[Performance] Faster clone

[vmoens]

Stack from ghstack (oldest at bottom):

• -> [Performance] Faster clone #1040
• [Refactor] Put values, lengths and offsets of NJTs together in storage #1023
• [Feature] TD+NJT to(device) support #1022

[github-actions]

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

Total Benchmarks: 216. Improved: $\large\color{#35bf28}22$. Worsened: $\large\color{#d91a1a}7$.

Expand to view detailed results

Name	Max	Mean	Ops	Ops on Repo HEAD	Change
test_plain_set_nested	69.2890μs	25.7869μs	38.7793 KOps/s	38.1941 KOps/s	$\color{#35bf28}+1.53\%$
test_plain_set_stack_nested	0.1289ms	26.0733μs	38.3534 KOps/s	38.4622 KOps/s	$\color{#d91a1a}-0.28\%$
test_plain_set_nested_inplace	68.5880μs	28.1292μs	35.5502 KOps/s	34.8843 KOps/s	$\color{#35bf28}+1.91\%$
test_plain_set_stack_nested_inplace	71.5940μs	28.1707μs	35.4979 KOps/s	35.4986 KOps/s	$-0.00\%$
test_items	28.9840μs	4.1711μs	239.7425 KOps/s	238.0274 KOps/s	$\color{#35bf28}+0.72\%$
test_items_nested	0.4657ms	0.3849ms	2.5978 KOps/s	2.6160 KOps/s	$\color{#d91a1a}-0.69\%$
test_items_nested_locked	0.4689ms	0.3848ms	2.5984 KOps/s	2.6169 KOps/s	$\color{#d91a1a}-0.71\%$
test_items_nested_leaf	0.1420ms	80.5680μs	12.4119 KOps/s	12.1863 KOps/s	$\color{#35bf28}+1.85\%$
test_items_stack_nested	0.7067ms	0.3953ms	2.5300 KOps/s	2.5880 KOps/s	$\color{#d91a1a}-2.24\%$
test_items_stack_nested_leaf	0.2818ms	85.4766μs	11.6991 KOps/s	11.8848 KOps/s	$\color{#d91a1a}-1.56\%$
test_items_stack_nested_locked	0.6304ms	0.3856ms	2.5933 KOps/s	2.5832 KOps/s	$\color{#35bf28}+0.39\%$
test_keys	19.8570μs	3.5769μs	279.5696 KOps/s	291.5474 KOps/s	$\color{#d91a1a}-4.11\%$
test_keys_nested	0.2197ms	0.1321ms	7.5706 KOps/s	7.4204 KOps/s	$\color{#35bf28}+2.02\%$
test_keys_nested_locked	0.7327ms	0.1373ms	7.2809 KOps/s	7.1266 KOps/s	$\color{#35bf28}+2.17\%$
test_keys_nested_leaf	0.2174ms	0.1160ms	8.6193 KOps/s	8.5042 KOps/s	$\color{#35bf28}+1.35\%$
test_keys_stack_nested	0.2245ms	0.1310ms	7.6328 KOps/s	7.6033 KOps/s	$\color{#35bf28}+0.39\%$
test_keys_stack_nested_leaf	0.3499ms	0.1145ms	8.7316 KOps/s	8.6760 KOps/s	$\color{#35bf28}+0.64\%$
test_keys_stack_nested_locked	0.2644ms	0.1362ms	7.3407 KOps/s	7.2505 KOps/s	$\color{#35bf28}+1.24\%$
test_values	10.7280μs	1.0517μs	950.8277 KOps/s	954.4168 KOps/s	$\color{#d91a1a}-0.38\%$
test_values_nested	0.1690ms	93.7003μs	10.6723 KOps/s	10.5723 KOps/s	$\color{#35bf28}+0.95\%$
test_values_nested_locked	0.1532ms	93.3841μs	10.7085 KOps/s	10.4487 KOps/s	$\color{#35bf28}+2.49\%$
test_values_nested_leaf	0.1484ms	77.2129μs	12.9512 KOps/s	12.5887 KOps/s	$\color{#35bf28}+2.88\%$
