[BugFix] auto-batch-size in dispatch

[vmoens]

Stack from ghstack (oldest at bottom):

• -> [BugFix] auto-batch-size in dispatch #1109
• [Feature] Better list casting in TensorDict.from_any #1108
• [Feature] NonTensorStack.from_list #1107
• [Feature] from_any with UserDict #1106

[github-actions]

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

Total Benchmarks: 229. Improved: $\large\color{#35bf28}31$. Worsened: $\large\color{#d91a1a}8$.

Expand to view detailed results

Name	Max	Mean	Ops	Ops on Repo HEAD	Change
test_plain_set_nested	41.6520μs	10.3883μs	96.2626 KOps/s	92.1210 KOps/s	$\color{#35bf28}+4.50\%$
test_plain_set_stack_nested	34.8920μs	10.4739μs	95.4757 KOps/s	92.3599 KOps/s	$\color{#35bf28}+3.37\%$
test_plain_set_nested_inplace	41.9120μs	11.3885μs	87.8083 KOps/s	85.3981 KOps/s	$\color{#35bf28}+2.82\%$
test_plain_set_stack_nested_inplace	37.3220μs	11.3691μs	87.9577 KOps/s	85.5795 KOps/s	$\color{#35bf28}+2.78\%$
test_items	37.2520μs	2.9160μs	342.9370 KOps/s	342.0708 KOps/s	$\color{#35bf28}+0.25\%$
test_items_nested	0.4079ms	0.3216ms	3.1091 KOps/s	3.0974 KOps/s	$\color{#35bf28}+0.38\%$
test_items_nested_locked	0.3676ms	0.3248ms	3.0791 KOps/s	3.0754 KOps/s	$\color{#35bf28}+0.12\%$
test_items_nested_leaf	0.1007ms	58.2977μs	17.1533 KOps/s	17.0492 KOps/s	$\color{#35bf28}+0.61\%$
test_items_stack_nested	0.3784ms	0.3244ms	3.0824 KOps/s	3.0986 KOps/s	$\color{#d91a1a}-0.52\%$
test_items_stack_nested_leaf	88.4250μs	60.0635μs	16.6490 KOps/s	16.7141 KOps/s	$\color{#d91a1a}-0.39\%$
test_items_stack_nested_locked	0.3838ms	0.3231ms	3.0948 KOps/s	3.1009 KOps/s	$\color{#d91a1a}-0.20\%$
test_keys	35.3820μs	3.5048μs	285.3225 KOps/s	288.5344 KOps/s	$\color{#d91a1a}-1.11\%$
test_keys_nested	0.1080ms	70.7479μs	14.1347 KOps/s	14.1423 KOps/s	$\color{#d91a1a}-0.05\%$
test_keys_nested_locked	0.7852ms	75.7746μs	13.1970 KOps/s	13.0538 KOps/s	$\color{#35bf28}+1.10\%$
test_keys_nested_leaf	0.1028ms	61.7607μs	16.1915 KOps/s	16.2281 KOps/s	$\color{#d91a1a}-0.23\%$
test_keys_stack_nested	0.1151ms	71.0634μs	14.0719 KOps/s	14.1995 KOps/s	$\color{#d91a1a}-0.90\%$
test_keys_stack_nested_leaf	94.9050μs	62.7813μs	15.9283 KOps/s	16.1101 KOps/s	$\color{#d91a1a}-1.13\%$
test_keys_stack_nested_locked	0.1150ms	76.9137μs	13.0016 KOps/s	13.1017 KOps/s	$\color{#d91a1a}-0.76\%$
test_values	4.9185μs	0.8468μs	1.1809 MOps/s	1.1793 MOps/s	$\color{#35bf28}+0.14\%$
test_values_nested	64.1130μs	31.3850μs	31.8623 KOps/s	32.0966 KOps/s	$\color{#d91a1a}-0.73\%$
test_values_nested_locked	72.4340μs	32.8884μs	30.4059 KOps/s	30.4367 KOps/s	$\color{#d91a1a}-0.10\%$
test_values_nested_leaf	91.3350μs	33.5721μs	29.7866 KOps/s	29.8642 KOps/s	$\color{#d91a1a}-0.26\%$
test_values_stack_nested	61.7340μs	31.8036μs	31.4429 KOps/s	31.5442 KOps/s	$\color{#d91a1a}-0.32\%$
test_values_stack_nested_leaf	57.1730μs	34.5041μs	28.9821 KOps/s	29.2624 KOps/s	$\color{#d91a1a}-0.96\%$
test_values_stack_nested_locked	65.8830μs	33.5546μs	29.8022 KOps/s	30.0766 KOps/s	$\color{#d91a1a}-0.91\%$
test_membership	1.9121μs	0.5078μs	1.9692 MOps/s	1.9871 MOps/s	$\color{#d91a1a}-0.90\%$
