[Performance] Faster _get_item

[vmoens]

Stack from ghstack (oldest at bottom):

• [BugFix] Better prop of args in update #1290
• [Test] Deactivate failing test on 2.6 #1289
• -> [Performance] Faster _get_item #1288

[github-actions]

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

Total Benchmarks: 233. Improved: $\large\color{#35bf28}16$. Worsened: $\large\color{#d91a1a}15$.

Expand to view detailed results

Name	Max	Mean	Ops	Ops on Repo HEAD	Change
test_plain_set_nested	48.0600μs	11.2876μs	88.5928 KOps/s	88.1983 KOps/s	$\color{#35bf28}+0.45\%$
test_plain_set_stack_nested	30.0800μs	11.4308μs	87.4830 KOps/s	87.9144 KOps/s	$\color{#d91a1a}-0.49\%$
test_plain_set_nested_inplace	42.5600μs	12.4488μs	80.3292 KOps/s	80.0719 KOps/s	$\color{#35bf28}+0.32\%$
test_plain_set_stack_nested_inplace	40.7510μs	12.3835μs	80.7528 KOps/s	80.6972 KOps/s	$\color{#35bf28}+0.07\%$
test_items	21.3600μs	2.9391μs	340.2434 KOps/s	343.4497 KOps/s	$\color{#d91a1a}-0.93\%$
test_items_nested	0.3881ms	0.3588ms	2.7867 KOps/s	2.7520 KOps/s	$\color{#35bf28}+1.26\%$
test_items_nested_locked	0.4416ms	0.3602ms	2.7765 KOps/s	2.7270 KOps/s	$\color{#35bf28}+1.81\%$
test_items_nested_leaf	88.2510μs	60.5628μs	16.5118 KOps/s	16.5350 KOps/s	$\color{#d91a1a}-0.14\%$
test_items_stack_nested	0.4217ms	0.3602ms	2.7760 KOps/s	2.7065 KOps/s	$\color{#35bf28}+2.57\%$
test_items_stack_nested_leaf	82.0110μs	60.6773μs	16.4806 KOps/s	16.4886 KOps/s	$\color{#d91a1a}-0.05\%$
test_items_stack_nested_locked	0.4203ms	0.3606ms	2.7732 KOps/s	2.7134 KOps/s	$\color{#35bf28}+2.20\%$
test_keys	66.6910μs	3.9286μs	254.5425 KOps/s	288.6116 KOps/s	$\textbf{\color{#d91a1a}-11.80\%}$
test_keys_nested	0.1254ms	87.8920μs	11.3776 KOps/s	11.2331 KOps/s	$\color{#35bf28}+1.29\%$
test_keys_nested_locked	0.8195ms	93.9604μs	10.6428 KOps/s	10.4902 KOps/s	$\color{#35bf28}+1.45\%$
test_keys_nested_leaf	0.1224ms	79.3350μs	12.6048 KOps/s	12.5225 KOps/s	$\color{#35bf28}+0.66\%$
test_keys_stack_nested	0.1299ms	87.4407μs	11.4363 KOps/s	11.1932 KOps/s	$\color{#35bf28}+2.17\%$
test_keys_stack_nested_leaf	0.1220ms	78.4779μs	12.7424 KOps/s	12.4492 KOps/s	$\color{#35bf28}+2.36\%$
test_keys_stack_nested_locked	0.1349ms	93.3558μs	10.7117 KOps/s	10.5507 KOps/s	$\color{#35bf28}+1.53\%$
test_values	6.6700μs	0.8547μs	1.1700 MOps/s	1.1672 MOps/s	$\color{#35bf28}+0.24\%$
test_values_nested	71.9310μs	37.7296μs	26.5044 KOps/s	26.5630 KOps/s	$\color{#d91a1a}-0.22\%$
test_values_nested_locked	97.4710μs	39.5888μs	25.2597 KOps/s	25.1484 KOps/s	$\color{#35bf28}+0.44\%$
test_values_nested_leaf	70.1110μs	42.7547μs	23.3893 KOps/s	23.3155 KOps/s	$\color{#35bf28}+0.32\%$
test_values_stack_nested	68.3210μs	37.8618μs	26.4119 KOps/s	26.3393 KOps/s	$\color{#35bf28}+0.28\%$
test_values_stack_nested_leaf	69.0310μs	42.9054μs	23.3071 KOps/s	23.2580 KOps/s	$\color{#35bf28}+0.21\%$
test_values_stack_nested_locked	70.6710μs	40.0552μs	24.9655 KOps/s	25.0675 KOps/s	$\color{#d91a1a}-0.41\%$
test_membership	3.9686μs	0.4988μs	2.0049 MOps/s	1.9950 MOps/s	$\color{#35bf28}+0.50\%$
test_membership_nested	14.3050μs	1.9872μs	503.2090 KOps/s	485.1320 KOps/s	$\color{#35bf28}+3.73\%$
