[Feature] flexible return type when indexing prob sequences

[vmoens]

Stack from ghstack (oldest at bottom):

• -> [Feature] flexible return type when indexing prob sequences #1189

[github-actions]

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

Total Benchmarks: 229. Improved: $\large\color{#35bf28}48$. Worsened: $\large\color{#d91a1a}11$.

Expand to view detailed results

Name	Max	Mean	Ops	Ops on Repo HEAD	Change
test_plain_set_nested	34.2710μs	11.4245μs	87.5314 KOps/s	74.0584 KOps/s	$\textbf{\color{#35bf28}+18.19\%}$
test_plain_set_stack_nested	34.1510μs	11.6590μs	85.7703 KOps/s	72.6472 KOps/s	$\textbf{\color{#35bf28}+18.06\%}$
test_plain_set_nested_inplace	52.6800μs	12.5686μs	79.5632 KOps/s	69.0652 KOps/s	$\textbf{\color{#35bf28}+15.20\%}$
test_plain_set_stack_nested_inplace	43.5010μs	12.5399μs	79.7454 KOps/s	68.9896 KOps/s	$\textbf{\color{#35bf28}+15.59\%}$
test_items	21.1400μs	2.9149μs	343.0705 KOps/s	340.7535 KOps/s	$\color{#35bf28}+0.68\%$
test_items_nested	0.4158ms	0.3618ms	2.7641 KOps/s	2.7452 KOps/s	$\color{#35bf28}+0.69\%$
test_items_nested_locked	0.5574ms	0.3688ms	2.7116 KOps/s	2.7397 KOps/s	$\color{#d91a1a}-1.03\%$
test_items_nested_leaf	84.3020μs	58.5174μs	17.0889 KOps/s	17.0415 KOps/s	$\color{#35bf28}+0.28\%$
test_items_stack_nested	0.4154ms	0.3646ms	2.7425 KOps/s	2.7356 KOps/s	$\color{#35bf28}+0.25\%$
test_items_stack_nested_leaf	90.6520μs	59.9455μs	16.6818 KOps/s	16.6472 KOps/s	$\color{#35bf28}+0.21\%$
test_items_stack_nested_locked	0.3920ms	0.3659ms	2.7333 KOps/s	2.7265 KOps/s	$\color{#35bf28}+0.25\%$
test_keys	30.0910μs	3.4713μs	288.0767 KOps/s	288.4604 KOps/s	$\color{#d91a1a}-0.13\%$
test_keys_nested	0.1269ms	87.8884μs	11.3781 KOps/s	11.4683 KOps/s	$\color{#d91a1a}-0.79\%$
test_keys_nested_locked	0.7249ms	93.9821μs	10.6403 KOps/s	10.7784 KOps/s	$\color{#d91a1a}-1.28\%$
test_keys_nested_leaf	0.1212ms	78.4432μs	12.7481 KOps/s	12.8161 KOps/s	$\color{#d91a1a}-0.53\%$
test_keys_stack_nested	0.1513ms	87.8316μs	11.3854 KOps/s	11.3594 KOps/s	$\color{#35bf28}+0.23\%$
test_keys_stack_nested_leaf	0.1176ms	78.8555μs	12.6814 KOps/s	12.6611 KOps/s	$\color{#35bf28}+0.16\%$
test_keys_stack_nested_locked	0.1291ms	93.1290μs	10.7378 KOps/s	10.6756 KOps/s	$\color{#35bf28}+0.58\%$
test_values	5.8633μs	0.8541μs	1.1708 MOps/s	1.1784 MOps/s	$\color{#d91a1a}-0.64\%$
test_values_nested	67.2910μs	37.9280μs	26.3658 KOps/s	26.9034 KOps/s	$\color{#d91a1a}-2.00\%$
test_values_nested_locked	65.8210μs	39.1954μs	25.5132 KOps/s	25.8360 KOps/s	$\color{#d91a1a}-1.25\%$
test_values_nested_leaf	88.8920μs	41.7198μs	23.9694 KOps/s	23.9851 KOps/s	$\color{#d91a1a}-0.07\%$
test_values_stack_nested	0.1086ms	38.5644μs	25.9307 KOps/s	26.4775 KOps/s	$\color{#d91a1a}-2.07\%$
test_values_stack_nested_leaf	75.6710μs	42.3412μs	23.6176 KOps/s	23.5387 KOps/s	$\color{#35bf28}+0.34\%$
test_values_stack_nested_locked	87.4610μs	39.8939μs	25.0665 KOps/s	25.4017 KOps/s	$\color{#d91a1a}-1.32\%$
test_membership	2.6170μs	0.5002μs	1.9993 MOps/s	1.9568 MOps/s	$\color{#35bf28}+2.17\%$
