[BugFix] Faster empty_like for MemoryMappedTensor

[vmoens]

No description provided.

[github-actions]

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

Total Benchmarks: 113. Improved: $\large\color{#35bf28}12$. Worsened: $\large\color{#d91a1a}4$.

Expand to view detailed results

Name	Max	Mean	Ops	Ops on Repo HEAD	Change
test_plain_set_nested	31.8500μs	15.7819μs	63.3636 KOps/s	63.0403 KOps/s	$\color{#35bf28}+0.51\%$
test_plain_set_stack_nested	0.1793ms	0.1435ms	6.9673 KOps/s	6.9051 KOps/s	$\color{#35bf28}+0.90\%$
test_plain_set_nested_inplace	43.5710μs	18.7853μs	53.2331 KOps/s	52.2175 KOps/s	$\color{#35bf28}+1.94\%$
test_plain_set_stack_nested_inplace	0.3254ms	0.1720ms	5.8125 KOps/s	5.8058 KOps/s	$\color{#35bf28}+0.12\%$
test_items	31.3580μs	2.4211μs	413.0283 KOps/s	402.7880 KOps/s	$\color{#35bf28}+2.54\%$
test_items_nested	0.4411ms	0.2838ms	3.5242 KOps/s	3.6204 KOps/s	$\color{#d91a1a}-2.66\%$
test_items_nested_locked	0.3233ms	0.2692ms	3.7150 KOps/s	3.5956 KOps/s	$\color{#35bf28}+3.32\%$
test_items_nested_leaf	0.8253ms	0.1686ms	5.9317 KOps/s	5.8707 KOps/s	$\color{#35bf28}+1.04\%$
test_items_stack_nested	1.6585ms	1.4835ms	674.0658 Ops/s	656.7914 Ops/s	$\color{#35bf28}+2.63\%$
test_items_stack_nested_leaf	1.7671ms	1.3481ms	741.7949 Ops/s	714.9756 Ops/s	$\color{#35bf28}+3.75\%$
test_items_stack_nested_locked	0.8474ms	0.7607ms	1.3146 KOps/s	1.2609 KOps/s	$\color{#35bf28}+4.26\%$
test_keys	52.4780μs	3.8851μs	257.3912 KOps/s	259.3479 KOps/s	$\color{#d91a1a}-0.75\%$
test_keys_nested	3.1677ms	0.1409ms	7.0960 KOps/s	6.7144 KOps/s	$\textbf{\color{#35bf28}+5.68\%}$
test_keys_nested_locked	0.1857ms	0.1390ms	7.1952 KOps/s	7.0608 KOps/s	$\color{#35bf28}+1.90\%$
test_keys_nested_leaf	0.3809ms	0.1389ms	7.2000 KOps/s	6.9710 KOps/s	$\color{#35bf28}+3.28\%$
test_keys_stack_nested	1.5095ms	1.4072ms	710.6479 Ops/s	695.0822 Ops/s	$\color{#35bf28}+2.24\%$
test_keys_stack_nested_leaf	1.7588ms	1.4070ms	710.7531 Ops/s	697.4718 Ops/s	$\color{#35bf28}+1.90\%$
test_keys_stack_nested_locked	0.7550ms	0.6704ms	1.4916 KOps/s	1.4541 KOps/s	$\color{#35bf28}+2.58\%$
test_values	6.5924μs	1.1730μs	852.5254 KOps/s	772.4969 KOps/s	$\textbf{\color{#35bf28}+10.36\%}$
test_values_nested	89.2960μs	49.2461μs	20.3062 KOps/s	20.0247 KOps/s	$\color{#35bf28}+1.41\%$
test_values_nested_locked	82.5730μs	49.3704μs	20.2551 KOps/s	19.8548 KOps/s	$\color{#35bf28}+2.02\%$
test_values_nested_leaf	99.2950μs	44.6909μs	22.3759 KOps/s	22.3729 KOps/s	$\color{#35bf28}+0.01\%$
test_values_stack_nested	1.4297ms	1.1966ms	835.7200 Ops/s	813.3827 Ops/s	$\color{#35bf28}+2.75\%$
test_values_stack_nested_leaf	1.4125ms	1.1913ms	839.4367 Ops/s	824.1280 Ops/s	$\color{#35bf28}+1.86\%$
test_values_stack_nested_locked	0.9665ms	0.5118ms	1.9540 KOps/s	1.9014 KOps/s	$\color{#35bf28}+2.77\%$
test_membership	9.7480μs	1.3523μs	739.4646 KOps/s	727.6418 KOps/s	$\color{#35bf28}+1.62\%$