test_values_stack_nested	0.3070ms	94.2557μs	10.6094 KOps/s	10.3395 KOps/s	$\color{#35bf28}+2.61\%$
test_values_stack_nested_leaf	0.1719ms	77.8732μs	12.8414 KOps/s	12.6447 KOps/s	$\color{#35bf28}+1.56\%$
test_values_stack_nested_locked	0.1680ms	94.2609μs	10.6089 KOps/s	10.4893 KOps/s	$\color{#35bf28}+1.14\%$
test_membership	6.2717μs	0.7165μs	1.3958 MOps/s	1.4127 MOps/s	$\color{#d91a1a}-1.20\%$
test_membership_nested	41.2570μs	2.7172μs	368.0244 KOps/s	361.0812 KOps/s	$\color{#35bf28}+1.92\%$
test_membership_nested_leaf	98.2630μs	2.7446μs	364.3501 KOps/s	358.2951 KOps/s	$\color{#35bf28}+1.69\%$
test_membership_stacked_nested	30.9580μs	2.6842μs	372.5510 KOps/s	366.1976 KOps/s	$\color{#35bf28}+1.73\%$
test_membership_stacked_nested_leaf	43.7920μs	2.7060μs	369.5501 KOps/s	357.7445 KOps/s	$\color{#35bf28}+3.30\%$
test_membership_nested_last	35.0050μs	4.1628μs	240.2215 KOps/s	239.2007 KOps/s	$\color{#35bf28}+0.43\%$
test_membership_nested_leaf_last	48.1900μs	4.1683μs	239.9056 KOps/s	238.1114 KOps/s	$\color{#35bf28}+0.75\%$
test_membership_stacked_nested_last	31.4690μs	5.0167μs	199.3334 KOps/s	76.2548 KOps/s	$\textbf{\color{#35bf28}+161.40\%}$
test_membership_stacked_nested_leaf_last	45.8460μs	5.0605μs	197.6078 KOps/s	75.8753 KOps/s	$\textbf{\color{#35bf28}+160.44\%}$
test_nested_getleaf	61.1240μs	10.9011μs	91.7342 KOps/s	93.7342 KOps/s	$\color{#d91a1a}-2.13\%$
test_nested_get	54.8830μs	10.5153μs	95.0999 KOps/s	98.1153 KOps/s	$\color{#d91a1a}-3.07\%$
test_stacked_getleaf	38.1410μs	10.8283μs	92.3506 KOps/s	93.8886 KOps/s	$\color{#d91a1a}-1.64\%$
test_stacked_get	35.7870μs	10.4399μs	95.7860 KOps/s	101.4237 KOps/s	$\textbf{\color{#d91a1a}-5.56\%}$
test_nested_getitemleaf	42.0390μs	11.4129μs	87.6202 KOps/s	88.8781 KOps/s	$\color{#d91a1a}-1.42\%$
test_nested_getitem	42.7500μs	10.6744μs	93.6817 KOps/s	95.3144 KOps/s	$\color{#d91a1a}-1.71\%$
test_stacked_getitemleaf	37.3290μs	11.2793μs	88.6584 KOps/s	89.9029 KOps/s	$\color{#d91a1a}-1.38\%$
test_stacked_getitem	52.3780μs	10.6005μs	94.3355 KOps/s	96.3238 KOps/s	$\color{#d91a1a}-2.06\%$
test_lock_nested	88.8736ms	0.5960ms	1.6778 KOps/s	1.9946 KOps/s	$\textbf{\color{#d91a1a}-15.88\%}$
test_lock_stack_nested	0.6997ms	0.4659ms	2.1463 KOps/s	2.2014 KOps/s	$\color{#d91a1a}-2.51\%$
test_unlock_nested	0.1094s	0.5357ms	1.8666 KOps/s	2.3629 KOps/s	$\textbf{\color{#d91a1a}-21.00\%}$
test_unlock_stack_nested	0.6426ms	0.3800ms	2.6316 KOps/s	2.6914 KOps/s	$\color{#d91a1a}-2.22\%$
test_flatten_speed	0.2159ms	0.1008ms	9.9226 KOps/s	9.8008 KOps/s	$\color{#35bf28}+1.24\%$
test_unflatten_speed	0.8974ms	0.5260ms	1.9011 KOps/s	1.8961 KOps/s	$\color{#35bf28}+0.27\%$
