[Feature] COMPOSITE_LP_AGGREGATE env variable

[vmoens]

Stack from ghstack (oldest at bottom):

• -> [Feature] COMPOSITE_LP_AGGREGATE env variable #1190

[github-actions]

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

Total Benchmarks: 229. Improved: $\large\color{#35bf28}28$. Worsened: $\large\color{#d91a1a}12$.

Expand to view detailed results

Name	Max	Mean	Ops	Ops on Repo HEAD	Change
test_plain_set_nested	74.0310μs	11.5786μs	86.3661 KOps/s	75.5462 KOps/s	$\textbf{\color{#35bf28}+14.32\%}$
test_plain_set_stack_nested	40.9800μs	11.8179μs	84.6172 KOps/s	75.1125 KOps/s	$\textbf{\color{#35bf28}+12.65\%}$
test_plain_set_nested_inplace	37.3610μs	12.6428μs	79.0963 KOps/s	70.5168 KOps/s	$\textbf{\color{#35bf28}+12.17\%}$
test_plain_set_stack_nested_inplace	40.3810μs	12.6590μs	78.9950 KOps/s	69.9810 KOps/s	$\textbf{\color{#35bf28}+12.88\%}$
test_items	28.0710μs	2.9066μs	344.0447 KOps/s	338.0267 KOps/s	$\color{#35bf28}+1.78\%$
test_items_nested	0.4283ms	0.3709ms	2.6959 KOps/s	2.7519 KOps/s	$\color{#d91a1a}-2.03\%$
test_items_nested_locked	0.4056ms	0.3682ms	2.7156 KOps/s	2.7514 KOps/s	$\color{#d91a1a}-1.30\%$
test_items_nested_leaf	0.1698ms	58.6559μs	17.0486 KOps/s	17.2545 KOps/s	$\color{#d91a1a}-1.19\%$
test_items_stack_nested	0.3972ms	0.3681ms	2.7165 KOps/s	2.7776 KOps/s	$\color{#d91a1a}-2.20\%$
test_items_stack_nested_leaf	85.7520μs	58.3924μs	17.1255 KOps/s	16.7508 KOps/s	$\color{#35bf28}+2.24\%$
test_items_stack_nested_locked	0.4353ms	0.3683ms	2.7151 KOps/s	2.7281 KOps/s	$\color{#d91a1a}-0.48\%$
test_keys	27.8900μs	3.4500μs	289.8557 KOps/s	291.4224 KOps/s	$\color{#d91a1a}-0.54\%$
test_keys_nested	0.1233ms	88.8541μs	11.2544 KOps/s	11.3383 KOps/s	$\color{#d91a1a}-0.74\%$
test_keys_nested_locked	0.7566ms	94.4797μs	10.5843 KOps/s	10.6340 KOps/s	$\color{#d91a1a}-0.47\%$
test_keys_nested_leaf	0.1122ms	79.3856μs	12.5967 KOps/s	12.6177 KOps/s	$\color{#d91a1a}-0.17\%$
test_keys_stack_nested	0.1277ms	89.2500μs	11.2045 KOps/s	11.1652 KOps/s	$\color{#35bf28}+0.35\%$
test_keys_stack_nested_leaf	0.1149ms	79.5920μs	12.5641 KOps/s	12.5367 KOps/s	$\color{#35bf28}+0.22\%$
test_keys_stack_nested_locked	0.1248ms	95.2600μs	10.4976 KOps/s	10.5631 KOps/s	$\color{#d91a1a}-0.62\%$
test_values	6.1585μs	0.8522μs	1.1734 MOps/s	1.1819 MOps/s	$\color{#d91a1a}-0.72\%$
test_values_nested	60.9210μs	37.6277μs	26.5762 KOps/s	26.5848 KOps/s	$\color{#d91a1a}-0.03\%$
test_values_nested_locked	68.4910μs	39.5044μs	25.3137 KOps/s	25.5128 KOps/s	$\color{#d91a1a}-0.78\%$
test_values_nested_leaf	65.3610μs	42.1983μs	23.6977 KOps/s	23.9674 KOps/s	$\color{#d91a1a}-1.13\%$
test_values_stack_nested	0.1878ms	38.3019μs	26.1084 KOps/s	25.8399 KOps/s	$\color{#35bf28}+1.04\%$
test_values_stack_nested_leaf	78.6010μs	43.0457μs	23.2311 KOps/s	23.3579 KOps/s	$\color{#d91a1a}-0.54\%$
test_values_stack_nested_locked	93.4920μs	39.6305μs	25.2331 KOps/s	24.7453 KOps/s	$\color{#35bf28}+1.97\%$
test_membership	1.8965μs	0.5203μs	1.9221 MOps/s	1.9810 MOps/s	$\color{#d91a1a}-2.97\%$
