[Feature] determinstic_sample for composite dist

[vmoens]

No description provided.

[github-actions]

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

Total Benchmarks: 144. Improved: $\large\color{#35bf28}8$. Worsened: $\large\color{#d91a1a}13$.

Expand to view detailed results

Name	Max	Mean	Ops	Ops on Repo HEAD	Change
test_plain_set_nested	36.7590μs	16.7688μs	59.6344 KOps/s	61.9688 KOps/s	$\color{#d91a1a}-3.77\%$
test_plain_set_stack_nested	38.2110μs	16.9339μs	59.0532 KOps/s	60.9168 KOps/s	$\color{#d91a1a}-3.06\%$
test_plain_set_nested_inplace	56.8260μs	19.4013μs	51.5429 KOps/s	53.8980 KOps/s	$\color{#d91a1a}-4.37\%$
test_plain_set_stack_nested_inplace	43.5410μs	18.9837μs	52.6769 KOps/s	54.3214 KOps/s	$\color{#d91a1a}-3.03\%$
test_items	37.9110μs	2.5030μs	399.5129 KOps/s	396.4423 KOps/s	$\color{#35bf28}+0.77\%$
test_items_nested	0.8117ms	0.2660ms	3.7591 KOps/s	3.7702 KOps/s	$\color{#d91a1a}-0.29\%$
test_items_nested_locked	0.5080ms	0.2637ms	3.7925 KOps/s	3.7868 KOps/s	$\color{#35bf28}+0.15\%$
test_items_nested_leaf	0.1301ms	77.7429μs	12.8629 KOps/s	12.9970 KOps/s	$\color{#d91a1a}-1.03\%$
test_items_stack_nested	0.5125ms	0.2665ms	3.7524 KOps/s	3.7800 KOps/s	$\color{#d91a1a}-0.73\%$
test_items_stack_nested_leaf	0.1651ms	76.6818μs	13.0409 KOps/s	13.0557 KOps/s	$\color{#d91a1a}-0.11\%$
test_items_stack_nested_locked	0.4487ms	0.2663ms	3.7546 KOps/s	3.7969 KOps/s	$\color{#d91a1a}-1.11\%$
test_keys	20.9990μs	3.8616μs	258.9621 KOps/s	172.9426 KOps/s	$\textbf{\color{#35bf28}+49.74\%}$
test_keys_nested	0.2623ms	0.1398ms	7.1549 KOps/s	7.2432 KOps/s	$\color{#d91a1a}-1.22\%$
test_keys_nested_locked	0.7095ms	0.1460ms	6.8487 KOps/s	6.9718 KOps/s	$\color{#d91a1a}-1.77\%$
test_keys_nested_leaf	0.2376ms	0.1182ms	8.4582 KOps/s	8.4915 KOps/s	$\color{#d91a1a}-0.39\%$
test_keys_stack_nested	0.2170ms	0.1399ms	7.1499 KOps/s	7.2724 KOps/s	$\color{#d91a1a}-1.69\%$
test_keys_stack_nested_leaf	0.2496ms	0.1188ms	8.4178 KOps/s	8.5408 KOps/s	$\color{#d91a1a}-1.44\%$
test_keys_stack_nested_locked	0.2366ms	0.1456ms	6.8675 KOps/s	7.0321 KOps/s	$\color{#d91a1a}-2.34\%$
test_values	9.8760μs	1.1744μs	851.5007 KOps/s	861.5391 KOps/s	$\color{#d91a1a}-1.17\%$
test_values_nested	0.1063ms	51.3600μs	19.4704 KOps/s	19.9910 KOps/s	$\color{#d91a1a}-2.60\%$
test_values_nested_locked	0.1085ms	50.6017μs	19.7622 KOps/s	19.7940 KOps/s	$\color{#d91a1a}-0.16\%$
test_values_nested_leaf	89.1660μs	46.0193μs	21.7300 KOps/s	21.8099 KOps/s	$\color{#d91a1a}-0.37\%$
test_values_stack_nested	96.9310μs	53.8271μs	18.5780 KOps/s	19.7355 KOps/s	$\textbf{\color{#d91a1a}-5.87\%}$
test_values_stack_nested_leaf	96.0090μs	46.9248μs	21.3107 KOps/s	21.9080 KOps/s	$\color{#d91a1a}-2.73\%$
test_values_stack_nested_locked	98.6030μs	52.7847μs	18.9449 KOps/s	19.6365 KOps/s	$\color{#d91a1a}-3.52\%$
test_membership	11.4010μs	1.3234μs	755.6380 KOps/s	759.4444 KOps/s	$\color{#d91a1a}-0.50\%$
test_membership_nested	34.6550μs	3.4157μs	292.7670 KOps/s	293.4854 KOps/s	$\color{#d91a1a}-0.24\%$
test_membership_nested_leaf	20.5080μs	3.4364μs	291.0010 KOps/s	291.1697 KOps/s	$\color{#d91a1a}-0.06\%$
test_membership_stacked_nested	37.3100μs	3.4140μs	292.9139 KOps/s	296.4372 KOps/s	$\color{#d91a1a}-1.19\%$
test_membership_stacked_nested_leaf	25.9290μs	3.3924μs	294.7757 KOps/s	296.0550 KOps/s	$\color{#d91a1a}-0.43\%$
test_membership_nested_last	52.1710μs	4.1253μs	242.4067 KOps/s	239.1063 KOps/s	$\color{#35bf28}+1.38\%$
test_membership_nested_leaf_last	22.3020μs	4.1565μs	240.5879 KOps/s	241.5540 KOps/s	$\color{#d91a1a}-0.40\%$
test_membership_stacked_nested_last	37.2860μs	5.2327μs	191.1050 KOps/s	239.8694 KOps/s	$\textbf{\color{#d91a1a}-20.33\%}$