test_membership_nested	16.1300μs	1.9998μs	500.0458 KOps/s	493.2413 KOps/s	$\color{#35bf28}+1.38\%$
test_membership_nested_leaf	17.4300μs	2.0234μs	494.2097 KOps/s	493.2794 KOps/s	$\color{#35bf28}+0.19\%$
test_membership_stacked_nested	29.6400μs	2.1269μs	470.1635 KOps/s	481.1856 KOps/s	$\color{#d91a1a}-2.29\%$
test_membership_stacked_nested_leaf	32.3010μs	2.0879μs	478.9574 KOps/s	481.1007 KOps/s	$\color{#d91a1a}-0.45\%$
test_membership_nested_last	36.4510μs	3.0844μs	324.2097 KOps/s	317.5868 KOps/s	$\color{#35bf28}+2.09\%$
test_membership_nested_leaf_last	39.7400μs	3.0835μs	324.3050 KOps/s	315.8241 KOps/s	$\color{#35bf28}+2.69\%$
test_membership_stacked_nested_last	54.7310μs	8.1967μs	122.0005 KOps/s	243.0257 KOps/s	$\textbf{\color{#d91a1a}-49.80\%}$
test_membership_stacked_nested_leaf_last	35.0000μs	8.2478μs	121.2447 KOps/s	242.1743 KOps/s	$\textbf{\color{#d91a1a}-49.93\%}$
test_nested_getleaf	37.0610μs	6.0865μs	164.2979 KOps/s	162.2235 KOps/s	$\color{#35bf28}+1.28\%$
test_nested_get	34.6510μs	5.8415μs	171.1877 KOps/s	171.2673 KOps/s	$\color{#d91a1a}-0.05\%$
test_stacked_getleaf	38.9610μs	6.1337μs	163.0344 KOps/s	163.8438 KOps/s	$\color{#d91a1a}-0.49\%$
test_stacked_get	33.1410μs	5.8095μs	172.1327 KOps/s	173.0299 KOps/s	$\color{#d91a1a}-0.52\%$
test_nested_getitemleaf	32.4410μs	6.4002μs	156.2440 KOps/s	155.5115 KOps/s	$\color{#35bf28}+0.47\%$
test_nested_getitem	32.9410μs	6.1759μs	161.9202 KOps/s	162.9436 KOps/s	$\color{#d91a1a}-0.63\%$
test_stacked_getitemleaf	37.5010μs	6.4437μs	155.1910 KOps/s	155.5388 KOps/s	$\color{#d91a1a}-0.22\%$
test_stacked_getitem	27.5900μs	6.0949μs	164.0707 KOps/s	163.9280 KOps/s	$\color{#35bf28}+0.09\%$
test_lock_nested	8.8157ms	0.3514ms	2.8455 KOps/s	2.8012 KOps/s	$\color{#35bf28}+1.58\%$
test_lock_stack_nested	0.3911ms	0.3377ms	2.9616 KOps/s	2.8578 KOps/s	$\color{#35bf28}+3.63\%$
test_unlock_nested	0.3503ms	0.2818ms	3.5485 KOps/s	3.4481 KOps/s	$\color{#35bf28}+2.91\%$
test_unlock_stack_nested	0.3184ms	0.2753ms	3.6321 KOps/s	3.4994 KOps/s	$\color{#35bf28}+3.79\%$
test_flatten_speed	0.1247ms	75.5420μs	13.2377 KOps/s	13.1493 KOps/s	$\color{#35bf28}+0.67\%$
test_unflatten_speed	0.4502ms	0.3225ms	3.1012 KOps/s	3.0708 KOps/s	$\color{#35bf28}+0.99\%$
test_common_ops	0.7388ms	0.5941ms	1.6832 KOps/s	1.4615 KOps/s	$\textbf{\color{#35bf28}+15.17\%}$
test_creation	0.1003ms	1.7405μs	574.5598 KOps/s	573.9256 KOps/s	$\color{#35bf28}+0.11\%$
test_creation_empty	32.3810μs	6.9871μs	143.1211 KOps/s	92.6450 KOps/s	$\textbf{\color{#35bf28}+54.48\%}$
test_creation_nested_1	42.1000μs	8.7205μs	114.6720 KOps/s	79.6560 KOps/s	$\textbf{\color{#35bf28}+43.96\%}$
test_creation_nested_2	33.5000μs	11.5260μs	86.7602 KOps/s	65.6566 KOps/s	$\textbf{\color{#35bf28}+32.14\%}$
test_clone	45.3510μs	10.7415μs	93.0971 KOps/s	87.7606 KOps/s	$\textbf{\color{#35bf28}+6.08\%}$
test_getitem[int]	1.1478ms	10.8619μs	92.0647 KOps/s	89.6719 KOps/s	$\color{#35bf28}+2.67\%$
test_getitem[slice_int]	0.1064ms	21.0626μs	47.4775 KOps/s	45.6785 KOps/s	$\color{#35bf28}+3.94\%$
test_getitem[range]	0.1245ms	38.8388μs	25.7474 KOps/s	25.0841 KOps/s	$\color{#35bf28}+2.64\%$