test_membership_nested_leaf	14.0455μs	2.0163μs	495.9602 KOps/s	495.4655 KOps/s	$\color{#35bf28}+0.10\%$
test_membership_stacked_nested	24.8800μs	2.0341μs	491.6189 KOps/s	483.9398 KOps/s	$\color{#35bf28}+1.59\%$
test_membership_stacked_nested_leaf	28.6900μs	2.0305μs	492.5003 KOps/s	491.4090 KOps/s	$\color{#35bf28}+0.22\%$
test_membership_nested_last	39.9300μs	3.0282μs	330.2330 KOps/s	324.1073 KOps/s	$\color{#35bf28}+1.89\%$
test_membership_nested_leaf_last	66.1510μs	3.0203μs	331.0962 KOps/s	326.2141 KOps/s	$\color{#35bf28}+1.50\%$
test_membership_stacked_nested_last	33.1110μs	3.0048μs	332.7999 KOps/s	324.9040 KOps/s	$\color{#35bf28}+2.43\%$
test_membership_stacked_nested_leaf_last	34.1210μs	3.0237μs	330.7230 KOps/s	326.3436 KOps/s	$\color{#35bf28}+1.34\%$
test_nested_getleaf	40.5010μs	12.9856μs	77.0082 KOps/s	76.3140 KOps/s	$\color{#35bf28}+0.91\%$
test_nested_get	78.4710μs	12.3679μs	80.8547 KOps/s	80.8452 KOps/s	$\color{#35bf28}+0.01\%$
test_stacked_getleaf	46.0510μs	13.0094μs	76.8673 KOps/s	76.9830 KOps/s	$\color{#d91a1a}-0.15\%$
test_stacked_get	31.0510μs	12.2091μs	81.9059 KOps/s	80.4098 KOps/s	$\color{#35bf28}+1.86\%$
test_nested_getitemleaf	37.1700μs	13.3882μs	74.6927 KOps/s	74.8070 KOps/s	$\color{#d91a1a}-0.15\%$
test_nested_getitem	37.4100μs	12.7524μs	78.4166 KOps/s	78.6500 KOps/s	$\color{#d91a1a}-0.30\%$
test_stacked_getitemleaf	34.9710μs	13.3846μs	74.7128 KOps/s	74.8303 KOps/s	$\color{#d91a1a}-0.16\%$
test_stacked_getitem	38.5000μs	12.4675μs	80.2084 KOps/s	78.4898 KOps/s	$\color{#35bf28}+2.19\%$
test_lock_nested	0.8313ms	0.3540ms	2.8246 KOps/s	2.8833 KOps/s	$\color{#d91a1a}-2.04\%$
test_lock_stack_nested	0.3830ms	0.3433ms	2.9127 KOps/s	2.9240 KOps/s	$\color{#d91a1a}-0.39\%$
test_unlock_nested	0.5067ms	0.2955ms	3.3846 KOps/s	3.4794 KOps/s	$\color{#d91a1a}-2.72\%$
test_unlock_stack_nested	0.3140ms	0.2831ms	3.5320 KOps/s	3.5803 KOps/s	$\color{#d91a1a}-1.35\%$
test_flatten_speed	0.1165ms	77.0379μs	12.9806 KOps/s	13.0580 KOps/s	$\color{#d91a1a}-0.59\%$
test_unflatten_speed	0.4656ms	0.3972ms	2.5177 KOps/s	2.5120 KOps/s	$\color{#35bf28}+0.23\%$
test_common_ops	0.8786ms	0.6284ms	1.5912 KOps/s	1.5989 KOps/s	$\color{#d91a1a}-0.48\%$
test_creation	0.1815ms	1.7097μs	584.8818 KOps/s	583.2323 KOps/s	$\color{#35bf28}+0.28\%$
test_creation_empty	0.8029ms	7.0054μs	142.7467 KOps/s	141.8592 KOps/s	$\color{#35bf28}+0.63\%$
test_creation_nested_1	0.1999ms	9.9381μs	100.6228 KOps/s	100.7540 KOps/s	$\color{#d91a1a}-0.13\%$
test_creation_nested_2	0.1036ms	12.7433μs	78.4723 KOps/s	77.6816 KOps/s	$\color{#35bf28}+1.02\%$
test_clone	80.8810μs	10.9530μs	91.2992 KOps/s	93.1783 KOps/s	$\color{#d91a1a}-2.02\%$
test_getitem[int]	0.1575ms	10.2912μs	97.1708 KOps/s	95.6069 KOps/s	$\color{#35bf28}+1.64\%$
test_getitem[slice_int]	0.1138ms	19.9519μs	50.1206 KOps/s	47.7606 KOps/s	$\color{#35bf28}+4.94\%$
test_getitem[range]	0.1368ms	37.2879μs	26.8183 KOps/s	25.1851 KOps/s	$\textbf{\color{#35bf28}+6.48\%}$
test_getitem[tuple]	0.1939ms	17.6590μs	56.6283 KOps/s	55.5713 KOps/s	$\color{#35bf28}+1.90\%$