test_membership_nested	16.1310μs	1.9032μs	525.4302 KOps/s	516.5602 KOps/s	$\color{#35bf28}+1.72\%$
test_membership_nested_leaf	13.9960μs	1.9335μs	517.1840 KOps/s	506.2981 KOps/s	$\color{#35bf28}+2.15\%$
test_membership_stacked_nested	32.6420μs	2.0123μs	496.9414 KOps/s	492.0776 KOps/s	$\color{#35bf28}+0.99\%$
test_membership_stacked_nested_leaf	45.9220μs	1.9384μs	515.8920 KOps/s	497.8925 KOps/s	$\color{#35bf28}+3.62\%$
test_membership_nested_last	38.1520μs	2.8460μs	351.3699 KOps/s	347.6808 KOps/s	$\color{#35bf28}+1.06\%$
test_membership_nested_leaf_last	29.3810μs	2.8678μs	348.6979 KOps/s	346.1113 KOps/s	$\color{#35bf28}+0.75\%$
test_membership_stacked_nested_last	31.8110μs	3.3054μs	302.5343 KOps/s	207.8609 KOps/s	$\textbf{\color{#35bf28}+45.55\%}$
test_membership_stacked_nested_leaf_last	31.9810μs	3.3061μs	302.4668 KOps/s	206.5862 KOps/s	$\textbf{\color{#35bf28}+46.41\%}$
test_nested_getleaf	49.0120μs	5.9844μs	167.1009 KOps/s	168.3840 KOps/s	$\color{#d91a1a}-0.76\%$
test_nested_get	33.6120μs	5.7214μs	174.7818 KOps/s	175.5190 KOps/s	$\color{#d91a1a}-0.42\%$
test_stacked_getleaf	44.1820μs	5.9654μs	167.6327 KOps/s	165.1576 KOps/s	$\color{#35bf28}+1.50\%$
test_stacked_get	29.6110μs	5.6885μs	175.7930 KOps/s	175.4637 KOps/s	$\color{#35bf28}+0.19\%$
test_nested_getitemleaf	28.7520μs	6.0727μs	164.6714 KOps/s	162.0122 KOps/s	$\color{#35bf28}+1.64\%$
test_nested_getitem	27.5210μs	5.7815μs	172.9659 KOps/s	171.3101 KOps/s	$\color{#35bf28}+0.97\%$
test_stacked_getitemleaf	35.3620μs	6.0783μs	164.5192 KOps/s	163.8056 KOps/s	$\color{#35bf28}+0.44\%$
test_stacked_getitem	28.7420μs	5.7421μs	174.1523 KOps/s	172.3546 KOps/s	$\color{#35bf28}+1.04\%$
test_lock_nested	9.1912ms	0.3724ms	2.6851 KOps/s	2.6829 KOps/s	$\color{#35bf28}+0.08\%$
test_lock_stack_nested	0.3900ms	0.3302ms	3.0284 KOps/s	2.9759 KOps/s	$\color{#35bf28}+1.76\%$
test_unlock_nested	0.6501ms	0.3069ms	3.2579 KOps/s	3.2323 KOps/s	$\color{#35bf28}+0.79\%$
test_unlock_stack_nested	0.3097ms	0.2727ms	3.6673 KOps/s	3.6307 KOps/s	$\color{#35bf28}+1.01\%$
test_flatten_speed	97.0450μs	72.5279μs	13.7878 KOps/s	13.4329 KOps/s	$\color{#35bf28}+2.64\%$
test_unflatten_speed	0.3334ms	0.2988ms	3.3462 KOps/s	3.3661 KOps/s	$\color{#d91a1a}-0.59\%$
test_common_ops	1.6569ms	0.5680ms	1.7605 KOps/s	1.6690 KOps/s	$\textbf{\color{#35bf28}+5.48\%}$
test_creation	96.9350μs	1.4656μs	682.2943 KOps/s	699.3221 KOps/s	$\color{#d91a1a}-2.43\%$
test_creation_empty	28.4520μs	6.9333μs	144.2324 KOps/s	130.7191 KOps/s	$\textbf{\color{#35bf28}+10.34\%}$
test_creation_nested_1	45.5720μs	8.5054μs	117.5729 KOps/s	107.5781 KOps/s	$\textbf{\color{#35bf28}+9.29\%}$
test_creation_nested_2	29.9820μs	11.1179μs	89.9451 KOps/s	85.0303 KOps/s	$\textbf{\color{#35bf28}+5.78\%}$
test_clone	93.7540μs	10.3722μs	96.4118 KOps/s	92.4463 KOps/s	$\color{#35bf28}+4.29\%$
test_getitem[int]	1.7352ms	10.4963μs	95.2713 KOps/s	92.7088 KOps/s	$\color{#35bf28}+2.76\%$
test_getitem[slice_int]	0.1159ms	20.0078μs	49.9806 KOps/s	48.5400 KOps/s	$\color{#35bf28}+2.97\%$