test_common_ops	4.5001ms	1.1707ms	854.2225 Ops/s	822.3672 Ops/s	$\color{#35bf28}+3.87\%$
test_creation	35.7860μs	2.2469μs	445.0538 KOps/s	467.4938 KOps/s	$\color{#d91a1a}-4.80\%$
test_creation_empty	83.8170μs	20.7442μs	48.2062 KOps/s	48.4069 KOps/s	$\color{#d91a1a}-0.41\%$
test_creation_nested_1	67.4370μs	23.6103μs	42.3544 KOps/s	40.9332 KOps/s	$\color{#35bf28}+3.47\%$
test_creation_nested_2	81.9930μs	27.7048μs	36.0948 KOps/s	34.5873 KOps/s	$\color{#35bf28}+4.36\%$
test_clone	59.4710μs	16.9931μs	58.8474 KOps/s	56.4298 KOps/s	$\color{#35bf28}+4.28\%$
test_getitem[int]	0.9929ms	16.6607μs	60.0215 KOps/s	59.2432 KOps/s	$\color{#35bf28}+1.31\%$
test_getitem[slice_int]	0.1655ms	30.3908μs	32.9047 KOps/s	31.2526 KOps/s	$\textbf{\color{#35bf28}+5.29\%}$
test_getitem[range]	0.6282ms	57.6739μs	17.3389 KOps/s	17.2275 KOps/s	$\color{#35bf28}+0.65\%$
test_getitem[tuple]	0.1495ms	24.8673μs	40.2134 KOps/s	39.8748 KOps/s	$\color{#35bf28}+0.85\%$
test_getitem[list]	0.6607ms	52.6619μs	18.9891 KOps/s	18.9487 KOps/s	$\color{#35bf28}+0.21\%$
test_setitem_dim[int]	72.0350μs	32.6555μs	30.6227 KOps/s	30.6579 KOps/s	$\color{#d91a1a}-0.11\%$
test_setitem_dim[slice_int]	0.1131ms	60.0821μs	16.6439 KOps/s	16.4877 KOps/s	$\color{#35bf28}+0.95\%$
test_setitem_dim[range]	0.1579ms	83.3302μs	12.0005 KOps/s	11.7073 KOps/s	$\color{#35bf28}+2.50\%$
test_setitem_dim[tuple]	0.1020ms	48.4645μs	20.6337 KOps/s	20.4205 KOps/s	$\color{#35bf28}+1.04\%$
test_setitem	0.2896ms	30.6741μs	32.6008 KOps/s	31.0880 KOps/s	$\color{#35bf28}+4.87\%$
test_set	90.1580μs	30.0167μs	33.3147 KOps/s	31.1699 KOps/s	$\textbf{\color{#35bf28}+6.88\%}$
test_set_shared	3.2885ms	0.2183ms	4.5807 KOps/s	4.5827 KOps/s	$\color{#d91a1a}-0.04\%$
test_update	0.1517ms	39.5963μs	25.2549 KOps/s	24.0129 KOps/s	$\textbf{\color{#35bf28}+5.17\%}$
test_update_nested	0.1871ms	50.5647μs	19.7766 KOps/s	18.9301 KOps/s	$\color{#35bf28}+4.47\%$
test_update__nested	0.7191ms	45.5917μs	21.9338 KOps/s	22.3579 KOps/s	$\color{#d91a1a}-1.90\%$
test_set_nested	0.3237ms	33.3098μs	30.0212 KOps/s	29.4487 KOps/s	$\color{#35bf28}+1.94\%$
test_set_nested_new	0.3274ms	38.5260μs	25.9565 KOps/s	25.1328 KOps/s	$\color{#35bf28}+3.28\%$
test_select	0.1219ms	57.0016μs	17.5434 KOps/s	17.0384 KOps/s	$\color{#35bf28}+2.96\%$
test_select_nested	0.1147ms	59.8925μs	16.6966 KOps/s	16.4894 KOps/s	$\color{#35bf28}+1.26\%$
test_exclude_nested	0.6498ms	74.8501μs	13.3600 KOps/s	13.0447 KOps/s	$\color{#35bf28}+2.42\%$
test_empty[True]	0.5408ms	0.3504ms	2.8542 KOps/s	2.8434 KOps/s	$\color{#35bf28}+0.38\%$
test_empty[False]	18.4375μs	1.2484μs	801.0289 KOps/s	827.1624 KOps/s	$\color{#d91a1a}-3.16\%$