test_membership_nested	20.7280μs	2.8023μs	356.8508 KOps/s	351.4277 KOps/s	$\color{#35bf28}+1.54\%$
test_membership_nested_leaf	28.1720μs	2.8511μs	350.7442 KOps/s	348.1738 KOps/s	$\color{#35bf28}+0.74\%$
test_membership_stacked_nested	46.7170μs	11.8225μs	84.5844 KOps/s	83.1906 KOps/s	$\color{#35bf28}+1.68\%$
test_membership_stacked_nested_leaf	34.9650μs	11.8419μs	84.4458 KOps/s	79.7554 KOps/s	$\textbf{\color{#35bf28}+5.88\%}$
test_membership_nested_last	34.5540μs	5.9203μs	168.9110 KOps/s	157.1748 KOps/s	$\textbf{\color{#35bf28}+7.47\%}$
test_membership_nested_leaf_last	35.3150μs	6.0254μs	165.9630 KOps/s	166.2569 KOps/s	$\color{#d91a1a}-0.18\%$
test_membership_stacked_nested_last	0.2391ms	0.1683ms	5.9428 KOps/s	5.9803 KOps/s	$\color{#d91a1a}-0.63\%$
test_membership_stacked_nested_leaf_last	47.4580μs	13.7307μs	72.8297 KOps/s	71.8509 KOps/s	$\color{#35bf28}+1.36\%$
test_nested_getleaf	39.7940μs	10.6570μs	93.8348 KOps/s	94.1917 KOps/s	$\color{#d91a1a}-0.38\%$
test_nested_get	41.1470μs	10.1225μs	98.7900 KOps/s	100.5221 KOps/s	$\color{#d91a1a}-1.72\%$
test_stacked_getleaf	1.0610ms	0.6459ms	1.5481 KOps/s	1.5174 KOps/s	$\color{#35bf28}+2.02\%$
test_stacked_get	1.1805ms	0.6138ms	1.6293 KOps/s	1.5753 KOps/s	$\color{#35bf28}+3.43\%$
test_nested_getitemleaf	29.9060μs	10.6520μs	93.8793 KOps/s	93.0424 KOps/s	$\color{#35bf28}+0.90\%$
test_nested_getitem	39.2730μs	10.0967μs	99.0420 KOps/s	97.6376 KOps/s	$\color{#35bf28}+1.44\%$
test_stacked_getitemleaf	0.7623ms	0.6427ms	1.5560 KOps/s	1.5005 KOps/s	$\color{#35bf28}+3.69\%$
test_stacked_getitem	1.3129ms	0.6098ms	1.6399 KOps/s	1.5617 KOps/s	$\textbf{\color{#35bf28}+5.01\%}$
test_lock_nested	7.1352ms	0.5680ms	1.7606 KOps/s	1.7704 KOps/s	$\color{#d91a1a}-0.56\%$
test_lock_stack_nested	7.6001ms	5.0558ms	197.7943 Ops/s	197.2205 Ops/s	$\color{#35bf28}+0.29\%$
test_unlock_nested	70.3042ms	0.5134ms	1.9478 KOps/s	2.2528 KOps/s	$\textbf{\color{#d91a1a}-13.54\%}$
test_unlock_stack_nested	66.0627ms	6.7768ms	147.5613 Ops/s	147.0538 Ops/s	$\color{#35bf28}+0.35\%$
test_flatten_speed	0.5727ms	0.2676ms	3.7364 KOps/s	3.7207 KOps/s	$\color{#35bf28}+0.42\%$
test_unflatten_speed	1.7455ms	0.4835ms	2.0684 KOps/s	2.1800 KOps/s	$\textbf{\color{#d91a1a}-5.12\%}$
test_common_ops	1.2089ms	0.6642ms	1.5056 KOps/s	1.4846 KOps/s	$\color{#35bf28}+1.42\%$
test_creation	58.8390μs	2.4758μs	403.9046 KOps/s	398.9798 KOps/s	$\color{#35bf28}+1.23\%$
test_creation_empty	30.2060μs	8.1581μs	122.5778 KOps/s	123.3529 KOps/s	$\color{#d91a1a}-0.63\%$
test_creation_nested_1	40.4450μs	11.2943μs	88.5406 KOps/s	87.6794 KOps/s	$\color{#35bf28}+0.98\%$
test_creation_nested_2	39.6440μs	15.0413μs	66.4835 KOps/s	66.1438 KOps/s	$\color{#35bf28}+0.51\%$
test_clone	87.8630μs	13.3905μs	74.6799 KOps/s	74.5813 KOps/s	$\color{#35bf28}+0.13\%$
test_getitem[int]	34.8650μs	13.0017μs	76.9133 KOps/s	76.7222 KOps/s	$\color{#35bf28}+0.25\%$