test_membership_nested	14.1950μs	2.0390μs	490.4313 KOps/s	472.4362 KOps/s	$\color{#35bf28}+3.81\%$
test_membership_nested_leaf	18.4305μs	2.0114μs	497.1682 KOps/s	491.0779 KOps/s	$\color{#35bf28}+1.24\%$
test_membership_stacked_nested	27.0100μs	2.0726μs	482.4764 KOps/s	478.2236 KOps/s	$\color{#35bf28}+0.89\%$
test_membership_stacked_nested_leaf	62.3910μs	2.0880μs	478.9344 KOps/s	481.2868 KOps/s	$\color{#d91a1a}-0.49\%$
test_membership_nested_last	31.8300μs	3.1074μs	321.8099 KOps/s	317.4649 KOps/s	$\color{#35bf28}+1.37\%$
test_membership_nested_leaf_last	37.0610μs	3.1299μs	319.4963 KOps/s	312.3128 KOps/s	$\color{#35bf28}+2.30\%$
test_membership_stacked_nested_last	0.1725ms	3.1276μs	319.7345 KOps/s	322.6004 KOps/s	$\color{#d91a1a}-0.89\%$
test_membership_stacked_nested_leaf_last	34.6410μs	3.1139μs	321.1415 KOps/s	314.3216 KOps/s	$\color{#35bf28}+2.17\%$
test_nested_getleaf	0.1747ms	6.1329μs	163.0547 KOps/s	161.5533 KOps/s	$\color{#35bf28}+0.93\%$
test_nested_get	0.1826ms	5.8826μs	169.9925 KOps/s	170.8474 KOps/s	$\color{#d91a1a}-0.50\%$
test_stacked_getleaf	32.3910μs	6.1905μs	161.5371 KOps/s	161.9238 KOps/s	$\color{#d91a1a}-0.24\%$
test_stacked_get	0.2025ms	5.9403μs	168.3430 KOps/s	171.0731 KOps/s	$\color{#d91a1a}-1.60\%$
test_nested_getitemleaf	38.2410μs	6.4714μs	154.5268 KOps/s	152.7337 KOps/s	$\color{#35bf28}+1.17\%$
test_nested_getitem	27.6800μs	6.1735μs	161.9836 KOps/s	160.7831 KOps/s	$\color{#35bf28}+0.75\%$
test_stacked_getitemleaf	41.4910μs	6.4437μs	155.1904 KOps/s	153.9642 KOps/s	$\color{#35bf28}+0.80\%$
test_stacked_getitem	43.8100μs	6.1193μs	163.4172 KOps/s	162.1157 KOps/s	$\color{#35bf28}+0.80\%$
test_lock_nested	8.9141ms	0.3515ms	2.8453 KOps/s	2.8754 KOps/s	$\color{#d91a1a}-1.05\%$
test_lock_stack_nested	0.4934ms	0.3471ms	2.8807 KOps/s	2.9013 KOps/s	$\color{#d91a1a}-0.71\%$
test_unlock_nested	0.3919ms	0.2865ms	3.4904 KOps/s	3.5728 KOps/s	$\color{#d91a1a}-2.31\%$
test_unlock_stack_nested	0.4122ms	0.2842ms	3.5185 KOps/s	3.5328 KOps/s	$\color{#d91a1a}-0.40\%$
test_flatten_speed	0.1136ms	75.9746μs	13.1623 KOps/s	13.2249 KOps/s	$\color{#d91a1a}-0.47\%$
test_unflatten_speed	0.3663ms	0.3247ms	3.0802 KOps/s	3.1206 KOps/s	$\color{#d91a1a}-1.29\%$
test_common_ops	0.7779ms	0.6122ms	1.6335 KOps/s	1.5412 KOps/s	$\textbf{\color{#35bf28}+5.99\%}$
test_creation	94.8810μs	1.7498μs	571.5097 KOps/s	570.3175 KOps/s	$\color{#35bf28}+0.21\%$
test_creation_empty	65.7710μs	7.0829μs	141.1846 KOps/s	98.9343 KOps/s	$\textbf{\color{#35bf28}+42.71\%}$
test_creation_nested_1	33.3000μs	8.8229μs	113.3410 KOps/s	84.1737 KOps/s	$\textbf{\color{#35bf28}+34.65\%}$
test_creation_nested_2	31.4810μs	11.7109μs	85.3904 KOps/s	69.1048 KOps/s	$\textbf{\color{#35bf28}+23.57\%}$
test_clone	50.3510μs	11.3220μs	88.3234 KOps/s	92.8127 KOps/s	$\color{#d91a1a}-4.84\%$
test_getitem[int]	1.3477ms	11.1769μs	89.4700 KOps/s	93.9234 KOps/s	$\color{#d91a1a}-4.74\%$
test_getitem[slice_int]	0.1637ms	21.7980μs	45.8758 KOps/s	47.8513 KOps/s	$\color{#d91a1a}-4.13\%$
test_getitem[range]	0.1357ms	39.4863μs	25.3252 KOps/s	26.2927 KOps/s	$\color{#d91a1a}-3.68\%$