test_getitem[range]	0.1599ms	36.5183μs	27.3835 KOps/s	26.9104 KOps/s	$\color{#35bf28}+1.76\%$
test_getitem[tuple]	0.1098ms	17.7100μs	56.4652 KOps/s	54.8308 KOps/s	$\color{#35bf28}+2.98\%$
test_getitem[list]	0.6004ms	34.3817μs	29.0852 KOps/s	30.5075 KOps/s	$\color{#d91a1a}-4.66\%$
test_setitem_dim[int]	43.2120μs	20.1826μs	49.5477 KOps/s	53.2414 KOps/s	$\textbf{\color{#d91a1a}-6.94\%}$
test_setitem_dim[slice_int]	90.2440μs	40.6043μs	24.6279 KOps/s	26.7571 KOps/s	$\textbf{\color{#d91a1a}-7.96\%}$
test_setitem_dim[range]	0.1222ms	56.7839μs	17.6106 KOps/s	18.7662 KOps/s	$\textbf{\color{#d91a1a}-6.16\%}$
test_setitem_dim[tuple]	82.0840μs	33.8517μs	29.5406 KOps/s	31.7355 KOps/s	$\textbf{\color{#d91a1a}-6.92\%}$
test_setitem	83.5640μs	15.4208μs	64.8474 KOps/s	63.9621 KOps/s	$\color{#35bf28}+1.38\%$
test_set	87.3240μs	14.8161μs	67.4941 KOps/s	65.5914 KOps/s	$\color{#35bf28}+2.90\%$
test_set_shared	1.6246ms	0.1452ms	6.8852 KOps/s	6.8037 KOps/s	$\color{#35bf28}+1.20\%$
test_update	0.3599ms	16.3002μs	61.3491 KOps/s	56.4068 KOps/s	$\textbf{\color{#35bf28}+8.76\%}$
test_update_nested	86.2350μs	21.4389μs	46.6442 KOps/s	44.3219 KOps/s	$\textbf{\color{#35bf28}+5.24\%}$
test_update__nested	0.8083ms	24.0806μs	41.5272 KOps/s	40.0244 KOps/s	$\color{#35bf28}+3.75\%$
test_set_nested	76.9540μs	15.0398μs	66.4905 KOps/s	60.9344 KOps/s	$\textbf{\color{#35bf28}+9.12\%}$
test_set_nested_new	94.4950μs	16.9898μs	58.8590 KOps/s	54.3965 KOps/s	$\textbf{\color{#35bf28}+8.20\%}$
test_select	90.0940μs	28.9199μs	34.5782 KOps/s	32.9701 KOps/s	$\color{#35bf28}+4.88\%$
test_select_nested	84.0250μs	41.6223μs	24.0256 KOps/s	23.5657 KOps/s	$\color{#35bf28}+1.95\%$
test_exclude_nested	90.1250μs	59.1752μs	16.8990 KOps/s	16.8304 KOps/s	$\color{#35bf28}+0.41\%$
test_empty[True]	0.3024ms	0.2573ms	3.8870 KOps/s	3.9096 KOps/s	$\color{#d91a1a}-0.58\%$
test_empty[False]	3.6942μs	0.7396μs	1.3520 MOps/s	1.3448 MOps/s	$\color{#35bf28}+0.54\%$
test_to	88.1450μs	54.7129μs	18.2772 KOps/s	18.5069 KOps/s	$\color{#d91a1a}-1.24\%$
test_to_nonblocking	90.9140μs	45.1148μs	22.1657 KOps/s	21.8980 KOps/s	$\color{#35bf28}+1.22\%$
test_unbind_speed	1.7570ms	0.2299ms	4.3499 KOps/s	4.2727 KOps/s	$\color{#35bf28}+1.81\%$
test_unbind_speed_stack0	0.2917ms	0.2287ms	4.3734 KOps/s	4.2744 KOps/s	$\color{#35bf28}+2.31\%$
test_unbind_speed_stack1	91.4211ms	0.6396ms	1.5634 KOps/s	1.5422 KOps/s	$\color{#35bf28}+1.37\%$
test_split	93.9186ms	1.5659ms	638.5958 Ops/s	622.3903 Ops/s	$\color{#35bf28}+2.60\%$
test_chunk	93.7139ms	1.5463ms	646.7209 Ops/s	626.2122 Ops/s	$\color{#35bf28}+3.28\%$
test_consolidate[False-None]	95.6074ms	2.8084ms	356.0797 Ops/s	349.4076 Ops/s	$\color{#35bf28}+1.91\%$
test_consolidate[default-None]	2.0265ms	1.6326ms	612.5155 Ops/s	589.7904 Ops/s	$\color{#35bf28}+3.85\%$
test_consolidate[reduce-overhead-None]	2.0633ms	1.6654ms	600.4581 Ops/s	578.4181 Ops/s	$\color{#35bf28}+3.81\%$
test_consolidate_njt[False-None]	6.8771ms	6.5066ms	153.6906 Ops/s	111.1635 Ops/s	$\textbf{\color{#35bf28}+38.26\%}$