test_getitem[tuple]	0.1067ms	18.6403μs	53.6473 KOps/s	52.5165 KOps/s	$\color{#35bf28}+2.15\%$
test_getitem[list]	0.1315ms	33.2375μs	30.0865 KOps/s	28.2871 KOps/s	$\textbf{\color{#35bf28}+6.36\%}$
test_setitem_dim[int]	41.2010μs	19.8500μs	50.3779 KOps/s	47.5882 KOps/s	$\textbf{\color{#35bf28}+5.86\%}$
test_setitem_dim[slice_int]	60.5610μs	39.3330μs	25.4239 KOps/s	25.0482 KOps/s	$\color{#35bf28}+1.50\%$
test_setitem_dim[range]	88.6320μs	54.8144μs	18.2434 KOps/s	18.0270 KOps/s	$\color{#35bf28}+1.20\%$
test_setitem_dim[tuple]	61.3010μs	33.2495μs	30.0756 KOps/s	29.2012 KOps/s	$\color{#35bf28}+2.99\%$
test_setitem	49.0810μs	14.7614μs	67.7442 KOps/s	58.2152 KOps/s	$\textbf{\color{#35bf28}+16.37\%}$
test_set	55.5510μs	14.2883μs	69.9874 KOps/s	58.8965 KOps/s	$\textbf{\color{#35bf28}+18.83\%}$
test_set_shared	0.5076ms	0.1621ms	6.1709 KOps/s	6.1777 KOps/s	$\color{#d91a1a}-0.11\%$
test_update	0.3693ms	16.3153μs	61.2921 KOps/s	47.6036 KOps/s	$\textbf{\color{#35bf28}+28.76\%}$
test_update_nested	49.5710μs	21.4561μs	46.6068 KOps/s	37.5994 KOps/s	$\textbf{\color{#35bf28}+23.96\%}$
test_update__nested	0.4961ms	25.8945μs	38.6183 KOps/s	36.6641 KOps/s	$\textbf{\color{#35bf28}+5.33\%}$
test_set_nested	64.3110μs	15.3908μs	64.9738 KOps/s	55.0176 KOps/s	$\textbf{\color{#35bf28}+18.10\%}$
test_set_nested_new	52.2500μs	17.6452μs	56.6728 KOps/s	48.5175 KOps/s	$\textbf{\color{#35bf28}+16.81\%}$
test_select	65.5910μs	29.3658μs	34.0532 KOps/s	30.6089 KOps/s	$\textbf{\color{#35bf28}+11.25\%}$
test_select_nested	88.5810μs	44.5962μs	22.4234 KOps/s	22.6758 KOps/s	$\color{#d91a1a}-1.11\%$
test_exclude_nested	0.1064ms	64.3675μs	15.5358 KOps/s	15.6699 KOps/s	$\color{#d91a1a}-0.86\%$
test_empty[True]	0.3448ms	0.2949ms	3.3905 KOps/s	3.3825 KOps/s	$\color{#35bf28}+0.24\%$
test_empty[False]	4.3211μs	0.8316μs	1.2024 MOps/s	1.2063 MOps/s	$\color{#d91a1a}-0.32\%$
test_to	84.9410μs	56.8937μs	17.5766 KOps/s	17.5322 KOps/s	$\color{#35bf28}+0.25\%$
test_to_nonblocking	88.6210μs	49.0766μs	20.3763 KOps/s	20.4592 KOps/s	$\color{#d91a1a}-0.41\%$
test_unbind_speed	0.2666ms	0.2413ms	4.1448 KOps/s	4.0604 KOps/s	$\color{#35bf28}+2.08\%$
test_unbind_speed_stack0	0.2771ms	0.2352ms	4.2509 KOps/s	4.0693 KOps/s	$\color{#35bf28}+4.46\%$
test_unbind_speed_stack1	92.3105ms	0.7264ms	1.3766 KOps/s	1.3700 KOps/s	$\color{#35bf28}+0.48\%$
test_split	93.4549ms	1.5934ms	627.6045 Ops/s	611.5255 Ops/s	$\color{#35bf28}+2.63\%$
test_chunk	95.8648ms	1.6076ms	622.0548 Ops/s	606.7582 Ops/s	$\color{#35bf28}+2.52\%$
test_consolidate[False-None]	3.4107ms	2.7565ms	362.7823 Ops/s	363.8128 Ops/s	$\color{#d91a1a}-0.28\%$
test_consolidate[default-None]	1.8169ms	1.7404ms	574.5760 Ops/s	579.7707 Ops/s	$\color{#d91a1a}-0.90\%$
test_consolidate[reduce-overhead-None]	1.8297ms	1.7605ms	568.0225 Ops/s	560.1837 Ops/s	$\color{#35bf28}+1.40\%$
test_consolidate_njt[False-None]	7.1986ms	6.9121ms	144.6743 Ops/s	145.6552 Ops/s	$\color{#d91a1a}-0.67\%$
test_to[False-False-None]	1.8789ms	1.7674ms	565.7933 Ops/s	559.1862 Ops/s	$\color{#35bf28}+1.18\%$