test_getitem[slice_int]	0.1301ms	26.3031μs	38.0183 KOps/s	38.9085 KOps/s	$\color{#d91a1a}-2.29\%$
test_getitem[range]	87.9140μs	44.0657μs	22.6934 KOps/s	22.3669 KOps/s	$\color{#35bf28}+1.46\%$
test_getitem[tuple]	63.3180μs	20.2697μs	49.3347 KOps/s	48.1991 KOps/s	$\color{#35bf28}+2.36\%$
test_getitem[list]	94.2850μs	38.6804μs	25.8529 KOps/s	25.2651 KOps/s	$\color{#35bf28}+2.33\%$
test_setitem_dim[int]	59.4300μs	28.7006μs	34.8425 KOps/s	35.0077 KOps/s	$\color{#d91a1a}-0.47\%$
test_setitem_dim[slice_int]	92.5210μs	52.2875μs	19.1250 KOps/s	18.8562 KOps/s	$\color{#35bf28}+1.43\%$
test_setitem_dim[range]	0.1487ms	73.1579μs	13.6691 KOps/s	13.6812 KOps/s	$\color{#d91a1a}-0.09\%$
test_setitem_dim[tuple]	84.6870μs	41.9583μs	23.8332 KOps/s	23.8323 KOps/s	$+0.00\%$
test_setitem	78.8260μs	18.4254μs	54.2729 KOps/s	54.1930 KOps/s	$\color{#35bf28}+0.15\%$
test_set	79.7080μs	17.5207μs	57.0754 KOps/s	55.5369 KOps/s	$\color{#35bf28}+2.77\%$
test_set_shared	1.6807ms	0.1411ms	7.0867 KOps/s	7.0358 KOps/s	$\color{#35bf28}+0.72\%$
test_update	0.1111ms	19.4289μs	51.4698 KOps/s	53.3670 KOps/s	$\color{#d91a1a}-3.56\%$
test_update_nested	89.2860μs	26.2988μs	38.0245 KOps/s	37.5058 KOps/s	$\color{#35bf28}+1.38\%$
test_set_nested	89.8680μs	19.3517μs	51.6751 KOps/s	51.2719 KOps/s	$\color{#35bf28}+0.79\%$
test_set_nested_new	78.2360μs	24.4348μs	40.9252 KOps/s	39.0178 KOps/s	$\color{#35bf28}+4.89\%$
test_select	0.1050ms	49.8648μs	20.0542 KOps/s	19.7768 KOps/s	$\color{#35bf28}+1.40\%$
test_unbind_speed	0.4551ms	0.3760ms	2.6597 KOps/s	2.6768 KOps/s	$\color{#d91a1a}-0.64\%$
test_unbind_speed_stack0	63.5713ms	4.7261ms	211.5898 Ops/s	224.1267 Ops/s	$\textbf{\color{#d91a1a}-5.59\%}$
test_unbind_speed_stack1	1.6471μs	0.6518μs	1.5342 MOps/s	1.5732 MOps/s	$\color{#d91a1a}-2.48\%$
test_split	54.9724ms	1.7616ms	567.6578 Ops/s	567.6658 Ops/s	$-0.00\%$
test_chunk	52.6340ms	1.7449ms	573.0871 Ops/s	549.7865 Ops/s	$\color{#35bf28}+4.24\%$
test_creation[device0]	0.6078ms	0.2940ms	3.4013 KOps/s	3.3334 KOps/s	$\color{#35bf28}+2.04\%$
test_creation_from_tensor	2.6767ms	0.3320ms	3.0121 KOps/s	3.0086 KOps/s	$\color{#35bf28}+0.11\%$
test_add_one[memmap_tensor0]	93.8840μs	25.3146μs	39.5029 KOps/s	39.4299 KOps/s	$\color{#35bf28}+0.19\%$
test_contiguous[memmap_tensor0]	30.4970μs	5.7286μs	174.5633 KOps/s	173.2443 KOps/s	$\color{#35bf28}+0.76\%$
test_stack[memmap_tensor0]	87.5230μs	19.5318μs	51.1985 KOps/s	53.9055 KOps/s	$\textbf{\color{#d91a1a}-5.02\%}$
test_memmaptd_index	1.2355ms	0.2251ms	4.4426 KOps/s	2.4373 KOps/s	$\textbf{\color{#35bf28}+82.27\%}$
test_memmaptd_index_astensor	0.3215ms	0.2573ms	3.8860 KOps/s	2.1348 KOps/s	$\textbf{\color{#35bf28}+82.03\%}$
test_memmaptd_index_op	0.6072ms	0.4956ms	2.0176 KOps/s	1.4201 KOps/s	$\textbf{\color{#35bf28}+42.07\%}$
test_reshape_pytree	55.8040μs	23.4572μs	42.6308 KOps/s	43.2654 KOps/s	$\color{#d91a1a}-1.47\%$