test_getitem[tuple]	0.1092ms	19.1773μs	52.1451 KOps/s	54.6604 KOps/s	$\color{#d91a1a}-4.60\%$
test_getitem[list]	0.1579ms	34.9088μs	28.6460 KOps/s	29.4729 KOps/s	$\color{#d91a1a}-2.81\%$
test_setitem_dim[int]	0.1255ms	21.3405μs	46.8592 KOps/s	50.3663 KOps/s	$\textbf{\color{#d91a1a}-6.96\%}$
test_setitem_dim[slice_int]	80.1310μs	40.3122μs	24.8064 KOps/s	25.2338 KOps/s	$\color{#d91a1a}-1.69\%$
test_setitem_dim[range]	80.4610μs	55.5216μs	18.0110 KOps/s	18.3701 KOps/s	$\color{#d91a1a}-1.95\%$
test_setitem_dim[tuple]	81.3610μs	34.6511μs	28.8592 KOps/s	29.6454 KOps/s	$\color{#d91a1a}-2.65\%$
test_setitem	0.1117ms	15.3776μs	65.0295 KOps/s	61.1153 KOps/s	$\textbf{\color{#35bf28}+6.40\%}$
test_set	53.8200μs	14.9242μs	67.0051 KOps/s	62.6621 KOps/s	$\textbf{\color{#35bf28}+6.93\%}$
test_set_shared	0.5097ms	0.1623ms	6.1611 KOps/s	6.1843 KOps/s	$\color{#d91a1a}-0.38\%$
test_update	0.3994ms	17.1597μs	58.2760 KOps/s	51.6086 KOps/s	$\textbf{\color{#35bf28}+12.92\%}$
test_update_nested	53.2800μs	22.7811μs	43.8961 KOps/s	39.8732 KOps/s	$\textbf{\color{#35bf28}+10.09\%}$
test_update__nested	0.5139ms	26.6324μs	37.5482 KOps/s	37.7311 KOps/s	$\color{#d91a1a}-0.48\%$
test_set_nested	54.6200μs	16.4916μs	60.6371 KOps/s	57.9622 KOps/s	$\color{#35bf28}+4.61\%$
test_set_nested_new	0.1762ms	18.8890μs	52.9410 KOps/s	52.4406 KOps/s	$\color{#35bf28}+0.95\%$
test_select	71.3210μs	31.2182μs	32.0326 KOps/s	31.5583 KOps/s	$\color{#35bf28}+1.50\%$
test_select_nested	0.1012ms	44.5745μs	22.4343 KOps/s	22.9861 KOps/s	$\color{#d91a1a}-2.40\%$
test_exclude_nested	94.0520μs	64.2857μs	15.5556 KOps/s	16.0405 KOps/s	$\color{#d91a1a}-3.02\%$
test_empty[True]	0.3260ms	0.2986ms	3.3489 KOps/s	3.4253 KOps/s	$\color{#d91a1a}-2.23\%$
test_empty[False]	3.8191μs	0.8332μs	1.2001 MOps/s	1.2056 MOps/s	$\color{#d91a1a}-0.45\%$
test_to	88.2110μs	57.5023μs	17.3906 KOps/s	17.3427 KOps/s	$\color{#35bf28}+0.28\%$
test_to_nonblocking	93.5710μs	48.5394μs	20.6018 KOps/s	20.8331 KOps/s	$\color{#d91a1a}-1.11\%$
test_unbind_speed	0.2862ms	0.2461ms	4.0627 KOps/s	4.1967 KOps/s	$\color{#d91a1a}-3.19\%$
test_unbind_speed_stack0	0.3270ms	0.2447ms	4.0860 KOps/s	4.2173 KOps/s	$\color{#d91a1a}-3.11\%$
test_unbind_speed_stack1	95.3441ms	0.7453ms	1.3417 KOps/s	1.3585 KOps/s	$\color{#d91a1a}-1.24\%$
test_split	95.4225ms	1.6266ms	614.7725 Ops/s	631.2088 Ops/s	$\color{#d91a1a}-2.60\%$
test_chunk	97.5497ms	1.6308ms	613.2066 Ops/s	627.7503 Ops/s	$\color{#d91a1a}-2.32\%$
test_consolidate[False-None]	2.8127ms	2.7046ms	369.7461 Ops/s	334.4725 Ops/s	$\textbf{\color{#35bf28}+10.55\%}$
test_consolidate[default-None]	1.9905ms	1.7419ms	574.0911 Ops/s	587.4905 Ops/s	$\color{#d91a1a}-2.28\%$
test_consolidate[reduce-overhead-None]	1.8930ms	1.7766ms	562.8751 Ops/s	571.1336 Ops/s	$\color{#d91a1a}-1.45\%$
test_consolidate_njt[False-None]	6.7969ms	6.6019ms	151.4725 Ops/s	151.0708 Ops/s	$\color{#35bf28}+0.27\%$
test_to[False-False-None]	1.9363ms	1.7113ms	584.3505 Ops/s	586.2574 Ops/s	$\color{#d91a1a}-0.33\%$