test_to[True-False-None]	1.7290ms	1.4065ms	710.9699 Ops/s	705.5864 Ops/s	$\color{#35bf28}+0.76\%$
test_to[within-False-None]	4.3574ms	4.1823ms	239.1013 Ops/s	229.9538 Ops/s	$\color{#35bf28}+3.98\%$
test_to[True-default-None]	5.9665ms	5.5067ms	181.5984 Ops/s	185.4693 Ops/s	$\color{#d91a1a}-2.09\%$
test_to_njt[False-False-None]	7.1762ms	7.0643ms	141.5559 Ops/s	139.1086 Ops/s	$\color{#35bf28}+1.76\%$
test_to_njt[True-False-None]	6.0786ms	5.6110ms	178.2222 Ops/s	170.5579 Ops/s	$\color{#35bf28}+4.49\%$
test_to_njt[within-False-None]	12.4929ms	12.3914ms	80.7011 Ops/s	78.3343 Ops/s	$\color{#35bf28}+3.02\%$
test_creation[device0]	0.2837ms	85.5903μs	11.6836 KOps/s	12.1382 KOps/s	$\color{#d91a1a}-3.75\%$
test_creation_from_tensor	0.5462ms	89.4112μs	11.1843 KOps/s	11.2308 KOps/s	$\color{#d91a1a}-0.41\%$
test_add_one[memmap_tensor0]	0.5776ms	6.9077μs	144.7661 KOps/s	136.3970 KOps/s	$\textbf{\color{#35bf28}+6.14\%}$
test_contiguous[memmap_tensor0]	2.4685μs	0.4203μs	2.3792 MOps/s	2.3489 MOps/s	$\color{#35bf28}+1.29\%$
test_stack[memmap_tensor0]	36.3310μs	4.4449μs	224.9782 KOps/s	215.1177 KOps/s	$\color{#35bf28}+4.58\%$
test_memmaptd_index	1.4640ms	0.2446ms	4.0883 KOps/s	3.9479 KOps/s	$\color{#35bf28}+3.55\%$
test_memmaptd_index_astensor	0.4545ms	0.3081ms	3.2458 KOps/s	3.1869 KOps/s	$\color{#35bf28}+1.85\%$
test_memmaptd_index_op	0.7049ms	0.5610ms	1.7826 KOps/s	1.5404 KOps/s	$\textbf{\color{#35bf28}+15.72\%}$
test_serialize_model	0.4291s	0.1731s	5.7778 Ops/s	7.6324 Ops/s	$\textbf{\color{#d91a1a}-24.30\%}$
test_serialize_model_pickle	1.3843s	1.2157s	0.8226 Ops/s	0.8248 Ops/s	$\color{#d91a1a}-0.26\%$
test_serialize_weights	0.1321s	0.1304s	7.6684 Ops/s	7.6814 Ops/s	$\color{#d91a1a}-0.17\%$
test_serialize_weights_returnearly	0.3295s	54.9902ms	18.1851 Ops/s	14.9987 Ops/s	$\textbf{\color{#35bf28}+21.24\%}$
test_serialize_weights_pickle	2.9377s	1.7565s	0.5693 Ops/s	0.8254 Ops/s	$\textbf{\color{#d91a1a}-31.02\%}$
test_reshape_pytree	52.5400μs	22.4919μs	44.4604 KOps/s	44.0636 KOps/s	$\color{#35bf28}+0.90\%$
test_reshape_td	69.6310μs	27.8680μs	35.8835 KOps/s	35.4466 KOps/s	$\color{#35bf28}+1.23\%$
test_view_pytree	61.4110μs	22.8158μs	43.8294 KOps/s	44.8365 KOps/s	$\color{#d91a1a}-2.25\%$
test_view_td	70.4810μs	32.9571μs	30.3425 KOps/s	28.7999 KOps/s	$\textbf{\color{#35bf28}+5.36\%}$
test_unbind_pytree	54.3700μs	28.3846μs	35.2304 KOps/s	34.7553 KOps/s	$\color{#35bf28}+1.37\%$
test_unbind_td	0.8051ms	37.2263μs	26.8627 KOps/s	26.5785 KOps/s	$\color{#35bf28}+1.07\%$
test_split_pytree	58.2610μs	30.8017μs	32.4658 KOps/s	31.8726 KOps/s	$\color{#35bf28}+1.86\%$
test_split_td	0.6490ms	39.1240μs	25.5598 KOps/s	25.0535 KOps/s	$\color{#35bf28}+2.02\%$
test_add_pytree	76.9410μs	35.3173μs	28.3147 KOps/s	27.0926 KOps/s	$\color{#35bf28}+4.51\%$
test_add_td	87.8320μs	47.5066μs	21.0497 KOps/s	18.4998 KOps/s	$\textbf{\color{#35bf28}+13.78\%}$
test_compile_add_one_nested[tensordict-compile]	0.1787ms	0.1253ms	7.9838 KOps/s	7.5456 KOps/s	$\textbf{\color{#35bf28}+5.81\%}$
test_compile_add_one_nested[tensordict-eager]	0.2222ms	0.1324ms	7.5555 KOps/s	7.4313 KOps/s	$\color{#35bf28}+1.67\%$