test_reshape_td	0.3969ms	31.6100μs	31.6355 KOps/s	30.6044 KOps/s	$\color{#35bf28}+3.37\%$
test_view_pytree	58.5290μs	23.4583μs	42.6289 KOps/s	43.0650 KOps/s	$\color{#d91a1a}-1.01\%$
test_view_td	20.3280μs	4.8598μs	205.7678 KOps/s	207.6055 KOps/s	$\color{#d91a1a}-0.89\%$
test_unbind_pytree	76.8330μs	26.8161μs	37.2911 KOps/s	38.1050 KOps/s	$\color{#d91a1a}-2.14\%$
test_unbind_td	0.1258ms	59.3170μs	16.8586 KOps/s	15.2398 KOps/s	$\textbf{\color{#35bf28}+10.62\%}$
test_split_pytree	55.5440μs	26.7533μs	37.3785 KOps/s	38.4671 KOps/s	$\color{#d91a1a}-2.83\%$
test_split_td	88.8260μs	46.3286μs	21.5849 KOps/s	20.9907 KOps/s	$\color{#35bf28}+2.83\%$
test_add_pytree	73.7570μs	32.2100μs	31.0462 KOps/s	27.3006 KOps/s	$\textbf{\color{#35bf28}+13.72\%}$
test_add_td	0.1308ms	45.4094μs	22.0219 KOps/s	21.8670 KOps/s	$\color{#35bf28}+0.71\%$
test_distributed	49.1220μs	5.9400μs	168.3509 KOps/s	167.9495 KOps/s	$\color{#35bf28}+0.24\%$
test_tdmodule	0.1728ms	20.8578μs	47.9438 KOps/s	44.3328 KOps/s	$\textbf{\color{#35bf28}+8.15\%}$
test_tdmodule_dispatch	0.1682ms	37.7168μs	26.5134 KOps/s	25.3878 KOps/s	$\color{#35bf28}+4.43\%$
test_tdseq	50.9540μs	23.7263μs	42.1474 KOps/s	40.6494 KOps/s	$\color{#35bf28}+3.69\%$
test_tdseq_dispatch	0.1372ms	42.6089μs	23.4693 KOps/s	22.9931 KOps/s	$\color{#35bf28}+2.07\%$
test_instantiation_functorch	1.4130ms	1.2995ms	769.5415 Ops/s	778.5909 Ops/s	$\color{#d91a1a}-1.16\%$
test_instantiation_td	1.6170ms	1.0298ms	971.0264 Ops/s	923.1350 Ops/s	$\textbf{\color{#35bf28}+5.19\%}$
test_exec_functorch	0.3696ms	0.1589ms	6.2947 KOps/s	6.1495 KOps/s	$\color{#35bf28}+2.36\%$
test_exec_functional_call	0.4108ms	0.1495ms	6.6896 KOps/s	6.6385 KOps/s	$\color{#35bf28}+0.77\%$
test_exec_td	0.3274ms	0.1435ms	6.9695 KOps/s	6.8402 KOps/s	$\color{#35bf28}+1.89\%$
test_exec_td_decorator	0.9495ms	0.1772ms	5.6420 KOps/s	5.4875 KOps/s	$\color{#35bf28}+2.82\%$
test_vmap_mlp_speed[True-True]	1.4163ms	0.8990ms	1.1124 KOps/s	1.1181 KOps/s	$\color{#d91a1a}-0.51\%$
test_vmap_mlp_speed[True-False]	0.9511ms	0.4808ms	2.0799 KOps/s	2.1178 KOps/s	$\color{#d91a1a}-1.79\%$
test_vmap_mlp_speed[False-True]	1.1643ms	0.7811ms	1.2802 KOps/s	1.2923 KOps/s	$\color{#d91a1a}-0.94\%$
test_vmap_mlp_speed[False-False]	0.5431ms	0.3814ms	2.6222 KOps/s	2.6003 KOps/s	$\color{#35bf28}+0.84\%$
test_vmap_mlp_speed_decorator[True-True]	2.6583ms	1.7782ms	562.3693 Ops/s	565.0756 Ops/s	$\color{#d91a1a}-0.48\%$
test_vmap_mlp_speed_decorator[True-False]	0.9007ms	0.5122ms	1.9523 KOps/s	1.9405 KOps/s	$\color{#35bf28}+0.61\%$
test_vmap_mlp_speed_decorator[False-True]	2.0033ms	1.4830ms	674.3256 Ops/s	681.4433 Ops/s	$\color{#d91a1a}-1.04\%$
test_vmap_mlp_speed_decorator[False-False]	0.7601ms	0.3961ms	2.5247 KOps/s	2.5088 KOps/s	$\color{#35bf28}+0.64\%$

[github-actions]

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

Total Benchmarks: 127. Improved: $\large\color{#35bf28}12$. Worsened: $\large\color{#d91a1a}2$.