test_to[True-False-None]	1.6301ms	1.3795ms	724.8760 Ops/s	724.4853 Ops/s	$\color{#35bf28}+0.05\%$
test_to[within-False-None]	4.4992ms	4.2090ms	237.5865 Ops/s	170.1876 Ops/s	$\textbf{\color{#35bf28}+39.60\%}$
test_to[True-default-None]	5.7700ms	5.5421ms	180.4357 Ops/s	187.6667 Ops/s	$\color{#d91a1a}-3.85\%$
test_to_njt[False-False-None]	7.1320ms	6.9389ms	144.1148 Ops/s	141.9668 Ops/s	$\color{#35bf28}+1.51\%$
test_to_njt[True-False-None]	6.0367ms	5.6168ms	178.0366 Ops/s	175.6529 Ops/s	$\color{#35bf28}+1.36\%$
test_to_njt[within-False-None]	12.5172ms	12.3466ms	80.9943 Ops/s	79.7008 Ops/s	$\color{#35bf28}+1.62\%$
test_creation[device0]	0.4589ms	81.7910μs	12.2263 KOps/s	12.2701 KOps/s	$\color{#d91a1a}-0.36\%$
test_creation_from_tensor	0.4460ms	84.6110μs	11.8188 KOps/s	11.7123 KOps/s	$\color{#35bf28}+0.91\%$
test_add_one[memmap_tensor0]	0.2323ms	7.4582μs	134.0802 KOps/s	141.4191 KOps/s	$\textbf{\color{#d91a1a}-5.19\%}$
test_contiguous[memmap_tensor0]	1.8396μs	0.4374μs	2.2865 MOps/s	2.3410 MOps/s	$\color{#d91a1a}-2.33\%$
test_stack[memmap_tensor0]	60.0110μs	4.6666μs	214.2885 KOps/s	230.2495 KOps/s	$\textbf{\color{#d91a1a}-6.93\%}$
test_memmaptd_index	1.5371ms	0.2559ms	3.9077 KOps/s	4.1530 KOps/s	$\textbf{\color{#d91a1a}-5.91\%}$
test_memmaptd_index_astensor	0.5240ms	0.3173ms	3.1519 KOps/s	3.3222 KOps/s	$\textbf{\color{#d91a1a}-5.13\%}$
test_memmaptd_index_op	0.7371ms	0.5898ms	1.6954 KOps/s	1.6367 KOps/s	$\color{#35bf28}+3.59\%$
test_serialize_model	0.1320s	0.1311s	7.6261 Ops/s	7.6655 Ops/s	$\color{#d91a1a}-0.51\%$
test_serialize_model_pickle	1.3740s	1.2171s	0.8216 Ops/s	0.8177 Ops/s	$\color{#35bf28}+0.48\%$
test_serialize_weights	0.1328s	0.1301s	7.6878 Ops/s	7.6618 Ops/s	$\color{#35bf28}+0.34\%$
test_serialize_weights_returnearly	0.3297s	55.3887ms	18.0542 Ops/s	14.5219 Ops/s	$\textbf{\color{#35bf28}+24.32\%}$
test_serialize_weights_pickle	1.3661s	1.2162s	0.8222 Ops/s	0.8219 Ops/s	$\color{#35bf28}+0.03\%$
test_reshape_pytree	0.1075ms	22.4736μs	44.4966 KOps/s	44.7761 KOps/s	$\color{#d91a1a}-0.62\%$
test_reshape_td	57.0710μs	27.7553μs	36.0292 KOps/s	35.9327 KOps/s	$\color{#35bf28}+0.27\%$
test_view_pytree	44.4910μs	22.2157μs	45.0132 KOps/s	45.9306 KOps/s	$\color{#d91a1a}-2.00\%$
test_view_td	66.3910μs	32.2319μs	31.0252 KOps/s	31.3709 KOps/s	$\color{#d91a1a}-1.10\%$
test_unbind_pytree	85.8810μs	28.4658μs	35.1299 KOps/s	34.8885 KOps/s	$\color{#35bf28}+0.69\%$
test_unbind_td	0.7439ms	37.4259μs	26.7194 KOps/s	27.2849 KOps/s	$\color{#d91a1a}-2.07\%$
test_split_pytree	63.1110μs	30.5425μs	32.7413 KOps/s	32.9106 KOps/s	$\color{#d91a1a}-0.51\%$
test_split_td	0.9642ms	39.6485μs	25.2217 KOps/s	26.1767 KOps/s	$\color{#d91a1a}-3.65\%$
test_add_pytree	0.1735ms	36.4203μs	27.4572 KOps/s	28.3919 KOps/s	$\color{#d91a1a}-3.29\%$
test_add_td	0.1866ms	47.5503μs	21.0303 KOps/s	18.7130 KOps/s	$\textbf{\color{#35bf28}+12.38\%}$
test_compile_add_one_nested[tensordict-compile]	0.2727ms	0.1244ms	8.0398 KOps/s	7.9319 KOps/s	$\color{#35bf28}+1.36\%$
test_compile_add_one_nested[tensordict-eager]	0.5403ms	0.1332ms	7.5079 KOps/s	7.5664 KOps/s	$\color{#d91a1a}-0.77\%$