test_getitem[list]	0.1288ms	33.2672μs	30.0596 KOps/s	29.5081 KOps/s	$\color{#35bf28}+1.87\%$
test_setitem_dim[int]	0.1347ms	18.9874μs	52.6665 KOps/s	50.3883 KOps/s	$\color{#35bf28}+4.52\%$
test_setitem_dim[slice_int]	60.4310μs	37.5852μs	26.6062 KOps/s	26.2278 KOps/s	$\color{#35bf28}+1.44\%$
test_setitem_dim[range]	98.3510μs	52.0962μs	19.1953 KOps/s	18.0956 KOps/s	$\textbf{\color{#35bf28}+6.08\%}$
test_setitem_dim[tuple]	53.2310μs	31.2331μs	32.0173 KOps/s	30.0832 KOps/s	$\textbf{\color{#35bf28}+6.43\%}$
test_setitem	0.2138ms	15.6434μs	63.9246 KOps/s	64.6506 KOps/s	$\color{#d91a1a}-1.12\%$
test_set	0.2127ms	15.2160μs	65.7205 KOps/s	67.2031 KOps/s	$\color{#d91a1a}-2.21\%$
test_set_shared	0.5192ms	0.1577ms	6.3393 KOps/s	6.3099 KOps/s	$\color{#35bf28}+0.47\%$
test_update	0.4176ms	18.4725μs	54.1345 KOps/s	55.2185 KOps/s	$\color{#d91a1a}-1.96\%$
test_update_nested	0.1187ms	28.8410μs	34.6728 KOps/s	34.9929 KOps/s	$\color{#d91a1a}-0.91\%$
test_update__nested	70.0710μs	25.4557μs	39.2839 KOps/s	38.9776 KOps/s	$\color{#35bf28}+0.79\%$
test_set_nested	0.1379ms	16.4721μs	60.7088 KOps/s	63.4043 KOps/s	$\color{#d91a1a}-4.25\%$
test_set_nested_new	0.1139ms	19.0783μs	52.4157 KOps/s	52.7058 KOps/s	$\color{#d91a1a}-0.55\%$
test_select	0.1274ms	31.1634μs	32.0889 KOps/s	32.1437 KOps/s	$\color{#d91a1a}-0.17\%$
test_select_nested	77.3910μs	42.7895μs	23.3702 KOps/s	22.8609 KOps/s	$\color{#35bf28}+2.23\%$
test_exclude_nested	89.9610μs	61.3031μs	16.3124 KOps/s	16.1029 KOps/s	$\color{#35bf28}+1.30\%$
test_empty[True]	0.3993ms	0.2920ms	3.4246 KOps/s	3.4049 KOps/s	$\color{#35bf28}+0.58\%$
test_empty[False]	3.4771μs	0.8080μs	1.2377 MOps/s	1.2223 MOps/s	$\color{#35bf28}+1.26\%$
test_to	88.4920μs	59.2080μs	16.8896 KOps/s	17.1870 KOps/s	$\color{#d91a1a}-1.73\%$
test_to_nonblocking	0.1961ms	49.5292μs	20.1901 KOps/s	19.1920 KOps/s	$\textbf{\color{#35bf28}+5.20\%}$
test_unbind_speed	0.8760ms	0.2401ms	4.1651 KOps/s	4.1923 KOps/s	$\color{#d91a1a}-0.65\%$
test_unbind_speed_stack0	0.2816ms	0.2409ms	4.1511 KOps/s	4.2661 KOps/s	$\color{#d91a1a}-2.70\%$
test_unbind_speed_stack1	95.9333ms	0.7433ms	1.3453 KOps/s	1.3464 KOps/s	$\color{#d91a1a}-0.08\%$
test_split	94.8847ms	1.5816ms	632.2764 Ops/s	617.6797 Ops/s	$\color{#35bf28}+2.36\%$
test_chunk	99.8207ms	1.5932ms	627.6745 Ops/s	614.4800 Ops/s	$\color{#35bf28}+2.15\%$
test_consolidate[False-None]	99.3116ms	3.0611ms	326.6794 Ops/s	328.5690 Ops/s	$\color{#d91a1a}-0.58\%$
test_consolidate[default-None]	1.8271ms	1.6785ms	595.7744 Ops/s	588.7234 Ops/s	$\color{#35bf28}+1.20\%$
test_consolidate[reduce-overhead-None]	1.8648ms	1.7163ms	582.6610 Ops/s	578.6402 Ops/s	$\color{#35bf28}+0.69\%$
test_consolidate_njt[False-None]	6.8052ms	6.4463ms	155.1271 Ops/s	155.2273 Ops/s	$\color{#d91a1a}-0.06\%$
test_to[False-False-None]	1.9607ms	1.7858ms	559.9790 Ops/s	564.0883 Ops/s	$\color{#d91a1a}-0.73\%$
test_to[True-False-None]	1.5046ms	1.3838ms	722.6697 Ops/s	719.3109 Ops/s	$\color{#35bf28}+0.47\%$
test_to[within-False-None]	4.4170ms	4.2706ms	234.1585 Ops/s	233.8602 Ops/s	$\color{#35bf28}+0.13\%$