Expand to view detailed results

Name	Max	Mean	Ops	Ops on Repo HEAD	Change
test_plain_set_nested	0.4615ms	12.6032μs	79.3450 KOps/s	78.3180 KOps/s	$\color{#35bf28}+1.31\%$
test_plain_set_stack_nested	0.1967ms	0.1146ms	8.7284 KOps/s	8.6301 KOps/s	$\color{#35bf28}+1.14\%$
test_plain_set_nested_inplace	33.4710μs	14.8612μs	67.2893 KOps/s	66.2095 KOps/s	$\color{#35bf28}+1.63\%$
test_plain_set_stack_nested_inplace	0.1805ms	0.1399ms	7.1486 KOps/s	7.0619 KOps/s	$\color{#35bf28}+1.23\%$
test_items	28.8110μs	4.6702μs	214.1250 KOps/s	212.5502 KOps/s	$\color{#35bf28}+0.74\%$
test_items_nested	0.3774ms	0.3381ms	2.9574 KOps/s	2.9861 KOps/s	$\color{#d91a1a}-0.96\%$
test_items_nested_locked	0.4287ms	0.3426ms	2.9185 KOps/s	2.9522 KOps/s	$\color{#d91a1a}-1.14\%$
test_items_nested_leaf	0.2438ms	0.1991ms	5.0218 KOps/s	5.0531 KOps/s	$\color{#d91a1a}-0.62\%$
test_items_stack_nested	1.5435ms	1.4793ms	675.9844 Ops/s	668.7865 Ops/s	$\color{#35bf28}+1.08\%$
test_items_stack_nested_leaf	1.3606ms	1.3065ms	765.4036 Ops/s	759.3990 Ops/s	$\color{#35bf28}+0.79\%$
test_items_stack_nested_locked	0.8497ms	0.8129ms	1.2302 KOps/s	1.2095 KOps/s	$\color{#35bf28}+1.71\%$
test_keys	41.3110μs	4.5990μs	217.4367 KOps/s	215.9634 KOps/s	$\color{#35bf28}+0.68\%$
test_keys_nested	3.2786ms	90.7523μs	11.0190 KOps/s	11.0600 KOps/s	$\color{#d91a1a}-0.37\%$
test_keys_nested_locked	0.1143ms	90.2784μs	11.0768 KOps/s	11.0639 KOps/s	$\color{#35bf28}+0.12\%$
test_keys_nested_leaf	41.3409ms	86.7979μs	11.5210 KOps/s	12.1759 KOps/s	$\textbf{\color{#d91a1a}-5.38\%}$
test_keys_stack_nested	1.5592ms	1.3013ms	768.4596 Ops/s	769.2222 Ops/s	$\color{#d91a1a}-0.10\%$
test_keys_stack_nested_leaf	1.3658ms	1.2875ms	776.7148 Ops/s	767.7453 Ops/s	$\color{#35bf28}+1.17\%$
test_keys_stack_nested_locked	0.6650ms	0.6234ms	1.6040 KOps/s	1.5779 KOps/s	$\color{#35bf28}+1.65\%$
test_values	14.9237μs	1.8928μs	528.3141 KOps/s	526.2636 KOps/s	$\color{#35bf28}+0.39\%$
test_values_nested	68.0830μs	43.1929μs	23.1519 KOps/s	23.0592 KOps/s	$\color{#35bf28}+0.40\%$
test_values_nested_locked	0.1073ms	45.3581μs	22.0468 KOps/s	21.8343 KOps/s	$\color{#35bf28}+0.97\%$
test_values_nested_leaf	58.0430μs	37.2890μs	26.8176 KOps/s	26.4450 KOps/s	$\color{#35bf28}+1.41\%$
test_values_stack_nested	1.2051ms	1.1445ms	873.7288 Ops/s	866.9907 Ops/s	$\color{#35bf28}+0.78\%$
test_values_stack_nested_leaf	1.1836ms	1.1219ms	891.3454 Ops/s	883.8879 Ops/s	$\color{#35bf28}+0.84\%$
test_values_stack_nested_locked	0.5468ms	0.4992ms	2.0032 KOps/s	1.9616 KOps/s	$\color{#35bf28}+2.12\%$
test_membership	6.1744μs	0.9373μs	1.0669 MOps/s	948.6238 KOps/s	$\textbf{\color{#35bf28}+12.47\%}$
test_membership_nested	18.6710μs	2.1887μs	456.8875 KOps/s	456.7165 KOps/s	$\color{#35bf28}+0.04\%$
test_membership_nested_leaf	14.0505μs	2.1169μs	472.3892 KOps/s	473.1406 KOps/s	$\color{#d91a1a}-0.16\%$
test_membership_stacked_nested	40.7510μs	11.0455μs	90.5344 KOps/s	91.2699 KOps/s	$\color{#d91a1a}-0.81\%$
test_membership_stacked_nested_leaf	73.6830μs	11.0465μs	90.5268 KOps/s	90.8871 KOps/s	$\color{#d91a1a}-0.40\%$
test_membership_nested_last	35.8820μs	4.6606μs	214.5650 KOps/s	218.4917 KOps/s	$\color{#d91a1a}-1.80\%$