test_compile_add_one_nested[pytree-compile]	0.1449ms	97.5140μs	10.2549 KOps/s	10.0130 KOps/s	$\color{#35bf28}+2.42\%$
test_compile_add_one_nested[pytree-eager]	1.3615ms	0.1485ms	6.7329 KOps/s	6.3580 KOps/s	$\textbf{\color{#35bf28}+5.90\%}$
test_compile_copy_nested[tensordict-compile]	65.3110μs	25.0195μs	39.9688 KOps/s	42.6738 KOps/s	$\textbf{\color{#d91a1a}-6.34\%}$
test_compile_copy_nested[tensordict-eager]	61.2510μs	29.6712μs	33.7027 KOps/s	33.8460 KOps/s	$\color{#d91a1a}-0.42\%$
test_compile_copy_nested[pytree-compile]	0.3915ms	66.5363μs	15.0294 KOps/s	15.2793 KOps/s	$\color{#d91a1a}-1.64\%$
test_compile_copy_nested[pytree-eager]	78.4210μs	49.2607μs	20.3002 KOps/s	20.6826 KOps/s	$\color{#d91a1a}-1.85\%$
test_compile_add_one_flat[tensordict-compile]	0.1839ms	0.1431ms	6.9857 KOps/s	7.0047 KOps/s	$\color{#d91a1a}-0.27\%$
test_compile_add_one_flat[tensordict-eager]	0.3195ms	0.2163ms	4.6229 KOps/s	4.6126 KOps/s	$\color{#35bf28}+0.22\%$
test_compile_add_one_flat[tensorclass-compile]	0.1610ms	98.9461μs	10.1065 KOps/s	10.0614 KOps/s	$\color{#35bf28}+0.45\%$
test_compile_add_one_flat[tensorclass-eager]	0.1175ms	55.8030μs	17.9202 KOps/s	16.5397 KOps/s	$\textbf{\color{#35bf28}+8.35\%}$
test_compile_add_one_flat[pytree-compile]	0.2075ms	0.1358ms	7.3646 KOps/s	7.3332 KOps/s	$\color{#35bf28}+0.43\%$
test_compile_add_one_flat[pytree-eager]	0.5297ms	0.4787ms	2.0891 KOps/s	1.9659 KOps/s	$\textbf{\color{#35bf28}+6.26\%}$
test_compile_add_self_flat[tensordict-eager]	0.3935ms	0.2606ms	3.8368 KOps/s	3.8060 KOps/s	$\color{#35bf28}+0.81\%$
test_compile_add_self_flat[tensordict-compile]	0.1829ms	0.1450ms	6.8960 KOps/s	6.8498 KOps/s	$\color{#35bf28}+0.67\%$
test_compile_add_self_flat[tensorclass-eager]	0.1570ms	70.6810μs	14.1481 KOps/s	13.6654 KOps/s	$\color{#35bf28}+3.53\%$
test_compile_add_self_flat[tensorclass-compile]	0.1487ms	0.1012ms	9.8856 KOps/s	9.6094 KOps/s	$\color{#35bf28}+2.87\%$
test_compile_add_self_flat[pytree-eager]	0.4548ms	0.4090ms	2.4448 KOps/s	2.3568 KOps/s	$\color{#35bf28}+3.73\%$
test_compile_add_self_flat[pytree-compile]	0.1745ms	0.1368ms	7.3086 KOps/s	7.2717 KOps/s	$\color{#35bf28}+0.51\%$
test_compile_copy_flat[tensordict-compile]	58.4810μs	19.5170μs	51.2375 KOps/s	55.2720 KOps/s	$\textbf{\color{#d91a1a}-7.30\%}$
test_compile_copy_flat[tensordict-eager]	69.5110μs	30.6016μs	32.6781 KOps/s	32.1756 KOps/s	$\color{#35bf28}+1.56\%$
test_compile_copy_flat[pytree-compile]	96.8810μs	70.3945μs	14.2056 KOps/s	14.2803 KOps/s	$\color{#d91a1a}-0.52\%$
test_compile_copy_flat[pytree-eager]	87.9210μs	51.3182μs	19.4863 KOps/s	19.3500 KOps/s	$\color{#35bf28}+0.70\%$
test_compile_assign_and_add[tensordict-compile]	1.6100ms	0.3899ms	2.5647 KOps/s	2.1873 KOps/s	$\textbf{\color{#35bf28}+17.26\%}$
test_compile_assign_and_add[tensordict-eager]	2.8665ms	2.6705ms	374.4641 Ops/s	364.4989 Ops/s	$\color{#35bf28}+2.73\%$
test_compile_assign_and_add[pytree-compile]	1.5841ms	0.4323ms	2.3131 KOps/s	2.2295 KOps/s	$\color{#35bf28}+3.75\%$