test_to[True-default-None]	5.4019ms	5.1849ms	192.8678 Ops/s	191.5057 Ops/s	$\color{#35bf28}+0.71\%$
test_to_njt[False-False-None]	7.2741ms	6.9208ms	144.4919 Ops/s	145.3772 Ops/s	$\color{#d91a1a}-0.61\%$
test_to_njt[True-False-None]	5.5314ms	5.3917ms	185.4703 Ops/s	185.1955 Ops/s	$\color{#35bf28}+0.15\%$
test_to_njt[within-False-None]	0.3233s	15.8378ms	63.1402 Ops/s	83.4748 Ops/s	$\textbf{\color{#d91a1a}-24.36\%}$
test_creation[device0]	0.4651ms	78.1292μs	12.7993 KOps/s	12.8573 KOps/s	$\color{#d91a1a}-0.45\%$
test_creation_from_tensor	0.5398ms	81.1089μs	12.3291 KOps/s	12.1720 KOps/s	$\color{#35bf28}+1.29\%$
test_add_one[memmap_tensor0]	0.4123ms	6.7107μs	149.0150 KOps/s	148.5790 KOps/s	$\color{#35bf28}+0.29\%$
test_contiguous[memmap_tensor0]	2.3695μs	0.3954μs	2.5292 MOps/s	2.4888 MOps/s	$\color{#35bf28}+1.62\%$
test_stack[memmap_tensor0]	30.7000μs	4.1730μs	239.6344 KOps/s	234.3215 KOps/s	$\color{#35bf28}+2.27\%$
test_memmaptd_index	1.7484ms	0.2378ms	4.2055 KOps/s	4.0802 KOps/s	$\color{#35bf28}+3.07\%$
test_memmaptd_index_astensor	0.4572ms	0.3006ms	3.3271 KOps/s	3.2644 KOps/s	$\color{#35bf28}+1.92\%$
test_memmaptd_index_op	0.9701ms	0.5453ms	1.8339 KOps/s	1.8125 KOps/s	$\color{#35bf28}+1.18\%$
test_serialize_model	0.1332s	0.1323s	7.5596 Ops/s	7.5573 Ops/s	$\color{#35bf28}+0.03\%$
test_serialize_model_pickle	1.3474s	1.1875s	0.8421 Ops/s	0.8214 Ops/s	$\color{#35bf28}+2.52\%$
test_serialize_weights	0.1316s	0.1310s	7.6365 Ops/s	7.5723 Ops/s	$\color{#35bf28}+0.85\%$
test_serialize_weights_returnearly	0.3105s	52.6205ms	19.0040 Ops/s	14.7155 Ops/s	$\textbf{\color{#35bf28}+29.14\%}$
test_serialize_weights_pickle	1.3874s	1.1931s	0.8382 Ops/s	0.8217 Ops/s	$\color{#35bf28}+2.01\%$
test_reshape_pytree	52.9410μs	22.1249μs	45.1979 KOps/s	45.0255 KOps/s	$\color{#35bf28}+0.38\%$
test_reshape_td	53.0510μs	28.2520μs	35.3958 KOps/s	38.5940 KOps/s	$\textbf{\color{#d91a1a}-8.29\%}$
test_view_pytree	89.2710μs	22.1137μs	45.2209 KOps/s	46.7526 KOps/s	$\color{#d91a1a}-3.28\%$
test_view_td	0.1606ms	34.1611μs	29.2730 KOps/s	29.2371 KOps/s	$\color{#35bf28}+0.12\%$
test_unbind_pytree	0.1175ms	27.7086μs	36.0899 KOps/s	35.8258 KOps/s	$\color{#35bf28}+0.74\%$
test_unbind_td	0.7988ms	37.0922μs	26.9598 KOps/s	25.5829 KOps/s	$\textbf{\color{#35bf28}+5.38\%}$
test_split_pytree	0.1127ms	29.1937μs	34.2539 KOps/s	32.5563 KOps/s	$\textbf{\color{#35bf28}+5.21\%}$
test_split_td	0.1836ms	38.0058μs	26.3118 KOps/s	23.0204 KOps/s	$\textbf{\color{#35bf28}+14.30\%}$
test_add_pytree	68.4710μs	36.8419μs	27.1430 KOps/s	27.9065 KOps/s	$\color{#d91a1a}-2.74\%$
test_add_td	0.2890ms	50.9628μs	19.6222 KOps/s	19.3048 KOps/s	$\color{#35bf28}+1.64\%$
test_compile_add_one_nested[tensordict-compile]	0.1859ms	0.1241ms	8.0587 KOps/s	7.9022 KOps/s	$\color{#35bf28}+1.98\%$
test_compile_add_one_nested[tensordict-eager]	0.2861ms	0.1434ms	6.9731 KOps/s	7.0170 KOps/s	$\color{#d91a1a}-0.63\%$
test_compile_add_one_nested[pytree-compile]	0.2666ms	97.2410μs	10.2837 KOps/s	10.4046 KOps/s	$\color{#d91a1a}-1.16\%$
test_compile_add_one_nested[pytree-eager]	1.7175ms	0.1570ms	6.3686 KOps/s	6.7199 KOps/s	$\textbf{\color{#d91a1a}-5.23\%}$