test_compile_add_one_nested[pytree-compile]	0.1331ms	96.3658μs	10.3771 KOps/s	10.1785 KOps/s	$\color{#35bf28}+1.95\%$
test_compile_add_one_nested[pytree-eager]	1.3125ms	0.1520ms	6.5774 KOps/s	6.6553 KOps/s	$\color{#d91a1a}-1.17\%$
test_compile_copy_nested[tensordict-compile]	0.1625ms	24.2081μs	41.3086 KOps/s	41.2118 KOps/s	$\color{#35bf28}+0.23\%$
test_compile_copy_nested[tensordict-eager]	0.1188ms	29.3167μs	34.1103 KOps/s	34.2833 KOps/s	$\color{#d91a1a}-0.50\%$
test_compile_copy_nested[pytree-compile]	0.4244ms	64.5303μs	15.4966 KOps/s	15.3549 KOps/s	$\color{#35bf28}+0.92\%$
test_compile_copy_nested[pytree-eager]	95.5220μs	48.8437μs	20.4735 KOps/s	20.3238 KOps/s	$\color{#35bf28}+0.74\%$
test_compile_add_one_flat[tensordict-compile]	0.1861ms	0.1419ms	7.0477 KOps/s	7.1976 KOps/s	$\color{#d91a1a}-2.08\%$
test_compile_add_one_flat[tensordict-eager]	0.3550ms	0.2202ms	4.5422 KOps/s	4.6032 KOps/s	$\color{#d91a1a}-1.32\%$
test_compile_add_one_flat[tensorclass-compile]	0.1434ms	98.8163μs	10.1198 KOps/s	10.3808 KOps/s	$\color{#d91a1a}-2.51\%$
test_compile_add_one_flat[tensorclass-eager]	0.2067ms	56.1166μs	17.8200 KOps/s	17.9129 KOps/s	$\color{#d91a1a}-0.52\%$
test_compile_add_one_flat[pytree-compile]	0.2685ms	0.1355ms	7.3784 KOps/s	7.5384 KOps/s	$\color{#d91a1a}-2.12\%$
test_compile_add_one_flat[pytree-eager]	0.6309ms	0.4912ms	2.0357 KOps/s	2.0890 KOps/s	$\color{#d91a1a}-2.55\%$
test_compile_add_self_flat[tensordict-eager]	0.3692ms	0.2643ms	3.7833 KOps/s	3.8194 KOps/s	$\color{#d91a1a}-0.94\%$
test_compile_add_self_flat[tensordict-compile]	0.2951ms	0.1432ms	6.9829 KOps/s	7.1858 KOps/s	$\color{#d91a1a}-2.82\%$
test_compile_add_self_flat[tensorclass-eager]	0.2156ms	67.8916μs	14.7294 KOps/s	14.6456 KOps/s	$\color{#35bf28}+0.57\%$
test_compile_add_self_flat[tensorclass-compile]	0.2437ms	99.7147μs	10.0286 KOps/s	10.3159 KOps/s	$\color{#d91a1a}-2.78\%$
test_compile_add_self_flat[pytree-eager]	0.5717ms	0.4050ms	2.4689 KOps/s	2.5024 KOps/s	$\color{#d91a1a}-1.34\%$
test_compile_add_self_flat[pytree-compile]	0.1809ms	0.1343ms	7.4436 KOps/s	7.5183 KOps/s	$\color{#d91a1a}-0.99\%$
test_compile_copy_flat[tensordict-compile]	0.2157ms	24.4546μs	40.8921 KOps/s	53.8403 KOps/s	$\textbf{\color{#d91a1a}-24.05\%}$
test_compile_copy_flat[tensordict-eager]	55.8910μs	31.2745μs	31.9749 KOps/s	32.0303 KOps/s	$\color{#d91a1a}-0.17\%$
test_compile_copy_flat[pytree-compile]	0.1047ms	71.4541μs	13.9950 KOps/s	14.0867 KOps/s	$\color{#d91a1a}-0.65\%$
test_compile_copy_flat[pytree-eager]	85.1010μs	52.0155μs	19.2250 KOps/s	19.3798 KOps/s	$\color{#d91a1a}-0.80\%$
test_compile_assign_and_add[tensordict-compile]	1.6634ms	0.3980ms	2.5124 KOps/s	2.1816 KOps/s	$\textbf{\color{#35bf28}+15.16\%}$
test_compile_assign_and_add[tensordict-eager]	3.0908ms	2.6794ms	373.2228 Ops/s	383.1160 Ops/s	$\color{#d91a1a}-2.58\%$
test_compile_assign_and_add[pytree-compile]	1.6214ms	0.4387ms	2.2796 KOps/s	2.2642 KOps/s	$\color{#35bf28}+0.68\%$