test_to[False-False-None]	2.0730ms	1.6645ms	600.7768 Ops/s	598.4191 Ops/s	$\color{#35bf28}+0.39\%$
test_to[True-False-None]	1.6442ms	1.2540ms	797.4741 Ops/s	792.9677 Ops/s	$\color{#35bf28}+0.57\%$
test_to[within-False-None]	4.3141ms	3.9528ms	252.9852 Ops/s	249.2839 Ops/s	$\color{#35bf28}+1.48\%$
test_to[True-default-None]	5.5071ms	5.1059ms	195.8534 Ops/s	185.5636 Ops/s	$\textbf{\color{#35bf28}+5.55\%}$
test_to_njt[False-False-None]	7.3496ms	6.9370ms	144.1540 Ops/s	140.0808 Ops/s	$\color{#35bf28}+2.91\%$
test_to_njt[True-False-None]	6.3057ms	5.4943ms	182.0079 Ops/s	180.7839 Ops/s	$\color{#35bf28}+0.68\%$
test_to_njt[within-False-None]	12.6000ms	12.0881ms	82.7260 Ops/s	81.4417 Ops/s	$\color{#35bf28}+1.58\%$
test_creation[device0]	0.4940ms	78.7996μs	12.6904 KOps/s	12.6300 KOps/s	$\color{#35bf28}+0.48\%$
test_creation_from_tensor	0.5164ms	82.4832μs	12.1237 KOps/s	12.2688 KOps/s	$\color{#d91a1a}-1.18\%$
test_add_one[memmap_tensor0]	0.2666ms	6.5760μs	152.0675 KOps/s	142.3964 KOps/s	$\textbf{\color{#35bf28}+6.79\%}$
test_contiguous[memmap_tensor0]	19.3470μs	0.3920μs	2.5511 MOps/s	2.3956 MOps/s	$\textbf{\color{#35bf28}+6.49\%}$
test_stack[memmap_tensor0]	44.3220μs	4.3320μs	230.8420 KOps/s	213.0591 KOps/s	$\textbf{\color{#35bf28}+8.35\%}$
test_memmaptd_index	1.7632ms	0.2475ms	4.0411 KOps/s	3.9677 KOps/s	$\color{#35bf28}+1.85\%$
test_memmaptd_index_astensor	0.6712ms	0.3037ms	3.2930 KOps/s	3.2234 KOps/s	$\color{#35bf28}+2.16\%$
test_memmaptd_index_op	0.9739ms	0.5620ms	1.7794 KOps/s	1.6937 KOps/s	$\textbf{\color{#35bf28}+5.06\%}$
test_serialize_model	0.1308s	0.1298s	7.7022 Ops/s	7.6943 Ops/s	$\color{#35bf28}+0.10\%$
test_serialize_model_pickle	1.3468s	1.2149s	0.8231 Ops/s	0.8113 Ops/s	$\color{#35bf28}+1.46\%$
test_serialize_weights	0.1306s	0.1291s	7.7456 Ops/s	7.7356 Ops/s	$\color{#35bf28}+0.13\%$
test_serialize_weights_returnearly	0.2961s	55.4824ms	18.0237 Ops/s	15.1530 Ops/s	$\textbf{\color{#35bf28}+18.94\%}$
test_serialize_weights_pickle	1.4047s	1.2038s	0.8307 Ops/s	0.8127 Ops/s	$\color{#35bf28}+2.21\%$
test_reshape_pytree	55.6920μs	22.5719μs	44.3029 KOps/s	45.3490 KOps/s	$\color{#d91a1a}-2.31\%$
test_reshape_td	0.4052ms	27.8941μs	35.8498 KOps/s	37.5842 KOps/s	$\color{#d91a1a}-4.61\%$
test_view_pytree	68.2330μs	21.9408μs	45.5772 KOps/s	45.5755 KOps/s	$+0.00\%$
test_view_td	0.1503ms	29.5266μs	33.8677 KOps/s	32.5920 KOps/s	$\color{#35bf28}+3.91\%$
test_unbind_pytree	84.5740μs	27.6489μs	36.1678 KOps/s	35.5400 KOps/s	$\color{#35bf28}+1.77\%$
test_unbind_td	0.5854ms	34.8455μs	28.6981 KOps/s	27.5321 KOps/s	$\color{#35bf28}+4.24\%$
test_split_pytree	72.2830μs	29.5134μs	33.8829 KOps/s	33.6460 KOps/s	$\color{#35bf28}+0.70\%$
test_split_td	0.7689ms	37.7650μs	26.4796 KOps/s	25.8262 KOps/s	$\color{#35bf28}+2.53\%$
test_add_pytree	69.5640μs	34.2981μs	29.1562 KOps/s	28.6855 KOps/s	$\color{#35bf28}+1.64\%$
test_add_td	78.5340μs	44.4317μs	22.5065 KOps/s	21.2116 KOps/s	$\textbf{\color{#35bf28}+6.10\%}$
test_compile_add_one_nested[tensordict-compile]	0.2364ms	0.1185ms	8.4418 KOps/s	7.9761 KOps/s	$\textbf{\color{#35bf28}+5.84\%}$