test_compile_copy_nested[tensordict-compile]	77.5510μs	24.9066μs	40.1501 KOps/s	39.9380 KOps/s	$\color{#35bf28}+0.53\%$
test_compile_copy_nested[tensordict-eager]	63.8900μs	34.5825μs	28.9164 KOps/s	28.6005 KOps/s	$\color{#35bf28}+1.10\%$
test_compile_copy_nested[pytree-compile]	0.3100ms	62.8078μs	15.9216 KOps/s	15.6459 KOps/s	$\color{#35bf28}+1.76\%$
test_compile_copy_nested[pytree-eager]	78.5710μs	48.0086μs	20.8296 KOps/s	20.1921 KOps/s	$\color{#35bf28}+3.16\%$
test_compile_add_one_flat[tensordict-compile]	0.2113ms	0.1417ms	7.0562 KOps/s	7.0435 KOps/s	$\color{#35bf28}+0.18\%$
test_compile_add_one_flat[tensordict-eager]	0.3307ms	0.2181ms	4.5860 KOps/s	4.5877 KOps/s	$\color{#d91a1a}-0.04\%$
test_compile_add_one_flat[tensorclass-compile]	0.1590ms	96.1986μs	10.3952 KOps/s	10.4351 KOps/s	$\color{#d91a1a}-0.38\%$
test_compile_add_one_flat[tensorclass-eager]	0.2514ms	62.0721μs	16.1103 KOps/s	17.6107 KOps/s	$\textbf{\color{#d91a1a}-8.52\%}$
test_compile_add_one_flat[pytree-compile]	0.1864ms	0.1358ms	7.3612 KOps/s	7.4172 KOps/s	$\color{#d91a1a}-0.76\%$
test_compile_add_one_flat[pytree-eager]	0.6333ms	0.4843ms	2.0648 KOps/s	2.0918 KOps/s	$\color{#d91a1a}-1.29\%$
test_compile_add_self_flat[tensordict-eager]	0.4544ms	0.2654ms	3.7685 KOps/s	3.7908 KOps/s	$\color{#d91a1a}-0.59\%$
test_compile_add_self_flat[tensordict-compile]	0.1911ms	0.1428ms	7.0020 KOps/s	7.0441 KOps/s	$\color{#d91a1a}-0.60\%$
test_compile_add_self_flat[tensorclass-eager]	0.2081ms	69.8316μs	14.3202 KOps/s	14.6125 KOps/s	$\color{#d91a1a}-2.00\%$
test_compile_add_self_flat[tensorclass-compile]	0.2545ms	0.1037ms	9.6429 KOps/s	10.4184 KOps/s	$\textbf{\color{#d91a1a}-7.44\%}$
test_compile_add_self_flat[pytree-eager]	0.5954ms	0.4258ms	2.3484 KOps/s	2.4357 KOps/s	$\color{#d91a1a}-3.59\%$
test_compile_add_self_flat[pytree-compile]	0.2091ms	0.1427ms	7.0067 KOps/s	7.5066 KOps/s	$\textbf{\color{#d91a1a}-6.66\%}$
test_compile_copy_flat[tensordict-compile]	0.1971ms	20.4188μs	48.9746 KOps/s	53.0272 KOps/s	$\textbf{\color{#d91a1a}-7.64\%}$
test_compile_copy_flat[tensordict-eager]	98.8510μs	31.9568μs	31.2923 KOps/s	31.3872 KOps/s	$\color{#d91a1a}-0.30\%$
test_compile_copy_flat[pytree-compile]	0.2332ms	69.3687μs	14.4157 KOps/s	14.4565 KOps/s	$\color{#d91a1a}-0.28\%$
test_compile_copy_flat[pytree-eager]	0.1436ms	51.8862μs	19.2729 KOps/s	19.2388 KOps/s	$\color{#35bf28}+0.18\%$
test_compile_assign_and_add[tensordict-compile]	1.6201ms	0.3969ms	2.5193 KOps/s	2.2340 KOps/s	$\textbf{\color{#35bf28}+12.77\%}$
test_compile_assign_and_add[tensordict-eager]	3.1890ms	2.7801ms	359.6938 Ops/s	363.3327 Ops/s	$\color{#d91a1a}-1.00\%$
test_compile_assign_and_add[pytree-compile]	1.5997ms	0.4336ms	2.3064 KOps/s	2.3019 KOps/s	$\color{#35bf28}+0.20\%$
test_compile_assign_and_add[pytree-eager]	2.8788ms	2.6634ms	375.4648 Ops/s	382.5620 Ops/s	$\color{#d91a1a}-1.86\%$
test_compile_indexing[tensor-tensordict-compile]	0.2574ms	0.1117ms	8.9551 KOps/s	9.0255 KOps/s	$\color{#d91a1a}-0.78\%$
test_compile_indexing[tensor-tensordict-eager]	0.5547ms	84.0941μs	11.8914 KOps/s	12.1575 KOps/s	$\color{#d91a1a}-2.19\%$