test_compile_assign_and_add[pytree-eager]	2.7823ms	2.6738ms	374.0030 Ops/s	359.1933 Ops/s	$\color{#35bf28}+4.12\%$
test_compile_indexing[tensor-tensordict-compile]	0.1733ms	0.1153ms	8.6756 KOps/s	8.6053 KOps/s	$\color{#35bf28}+0.82\%$
test_compile_indexing[tensor-tensordict-eager]	0.5594ms	79.8443μs	12.5244 KOps/s	11.7084 KOps/s	$\textbf{\color{#35bf28}+6.97\%}$
test_compile_indexing[tensor-tensorclass-compile]	0.2200ms	0.1128ms	8.8624 KOps/s	8.5911 KOps/s	$\color{#35bf28}+3.16\%$
test_compile_indexing[tensor-tensorclass-eager]	0.1403ms	72.4982μs	13.7935 KOps/s	14.0748 KOps/s	$\color{#d91a1a}-2.00\%$
test_compile_indexing[tensor-pytree-compile]	0.1661ms	0.1136ms	8.8046 KOps/s	8.8039 KOps/s	$+0.01\%$
test_compile_indexing[tensor-pytree-eager]	0.1438ms	72.3465μs	13.8224 KOps/s	14.2002 KOps/s	$\color{#d91a1a}-2.66\%$
test_compile_indexing[slice-tensordict-compile]	0.1561ms	0.1063ms	9.4094 KOps/s	9.7431 KOps/s	$\color{#d91a1a}-3.42\%$
test_compile_indexing[slice-tensordict-eager]	0.1473ms	19.5044μs	51.2705 KOps/s	53.2010 KOps/s	$\color{#d91a1a}-3.63\%$
test_compile_indexing[slice-tensorclass-compile]	0.1504ms	0.1022ms	9.7856 KOps/s	10.1775 KOps/s	$\color{#d91a1a}-3.85\%$
test_compile_indexing[slice-tensorclass-eager]	70.9010μs	17.0202μs	58.7536 KOps/s	59.6066 KOps/s	$\color{#d91a1a}-1.43\%$
test_compile_indexing[slice-pytree-compile]	0.1779ms	0.1026ms	9.7508 KOps/s	10.1511 KOps/s	$\color{#d91a1a}-3.94\%$
test_compile_indexing[slice-pytree-eager]	93.6710μs	17.0603μs	58.6156 KOps/s	59.9884 KOps/s	$\color{#d91a1a}-2.29\%$
test_compile_indexing[int-tensordict-compile]	0.1794ms	0.1072ms	9.3297 KOps/s	9.7285 KOps/s	$\color{#d91a1a}-4.10\%$
test_compile_indexing[int-tensordict-eager]	0.5578ms	19.0980μs	52.3615 KOps/s	53.8502 KOps/s	$\color{#d91a1a}-2.76\%$
test_compile_indexing[int-tensorclass-compile]	0.1942ms	99.3713μs	10.0633 KOps/s	10.1648 KOps/s	$\color{#d91a1a}-1.00\%$
test_compile_indexing[int-tensorclass-eager]	59.8210μs	17.1065μs	58.4575 KOps/s	60.3485 KOps/s	$\color{#d91a1a}-3.13\%$
test_compile_indexing[int-pytree-compile]	0.1543ms	0.1027ms	9.7403 KOps/s	10.1657 KOps/s	$\color{#d91a1a}-4.18\%$
test_compile_indexing[int-pytree-eager]	90.3710μs	16.6330μs	60.1215 KOps/s	60.0080 KOps/s	$\color{#35bf28}+0.19\%$
test_mod_add[eager]	92.6510μs	37.7048μs	26.5218 KOps/s	23.0083 KOps/s	$\textbf{\color{#35bf28}+15.27\%}$
test_mod_add[compile]	0.1140ms	82.0765μs	12.1838 KOps/s	12.0710 KOps/s	$\color{#35bf28}+0.93\%$
test_mod_add[compile-overhead]	0.3270ms	0.1680ms	5.9529 KOps/s	5.6156 KOps/s	$\textbf{\color{#35bf28}+6.01\%}$
test_mod_wrap[eager]	0.3288ms	0.2578ms	3.8785 KOps/s	3.7979 KOps/s	$\color{#35bf28}+2.12\%$
test_mod_wrap[compile]	0.3590ms	0.2877ms	3.4755 KOps/s	3.3833 KOps/s	$\color{#35bf28}+2.73\%$
test_mod_wrap[compile-overhead]	6.4095ms	3.5546ms	281.3288 Ops/s	272.9331 Ops/s	$\color{#35bf28}+3.08\%$
test_mod_wrap_and_backward[eager]	1.4979ms	1.3573ms	736.7450 Ops/s	661.1178 Ops/s	$\textbf{\color{#35bf28}+11.44\%}$
test_mod_wrap_and_backward[compile]	1.3621ms	1.2759ms	783.7872 Ops/s	711.4294 Ops/s	$\textbf{\color{#35bf28}+10.17\%}$