test_compile_add_one_nested[tensordict-eager]	0.2176ms	0.1236ms	8.0881 KOps/s	8.0509 KOps/s	$\color{#35bf28}+0.46\%$
test_compile_add_one_nested[pytree-compile]	0.1728ms	95.9050μs	10.4270 KOps/s	9.9900 KOps/s	$\color{#35bf28}+4.37\%$
test_compile_add_one_nested[pytree-eager]	1.3799ms	0.1461ms	6.8462 KOps/s	6.5996 KOps/s	$\color{#35bf28}+3.74\%$
test_compile_copy_nested[tensordict-compile]	62.1330μs	23.1799μs	43.1409 KOps/s	44.7382 KOps/s	$\color{#d91a1a}-3.57\%$
test_compile_copy_nested[tensordict-eager]	0.3895ms	26.2702μs	38.0660 KOps/s	37.3399 KOps/s	$\color{#35bf28}+1.94\%$
test_compile_copy_nested[pytree-compile]	0.4444ms	64.0826μs	15.6048 KOps/s	15.1943 KOps/s	$\color{#35bf28}+2.70\%$
test_compile_copy_nested[pytree-eager]	87.0140μs	49.6570μs	20.1382 KOps/s	19.8969 KOps/s	$\color{#35bf28}+1.21\%$
test_compile_add_one_flat[tensordict-compile]	0.2285ms	0.1424ms	7.0227 KOps/s	7.0886 KOps/s	$\color{#d91a1a}-0.93\%$
test_compile_add_one_flat[tensordict-eager]	0.2981ms	0.2063ms	4.8479 KOps/s	4.8922 KOps/s	$\color{#d91a1a}-0.91\%$
test_compile_add_one_flat[tensorclass-compile]	0.2468ms	97.0813μs	10.3006 KOps/s	10.0749 KOps/s	$\color{#35bf28}+2.24\%$
test_compile_add_one_flat[tensorclass-eager]	0.1779ms	51.3157μs	19.4872 KOps/s	18.9075 KOps/s	$\color{#35bf28}+3.07\%$
test_compile_add_one_flat[pytree-compile]	0.2814ms	0.1359ms	7.3596 KOps/s	7.3744 KOps/s	$\color{#d91a1a}-0.20\%$
test_compile_add_one_flat[pytree-eager]	0.6159ms	0.4681ms	2.1364 KOps/s	1.8865 KOps/s	$\textbf{\color{#35bf28}+13.25\%}$
test_compile_add_self_flat[tensordict-eager]	0.3621ms	0.2460ms	4.0657 KOps/s	4.0701 KOps/s	$\color{#d91a1a}-0.11\%$
test_compile_add_self_flat[tensordict-compile]	0.2529ms	0.1434ms	6.9723 KOps/s	7.0714 KOps/s	$\color{#d91a1a}-1.40\%$
test_compile_add_self_flat[tensorclass-eager]	0.1444ms	60.6676μs	16.4833 KOps/s	16.3224 KOps/s	$\color{#35bf28}+0.99\%$
test_compile_add_self_flat[tensorclass-compile]	0.1592ms	98.1524μs	10.1882 KOps/s	10.3731 KOps/s	$\color{#d91a1a}-1.78\%$
test_compile_add_self_flat[pytree-eager]	0.5728ms	0.4046ms	2.4713 KOps/s	2.5209 KOps/s	$\color{#d91a1a}-1.97\%$
test_compile_add_self_flat[pytree-compile]	0.1862ms	0.1359ms	7.3570 KOps/s	7.4326 KOps/s	$\color{#d91a1a}-1.02\%$
test_compile_copy_flat[tensordict-compile]	71.5840μs	19.5771μs	51.0801 KOps/s	53.7579 KOps/s	$\color{#d91a1a}-4.98\%$
test_compile_copy_flat[tensordict-eager]	87.8340μs	27.4783μs	36.3923 KOps/s	37.7465 KOps/s	$\color{#d91a1a}-3.59\%$
test_compile_copy_flat[pytree-compile]	0.1118ms	69.6306μs	14.3615 KOps/s	14.3353 KOps/s	$\color{#35bf28}+0.18\%$
test_compile_copy_flat[pytree-eager]	0.1594ms	51.2070μs	19.5286 KOps/s	19.3939 KOps/s	$\color{#35bf28}+0.69\%$
test_compile_assign_and_add[tensordict-compile]	1.6082ms	0.3891ms	2.5701 KOps/s	2.1386 KOps/s	$\textbf{\color{#35bf28}+20.18\%}$
test_compile_assign_and_add[tensordict-eager]	2.7019ms	2.5636ms	390.0778 Ops/s	371.7801 Ops/s	$\color{#35bf28}+4.92\%$
test_compile_assign_and_add[pytree-compile]	1.5960ms	0.4345ms	2.3013 KOps/s	2.2666 KOps/s	$\color{#35bf28}+1.53\%$