test_membership_nested_leaf_last	22.7810μs	4.6893μs	213.2503 KOps/s	218.3642 KOps/s	$\color{#d91a1a}-2.34\%$
test_membership_stacked_nested_last	0.1866ms	0.1342ms	7.4530 KOps/s	7.3967 KOps/s	$\color{#35bf28}+0.76\%$
test_membership_stacked_nested_leaf_last	51.6820μs	12.7416μs	78.4830 KOps/s	78.2578 KOps/s	$\color{#35bf28}+0.29\%$
test_nested_getleaf	29.2310μs	8.3546μs	119.6950 KOps/s	118.8351 KOps/s	$\color{#35bf28}+0.72\%$
test_nested_get	28.6510μs	7.9278μs	126.1391 KOps/s	125.9206 KOps/s	$\color{#35bf28}+0.17\%$
test_stacked_getleaf	0.6987ms	0.5690ms	1.7576 KOps/s	1.7796 KOps/s	$\color{#d91a1a}-1.24\%$
test_stacked_get	0.6061ms	0.5440ms	1.8384 KOps/s	1.8715 KOps/s	$\color{#d91a1a}-1.77\%$
test_nested_getitemleaf	27.7610μs	8.5018μs	117.6219 KOps/s	118.0450 KOps/s	$\color{#d91a1a}-0.36\%$
test_nested_getitem	31.2610μs	8.0607μs	124.0594 KOps/s	124.8637 KOps/s	$\color{#d91a1a}-0.64\%$
test_stacked_getitemleaf	0.8116ms	0.5668ms	1.7643 KOps/s	1.7569 KOps/s	$\color{#35bf28}+0.42\%$
test_stacked_getitem	0.5586ms	0.5328ms	1.8768 KOps/s	1.8543 KOps/s	$\color{#35bf28}+1.21\%$
test_lock_nested	3.2113ms	0.5539ms	1.8053 KOps/s	1.7612 KOps/s	$\color{#35bf28}+2.51\%$
test_lock_stack_nested	82.9370ms	7.2093ms	138.7105 Ops/s	137.9740 Ops/s	$\color{#35bf28}+0.53\%$
test_unlock_nested	2.3688ms	0.4265ms	2.3444 KOps/s	2.3029 KOps/s	$\color{#35bf28}+1.80\%$
test_unlock_stack_nested	66.5467ms	6.2426ms	160.1906 Ops/s	158.0031 Ops/s	$\color{#35bf28}+1.38\%$
test_flatten_speed	0.2246ms	0.1861ms	5.3732 KOps/s	5.3233 KOps/s	$\color{#35bf28}+0.94\%$
test_unflatten_speed	0.4375ms	0.3647ms	2.7420 KOps/s	2.7481 KOps/s	$\color{#d91a1a}-0.22\%$
test_common_ops	1.1255ms	0.5933ms	1.6856 KOps/s	1.6475 KOps/s	$\color{#35bf28}+2.31\%$
test_creation	64.1220μs	2.1034μs	475.4099 KOps/s	474.4200 KOps/s	$\color{#35bf28}+0.21\%$
test_creation_empty	27.5710μs	6.7310μs	148.5669 KOps/s	138.8676 KOps/s	$\textbf{\color{#35bf28}+6.98\%}$
test_creation_nested_1	42.2310μs	9.1010μs	109.8778 KOps/s	104.8260 KOps/s	$\color{#35bf28}+4.82\%$
test_creation_nested_2	41.1420μs	11.8089μs	84.6816 KOps/s	82.0787 KOps/s	$\color{#35bf28}+3.17\%$
test_clone	97.3440μs	14.2661μs	70.0965 KOps/s	68.7352 KOps/s	$\color{#35bf28}+1.98\%$
test_getitem[int]	30.3310μs	12.1694μs	82.1736 KOps/s	81.2715 KOps/s	$\color{#35bf28}+1.11\%$
test_getitem[slice_int]	50.9020μs	23.7061μs	42.1833 KOps/s	42.0469 KOps/s	$\color{#35bf28}+0.32\%$
test_getitem[range]	0.2405ms	40.1628μs	24.8986 KOps/s	24.8468 KOps/s	$\color{#35bf28}+0.21\%$
test_getitem[tuple]	40.4810μs	20.0523μs	49.8696 KOps/s	48.8674 KOps/s	$\color{#35bf28}+2.05\%$
test_getitem[list]	0.2554ms	36.5407μs	27.3668 KOps/s	26.4924 KOps/s	$\color{#35bf28}+3.30\%$
test_setitem_dim[int]	56.7230μs	25.3252μs	39.4863 KOps/s	37.9526 KOps/s	$\color{#35bf28}+4.04\%$
test_setitem_dim[slice_int]	61.6930μs	45.3231μs	22.0638 KOps/s	21.4143 KOps/s	$\color{#35bf28}+3.03\%$
test_setitem_dim[range]	97.3540μs	62.7407μs	15.9386 KOps/s	15.7419 KOps/s	$\color{#35bf28}+1.25\%$
test_setitem_dim[tuple]	59.4820μs	38.9700μs	25.6608 KOps/s	25.7412 KOps/s	$\color{#d91a1a}-0.31\%$
test_setitem	94.3340μs	17.9339μs	55.7603 KOps/s	53.9342 KOps/s	$\color{#35bf28}+3.39\%$