test_membership_stacked_nested_leaf_last	38.9520μs	5.2248μs	191.3966 KOps/s	237.0256 KOps/s	$\textbf{\color{#d91a1a}-19.25\%}$
test_nested_getleaf	50.9150μs	10.6291μs	94.0812 KOps/s	94.4616 KOps/s	$\color{#d91a1a}-0.40\%$
test_nested_get	51.1960μs	9.9808μs	100.1923 KOps/s	100.1431 KOps/s	$\color{#35bf28}+0.05\%$
test_stacked_getleaf	35.4060μs	10.3435μs	96.6794 KOps/s	95.4085 KOps/s	$\color{#35bf28}+1.33\%$
test_stacked_get	47.2380μs	9.7865μs	102.1815 KOps/s	100.3556 KOps/s	$\color{#35bf28}+1.82\%$
test_nested_getitemleaf	54.3520μs	11.1303μs	89.8449 KOps/s	88.0925 KOps/s	$\color{#35bf28}+1.99\%$
test_nested_getitem	38.8320μs	10.3881μs	96.2644 KOps/s	96.8074 KOps/s	$\color{#d91a1a}-0.56\%$
test_stacked_getitemleaf	52.8780μs	10.9969μs	90.9343 KOps/s	90.7921 KOps/s	$\color{#35bf28}+0.16\%$
test_stacked_getitem	54.4710μs	10.2336μs	97.7177 KOps/s	98.4071 KOps/s	$\color{#d91a1a}-0.70\%$
test_lock_nested	50.6845ms	0.3885ms	2.5743 KOps/s	2.9692 KOps/s	$\textbf{\color{#d91a1a}-13.30\%}$
test_lock_stack_nested	0.4587ms	0.2996ms	3.3373 KOps/s	3.2367 KOps/s	$\color{#35bf28}+3.11\%$
test_unlock_nested	0.6806ms	0.3390ms	2.9497 KOps/s	2.8940 KOps/s	$\color{#35bf28}+1.92\%$
test_unlock_stack_nested	0.4014ms	0.3076ms	3.2511 KOps/s	3.1498 KOps/s	$\color{#35bf28}+3.22\%$
test_flatten_speed	0.1887ms	94.0362μs	10.6342 KOps/s	10.4412 KOps/s	$\color{#35bf28}+1.85\%$
test_unflatten_speed	0.5050ms	0.4079ms	2.4516 KOps/s	2.4834 KOps/s	$\color{#d91a1a}-1.28\%$
test_common_ops	5.5372ms	0.7258ms	1.3777 KOps/s	1.4508 KOps/s	$\textbf{\color{#d91a1a}-5.03\%}$
test_creation	36.9890μs	1.8525μs	539.8020 KOps/s	529.4516 KOps/s	$\color{#35bf28}+1.95\%$
test_creation_empty	26.0690μs	10.6905μs	93.5411 KOps/s	106.2533 KOps/s	$\textbf{\color{#d91a1a}-11.96\%}$
test_creation_nested_1	43.5920μs	13.4427μs	74.3897 KOps/s	82.4778 KOps/s	$\textbf{\color{#d91a1a}-9.81\%}$
test_creation_nested_2	66.3140μs	16.8214μs	59.4482 KOps/s	65.1451 KOps/s	$\textbf{\color{#d91a1a}-8.74\%}$
test_clone	1.3117ms	13.3476μs	74.9201 KOps/s	74.4311 KOps/s	$\color{#35bf28}+0.66\%$
test_getitem[int]	34.3740μs	11.4478μs	87.3530 KOps/s	88.9394 KOps/s	$\color{#d91a1a}-1.78\%$
test_getitem[slice_int]	66.3940μs	22.8291μs	43.8037 KOps/s	43.5335 KOps/s	$\color{#35bf28}+0.62\%$
test_getitem[range]	77.8040μs	60.3202μs	16.5782 KOps/s	17.0966 KOps/s	$\color{#d91a1a}-3.03\%$
test_getitem[tuple]	58.4790μs	18.4156μs	54.3017 KOps/s	52.5147 KOps/s	$\color{#35bf28}+3.40\%$
test_getitem[list]	0.1010ms	39.8187μs	25.1138 KOps/s	24.2512 KOps/s	$\color{#35bf28}+3.56\%$
test_setitem_dim[int]	72.5250μs	34.9592μs	28.6048 KOps/s	30.1101 KOps/s	$\color{#d91a1a}-5.00\%$
test_setitem_dim[slice_int]	0.1065ms	60.7791μs	16.4530 KOps/s	16.5049 KOps/s	$\color{#d91a1a}-0.31\%$
test_setitem_dim[range]	0.1421ms	81.6169μs	12.2524 KOps/s	12.2710 KOps/s	$\color{#d91a1a}-0.15\%$
test_setitem_dim[tuple]	96.9200μs	48.9510μs	20.4286 KOps/s	21.0836 KOps/s	$\color{#d91a1a}-3.11\%$
test_setitem	59.1810μs	20.1030μs	49.7438 KOps/s	51.2136 KOps/s	$\color{#d91a1a}-2.87\%$
test_set	61.4940μs	19.8021μs	50.4997 KOps/s	52.6100 KOps/s	$\color{#d91a1a}-4.01\%$
test_set_shared	1.6563ms	0.1408ms	7.1034 KOps/s	7.0804 KOps/s	$\color{#35bf28}+0.32\%$
test_update	0.1291ms	21.6985μs	46.0861 KOps/s	48.0211 KOps/s	$\color{#d91a1a}-4.03\%$
test_update_nested	93.1640μs	30.9521μs	32.3079 KOps/s	34.1054 KOps/s	$\textbf{\color{#d91a1a}-5.27\%}$
test_update__nested	58.6990μs	24.8234μs	40.2845 KOps/s	40.0234 KOps/s	$\color{#35bf28}+0.65\%$
test_set_nested	72.6850μs	21.3275μs	46.8878 KOps/s	47.7203 KOps/s	$\color{#d91a1a}-1.74\%$