test_compile_indexing[tensor-tensorclass-compile]	0.6462ms	0.1097ms	9.1149 KOps/s	9.4452 KOps/s	$\color{#d91a1a}-3.50\%$
test_compile_indexing[tensor-tensorclass-eager]	0.2332ms	67.3116μs	14.8563 KOps/s	14.7726 KOps/s	$\color{#35bf28}+0.57\%$
test_compile_indexing[tensor-pytree-compile]	0.2683ms	0.1088ms	9.1924 KOps/s	9.4240 KOps/s	$\color{#d91a1a}-2.46\%$
test_compile_indexing[tensor-pytree-eager]	0.2119ms	67.4674μs	14.8220 KOps/s	14.2772 KOps/s	$\color{#35bf28}+3.82\%$
test_compile_indexing[slice-tensordict-compile]	0.1666ms	98.9796μs	10.1031 KOps/s	10.1640 KOps/s	$\color{#d91a1a}-0.60\%$
test_compile_indexing[slice-tensordict-eager]	0.2551ms	19.3421μs	51.7006 KOps/s	52.1869 KOps/s	$\color{#d91a1a}-0.93\%$
test_compile_indexing[slice-tensorclass-compile]	0.1476ms	94.9503μs	10.5318 KOps/s	10.4056 KOps/s	$\color{#35bf28}+1.21\%$
test_compile_indexing[slice-tensorclass-eager]	52.1410μs	15.4739μs	64.6248 KOps/s	64.2434 KOps/s	$\color{#35bf28}+0.59\%$
test_compile_indexing[slice-pytree-compile]	0.2482ms	96.1522μs	10.4002 KOps/s	10.4912 KOps/s	$\color{#d91a1a}-0.87\%$
test_compile_indexing[slice-pytree-eager]	45.1800μs	15.2668μs	65.5018 KOps/s	64.1249 KOps/s	$\color{#35bf28}+2.15\%$
test_compile_indexing[int-tensordict-compile]	0.2506ms	99.2241μs	10.0782 KOps/s	10.1359 KOps/s	$\color{#d91a1a}-0.57\%$
test_compile_indexing[int-tensordict-eager]	0.7854ms	18.7032μs	53.4669 KOps/s	52.4954 KOps/s	$\color{#35bf28}+1.85\%$
test_compile_indexing[int-tensorclass-compile]	0.2447ms	95.6347μs	10.4565 KOps/s	10.4861 KOps/s	$\color{#d91a1a}-0.28\%$
test_compile_indexing[int-tensorclass-eager]	47.5210μs	15.8352μs	63.1504 KOps/s	64.4681 KOps/s	$\color{#d91a1a}-2.04\%$
test_compile_indexing[int-pytree-compile]	0.2547ms	99.7361μs	10.0265 KOps/s	10.5100 KOps/s	$\color{#d91a1a}-4.60\%$
test_compile_indexing[int-pytree-eager]	79.4110μs	15.1334μs	66.0789 KOps/s	65.0284 KOps/s	$\color{#35bf28}+1.62\%$
test_mod_add[eager]	76.0610μs	37.2148μs	26.8711 KOps/s	25.5133 KOps/s	$\textbf{\color{#35bf28}+5.32\%}$
test_mod_add[compile]	0.2927ms	80.3199μs	12.4502 KOps/s	12.2438 KOps/s	$\color{#35bf28}+1.69\%$
test_mod_add[compile-overhead]	0.3330ms	0.1681ms	5.9488 KOps/s	5.7555 KOps/s	$\color{#35bf28}+3.36\%$
test_mod_wrap[eager]	0.3325ms	0.2465ms	4.0567 KOps/s	3.9560 KOps/s	$\color{#35bf28}+2.55\%$
test_mod_wrap[compile]	1.1198ms	0.2928ms	3.4156 KOps/s	3.3430 KOps/s	$\color{#35bf28}+2.17\%$
test_mod_wrap[compile-overhead]	7.3365ms	3.8564ms	259.3078 Ops/s	267.9198 Ops/s	$\color{#d91a1a}-3.21\%$
test_mod_wrap_and_backward[eager]	1.6142ms	1.3668ms	731.6338 Ops/s	680.5066 Ops/s	$\textbf{\color{#35bf28}+7.51\%}$
test_mod_wrap_and_backward[compile]	1.8173ms	1.2620ms	792.4098 Ops/s	728.3212 Ops/s	$\textbf{\color{#35bf28}+8.80\%}$
test_mod_wrap_and_backward[compile-overhead]	1.3841ms	0.9155ms	1.0923 KOps/s	959.5106 Ops/s	$\textbf{\color{#35bf28}+13.84\%}$
test_seq_add[eager]	0.3104ms	0.1267ms	7.8944 KOps/s	7.9219 KOps/s	$\color{#d91a1a}-0.35\%$
test_seq_add[compile]	0.2388ms	88.6392μs	11.2817 KOps/s	11.1878 KOps/s	$\color{#35bf28}+0.84\%$