test_set	88.2050μs	17.4950μs	57.1593 KOps/s	56.3603 KOps/s	$\color{#35bf28}+1.42\%$
test_set_shared	2.8966ms	0.1047ms	9.5485 KOps/s	8.6112 KOps/s	$\textbf{\color{#35bf28}+10.88\%}$
test_update	94.9040μs	18.4243μs	54.2762 KOps/s	52.2170 KOps/s	$\color{#35bf28}+3.94\%$
test_update_nested	0.1077ms	25.2293μs	39.6364 KOps/s	38.9236 KOps/s	$\color{#35bf28}+1.83\%$
test_set_nested	98.4540μs	18.7529μs	53.3250 KOps/s	52.2182 KOps/s	$\color{#35bf28}+2.12\%$
test_set_nested_new	96.3240μs	22.9429μs	43.5864 KOps/s	42.5748 KOps/s	$\color{#35bf28}+2.38\%$
test_select	73.8830μs	46.3572μs	21.5716 KOps/s	21.2255 KOps/s	$\color{#35bf28}+1.63\%$
test_to	74.6230μs	54.8299μs	18.2382 KOps/s	18.2716 KOps/s	$\color{#d91a1a}-0.18\%$
test_to_nonblocking	68.5530μs	34.8684μs	28.6793 KOps/s	25.7870 KOps/s	$\textbf{\color{#35bf28}+11.22\%}$
test_unbind_speed	0.4077ms	0.3618ms	2.7636 KOps/s	2.7548 KOps/s	$\color{#35bf28}+0.32\%$
test_unbind_speed_stack0	63.2311ms	4.3730ms	228.6754 Ops/s	233.1590 Ops/s	$\color{#d91a1a}-1.92\%$
test_unbind_speed_stack1	1.6431μs	0.5259μs	1.9016 MOps/s	1.8929 MOps/s	$\color{#35bf28}+0.46\%$
test_split	54.0366ms	1.8082ms	553.0379 Ops/s	541.5858 Ops/s	$\color{#35bf28}+2.11\%$
test_chunk	54.3971ms	1.7994ms	555.7271 Ops/s	544.9208 Ops/s	$\color{#35bf28}+1.98\%$
test_creation[device0]	0.5046ms	0.3092ms	3.2346 KOps/s	3.2283 KOps/s	$\color{#35bf28}+0.20\%$
test_creation[device1]	0.8266ms	0.3145ms	3.1797 KOps/s	3.1659 KOps/s	$\color{#35bf28}+0.44\%$
test_creation_from_tensor	57.3963ms	0.3675ms	2.7208 KOps/s	2.9052 KOps/s	$\textbf{\color{#d91a1a}-6.35\%}$
test_add_one[memmap_tensor0]	91.0340μs	23.9931μs	41.6787 KOps/s	40.5487 KOps/s	$\color{#35bf28}+2.79\%$
test_add_one[memmap_tensor1]	0.2175ms	72.9038μs	13.7167 KOps/s	13.3934 KOps/s	$\color{#35bf28}+2.41\%$
test_contiguous[memmap_tensor0]	20.4710μs	5.7952μs	172.5577 KOps/s	172.1464 KOps/s	$\color{#35bf28}+0.24\%$
test_contiguous[memmap_tensor1]	61.6230μs	21.9253μs	45.6094 KOps/s	45.4119 KOps/s	$\color{#35bf28}+0.43\%$
test_stack[memmap_tensor0]	88.9550μs	19.2375μs	51.9817 KOps/s	50.8372 KOps/s	$\color{#35bf28}+2.25\%$
test_stack[memmap_tensor1]	0.1515ms	73.3245μs	13.6380 KOps/s	13.7088 KOps/s	$\color{#d91a1a}-0.52\%$
test_memmaptd_index	0.2721ms	0.2350ms	4.2548 KOps/s	2.2685 KOps/s	$\textbf{\color{#35bf28}+87.56\%}$
test_memmaptd_index_astensor	0.3734ms	0.2938ms	3.4039 KOps/s	2.0154 KOps/s	$\textbf{\color{#35bf28}+68.90\%}$
test_memmaptd_index_op	0.5966ms	0.5452ms	1.8341 KOps/s	1.3118 KOps/s	$\textbf{\color{#35bf28}+39.82\%}$
test_reshape_pytree	39.7710μs	20.6530μs	48.4192 KOps/s	47.2713 KOps/s	$\color{#35bf28}+2.43\%$
test_reshape_td	59.2020μs	30.6095μs	32.6696 KOps/s	31.9493 KOps/s	$\color{#35bf28}+2.25\%$
test_view_pytree	36.4320μs	20.6473μs	48.4324 KOps/s	47.8569 KOps/s	$\color{#35bf28}+1.20\%$
test_view_td	17.7210μs	4.0879μs	244.6240 KOps/s	243.0415 KOps/s	$\color{#35bf28}+0.65\%$
test_unbind_pytree	55.5120μs	25.8276μs	38.7183 KOps/s	37.6220 KOps/s	$\color{#35bf28}+2.91\%$
test_unbind_td	96.3440μs	56.0317μs	17.8470 KOps/s	17.4211 KOps/s	$\color{#35bf28}+2.45\%$
test_split_pytree	52.9920μs	23.8551μs	41.9197 KOps/s	41.3863 KOps/s	$\color{#35bf28}+1.29\%$