test_mod_wrap_and_backward[compile-overhead]	1.3861ms	0.9246ms	1.0815 KOps/s	955.2525 Ops/s	$\textbf{\color{#35bf28}+13.22\%}$
test_seq_add[eager]	0.1605ms	0.1149ms	8.7036 KOps/s	8.1337 KOps/s	$\textbf{\color{#35bf28}+7.01\%}$
test_seq_add[compile]	0.1230ms	89.0996μs	11.2234 KOps/s	11.0938 KOps/s	$\color{#35bf28}+1.17\%$
test_seq_add[compile-overhead]	0.1820ms	0.1287ms	7.7673 KOps/s	7.5578 KOps/s	$\color{#35bf28}+2.77\%$
test_seq_wrap[eager]	0.4919ms	0.4180ms	2.3923 KOps/s	2.2376 KOps/s	$\textbf{\color{#35bf28}+6.91\%}$
test_seq_wrap[compile]	0.3453ms	0.3041ms	3.2887 KOps/s	3.0459 KOps/s	$\textbf{\color{#35bf28}+7.97\%}$
test_seq_wrap[compile-overhead]	0.2735ms	0.2262ms	4.4216 KOps/s	4.3617 KOps/s	$\color{#35bf28}+1.37\%$
test_func_call_runtime[False-eager]	0.8537ms	0.7445ms	1.3432 KOps/s	1.2971 KOps/s	$\color{#35bf28}+3.56\%$
test_func_call_runtime[False-compile]	0.8185ms	0.7536ms	1.3270 KOps/s	1.2962 KOps/s	$\color{#35bf28}+2.38\%$
test_func_call_runtime[False-compile-overhead]	0.4110ms	0.3669ms	2.7256 KOps/s	2.7034 KOps/s	$\color{#35bf28}+0.82\%$
test_func_call_runtime[True-eager]	1.0553ms	0.9103ms	1.0985 KOps/s	1.0683 KOps/s	$\color{#35bf28}+2.83\%$
test_func_call_runtime[True-compile]	0.8752ms	0.7917ms	1.2631 KOps/s	1.2742 KOps/s	$\color{#d91a1a}-0.87\%$
test_func_call_runtime[True-compile-overhead]	0.4365ms	0.3854ms	2.5945 KOps/s	2.5570 KOps/s	$\color{#35bf28}+1.47\%$
test_func_call_cm_runtime[False-eager]	1.1580ms	0.7418ms	1.3481 KOps/s	1.3048 KOps/s	$\color{#35bf28}+3.32\%$
test_func_call_cm_runtime[False-compile]	1.1744ms	0.7562ms	1.3224 KOps/s	1.2963 KOps/s	$\color{#35bf28}+2.01\%$
test_func_call_cm_runtime[False-compile-overhead]	0.4353ms	0.3705ms	2.6991 KOps/s	2.6738 KOps/s	$\color{#35bf28}+0.94\%$
test_func_call_cm_runtime[True-eager]	1.4199ms	1.0179ms	982.3745 Ops/s	959.6950 Ops/s	$\color{#35bf28}+2.36\%$
test_func_call_cm_runtime[True-compile]	1.4144ms	1.0267ms	973.9678 Ops/s	919.1223 Ops/s	$\textbf{\color{#35bf28}+5.97\%}$
test_func_call_cm_runtime[True-compile-overhead]	1.1355ms	0.9974ms	1.0026 KOps/s	976.0718 Ops/s	$\color{#35bf28}+2.71\%$
test_vmap_func_call_cm_runtime[eager]	2.5527ms	2.1151ms	472.7818 Ops/s	464.7163 Ops/s	$\color{#35bf28}+1.74\%$
test_vmap_func_call_cm_runtime[compile]	0.9353ms	0.8241ms	1.2135 KOps/s	1.1900 KOps/s	$\color{#35bf28}+1.97\%$
test_vmap_func_call_cm_runtime[compile-overhead]	0.8349ms	0.4172ms	2.3972 KOps/s	2.3501 KOps/s	$\color{#35bf28}+2.00\%$
test_distributed	3.0278ms	0.2085ms	4.7958 KOps/s	8.4933 KOps/s	$\textbf{\color{#d91a1a}-43.53\%}$
test_tdmodule	79.4310μs	19.4327μs	51.4597 KOps/s	46.8217 KOps/s	$\textbf{\color{#35bf28}+9.91\%}$
test_tdmodule_dispatch	0.3019ms	34.9475μs	28.6143 KOps/s	25.8064 KOps/s	$\textbf{\color{#35bf28}+10.88\%}$
test_tdseq	41.9310μs	19.9860μs	50.0349 KOps/s	43.9210 KOps/s	$\textbf{\color{#35bf28}+13.92\%}$
test_tdseq_dispatch	61.1110μs	36.8402μs	27.1442 KOps/s	23.8873 KOps/s	$\textbf{\color{#35bf28}+13.63\%}$
test_instantiation_functorch	1.6883ms	1.5761ms	634.4784 Ops/s	623.3938 Ops/s	$\color{#35bf28}+1.78\%$