test_set_nested_new	77.0430μs	25.4366μs	39.3134 KOps/s	39.9602 KOps/s	$\color{#d91a1a}-1.62\%$
test_select	0.1007ms	40.0620μs	24.9613 KOps/s	24.8976 KOps/s	$\color{#35bf28}+0.26\%$
test_select_nested	0.1134ms	59.0599μs	16.9320 KOps/s	17.1073 KOps/s	$\color{#d91a1a}-1.02\%$
test_exclude_nested	0.2205ms	0.1181ms	8.4684 KOps/s	8.4910 KOps/s	$\color{#d91a1a}-0.27\%$
test_empty[True]	1.1521ms	0.3924ms	2.5482 KOps/s	2.5748 KOps/s	$\color{#d91a1a}-1.03\%$
test_empty[False]	8.8506μs	1.1328μs	882.7782 KOps/s	859.1529 KOps/s	$\color{#35bf28}+2.75\%$
test_unbind_speed	0.5102ms	0.2573ms	3.8867 KOps/s	3.9650 KOps/s	$\color{#d91a1a}-1.97\%$
test_unbind_speed_stack0	0.4229ms	0.2498ms	4.0038 KOps/s	3.9602 KOps/s	$\color{#35bf28}+1.10\%$
test_unbind_speed_stack1	68.3351ms	0.7148ms	1.3989 KOps/s	1.3572 KOps/s	$\color{#35bf28}+3.08\%$
test_split	65.8132ms	1.5941ms	627.3128 Ops/s	625.8951 Ops/s	$\color{#35bf28}+0.23\%$
test_chunk	70.0107ms	1.6550ms	604.2236 Ops/s	624.2109 Ops/s	$\color{#d91a1a}-3.20\%$
test_creation[device0]	0.2319ms	84.0966μs	11.8911 KOps/s	12.2035 KOps/s	$\color{#d91a1a}-2.56\%$
test_creation_from_tensor	0.2629ms	82.6097μs	12.1051 KOps/s	11.6829 KOps/s	$\color{#35bf28}+3.61\%$
test_add_one[memmap_tensor0]	77.9850μs	5.2730μs	189.6460 KOps/s	187.0717 KOps/s	$\color{#35bf28}+1.38\%$
test_contiguous[memmap_tensor0]	13.3450μs	0.6448μs	1.5508 MOps/s	1.5741 MOps/s	$\color{#d91a1a}-1.48\%$
test_stack[memmap_tensor0]	25.3170μs	3.5464μs	281.9778 KOps/s	279.3187 KOps/s	$\color{#35bf28}+0.95\%$
test_memmaptd_index	1.0340ms	0.2459ms	4.0665 KOps/s	3.9169 KOps/s	$\color{#35bf28}+3.82\%$
test_memmaptd_index_astensor	0.7527ms	0.3189ms	3.1354 KOps/s	3.0373 KOps/s	$\color{#35bf28}+3.23\%$
test_memmaptd_index_op	1.0372ms	0.6021ms	1.6608 KOps/s	1.7262 KOps/s	$\color{#d91a1a}-3.78\%$
test_serialize_model	0.1770s	0.1125s	8.8897 Ops/s	8.3393 Ops/s	$\textbf{\color{#35bf28}+6.60\%}$
test_serialize_model_pickle	0.4493s	0.3779s	2.6462 Ops/s	2.6295 Ops/s	$\color{#35bf28}+0.64\%$
test_serialize_weights	0.1654s	0.1090s	9.1714 Ops/s	8.7869 Ops/s	$\color{#35bf28}+4.38\%$
test_serialize_weights_returnearly	0.1843s	0.1328s	7.5275 Ops/s	7.1323 Ops/s	$\textbf{\color{#35bf28}+5.54\%}$
test_serialize_weights_pickle	1.1361s	0.6087s	1.6428 Ops/s	1.4547 Ops/s	$\textbf{\color{#35bf28}+12.93\%}$
test_serialize_weights_filesystem	0.1024s	90.5629ms	11.0420 Ops/s	10.7003 Ops/s	$\color{#35bf28}+3.19\%$
test_serialize_model_filesystem	0.1533s	98.9581ms	10.1053 Ops/s	10.0982 Ops/s	$\color{#35bf28}+0.07\%$
test_reshape_pytree	67.4860μs	25.6125μs	39.0434 KOps/s	38.5157 KOps/s	$\color{#35bf28}+1.37\%$
test_reshape_td	65.6120μs	34.5695μs	28.9272 KOps/s	28.5968 KOps/s	$\color{#35bf28}+1.16\%$
test_view_pytree	59.1800μs	25.3340μs	39.4727 KOps/s	39.4012 KOps/s	$\color{#35bf28}+0.18\%$
test_view_td	82.3430μs	37.7298μs	26.5042 KOps/s	25.5907 KOps/s	$\color{#35bf28}+3.57\%$
test_unbind_pytree	60.4730μs	28.3480μs	35.2759 KOps/s	33.7152 KOps/s	$\color{#35bf28}+4.63\%$
test_unbind_td	0.3899ms	37.1182μs	26.9410 KOps/s	26.4827 KOps/s	$\color{#35bf28}+1.73\%$
test_split_pytree	74.4890μs	28.8684μs	34.6399 KOps/s	33.5916 KOps/s	$\color{#35bf28}+3.12\%$
test_split_td	0.1223ms	40.3764μs	24.7669 KOps/s	24.5520 KOps/s	$\color{#35bf28}+0.88\%$
test_add_pytree	87.4920μs	33.9013μs	29.4974 KOps/s	28.7973 KOps/s	$\color{#35bf28}+2.43\%$
test_add_td	0.1168ms	56.6874μs	17.6406 KOps/s	19.4693 KOps/s	$\textbf{\color{#d91a1a}-9.39\%}$
test_distributed	0.1920ms	0.1005ms	9.9472 KOps/s	9.7789 KOps/s	$\color{#35bf28}+1.72\%$
test_tdmodule	39.3540μs	17.2312μs	58.0344 KOps/s	59.3356 KOps/s	$\color{#d91a1a}-2.19\%$