test_compile_assign_and_add[pytree-eager]	3.4866ms	2.7100ms	368.9976 Ops/s	383.1150 Ops/s	$\color{#d91a1a}-3.68\%$
test_compile_indexing[tensor-tensordict-compile]	0.5288ms	0.1178ms	8.4919 KOps/s	8.6215 KOps/s	$\color{#d91a1a}-1.50\%$
test_compile_indexing[tensor-tensordict-eager]	0.5537ms	78.9303μs	12.6694 KOps/s	12.3238 KOps/s	$\color{#35bf28}+2.80\%$
test_compile_indexing[tensor-tensorclass-compile]	0.6356ms	0.1105ms	9.0504 KOps/s	8.9528 KOps/s	$\color{#35bf28}+1.09\%$
test_compile_indexing[tensor-tensorclass-eager]	0.3149ms	71.8100μs	13.9256 KOps/s	14.4259 KOps/s	$\color{#d91a1a}-3.47\%$
test_compile_indexing[tensor-pytree-compile]	0.2997ms	0.1124ms	8.8977 KOps/s	8.8610 KOps/s	$\color{#35bf28}+0.41\%$
test_compile_indexing[tensor-pytree-eager]	0.2492ms	71.4343μs	13.9989 KOps/s	13.8179 KOps/s	$\color{#35bf28}+1.31\%$
test_compile_indexing[slice-tensordict-compile]	0.2403ms	0.1055ms	9.4795 KOps/s	10.0088 KOps/s	$\textbf{\color{#d91a1a}-5.29\%}$
test_compile_indexing[slice-tensordict-eager]	0.1426ms	17.8998μs	55.8666 KOps/s	47.4486 KOps/s	$\textbf{\color{#35bf28}+17.74\%}$
test_compile_indexing[slice-tensorclass-compile]	0.2754ms	0.1001ms	9.9894 KOps/s	10.4603 KOps/s	$\color{#d91a1a}-4.50\%$
test_compile_indexing[slice-tensorclass-eager]	0.1628ms	16.2781μs	61.4322 KOps/s	62.9642 KOps/s	$\color{#d91a1a}-2.43\%$
test_compile_indexing[slice-pytree-compile]	0.2785ms	0.1006ms	9.9388 KOps/s	10.4059 KOps/s	$\color{#d91a1a}-4.49\%$
test_compile_indexing[slice-pytree-eager]	0.1539ms	16.1423μs	61.9489 KOps/s	62.9429 KOps/s	$\color{#d91a1a}-1.58\%$
test_compile_indexing[int-tensordict-compile]	0.2795ms	0.1076ms	9.2943 KOps/s	9.8678 KOps/s	$\textbf{\color{#d91a1a}-5.81\%}$
test_compile_indexing[int-tensordict-eager]	0.5589ms	17.6074μs	56.7942 KOps/s	54.6238 KOps/s	$\color{#35bf28}+3.97\%$
test_compile_indexing[int-tensorclass-compile]	0.2818ms	0.1011ms	9.8958 KOps/s	9.9757 KOps/s	$\color{#d91a1a}-0.80\%$
test_compile_indexing[int-tensorclass-eager]	0.2074ms	19.2466μs	51.9572 KOps/s	63.9761 KOps/s	$\textbf{\color{#d91a1a}-18.79\%}$
test_compile_indexing[int-pytree-compile]	0.2740ms	0.1009ms	9.9073 KOps/s	10.0446 KOps/s	$\color{#d91a1a}-1.37\%$
test_compile_indexing[int-pytree-eager]	0.4424ms	16.2562μs	61.5150 KOps/s	63.1534 KOps/s	$\color{#d91a1a}-2.59\%$
test_mod_add[eager]	0.1761ms	37.6529μs	26.5584 KOps/s	23.9173 KOps/s	$\textbf{\color{#35bf28}+11.04\%}$
test_mod_add[compile]	0.4056ms	83.8949μs	11.9197 KOps/s	12.2626 KOps/s	$\color{#d91a1a}-2.80\%$
test_mod_add[compile-overhead]	0.3312ms	0.1699ms	5.8843 KOps/s	5.6902 KOps/s	$\color{#35bf28}+3.41\%$
test_mod_wrap[eager]	0.4144ms	0.2525ms	3.9603 KOps/s	3.8981 KOps/s	$\color{#35bf28}+1.60\%$
test_mod_wrap[compile]	0.4442ms	0.2845ms	3.5146 KOps/s	3.4791 KOps/s	$\color{#35bf28}+1.02\%$
test_mod_wrap[compile-overhead]	7.1310ms	3.7177ms	268.9816 Ops/s	268.8193 Ops/s	$\color{#35bf28}+0.06\%$
test_mod_wrap_and_backward[eager]	1.7011ms	1.5095ms	662.4854 Ops/s	686.0508 Ops/s	$\color{#d91a1a}-3.43\%$
test_mod_wrap_and_backward[compile]	1.8224ms	1.2724ms	785.9321 Ops/s	727.5407 Ops/s	$\textbf{\color{#35bf28}+8.03\%}$
