat	50.5282ms	50.2747ms	19.8907 Ops/s	23.6898 Ops/s	$\textbf{\color{#d91a1a}-16.04\%}$