test_seq_add[compile-overhead]	0.2882ms	0.1281ms	7.8057 KOps/s	7.7561 KOps/s	$\color{#35bf28}+0.64\%$
test_seq_wrap[eager]	1.0253ms	0.4320ms	2.3147 KOps/s	2.2940 KOps/s	$\color{#35bf28}+0.90\%$
test_seq_wrap[compile]	1.1018ms	0.3016ms	3.3154 KOps/s	3.2948 KOps/s	$\color{#35bf28}+0.63\%$
test_seq_wrap[compile-overhead]	0.3568ms	0.2231ms	4.4814 KOps/s	4.4256 KOps/s	$\color{#35bf28}+1.26\%$
test_func_call_runtime[False-eager]	0.8118ms	0.7373ms	1.3563 KOps/s	1.3336 KOps/s	$\color{#35bf28}+1.70\%$
test_func_call_runtime[False-compile]	0.7890ms	0.7369ms	1.3571 KOps/s	1.3568 KOps/s	$\color{#35bf28}+0.02\%$
test_func_call_runtime[False-compile-overhead]	0.4021ms	0.3565ms	2.8051 KOps/s	2.7741 KOps/s	$\color{#35bf28}+1.12\%$
test_func_call_runtime[True-eager]	1.0549ms	0.8975ms	1.1141 KOps/s	1.0427 KOps/s	$\textbf{\color{#35bf28}+6.86\%}$
test_func_call_runtime[True-compile]	1.0401ms	0.7715ms	1.2962 KOps/s	1.3140 KOps/s	$\color{#d91a1a}-1.35\%$
test_func_call_runtime[True-compile-overhead]	0.4345ms	0.3807ms	2.6267 KOps/s	2.6266 KOps/s	$+0.00\%$
test_func_call_cm_runtime[False-eager]	0.8740ms	0.7318ms	1.3664 KOps/s	1.3434 KOps/s	$\color{#35bf28}+1.71\%$
test_func_call_cm_runtime[False-compile]	0.8281ms	0.7425ms	1.3468 KOps/s	1.3463 KOps/s	$\color{#35bf28}+0.04\%$
test_func_call_cm_runtime[False-compile-overhead]	0.4955ms	0.3591ms	2.7846 KOps/s	2.7760 KOps/s	$\color{#35bf28}+0.31\%$
test_func_call_cm_runtime[True-eager]	1.2710ms	1.0019ms	998.0660 Ops/s	987.7072 Ops/s	$\color{#35bf28}+1.05\%$
test_func_call_cm_runtime[True-compile]	1.1260ms	0.9855ms	1.0147 KOps/s	964.8854 Ops/s	$\textbf{\color{#35bf28}+5.16\%}$
test_func_call_cm_runtime[True-compile-overhead]	1.1420ms	0.9921ms	1.0080 KOps/s	991.0679 Ops/s	$\color{#35bf28}+1.71\%$
test_vmap_func_call_cm_runtime[eager]	2.5179ms	2.0923ms	477.9463 Ops/s	474.6373 Ops/s	$\color{#35bf28}+0.70\%$
test_vmap_func_call_cm_runtime[compile]	0.9030ms	0.8049ms	1.2424 KOps/s	1.2379 KOps/s	$\color{#35bf28}+0.37\%$
test_vmap_func_call_cm_runtime[compile-overhead]	0.5694ms	0.4084ms	2.4487 KOps/s	2.3991 KOps/s	$\color{#35bf28}+2.07\%$
test_distributed	2.7141ms	0.3569ms	2.8018 KOps/s	8.7425 KOps/s	$\textbf{\color{#d91a1a}-67.95\%}$
test_tdmodule	38.5410μs	19.9003μs	50.2506 KOps/s	49.8852 KOps/s	$\color{#35bf28}+0.73\%$
test_tdmodule_dispatch	58.2600μs	37.7036μs	26.5227 KOps/s	26.4505 KOps/s	$\color{#35bf28}+0.27\%$
test_tdseq	42.4100μs	20.0159μs	49.9603 KOps/s	50.0190 KOps/s	$\color{#d91a1a}-0.12\%$
test_tdseq_dispatch	59.7610μs	39.6591μs	25.2149 KOps/s	25.0528 KOps/s	$\color{#35bf28}+0.65\%$
test_instantiation_functorch	1.6266ms	1.5360ms	651.0292 Ops/s	635.9737 Ops/s	$\color{#35bf28}+2.37\%$
test_exec_functorch	0.2767ms	0.1428ms	7.0030 KOps/s	6.8344 KOps/s	$\color{#35bf28}+2.47\%$
test_exec_functional_call	0.1861ms	0.1382ms	7.2339 KOps/s	7.2609 KOps/s	$\color{#d91a1a}-0.37\%$
test_exec_td_decorator	0.3790ms	0.1875ms	5.3327 KOps/s	5.2657 KOps/s	$\color{#35bf28}+1.27\%$
test_vmap_mlp_speed_decorator[True-True]	0.8969ms	0.6922ms	1.4446 KOps/s	1.4318 KOps/s	$\color{#35bf28}+0.90\%$