test_mod_wrap_and_backward[compile-overhead]	1.4901ms	0.9532ms	1.0491 KOps/s	978.2418 Ops/s	$\textbf{\color{#35bf28}+7.24\%}$
test_seq_add[eager]	0.2595ms	0.1152ms	8.6774 KOps/s	8.2213 KOps/s	$\textbf{\color{#35bf28}+5.55\%}$
test_seq_add[compile]	0.2384ms	89.5902μs	11.1619 KOps/s	10.9887 KOps/s	$\color{#35bf28}+1.58\%$
test_seq_add[compile-overhead]	0.2743ms	0.1300ms	7.6932 KOps/s	7.7726 KOps/s	$\color{#d91a1a}-1.02\%$
test_seq_wrap[eager]	0.5880ms	0.4192ms	2.3852 KOps/s	2.3135 KOps/s	$\color{#35bf28}+3.10\%$
test_seq_wrap[compile]	0.4511ms	0.3003ms	3.3296 KOps/s	3.2827 KOps/s	$\color{#35bf28}+1.43\%$
test_seq_wrap[compile-overhead]	0.3008ms	0.2247ms	4.4511 KOps/s	4.3623 KOps/s	$\color{#35bf28}+2.04\%$
test_func_call_runtime[False-eager]	0.8981ms	0.7467ms	1.3392 KOps/s	1.3225 KOps/s	$\color{#35bf28}+1.27\%$
test_func_call_runtime[False-compile]	0.9582ms	0.7804ms	1.2813 KOps/s	1.3459 KOps/s	$\color{#d91a1a}-4.80\%$
test_func_call_runtime[False-compile-overhead]	0.5005ms	0.3661ms	2.7313 KOps/s	2.7133 KOps/s	$\color{#35bf28}+0.66\%$
test_func_call_runtime[True-eager]	1.0752ms	0.9197ms	1.0873 KOps/s	1.0771 KOps/s	$\color{#35bf28}+0.95\%$
test_func_call_runtime[True-compile]	0.9123ms	0.7663ms	1.3050 KOps/s	1.3029 KOps/s	$\color{#35bf28}+0.17\%$
test_func_call_runtime[True-compile-overhead]	0.4360ms	0.3874ms	2.5815 KOps/s	2.5865 KOps/s	$\color{#d91a1a}-0.19\%$
test_func_call_cm_runtime[False-eager]	0.9350ms	0.7565ms	1.3218 KOps/s	1.3460 KOps/s	$\color{#d91a1a}-1.80\%$
test_func_call_cm_runtime[False-compile]	0.8869ms	0.7530ms	1.3281 KOps/s	1.3111 KOps/s	$\color{#35bf28}+1.29\%$
test_func_call_cm_runtime[False-compile-overhead]	0.4279ms	0.3697ms	2.7047 KOps/s	2.7180 KOps/s	$\color{#d91a1a}-0.49\%$
test_func_call_cm_runtime[True-eager]	1.1586ms	1.0151ms	985.1447 Ops/s	985.1765 Ops/s	$-0.00\%$
test_func_call_cm_runtime[True-compile]	1.2218ms	1.0052ms	994.8649 Ops/s	976.6549 Ops/s	$\color{#35bf28}+1.86\%$
test_func_call_cm_runtime[True-compile-overhead]	1.1546ms	1.0043ms	995.7042 Ops/s	977.8164 Ops/s	$\color{#35bf28}+1.83\%$
test_vmap_func_call_cm_runtime[eager]	2.5063ms	2.1082ms	474.3332 Ops/s	466.2647 Ops/s	$\color{#35bf28}+1.73\%$
test_vmap_func_call_cm_runtime[compile]	0.9931ms	0.8213ms	1.2176 KOps/s	1.2215 KOps/s	$\color{#d91a1a}-0.32\%$
test_vmap_func_call_cm_runtime[compile-overhead]	0.5858ms	0.4167ms	2.3997 KOps/s	2.3802 KOps/s	$\color{#35bf28}+0.82\%$
test_distributed	6.7719ms	0.2461ms	4.0631 KOps/s	8.4867 KOps/s	$\textbf{\color{#d91a1a}-52.12\%}$
test_tdmodule	0.1862ms	20.5206μs	48.7315 KOps/s	45.5527 KOps/s	$\textbf{\color{#35bf28}+6.98\%}$
test_tdmodule_dispatch	52.5110μs	35.3603μs	28.2803 KOps/s	26.1973 KOps/s	$\textbf{\color{#35bf28}+7.95\%}$
test_tdseq	0.1939ms	21.0166μs	47.5815 KOps/s	45.6766 KOps/s	$\color{#35bf28}+4.17\%$
test_tdseq_dispatch	58.8610μs	37.1978μs	26.8833 KOps/s	24.3800 KOps/s	$\textbf{\color{#35bf28}+10.27\%}$
test_instantiation_functorch	1.7640ms	1.6165ms	618.6127 Ops/s	635.1508 Ops/s	$\color{#d91a1a}-2.60\%$