test_compile_assign_and_add[pytree-eager]	2.7757ms	2.5872ms	386.5157 Ops/s	378.0587 Ops/s	$\color{#35bf28}+2.24\%$
test_compile_indexing[tensor-tensordict-compile]	0.1536ms	0.1117ms	8.9490 KOps/s	8.8309 KOps/s	$\color{#35bf28}+1.34\%$
test_compile_indexing[tensor-tensordict-eager]	0.5618ms	78.7741μs	12.6945 KOps/s	12.1089 KOps/s	$\color{#35bf28}+4.84\%$
test_compile_indexing[tensor-tensorclass-compile]	0.1930ms	0.1114ms	8.9732 KOps/s	9.4243 KOps/s	$\color{#d91a1a}-4.79\%$
test_compile_indexing[tensor-tensorclass-eager]	0.4893ms	68.5638μs	14.5849 KOps/s	14.7420 KOps/s	$\color{#d91a1a}-1.07\%$
test_compile_indexing[tensor-pytree-compile]	0.5031ms	0.1108ms	9.0219 KOps/s	9.5687 KOps/s	$\textbf{\color{#d91a1a}-5.71\%}$
test_compile_indexing[tensor-pytree-eager]	0.4556ms	69.5702μs	14.3740 KOps/s	14.9355 KOps/s	$\color{#d91a1a}-3.76\%$
test_compile_indexing[slice-tensordict-compile]	0.1362ms	0.1002ms	9.9831 KOps/s	9.6562 KOps/s	$\color{#35bf28}+3.39\%$
test_compile_indexing[slice-tensordict-eager]	0.4234ms	18.6077μs	53.7411 KOps/s	55.0598 KOps/s	$\color{#d91a1a}-2.40\%$
test_compile_indexing[slice-tensorclass-compile]	0.4921ms	98.4844μs	10.1539 KOps/s	10.1673 KOps/s	$\color{#d91a1a}-0.13\%$
test_compile_indexing[slice-tensorclass-eager]	0.4017ms	16.5786μs	60.3189 KOps/s	63.1165 KOps/s	$\color{#d91a1a}-4.43\%$
test_compile_indexing[slice-pytree-compile]	0.5028ms	98.6277μs	10.1391 KOps/s	9.9578 KOps/s	$\color{#35bf28}+1.82\%$
test_compile_indexing[slice-pytree-eager]	56.7830μs	16.3186μs	61.2798 KOps/s	62.9969 KOps/s	$\color{#d91a1a}-2.73\%$
test_compile_indexing[int-tensordict-compile]	0.5165ms	0.1050ms	9.5204 KOps/s	9.7360 KOps/s	$\color{#d91a1a}-2.21\%$
test_compile_indexing[int-tensordict-eager]	0.5792ms	18.0808μs	55.3072 KOps/s	56.8424 KOps/s	$\color{#d91a1a}-2.70\%$
test_compile_indexing[int-tensorclass-compile]	0.4953ms	0.1013ms	9.8760 KOps/s	10.0660 KOps/s	$\color{#d91a1a}-1.89\%$
test_compile_indexing[int-tensorclass-eager]	0.4223ms	16.2312μs	61.6099 KOps/s	63.7013 KOps/s	$\color{#d91a1a}-3.28\%$
test_compile_indexing[int-pytree-compile]	0.4889ms	99.5374μs	10.0465 KOps/s	9.9563 KOps/s	$\color{#35bf28}+0.91\%$
test_compile_indexing[int-pytree-eager]	0.3993ms	15.9177μs	62.8230 KOps/s	64.0035 KOps/s	$\color{#d91a1a}-1.84\%$
test_mod_add[eager]	84.6940μs	33.0938μs	30.2172 KOps/s	28.5809 KOps/s	$\textbf{\color{#35bf28}+5.72\%}$
test_mod_add[compile]	0.4833ms	79.1903μs	12.6278 KOps/s	12.3792 KOps/s	$\color{#35bf28}+2.01\%$
test_mod_add[compile-overhead]	0.3225ms	0.1653ms	6.0512 KOps/s	5.7853 KOps/s	$\color{#35bf28}+4.60\%$
test_mod_wrap[eager]	0.6415ms	0.2399ms	4.1691 KOps/s	4.0808 KOps/s	$\color{#35bf28}+2.16\%$
test_mod_wrap[compile]	0.3556ms	0.2952ms	3.3880 KOps/s	3.5094 KOps/s	$\color{#d91a1a}-3.46\%$
test_mod_wrap[compile-overhead]	7.2048ms	3.7289ms	268.1740 Ops/s	262.2447 Ops/s	$\color{#35bf28}+2.26\%$
test_mod_wrap_and_backward[eager]	1.6142ms	1.4743ms	678.2945 Ops/s	693.1764 Ops/s	$\color{#d91a1a}-2.15\%$
test_mod_wrap_and_backward[compile]	1.4305ms	1.3503ms	740.5621 Ops/s	730.9096 Ops/s	$\color{#35bf28}+1.32\%$