test_tdmodule_dispatch	59.8620μs	34.5874μs	28.9123 KOps/s	30.2210 KOps/s	$\color{#d91a1a}-4.33\%$
test_tdseq	39.5730μs	20.1639μs	49.5936 KOps/s	49.6319 KOps/s	$\color{#d91a1a}-0.08\%$
test_tdseq_dispatch	68.6880μs	39.4473μs	25.3503 KOps/s	25.8873 KOps/s	$\color{#d91a1a}-2.07\%$
test_instantiation_functorch	2.0526ms	1.3463ms	742.7785 Ops/s	767.9545 Ops/s	$\color{#d91a1a}-3.28\%$
test_instantiation_td	1.5932ms	1.0051ms	994.9084 Ops/s	991.9273 Ops/s	$\color{#35bf28}+0.30\%$
test_exec_functorch	0.3473ms	0.1579ms	6.3339 KOps/s	6.2268 KOps/s	$\color{#35bf28}+1.72\%$
test_exec_functional_call	0.2953ms	0.1482ms	6.7466 KOps/s	6.8132 KOps/s	$\color{#d91a1a}-0.98\%$
test_exec_td	0.3503ms	0.1431ms	6.9863 KOps/s	7.0037 KOps/s	$\color{#d91a1a}-0.25\%$
test_exec_td_decorator	0.7028ms	0.2166ms	4.6165 KOps/s	4.6030 KOps/s	$\color{#35bf28}+0.29\%$
test_vmap_mlp_speed[True-True]	0.9156ms	0.4783ms	2.0906 KOps/s	2.1135 KOps/s	$\color{#d91a1a}-1.08\%$
test_vmap_mlp_speed[True-False]	0.7767ms	0.4716ms	2.1203 KOps/s	2.1588 KOps/s	$\color{#d91a1a}-1.78\%$
test_vmap_mlp_speed[False-True]	0.4936ms	0.3824ms	2.6154 KOps/s	2.6273 KOps/s	$\color{#d91a1a}-0.45\%$
test_vmap_mlp_speed[False-False]	8.8512ms	0.3892ms	2.5693 KOps/s	2.6324 KOps/s	$\color{#d91a1a}-2.40\%$
test_vmap_mlp_speed_decorator[True-True]	1.0910ms	0.5431ms	1.8413 KOps/s	1.8533 KOps/s	$\color{#d91a1a}-0.65\%$
test_vmap_mlp_speed_decorator[True-False]	0.6803ms	0.5429ms	1.8420 KOps/s	1.8772 KOps/s	$\color{#d91a1a}-1.88\%$
test_vmap_mlp_speed_decorator[False-True]	0.5633ms	0.4447ms	2.2488 KOps/s	2.2487 KOps/s	$+0.01\%$
test_vmap_mlp_speed_decorator[False-False]	0.8248ms	0.4466ms	2.2390 KOps/s	2.2594 KOps/s	$\color{#d91a1a}-0.90\%$
test_to_module_speed[True]	2.5548ms	1.6682ms	599.4326 Ops/s	596.3715 Ops/s	$\color{#35bf28}+0.51\%$
test_to_module_speed[False]	1.6959ms	1.6263ms	614.8925 Ops/s	541.6151 Ops/s	$\textbf{\color{#35bf28}+13.53\%}$
test_tc_init	61.2040μs	28.8685μs	34.6399 KOps/s	38.8134 KOps/s	$\textbf{\color{#d91a1a}-10.75\%}$
test_tc_init_nested	0.1378ms	59.8105μs	16.7195 KOps/s	19.1246 KOps/s	$\textbf{\color{#d91a1a}-12.58\%}$
test_tc_first_layer_tensor	4.3710μs	0.6877μs	1.4541 MOps/s	1.4521 MOps/s	$\color{#35bf28}+0.13\%$
test_tc_first_layer_nontensor	4.5915μs	0.6984μs	1.4319 MOps/s	1.4476 MOps/s	$\color{#d91a1a}-1.08\%$
test_tc_second_layer_tensor	69.1390μs	1.9268μs	519.0057 KOps/s	538.6658 KOps/s	$\color{#d91a1a}-3.65\%$
test_tc_second_layer_nontensor	52.5847μs	1.5189μs	658.3568 KOps/s	592.9207 KOps/s	$\textbf{\color{#35bf28}+11.04\%}$
test_unbind	82.3391ms	7.1538ms	139.7862 Ops/s	151.9005 Ops/s	$\textbf{\color{#d91a1a}-7.98\%}$
test_full_like	18.5020ms	10.6193ms	94.1678 Ops/s	91.8939 Ops/s	$\color{#35bf28}+2.47\%$
test_zeros_like	11.5032ms	5.5313ms	180.7900 Ops/s	171.8965 Ops/s	$\textbf{\color{#35bf28}+5.17\%}$
test_ones_like	12.0021ms	5.9599ms	167.7886 Ops/s	159.1026 Ops/s	$\textbf{\color{#35bf28}+5.46\%}$
test_clone	14.0439ms	7.7876ms	128.4097 Ops/s	123.8621 Ops/s	$\color{#35bf28}+3.67\%$
test_squeeze	60.0410μs	14.0664μs	71.0915 KOps/s	72.3519 KOps/s	$\color{#d91a1a}-1.74\%$
test_unsqueeze	0.1353ms	59.8462μs	16.7095 KOps/s	16.3973 KOps/s	$\color{#35bf28}+1.90\%$
test_split	0.2763ms	0.1132ms	8.8335 KOps/s	8.8799 KOps/s	$\color{#d91a1a}-0.52\%$
test_permute	0.2291ms	0.1243ms	8.0438 KOps/s	7.8524 KOps/s	$\color{#35bf28}+2.44\%$
test_stack	29.4246ms	22.2994ms	44.8442 Ops/s	44.3634 Ops/s	$\color{#35bf28}+1.08\%$
test_cat	29.8169ms	22.0615ms	45.3278 Ops/s	45.3248 Ops/s	$+0.01\%$

[github-actions]

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

Total Benchmarks: 152. Improved: $\large\color{#35bf28}6$. Worsened: $\large\color{#d91a1a}18$.