test_exec_functorch	0.1806ms	0.1460ms	6.8476 KOps/s	6.6980 KOps/s	$\color{#35bf28}+2.23\%$
test_exec_functional_call	0.1840ms	0.1400ms	7.1426 KOps/s	6.9520 KOps/s	$\color{#35bf28}+2.74\%$
test_exec_td_decorator	0.3888ms	0.1899ms	5.2660 KOps/s	5.2187 KOps/s	$\color{#35bf28}+0.91\%$
test_vmap_mlp_speed_decorator[True-True]	0.8112ms	0.6874ms	1.4548 KOps/s	1.4223 KOps/s	$\color{#35bf28}+2.28\%$
test_vmap_mlp_speed_decorator[True-False]	0.8274ms	0.6893ms	1.4508 KOps/s	1.3903 KOps/s	$\color{#35bf28}+4.35\%$
test_vmap_mlp_speed_decorator[False-True]	0.7236ms	0.6025ms	1.6598 KOps/s	1.5701 KOps/s	$\textbf{\color{#35bf28}+5.71\%}$
test_vmap_mlp_speed_decorator[False-False]	0.7337ms	0.6044ms	1.6546 KOps/s	1.6046 KOps/s	$\color{#35bf28}+3.11\%$
test_vmap_transformer_speed_decorator[True-True]	19.9721ms	19.3445ms	51.6944 Ops/s	51.1123 Ops/s	$\color{#35bf28}+1.14\%$
test_vmap_transformer_speed_decorator[True-False]	19.4172ms	19.3114ms	51.7828 Ops/s	51.0983 Ops/s	$\color{#35bf28}+1.34\%$
test_vmap_transformer_speed_decorator[False-True]	19.9622ms	19.6400ms	50.9164 Ops/s	51.1447 Ops/s	$\color{#d91a1a}-0.45\%$
test_vmap_transformer_speed_decorator[False-False]	19.9242ms	19.6367ms	50.9250 Ops/s	51.6029 Ops/s	$\color{#d91a1a}-1.31\%$
test_to_module_speed[True]	1.4901ms	0.9831ms	1.0172 KOps/s	1.0390 KOps/s	$\color{#d91a1a}-2.09\%$
test_to_module_speed[False]	1.0384ms	0.9650ms	1.0362 KOps/s	1.0561 KOps/s	$\color{#d91a1a}-1.88\%$
test_tc_init	72.6510μs	35.2020μs	28.4075 KOps/s	24.8831 KOps/s	$\textbf{\color{#35bf28}+14.16\%}$
test_tc_init_nested	0.1033ms	68.8254μs	14.5295 KOps/s	12.4574 KOps/s	$\textbf{\color{#35bf28}+16.63\%}$
test_tc_first_layer_tensor	30.5610μs	0.8179μs	1.2227 MOps/s	1.3976 MOps/s	$\textbf{\color{#d91a1a}-12.52\%}$
test_tc_first_layer_nontensor	27.9510μs	2.2619μs	442.1042 KOps/s	445.0371 KOps/s	$\color{#d91a1a}-0.66\%$
test_tc_second_layer_tensor	8.5953μs	1.4425μs	693.2215 KOps/s	694.6681 KOps/s	$\color{#d91a1a}-0.21\%$
test_tc_second_layer_nontensor	33.5510μs	2.9688μs	336.8318 KOps/s	329.7965 KOps/s	$\color{#35bf28}+2.13\%$
test_unbind	0.2215s	12.2836ms	81.4091 Ops/s	140.7518 Ops/s	$\textbf{\color{#d91a1a}-42.16\%}$
test_full_like	9.5256ms	9.2295ms	108.3482 Ops/s	108.1443 Ops/s	$\color{#35bf28}+0.19\%$
test_zeros_like	9.1963ms	7.2763ms	137.4317 Ops/s	231.2600 Ops/s	$\textbf{\color{#d91a1a}-40.57\%}$
test_ones_like	4.9693ms	4.3335ms	230.7620 Ops/s	230.4945 Ops/s	$\color{#35bf28}+0.12\%$
test_clone	11.6548ms	9.2550ms	108.0494 Ops/s	153.7866 Ops/s	$\textbf{\color{#d91a1a}-29.74\%}$
test_squeeze	67.1310μs	9.9210μs	100.7967 KOps/s	89.8484 KOps/s	$\textbf{\color{#35bf28}+12.19\%}$
test_unsqueeze	0.1237ms	74.3121μs	13.4568 KOps/s	12.8219 KOps/s	$\color{#35bf28}+4.95\%$
test_split	0.3825ms	0.1628ms	6.1424 KOps/s	6.0134 KOps/s	$\color{#35bf28}+2.15\%$
test_permute	0.2290ms	0.1780ms	5.6188 KOps/s	5.2783 KOps/s	$\textbf{\color{#35bf28}+6.45\%}$
test_stack	51.3315ms	50.7410ms	19.7079 Ops/s	19.5761 Ops/s	$\color{#35bf28}+0.67\%$
test_cat	51.3484ms	50.7894ms	19.6892 Ops/s	19.7650 Ops/s	$\color{#d91a1a}-0.38\%$