test_mod_wrap_and_backward[compile-overhead]	1.5262ms	1.0305ms	970.4423 Ops/s	967.0572 Ops/s	$\color{#35bf28}+0.35\%$
test_seq_add[eager]	0.2252ms	98.1831μs	10.1851 KOps/s	9.4774 KOps/s	$\textbf{\color{#35bf28}+7.47\%}$
test_seq_add[compile]	0.1499ms	91.6416μs	10.9121 KOps/s	11.3923 KOps/s	$\color{#d91a1a}-4.22\%$
test_seq_add[compile-overhead]	0.2740ms	0.1299ms	7.6958 KOps/s	7.7799 KOps/s	$\color{#d91a1a}-1.08\%$
test_seq_wrap[eager]	0.4727ms	0.3987ms	2.5080 KOps/s	2.5615 KOps/s	$\color{#d91a1a}-2.09\%$
test_seq_wrap[compile]	0.5208ms	0.3124ms	3.2010 KOps/s	3.3240 KOps/s	$\color{#d91a1a}-3.70\%$
test_seq_wrap[compile-overhead]	0.2918ms	0.2271ms	4.4032 KOps/s	4.3284 KOps/s	$\color{#35bf28}+1.73\%$
test_func_call_runtime[False-eager]	0.9157ms	0.7628ms	1.3110 KOps/s	1.2466 KOps/s	$\textbf{\color{#35bf28}+5.16\%}$
test_func_call_runtime[False-compile]	0.7749ms	0.7337ms	1.3629 KOps/s	1.3278 KOps/s	$\color{#35bf28}+2.64\%$
test_func_call_runtime[False-compile-overhead]	0.4424ms	0.3569ms	2.8018 KOps/s	2.7472 KOps/s	$\color{#35bf28}+1.99\%$
test_func_call_runtime[True-eager]	1.0446ms	0.8819ms	1.1339 KOps/s	1.1072 KOps/s	$\color{#35bf28}+2.41\%$
test_func_call_runtime[True-compile]	0.8699ms	0.7541ms	1.3261 KOps/s	1.2963 KOps/s	$\color{#35bf28}+2.30\%$
test_func_call_runtime[True-compile-overhead]	0.4558ms	0.3783ms	2.6433 KOps/s	2.6013 KOps/s	$\color{#35bf28}+1.61\%$
test_func_call_cm_runtime[False-eager]	0.8139ms	0.7199ms	1.3890 KOps/s	1.3543 KOps/s	$\color{#35bf28}+2.56\%$
test_func_call_cm_runtime[False-compile]	1.1293ms	0.7377ms	1.3555 KOps/s	1.3272 KOps/s	$\color{#35bf28}+2.13\%$
test_func_call_cm_runtime[False-compile-overhead]	0.4464ms	0.3593ms	2.7833 KOps/s	2.7424 KOps/s	$\color{#35bf28}+1.49\%$
test_func_call_cm_runtime[True-eager]	1.3994ms	0.9860ms	1.0142 KOps/s	995.2619 Ops/s	$\color{#35bf28}+1.90\%$
test_func_call_cm_runtime[True-compile]	1.1737ms	0.7849ms	1.2740 KOps/s	1.2499 KOps/s	$\color{#35bf28}+1.93\%$
test_func_call_cm_runtime[True-compile-overhead]	0.8084ms	0.4063ms	2.4610 KOps/s	2.4397 KOps/s	$\color{#35bf28}+0.87\%$
test_vmap_func_call_cm_runtime[eager]	2.5742ms	2.0606ms	485.2909 Ops/s	481.1133 Ops/s	$\color{#35bf28}+0.87\%$
test_vmap_func_call_cm_runtime[compile]	0.9474ms	0.7919ms	1.2628 KOps/s	1.1964 KOps/s	$\textbf{\color{#35bf28}+5.55\%}$
test_vmap_func_call_cm_runtime[compile-overhead]	0.4635ms	0.4072ms	2.4559 KOps/s	2.4209 KOps/s	$\color{#35bf28}+1.45\%$
test_distributed	1.7970ms	0.1995ms	5.0131 KOps/s	8.8136 KOps/s	$\textbf{\color{#d91a1a}-43.12\%}$
test_tdmodule	0.3268ms	15.3518μs	65.1391 KOps/s	62.7260 KOps/s	$\color{#35bf28}+3.85\%$
test_tdmodule_dispatch	47.6520μs	28.9899μs	34.4947 KOps/s	28.5413 KOps/s	$\textbf{\color{#35bf28}+20.86\%}$
test_tdseq	38.1720μs	15.0190μs	66.5825 KOps/s	63.6091 KOps/s	$\color{#35bf28}+4.67\%$
test_tdseq_dispatch	57.9530μs	31.0929μs	32.1617 KOps/s	30.2727 KOps/s	$\textbf{\color{#35bf28}+6.24\%}$
test_instantiation_functorch	1.6672ms	1.5566ms	642.4258 Ops/s	636.0971 Ops/s	$\color{#35bf28}+0.99\%$