Expand to view detailed results

Name	Max	Mean	Ops	Ops on Repo HEAD	Change
test_plain_set_nested	31.1400μs	13.6011μs	73.5234 KOps/s	79.5879 KOps/s	$\textbf{\color{#d91a1a}-7.62\%}$
test_plain_set_stack_nested	28.3210μs	13.7056μs	72.9628 KOps/s	78.7725 KOps/s	$\textbf{\color{#d91a1a}-7.38\%}$
test_plain_set_nested_inplace	37.1300μs	14.8804μs	67.2024 KOps/s	72.1146 KOps/s	$\textbf{\color{#d91a1a}-6.81\%}$
test_plain_set_stack_nested_inplace	36.0310μs	14.8955μs	67.1346 KOps/s	71.7004 KOps/s	$\textbf{\color{#d91a1a}-6.37\%}$
test_items	23.4500μs	4.7558μs	210.2703 KOps/s	208.1048 KOps/s	$\color{#35bf28}+1.04\%$
test_items_nested	0.3714ms	0.3379ms	2.9595 KOps/s	3.0103 KOps/s	$\color{#d91a1a}-1.69\%$
test_items_nested_locked	0.3812ms	0.3535ms	2.8287 KOps/s	2.9521 KOps/s	$\color{#d91a1a}-4.18\%$
test_items_nested_leaf	0.1058ms	83.5106μs	11.9745 KOps/s	12.1461 KOps/s	$\color{#d91a1a}-1.41\%$
test_items_stack_nested	0.3644ms	0.3389ms	2.9509 KOps/s	2.9544 KOps/s	$\color{#d91a1a}-0.12\%$
test_items_stack_nested_leaf	0.1009ms	84.5405μs	11.8287 KOps/s	11.9115 KOps/s	$\color{#d91a1a}-0.70\%$
test_items_stack_nested_locked	0.3687ms	0.3393ms	2.9475 KOps/s	2.9447 KOps/s	$\color{#35bf28}+0.10\%$
test_keys	22.3200μs	4.3264μs	231.1400 KOps/s	229.6119 KOps/s	$\color{#35bf28}+0.67\%$
test_keys_nested	96.1910μs	68.8103μs	14.5327 KOps/s	14.7027 KOps/s	$\color{#d91a1a}-1.16\%$
test_keys_nested_locked	2.1207ms	73.2431μs	13.6532 KOps/s	13.7518 KOps/s	$\color{#d91a1a}-0.72\%$
test_keys_nested_leaf	78.7710μs	58.8983μs	16.9784 KOps/s	17.0655 KOps/s	$\color{#d91a1a}-0.51\%$
test_keys_stack_nested	92.8510μs	68.0618μs	14.6925 KOps/s	14.7856 KOps/s	$\color{#d91a1a}-0.63\%$
test_keys_stack_nested_leaf	81.8210μs	58.7078μs	17.0335 KOps/s	17.0779 KOps/s	$\color{#d91a1a}-0.26\%$
test_keys_stack_nested_locked	0.1059ms	73.4163μs	13.6209 KOps/s	13.7437 KOps/s	$\color{#d91a1a}-0.89\%$
test_values	8.7367μs	1.8483μs	541.0489 KOps/s	550.9556 KOps/s	$\color{#d91a1a}-1.80\%$
test_values_nested	67.9610μs	35.3198μs	28.3127 KOps/s	28.0344 KOps/s	$\color{#35bf28}+0.99\%$
test_values_nested_locked	55.1410μs	37.1544μs	26.9147 KOps/s	26.9196 KOps/s	$\color{#d91a1a}-0.02\%$
test_values_nested_leaf	53.0600μs	31.5596μs	31.6860 KOps/s	31.8195 KOps/s	$\color{#d91a1a}-0.42\%$
test_values_stack_nested	55.2500μs	36.2392μs	27.5945 KOps/s	27.5365 KOps/s	$\color{#35bf28}+0.21\%$
test_values_stack_nested_leaf	56.2600μs	32.2400μs	31.0174 KOps/s	31.2157 KOps/s	$\color{#d91a1a}-0.64\%$
test_values_stack_nested_locked	59.5800μs	38.0108μs	26.3083 KOps/s	26.5271 KOps/s	$\color{#d91a1a}-0.82\%$
test_membership	1.7350μs	0.7049μs	1.4186 MOps/s	1.4170 MOps/s	$\color{#35bf28}+0.12\%$
test_membership_nested	27.5200μs	2.5772μs	388.0153 KOps/s	395.0789 KOps/s	$\color{#d91a1a}-1.79\%$
test_membership_nested_leaf	18.0600μs	2.5791μs	387.7374 KOps/s	389.0102 KOps/s	$\color{#d91a1a}-0.33\%$
test_membership_stacked_nested	15.3100μs	2.5460μs	392.7739 KOps/s	388.6895 KOps/s	$\color{#35bf28}+1.05\%$
test_membership_stacked_nested_leaf	33.2600μs	2.6009μs	384.4857 KOps/s	388.6486 KOps/s	$\color{#d91a1a}-1.07\%$
test_membership_nested_last	21.3710μs	3.1411μs	318.3549 KOps/s	322.4275 KOps/s	$\color{#d91a1a}-1.26\%$
test_membership_nested_leaf_last	57.4000μs	3.1072μs	321.8281 KOps/s	322.0654 KOps/s	$\color{#d91a1a}-0.07\%$
test_membership_stacked_nested_last	14.4500μs	3.0851μs	324.1425 KOps/s	324.1752 KOps/s	$\color{#d91a1a}-0.01\%$
test_membership_stacked_nested_leaf_last	21.1300μs	3.1227μs	320.2376 KOps/s	323.0815 KOps/s	$\color{#d91a1a}-0.88\%$
test_nested_getleaf	38.3900μs	8.4275μs	118.6585 KOps/s	118.9320 KOps/s	$\color{#d91a1a}-0.23\%$
test_nested_get	31.6710μs	7.9868μs	125.2068 KOps/s	126.2895 KOps/s	$\color{#d91a1a}-0.86\%$