test_vmap_mlp_speed_decorator[True-False]	0.8923ms	0.6903ms	1.4487 KOps/s	1.4340 KOps/s	$\color{#35bf28}+1.03\%$
test_vmap_mlp_speed_decorator[False-True]	0.7449ms	0.5989ms	1.6697 KOps/s	1.5746 KOps/s	$\textbf{\color{#35bf28}+6.04\%}$
test_vmap_mlp_speed_decorator[False-False]	0.7646ms	0.6138ms	1.6291 KOps/s	1.6694 KOps/s	$\color{#d91a1a}-2.41\%$
test_vmap_transformer_speed_decorator[True-True]	20.0109ms	19.3911ms	51.5700 Ops/s	52.1031 Ops/s	$\color{#d91a1a}-1.02\%$
test_vmap_transformer_speed_decorator[True-False]	19.8340ms	19.3468ms	51.6883 Ops/s	52.3257 Ops/s	$\color{#d91a1a}-1.22\%$
test_vmap_transformer_speed_decorator[False-True]	20.0566ms	19.2090ms	52.0591 Ops/s	52.6514 Ops/s	$\color{#d91a1a}-1.13\%$
test_vmap_transformer_speed_decorator[False-False]	20.2184ms	19.2792ms	51.8694 Ops/s	52.0193 Ops/s	$\color{#d91a1a}-0.29\%$
test_to_module_speed[True]	1.3692ms	0.9574ms	1.0445 KOps/s	1.0293 KOps/s	$\color{#35bf28}+1.47\%$
test_to_module_speed[False]	1.3888ms	0.9449ms	1.0583 KOps/s	1.0568 KOps/s	$\color{#35bf28}+0.13\%$
test_tc_init	0.1570ms	34.6194μs	28.8855 KOps/s	29.0144 KOps/s	$\color{#d91a1a}-0.44\%$
test_tc_init_tensor_only	0.1053ms	10.4634μs	95.5716 KOps/s	93.0081 KOps/s	$\color{#35bf28}+2.76\%$
test_tc_init_nested	0.1817ms	69.5764μs	14.3727 KOps/s	14.4986 KOps/s	$\color{#d91a1a}-0.87\%$
test_tc_first_layer_tensor	25.2900μs	0.8980μs	1.1136 MOps/s	1.1131 MOps/s	$\color{#35bf28}+0.04\%$
test_tc_first_layer_tensor_only	1.9010μs	0.4194μs	2.3845 MOps/s	2.3553 MOps/s	$\color{#35bf28}+1.24\%$
test_tc_first_layer_tensor_set	34.0800μs	2.8642μs	349.1335 KOps/s	336.9807 KOps/s	$\color{#35bf28}+3.61\%$
test_tc_first_layer_tensor_only_set	16.4655μs	1.8194μs	549.6212 KOps/s	550.5454 KOps/s	$\color{#d91a1a}-0.17\%$
test_tc_first_layer_nontensor	32.1410μs	2.3346μs	428.3348 KOps/s	431.5547 KOps/s	$\color{#d91a1a}-0.75\%$
test_tc_second_layer_tensor	31.6000μs	1.7358μs	576.0978 KOps/s	576.2821 KOps/s	$\color{#d91a1a}-0.03\%$
test_tc_second_layer_nontensor	30.0600μs	3.1584μs	316.6112 KOps/s	319.1979 KOps/s	$\color{#d91a1a}-0.81\%$
test_unbind	0.2393s	10.8003ms	92.5897 Ops/s	145.2140 Ops/s	$\textbf{\color{#d91a1a}-36.24\%}$
test_full_like	11.9596ms	8.8051ms	113.5706 Ops/s	221.6308 Ops/s	$\textbf{\color{#d91a1a}-48.76\%}$
test_zeros_like	9.6537ms	7.3313ms	136.4023 Ops/s	138.2435 Ops/s	$\color{#d91a1a}-1.33\%$
test_ones_like	9.5005ms	6.4866ms	154.1647 Ops/s	227.9121 Ops/s	$\textbf{\color{#d91a1a}-32.36\%}$
test_clone	12.1580ms	9.5025ms	105.2350 Ops/s	149.4636 Ops/s	$\textbf{\color{#d91a1a}-29.59\%}$
test_squeeze	79.7410μs	9.8312μs	101.7166 KOps/s	102.8951 KOps/s	$\color{#d91a1a}-1.15\%$
test_unsqueeze	0.1679ms	73.3897μs	13.6259 KOps/s	13.7453 KOps/s	$\color{#d91a1a}-0.87\%$
test_split	0.4227ms	0.1582ms	6.3211 KOps/s	6.0272 KOps/s	$\color{#35bf28}+4.88\%$
test_permute	0.2262ms	0.1786ms	5.6003 KOps/s	5.5529 KOps/s	$\color{#35bf28}+0.85\%$
test_stack	53.6563ms	53.3219ms	18.7540 Ops/s	36.7587 Ops/s	$\textbf{\color{#d91a1a}-48.98\%}$
test_cat	53.4677ms	53.0698ms	18.8431 Ops/s	23.3215 Ops/s	$\textbf{\color{#d91a1a}-19.20\%}$