test_split_td	71.4130μs	45.0372μs	22.2039 KOps/s	22.0811 KOps/s	$\color{#35bf28}+0.56\%$
test_add_pytree	68.3730μs	31.7818μs	31.4646 KOps/s	30.9881 KOps/s	$\color{#35bf28}+1.54\%$
test_add_td	73.6630μs	43.8924μs	22.7830 KOps/s	21.1781 KOps/s	$\textbf{\color{#35bf28}+7.58\%}$
test_distributed	18.1410μs	5.7663μs	173.4202 KOps/s	180.5504 KOps/s	$\color{#d91a1a}-3.95\%$
test_tdmodule	32.5110μs	16.7638μs	59.6523 KOps/s	58.6250 KOps/s	$\color{#35bf28}+1.75\%$
test_tdmodule_dispatch	0.1200ms	32.8489μs	30.4424 KOps/s	30.0855 KOps/s	$\color{#35bf28}+1.19\%$
test_tdseq	35.8810μs	19.6918μs	50.7826 KOps/s	48.9194 KOps/s	$\color{#35bf28}+3.81\%$
test_tdseq_dispatch	50.9820μs	35.6687μs	28.0358 KOps/s	27.6472 KOps/s	$\color{#35bf28}+1.41\%$
test_instantiation_functorch	1.7997ms	1.6697ms	598.9247 Ops/s	597.3339 Ops/s	$\color{#35bf28}+0.27\%$
test_instantiation_td	1.6844ms	1.1908ms	839.7842 Ops/s	843.6504 Ops/s	$\color{#d91a1a}-0.46\%$
test_exec_functorch	0.2141ms	0.1600ms	6.2499 KOps/s	6.1964 KOps/s	$\color{#35bf28}+0.86\%$
test_exec_functional_call	0.2113ms	0.1618ms	6.1791 KOps/s	6.3270 KOps/s	$\color{#d91a1a}-2.34\%$
test_exec_td	0.2088ms	0.1516ms	6.5962 KOps/s	6.7300 KOps/s	$\color{#d91a1a}-1.99\%$
test_exec_td_decorator	0.7449ms	0.1922ms	5.2030 KOps/s	5.3099 KOps/s	$\color{#d91a1a}-2.01\%$
test_vmap_mlp_speed[True-True]	1.1842ms	1.0811ms	925.0141 Ops/s	881.7027 Ops/s	$\color{#35bf28}+4.91\%$
test_vmap_mlp_speed[True-False]	0.6555ms	0.6194ms	1.6145 KOps/s	1.5536 KOps/s	$\color{#35bf28}+3.92\%$
test_vmap_mlp_speed[False-True]	1.0816ms	0.9977ms	1.0023 KOps/s	970.9010 Ops/s	$\color{#35bf28}+3.24\%$
test_vmap_mlp_speed[False-False]	0.6074ms	0.5470ms	1.8280 KOps/s	1.7407 KOps/s	$\textbf{\color{#35bf28}+5.02\%}$
test_vmap_mlp_speed_decorator[True-True]	3.0028ms	2.0612ms	485.1652 Ops/s	467.5803 Ops/s	$\color{#35bf28}+3.76\%$
test_vmap_mlp_speed_decorator[True-False]	1.1849ms	0.6629ms	1.5085 KOps/s	1.4614 KOps/s	$\color{#35bf28}+3.22\%$
test_vmap_mlp_speed_decorator[False-True]	2.2447ms	1.7828ms	560.9020 Ops/s	531.9862 Ops/s	$\textbf{\color{#35bf28}+5.44\%}$
test_vmap_mlp_speed_decorator[False-False]	0.9809ms	0.5640ms	1.7731 KOps/s	1.6935 KOps/s	$\color{#35bf28}+4.70\%$
test_vmap_transformer_speed[True-True]	12.8824ms	12.7318ms	78.5434 Ops/s	76.7213 Ops/s	$\color{#35bf28}+2.37\%$
test_vmap_transformer_speed[True-False]	8.3775ms	8.2961ms	120.5385 Ops/s	118.3804 Ops/s	$\color{#35bf28}+1.82\%$
test_vmap_transformer_speed[False-True]	12.7165ms	12.6251ms	79.2075 Ops/s	77.4830 Ops/s	$\color{#35bf28}+2.23\%$
test_vmap_transformer_speed[False-False]	8.2684ms	8.1915ms	122.0779 Ops/s	119.0515 Ops/s	$\color{#35bf28}+2.54\%$
test_vmap_transformer_speed_decorator[True-True]	66.1297ms	65.0234ms	15.3791 Ops/s	14.0207 Ops/s	$\textbf{\color{#35bf28}+9.69\%}$
test_vmap_transformer_speed_decorator[True-False]	22.0925ms	20.0479ms	49.8805 Ops/s	48.9901 Ops/s	$\color{#35bf28}+1.82\%$
test_vmap_transformer_speed_decorator[False-True]	0.1380s	63.3128ms	15.7946 Ops/s	16.5634 Ops/s	$\color{#d91a1a}-4.64\%$
test_vmap_transformer_speed_decorator[False-False]	21.6904ms	19.6205ms	50.9670 Ops/s	46.2853 Ops/s	$\textbf{\color{#35bf28}+10.11\%}$