test_stacked_getleaf	25.6810μs	8.4666μs	118.1113 KOps/s	118.7480 KOps/s	$\color{#d91a1a}-0.54\%$
test_stacked_get	36.2510μs	7.9608μs	125.6153 KOps/s	125.5631 KOps/s	$\color{#35bf28}+0.04\%$
test_nested_getitemleaf	33.2400μs	8.6093μs	116.1529 KOps/s	116.4313 KOps/s	$\color{#d91a1a}-0.24\%$
test_nested_getitem	22.3410μs	8.1261μs	123.0607 KOps/s	123.8858 KOps/s	$\color{#d91a1a}-0.67\%$
test_stacked_getitemleaf	35.5200μs	8.6354μs	115.8022 KOps/s	115.6223 KOps/s	$\color{#35bf28}+0.16\%$
test_stacked_getitem	23.2400μs	8.0920μs	123.5788 KOps/s	123.6282 KOps/s	$\color{#d91a1a}-0.04\%$
test_lock_nested	57.6021ms	0.4025ms	2.4845 KOps/s	2.4782 KOps/s	$\color{#35bf28}+0.26\%$
test_lock_stack_nested	0.3438ms	0.2979ms	3.3565 KOps/s	3.3136 KOps/s	$\color{#35bf28}+1.29\%$
test_unlock_nested	59.5620ms	0.4010ms	2.4936 KOps/s	2.4601 KOps/s	$\color{#35bf28}+1.36\%$
test_unlock_stack_nested	0.3302ms	0.3059ms	3.2687 KOps/s	3.2277 KOps/s	$\color{#35bf28}+1.27\%$
test_flatten_speed	0.3332ms	0.1003ms	9.9711 KOps/s	9.7761 KOps/s	$\color{#35bf28}+2.00\%$
test_unflatten_speed	0.3332ms	0.2894ms	3.4558 KOps/s	3.4528 KOps/s	$\color{#35bf28}+0.09\%$
test_common_ops	1.0617ms	0.6113ms	1.6360 KOps/s	1.7071 KOps/s	$\color{#d91a1a}-4.17\%$
test_creation	15.8400μs	1.6664μs	600.0833 KOps/s	592.4725 KOps/s	$\color{#35bf28}+1.28\%$
test_creation_empty	30.0300μs	10.0362μs	99.6396 KOps/s	121.7439 KOps/s	$\textbf{\color{#d91a1a}-18.16\%}$
test_creation_nested_1	40.9610μs	11.8528μs	84.3683 KOps/s	100.2829 KOps/s	$\textbf{\color{#d91a1a}-15.87\%}$
test_creation_nested_2	30.7400μs	13.9572μs	71.6477 KOps/s	81.0803 KOps/s	$\textbf{\color{#d91a1a}-11.63\%}$
test_clone	73.3510μs	11.9976μs	83.3499 KOps/s	82.2989 KOps/s	$\color{#35bf28}+1.28\%$
test_getitem[int]	34.2010μs	10.7197μs	93.2861 KOps/s	93.5129 KOps/s	$\color{#d91a1a}-0.24\%$
test_getitem[slice_int]	37.6600μs	20.5519μs	48.6574 KOps/s	44.4520 KOps/s	$\textbf{\color{#35bf28}+9.46\%}$
test_getitem[range]	65.4010μs	47.2664μs	21.1567 KOps/s	20.9633 KOps/s	$\color{#35bf28}+0.92\%$
test_getitem[tuple]	57.9210μs	18.6477μs	53.6260 KOps/s	55.3036 KOps/s	$\color{#d91a1a}-3.03\%$
test_getitem[list]	0.1213ms	34.6526μs	28.8579 KOps/s	28.7955 KOps/s	$\color{#35bf28}+0.22\%$
test_setitem_dim[int]	46.5000μs	30.3392μs	32.9607 KOps/s	35.0590 KOps/s	$\textbf{\color{#d91a1a}-5.98\%}$
test_setitem_dim[slice_int]	68.0310μs	50.2780μs	19.8894 KOps/s	19.9888 KOps/s	$\color{#d91a1a}-0.50\%$
test_setitem_dim[range]	91.4110μs	68.7103μs	14.5539 KOps/s	15.0561 KOps/s	$\color{#d91a1a}-3.34\%$
test_setitem_dim[tuple]	61.3610μs	44.3869μs	22.5292 KOps/s	23.5986 KOps/s	$\color{#d91a1a}-4.53\%$
test_setitem	55.4700μs	17.8116μs	56.1431 KOps/s	60.1325 KOps/s	$\textbf{\color{#d91a1a}-6.63\%}$
test_set	47.7210μs	16.8989μs	59.1753 KOps/s	61.6229 KOps/s	$\color{#d91a1a}-3.97\%$
test_set_shared	1.3410ms	99.4934μs	10.0509 KOps/s	9.9728 KOps/s	$\color{#35bf28}+0.78\%$
test_update	90.2710μs	20.8175μs	48.0366 KOps/s	51.7720 KOps/s	$\textbf{\color{#d91a1a}-7.22\%}$
test_update_nested	65.6800μs	26.8669μs	37.2205 KOps/s	37.9202 KOps/s	$\color{#d91a1a}-1.85\%$
test_update__nested	64.0710μs	23.1564μs	43.1845 KOps/s	43.5216 KOps/s	$\color{#d91a1a}-0.77\%$
test_set_nested	75.2110μs	18.4582μs	54.1763 KOps/s	54.9088 KOps/s	$\color{#d91a1a}-1.33\%$
test_set_nested_new	54.3810μs	21.2430μs	47.0744 KOps/s	47.7396 KOps/s	$\color{#d91a1a}-1.39\%$
test_select	75.2010μs	34.6907μs	28.8262 KOps/s	30.1349 KOps/s	$\color{#d91a1a}-4.34\%$
test_select_nested	0.5340ms	55.7732μs	17.9297 KOps/s	17.7511 KOps/s	$\color{#35bf28}+1.01\%$
test_exclude_nested	0.1363ms	0.1118ms	8.9460 KOps/s	8.9434 KOps/s	$\color{#35bf28}+0.03\%$
test_empty[True]	0.3681ms	0.3462ms	2.8885 KOps/s	2.8651 KOps/s	$\color{#35bf28}+0.82\%$
test_empty[False]	2.8081μs	0.9147μs	1.0933 MOps/s	1.0731 MOps/s	$\color{#35bf28}+1.88\%$