test_exec_functorch	0.2845ms	0.1516ms	6.5944 KOps/s	6.8011 KOps/s	$\color{#d91a1a}-3.04\%$
test_exec_functional_call	0.2394ms	0.1424ms	7.0211 KOps/s	7.0928 KOps/s	$\color{#d91a1a}-1.01\%$
test_exec_td_decorator	0.3820ms	0.1930ms	5.1803 KOps/s	5.2306 KOps/s	$\color{#d91a1a}-0.96\%$
test_vmap_mlp_speed_decorator[True-True]	0.8366ms	0.6930ms	1.4430 KOps/s	1.4350 KOps/s	$\color{#35bf28}+0.56\%$
test_vmap_mlp_speed_decorator[True-False]	0.8305ms	0.6923ms	1.4445 KOps/s	1.3899 KOps/s	$\color{#35bf28}+3.93\%$
test_vmap_mlp_speed_decorator[False-True]	0.7907ms	0.6065ms	1.6489 KOps/s	1.6602 KOps/s	$\color{#d91a1a}-0.68\%$
test_vmap_mlp_speed_decorator[False-False]	0.7593ms	0.6250ms	1.6000 KOps/s	1.6528 KOps/s	$\color{#d91a1a}-3.19\%$
test_vmap_transformer_speed_decorator[True-True]	20.1017ms	19.3565ms	51.6623 Ops/s	51.6597 Ops/s	$+0.01\%$
test_vmap_transformer_speed_decorator[True-False]	19.5181ms	19.3668ms	51.6347 Ops/s	51.7428 Ops/s	$\color{#d91a1a}-0.21\%$
test_vmap_transformer_speed_decorator[False-True]	19.4166ms	19.2350ms	51.9887 Ops/s	52.1505 Ops/s	$\color{#d91a1a}-0.31\%$
test_vmap_transformer_speed_decorator[False-False]	19.4592ms	19.2278ms	52.0081 Ops/s	52.2124 Ops/s	$\color{#d91a1a}-0.39\%$
test_to_module_speed[True]	1.4350ms	0.9737ms	1.0270 KOps/s	1.0366 KOps/s	$\color{#d91a1a}-0.92\%$
test_to_module_speed[False]	1.0383ms	0.9543ms	1.0478 KOps/s	1.0437 KOps/s	$\color{#35bf28}+0.40\%$
test_tc_init	67.2310μs	33.9618μs	29.4449 KOps/s	26.2759 KOps/s	$\textbf{\color{#35bf28}+12.06\%}$
test_tc_init_nested	0.1044ms	68.3820μs	14.6237 KOps/s	12.8944 KOps/s	$\textbf{\color{#35bf28}+13.41\%}$
test_tc_first_layer_tensor	28.8200μs	0.8138μs	1.2288 MOps/s	1.4211 MOps/s	$\textbf{\color{#d91a1a}-13.53\%}$
test_tc_first_layer_nontensor	20.2500μs	2.2426μs	445.9161 KOps/s	439.7544 KOps/s	$\color{#35bf28}+1.40\%$
test_tc_second_layer_tensor	8.9000μs	1.4330μs	697.8513 KOps/s	706.5930 KOps/s	$\color{#d91a1a}-1.24\%$
test_tc_second_layer_nontensor	31.1600μs	3.0236μs	330.7340 KOps/s	328.4114 KOps/s	$\color{#35bf28}+0.71\%$
test_unbind	0.2182s	12.1892ms	82.0397 Ops/s	141.8902 Ops/s	$\textbf{\color{#d91a1a}-42.18\%}$
test_full_like	10.8442ms	9.6664ms	103.4512 Ops/s	103.6403 Ops/s	$\color{#d91a1a}-0.18\%$
test_zeros_like	4.9704ms	4.3854ms	228.0279 Ops/s	113.4273 Ops/s	$\textbf{\color{#35bf28}+101.03\%}$
test_ones_like	5.6214ms	4.3971ms	227.4230 Ops/s	225.4611 Ops/s	$\color{#35bf28}+0.87\%$
test_clone	7.5119ms	6.8365ms	146.2726 Ops/s	147.9279 Ops/s	$\color{#d91a1a}-1.12\%$
test_squeeze	0.1273ms	10.0111μs	99.8889 KOps/s	102.2445 KOps/s	$\color{#d91a1a}-2.30\%$
test_unsqueeze	0.1350ms	72.5473μs	13.7841 KOps/s	13.3165 KOps/s	$\color{#35bf28}+3.51\%$
test_split	0.3707ms	0.1595ms	6.2715 KOps/s	6.0887 KOps/s	$\color{#35bf28}+3.00\%$
test_permute	0.3289ms	0.1780ms	5.6178 KOps/s	5.3594 KOps/s	$\color{#35bf28}+4.82\%$
test_stack	52.3677ms	51.4358ms	19.4417 Ops/s	19.2893 Ops/s	$\color{#35bf28}+0.79\%$
test_cat	52.9824ms	51.5133ms	19.4124 Ops/s	19.4026 Ops/s	$\color{#35bf28}+0.05\%$