test_exec_functorch	0.2133ms	0.1408ms	7.1010 KOps/s	6.7190 KOps/s	$\textbf{\color{#35bf28}+5.69\%}$
test_exec_functional_call	0.1851ms	0.1347ms	7.4218 KOps/s	6.9495 KOps/s	$\textbf{\color{#35bf28}+6.80\%}$
test_exec_td_decorator	0.3758ms	0.1812ms	5.5173 KOps/s	5.2392 KOps/s	$\textbf{\color{#35bf28}+5.31\%}$
test_vmap_mlp_speed_decorator[True-True]	0.7510ms	0.6704ms	1.4917 KOps/s	1.4819 KOps/s	$\color{#35bf28}+0.66\%$
test_vmap_mlp_speed_decorator[True-False]	0.8746ms	0.6739ms	1.4838 KOps/s	1.4750 KOps/s	$\color{#35bf28}+0.60\%$
test_vmap_mlp_speed_decorator[False-True]	0.7574ms	0.5900ms	1.6950 KOps/s	1.6869 KOps/s	$\color{#35bf28}+0.48\%$
test_vmap_mlp_speed_decorator[False-False]	0.7307ms	0.5928ms	1.6869 KOps/s	1.6935 KOps/s	$\color{#d91a1a}-0.39\%$
test_vmap_transformer_speed_decorator[True-True]	19.0358ms	18.9365ms	52.8081 Ops/s	52.4536 Ops/s	$\color{#35bf28}+0.68\%$
test_vmap_transformer_speed_decorator[True-False]	19.6687ms	19.0745ms	52.4259 Ops/s	52.3682 Ops/s	$\color{#35bf28}+0.11\%$
test_vmap_transformer_speed_decorator[False-True]	18.9837ms	18.8532ms	53.0414 Ops/s	52.8819 Ops/s	$\color{#35bf28}+0.30\%$
test_vmap_transformer_speed_decorator[False-False]	19.4567ms	18.8810ms	52.9633 Ops/s	52.7297 Ops/s	$\color{#35bf28}+0.44\%$
test_to_module_speed[True]	1.0358ms	0.9376ms	1.0665 KOps/s	1.0659 KOps/s	$\color{#35bf28}+0.06\%$
test_to_module_speed[False]	1.3907ms	0.9184ms	1.0888 KOps/s	1.0852 KOps/s	$\color{#35bf28}+0.33\%$
test_tc_init	72.5940μs	35.3663μs	28.2755 KOps/s	27.7717 KOps/s	$\color{#35bf28}+1.81\%$
test_tc_init_nested	0.1116ms	74.3781μs	13.4448 KOps/s	13.5339 KOps/s	$\color{#d91a1a}-0.66\%$
test_tc_first_layer_tensor	6.5161μs	0.7027μs	1.4230 MOps/s	1.4228 MOps/s	$\color{#35bf28}+0.01\%$
test_tc_first_layer_nontensor	26.3810μs	2.3226μs	430.5588 KOps/s	430.3570 KOps/s	$\color{#35bf28}+0.05\%$
test_tc_second_layer_tensor	8.7580μs	1.4170μs	705.7208 KOps/s	688.2588 KOps/s	$\color{#35bf28}+2.54\%$
test_tc_second_layer_nontensor	27.2210μs	3.0458μs	328.3177 KOps/s	327.0112 KOps/s	$\color{#35bf28}+0.40\%$
test_unbind	0.2225s	9.8275ms	101.7550 Ops/s	151.3914 Ops/s	$\textbf{\color{#d91a1a}-32.79\%}$
test_full_like	10.1885ms	9.1851ms	108.8717 Ops/s	108.9215 Ops/s	$\color{#d91a1a}-0.05\%$
test_zeros_like	9.1577ms	7.1231ms	140.3875 Ops/s	115.4698 Ops/s	$\textbf{\color{#35bf28}+21.58\%}$
test_ones_like	5.2000ms	4.3285ms	231.0251 Ops/s	231.9148 Ops/s	$\color{#d91a1a}-0.38\%$
test_clone	6.6500ms	6.3953ms	156.3656 Ops/s	156.6784 Ops/s	$\color{#d91a1a}-0.20\%$
test_squeeze	62.0930μs	9.6295μs	103.8475 KOps/s	107.5779 KOps/s	$\color{#d91a1a}-3.47\%$
test_unsqueeze	0.1184ms	70.1767μs	14.2498 KOps/s	14.0573 KOps/s	$\color{#35bf28}+1.37\%$
test_split	0.4207ms	0.1566ms	6.3852 KOps/s	6.3348 KOps/s	$\color{#35bf28}+0.80\%$
test_permute	0.2086ms	0.1748ms	5.7219 KOps/s	5.5119 KOps/s	$\color{#35bf28}+3.81\%$
test_stack	50.7552ms	50.4560ms	19.8193 Ops/s	20.0543 Ops/s	$\color{#d91a1a}-1.17\%$
test_cat	50.9543ms	50.3177ms	19.8737 Ops/s	23.8375 Ops/s	$\textbf{\color{#d91a1a}-16.63\%}$