test_to	0.1034ms	76.7541μs	13.0286 KOps/s	13.2109 KOps/s	$\color{#d91a1a}-1.38\%$
test_to_nonblocking	94.2810μs	62.1432μs	16.0919 KOps/s	15.0977 KOps/s	$\textbf{\color{#35bf28}+6.58\%}$
test_unbind_speed	0.2996ms	0.2627ms	3.8067 KOps/s	3.8044 KOps/s	$\color{#35bf28}+0.06\%$
test_unbind_speed_stack0	0.3021ms	0.2650ms	3.7740 KOps/s	3.7951 KOps/s	$\color{#d91a1a}-0.56\%$
test_unbind_speed_stack1	74.3637ms	0.7960ms	1.2563 KOps/s	1.2410 KOps/s	$\color{#35bf28}+1.24\%$
test_split	75.1542ms	1.6724ms	597.9311 Ops/s	602.0993 Ops/s	$\color{#d91a1a}-0.69\%$
test_chunk	1.6123ms	1.5584ms	641.6693 Ops/s	600.9901 Ops/s	$\textbf{\color{#35bf28}+6.77\%}$
test_creation[device0]	0.1280ms	57.6364μs	17.3502 KOps/s	17.0750 KOps/s	$\color{#35bf28}+1.61\%$
test_creation_from_tensor	0.1504ms	54.0414μs	18.5043 KOps/s	17.2961 KOps/s	$\textbf{\color{#35bf28}+6.99\%}$
test_add_one[memmap_tensor0]	76.5620μs	7.1016μs	140.8129 KOps/s	140.6538 KOps/s	$\color{#35bf28}+0.11\%$
test_contiguous[memmap_tensor0]	9.7500μs	0.6547μs	1.5275 MOps/s	1.5123 MOps/s	$\color{#35bf28}+1.01\%$
test_stack[memmap_tensor0]	32.4610μs	4.9535μs	201.8777 KOps/s	205.0057 KOps/s	$\color{#d91a1a}-1.53\%$
test_memmaptd_index	1.3183ms	0.2861ms	3.4954 KOps/s	3.5103 KOps/s	$\color{#d91a1a}-0.42\%$
test_memmaptd_index_astensor	74.7010ms	0.3909ms	2.5584 KOps/s	2.8267 KOps/s	$\textbf{\color{#d91a1a}-9.49\%}$
test_memmaptd_index_op	1.2411ms	0.6862ms	1.4572 KOps/s	1.5197 KOps/s	$\color{#d91a1a}-4.11\%$
test_serialize_model	0.1838s	0.1102s	9.0737 Ops/s	8.7096 Ops/s	$\color{#35bf28}+4.18\%$
test_serialize_model_pickle	1.3666s	1.2377s	0.8079 Ops/s	0.8065 Ops/s	$\color{#35bf28}+0.18\%$
test_serialize_weights	0.1765s	0.1076s	9.2978 Ops/s	8.7909 Ops/s	$\textbf{\color{#35bf28}+5.77\%}$
test_serialize_weights_returnearly	0.2842s	0.1040s	9.6144 Ops/s	10.0374 Ops/s	$\color{#d91a1a}-4.21\%$
test_serialize_weights_pickle	1.3520s	1.2480s	0.8013 Ops/s	0.8010 Ops/s	$\color{#35bf28}+0.04\%$
test_reshape_pytree	62.5900μs	26.3363μs	37.9704 KOps/s	38.4287 KOps/s	$\color{#d91a1a}-1.19\%$
test_reshape_td	68.1010μs	30.5651μs	32.7170 KOps/s	32.2472 KOps/s	$\color{#35bf28}+1.46\%$
test_view_pytree	53.7210μs	26.0632μs	38.3683 KOps/s	39.3884 KOps/s	$\color{#d91a1a}-2.59\%$
test_view_td	66.5910μs	35.4348μs	28.2208 KOps/s	28.6055 KOps/s	$\color{#d91a1a}-1.34\%$
test_unbind_pytree	59.9200μs	31.6995μs	31.5463 KOps/s	31.5395 KOps/s	$\color{#35bf28}+0.02\%$
test_unbind_td	0.4264ms	39.6862μs	25.1977 KOps/s	24.8462 KOps/s	$\color{#35bf28}+1.41\%$
test_split_pytree	65.2610μs	34.1991μs	29.2406 KOps/s	29.2216 KOps/s	$\color{#35bf28}+0.07\%$
test_split_td	0.2661ms	38.9969μs	25.6431 KOps/s	25.2035 KOps/s	$\color{#35bf28}+1.74\%$
test_add_pytree	72.2010μs	38.3644μs	26.0658 KOps/s	26.4564 KOps/s	$\color{#d91a1a}-1.48\%$
test_add_td	93.9210μs	52.3284μs	19.1101 KOps/s	19.6840 KOps/s	$\color{#d91a1a}-2.92\%$
test_distributed	1.8805ms	68.4335μs	14.6127 KOps/s	14.8248 KOps/s	$\color{#d91a1a}-1.43\%$
test_tdmodule	84.8910μs	15.5024μs	64.5061 KOps/s	70.3843 KOps/s	$\textbf{\color{#d91a1a}-8.35\%}$
test_tdmodule_dispatch	46.9710μs	30.7803μs	32.4883 KOps/s	35.0177 KOps/s	$\textbf{\color{#d91a1a}-7.22\%}$
test_tdseq	33.5610μs	17.3278μs	57.7107 KOps/s	61.1428 KOps/s	$\textbf{\color{#d91a1a}-5.61\%}$
test_tdseq_dispatch	58.1200μs	34.6527μs	28.8577 KOps/s	31.2830 KOps/s	$\textbf{\color{#d91a1a}-7.75\%}$
test_instantiation_functorch	1.6301ms	1.5313ms	653.0482 Ops/s	653.1604 Ops/s	$\color{#d91a1a}-0.02\%$
test_instantiation_td	1.5587ms	1.0533ms	949.3808 Ops/s	957.4899 Ops/s	$\color{#d91a1a}-0.85\%$
test_exec_functorch	0.1831ms	0.1501ms	6.6619 KOps/s	6.5833 KOps/s	$\color{#35bf28}+1.19\%$
test_exec_functional_call	0.1873ms	0.1424ms	7.0201 KOps/s	6.9979 KOps/s	$\color{#35bf28}+0.32\%$
test_exec_td	0.1696ms	0.1386ms	7.2152 KOps/s	6.9613 KOps/s	$\color{#35bf28}+3.65\%$
test_exec_td_decorator	0.8130ms	0.2149ms	4.6527 KOps/s	4.6276 KOps/s	$\color{#35bf28}+0.54\%$
test_vmap_mlp_speed[True-True]	0.7767ms	0.5887ms	1.6987 KOps/s	1.6701 KOps/s	$\color{#35bf28}+1.71\%$
test_vmap_mlp_speed[True-False]	0.6688ms	0.5883ms	1.6998 KOps/s	1.6995 KOps/s	$\color{#35bf28}+0.02\%$
test_vmap_mlp_speed[False-True]	0.6131ms	0.5353ms	1.8680 KOps/s	1.9339 KOps/s	$\color{#d91a1a}-3.41\%$
test_vmap_mlp_speed[False-False]	0.5680ms	0.5155ms	1.9398 KOps/s	1.9462 KOps/s	$\color{#d91a1a}-0.33\%$
test_vmap_mlp_speed_decorator[True-True]	1.0188ms	0.6561ms	1.5242 KOps/s	1.5418 KOps/s	$\color{#d91a1a}-1.14\%$
test_vmap_mlp_speed_decorator[True-False]	0.7358ms	0.6531ms	1.5311 KOps/s	1.5590 KOps/s	$\color{#d91a1a}-1.79\%$
test_vmap_mlp_speed_decorator[False-True]	0.7494ms	0.5860ms	1.7064 KOps/s	1.7440 KOps/s	$\color{#d91a1a}-2.15\%$
test_vmap_mlp_speed_decorator[False-False]	0.7157ms	0.5868ms	1.7041 KOps/s	1.7421 KOps/s	$\color{#d91a1a}-2.18\%$
test_vmap_transformer_speed[True-True]	8.0426ms	7.6829ms	130.1585 Ops/s	129.6139 Ops/s	$\color{#35bf28}+0.42\%$
test_vmap_transformer_speed[True-False]	8.0609ms	7.6827ms	130.1628 Ops/s	130.0732 Ops/s	$\color{#35bf28}+0.07\%$
test_vmap_transformer_speed[False-True]	8.0864ms	7.7658ms	128.7702 Ops/s	130.5461 Ops/s	$\color{#d91a1a}-1.36\%$
test_vmap_transformer_speed[False-False]	7.7545ms	7.5909ms	131.7367 Ops/s	130.4334 Ops/s	$\color{#35bf28}+1.00\%$
test_vmap_transformer_speed_decorator[True-True]	19.1351ms	18.6795ms	53.5346 Ops/s	53.3123 Ops/s	$\color{#35bf28}+0.42\%$
test_vmap_transformer_speed_decorator[True-False]	19.6616ms	18.7477ms	53.3399 Ops/s	53.2183 Ops/s	$\color{#35bf28}+0.23\%$
test_vmap_transformer_speed_decorator[False-True]	19.4852ms	18.6032ms	53.7541 Ops/s	53.5445 Ops/s	$\color{#35bf28}+0.39\%$
test_vmap_transformer_speed_decorator[False-False]	19.4327ms	18.5969ms	53.7723 Ops/s	53.5799 Ops/s	$\color{#35bf28}+0.36\%$
test_to_module_speed[True]	1.7112ms	1.5907ms	628.6642 Ops/s	632.7236 Ops/s	$\color{#d91a1a}-0.64\%$
test_to_module_speed[False]	2.5889ms	1.5680ms	637.7448 Ops/s	650.1753 Ops/s	$\color{#d91a1a}-1.91\%$
test_tc_init	48.6100μs	28.7890μs	34.7355 KOps/s	41.7950 KOps/s	$\textbf{\color{#d91a1a}-16.89\%}$
test_tc_init_nested	89.8710μs	63.1593μs	15.8330 KOps/s	20.4764 KOps/s	$\textbf{\color{#d91a1a}-22.68\%}$
test_tc_first_layer_tensor	0.9513μs	0.3583μs	2.7912 MOps/s	2.7490 MOps/s	$\color{#35bf28}+1.53\%$
test_tc_first_layer_nontensor	2.0377μs	0.3901μs	2.5632 MOps/s	2.5483 MOps/s	$\color{#35bf28}+0.59\%$
test_tc_second_layer_tensor	5.8800μs	0.9673μs	1.0338 MOps/s	1.0215 MOps/s	$\color{#35bf28}+1.20\%$
test_tc_second_layer_nontensor	2.3435μs	0.8051μs	1.2420 MOps/s	1.1909 MOps/s	$\color{#35bf28}+4.30\%$
test_unbind	0.1106s	7.5044ms	133.2545 Ops/s	142.7374 Ops/s	$\textbf{\color{#d91a1a}-6.64\%}$
test_full_like	11.9314ms	11.2422ms	88.9505 Ops/s	75.6575 Ops/s	$\textbf{\color{#35bf28}+17.57\%}$
test_zeros_like	8.4526ms	7.8921ms	126.7085 Ops/s	127.8413 Ops/s	$\color{#d91a1a}-0.89\%$
test_ones_like	8.3949ms	7.8577ms	127.2631 Ops/s	127.5286 Ops/s	$\color{#d91a1a}-0.21\%$
test_clone	9.5048ms	9.2941ms	107.5950 Ops/s	108.0755 Ops/s	$\color{#d91a1a}-0.44\%$
test_squeeze	53.5800μs	11.0309μs	90.6544 KOps/s	90.4437 KOps/s	$\color{#35bf28}+0.23\%$
test_unsqueeze	93.3010μs	51.6321μs	19.3678 KOps/s	19.1641 KOps/s	$\color{#35bf28}+1.06\%$
test_split	0.1326ms	0.1002ms	9.9758 KOps/s	10.3396 KOps/s	$\color{#d91a1a}-3.52\%$
test_permute	0.1458ms	0.1130ms	8.8512 KOps/s	9.0495 KOps/s	$\color{#d91a1a}-2.19\%$
test_stack	27.2375ms	26.8940ms	37.1830 Ops/s	37.0562 Ops/s	$\color{#35bf28}+0.34\%$
test_cat	27.7132ms	26.9103ms	37.1605 Ops/s	37.2262 Ops/s	$\color{#d91a1a}-0.18\%$