test_unbind_speed	0.3839ms	0.2970ms	3.3675 KOps/s	3.3754 KOps/s	$\color{#d91a1a}-0.23\%$
test_unbind_speed_stack0	0.4519ms	0.2898ms	3.4504 KOps/s	3.4838 KOps/s	$\color{#d91a1a}-0.96\%$
test_unbind_speed_stack1	0.1125s	0.9303ms	1.0749 KOps/s	1.3644 KOps/s	$\textbf{\color{#d91a1a}-21.22\%}$
test_split	3.2813ms	2.0000ms	500.0072 Ops/s	448.0534 Ops/s	$\textbf{\color{#35bf28}+11.60\%}$
test_chunk	0.1025s	2.2056ms	453.3918 Ops/s	445.4885 Ops/s	$\color{#35bf28}+1.77\%$
test_creation[device0]	0.2040ms	0.1151ms	8.6885 KOps/s	8.5011 KOps/s	$\color{#35bf28}+2.20\%$
test_creation_from_tensor	3.2607ms	0.1164ms	8.5903 KOps/s	8.4540 KOps/s	$\color{#35bf28}+1.61\%$
test_add_one[memmap_tensor0]	0.4741ms	7.0735μs	141.3727 KOps/s	129.4769 KOps/s	$\textbf{\color{#35bf28}+9.19\%}$
test_contiguous[memmap_tensor0]	24.0540μs	1.8974μs	527.0373 KOps/s	518.9374 KOps/s	$\color{#35bf28}+1.56\%$
test_stack[memmap_tensor0]	83.9170μs	5.6418μs	177.2483 KOps/s	174.6839 KOps/s	$\color{#35bf28}+1.47\%$
test_memmaptd_index	1.2059ms	0.4135ms	2.4182 KOps/s	2.4245 KOps/s	$\color{#d91a1a}-0.26\%$
test_memmaptd_index_astensor	0.9889ms	0.5085ms	1.9664 KOps/s	1.9360 KOps/s	$\color{#35bf28}+1.57\%$
test_memmaptd_index_op	1.7647ms	1.0622ms	941.4631 Ops/s	920.8502 Ops/s	$\color{#35bf28}+2.24\%$
test_serialize_model	0.1351s	0.1248s	8.0135 Ops/s	8.5637 Ops/s	$\textbf{\color{#d91a1a}-6.42\%}$
test_serialize_model_pickle	0.4324s	0.3929s	2.5450 Ops/s	2.5104 Ops/s	$\color{#35bf28}+1.38\%$
test_serialize_weights	0.1318s	0.1205s	8.2978 Ops/s	8.0192 Ops/s	$\color{#35bf28}+3.47\%$
test_serialize_weights_returnearly	0.2567s	0.1750s	5.7146 Ops/s	6.1465 Ops/s	$\textbf{\color{#d91a1a}-7.03\%}$
test_serialize_weights_pickle	0.4554s	0.4055s	2.4661 Ops/s	2.4289 Ops/s	$\color{#35bf28}+1.53\%$
test_serialize_weights_filesystem	0.1538s	0.1416s	7.0634 Ops/s	7.1450 Ops/s	$\color{#d91a1a}-1.14\%$
test_serialize_model_filesystem	0.1581s	0.1510s	6.6225 Ops/s	6.5821 Ops/s	$\color{#35bf28}+0.61\%$
test_reshape_pytree	83.8070μs	38.0874μs	26.2554 KOps/s	25.7606 KOps/s	$\color{#35bf28}+1.92\%$
test_reshape_td	91.0600μs	45.1273μs	22.1595 KOps/s	21.0053 KOps/s	$\textbf{\color{#35bf28}+5.50\%}$
test_view_pytree	0.1006ms	38.3607μs	26.0683 KOps/s	25.9838 KOps/s	$\color{#35bf28}+0.33\%$
test_view_td	0.1060ms	51.8040μs	19.3035 KOps/s	19.0150 KOps/s	$\color{#35bf28}+1.52\%$
test_unbind_pytree	75.7920μs	35.4909μs	28.1762 KOps/s	28.0567 KOps/s	$\color{#35bf28}+0.43\%$
test_unbind_td	0.3668ms	45.1200μs	22.1631 KOps/s	22.4875 KOps/s	$\color{#d91a1a}-1.44\%$
test_split_pytree	85.5400μs	37.2904μs	26.8166 KOps/s	26.6019 KOps/s	$\color{#35bf28}+0.81\%$
test_split_td	0.5266ms	57.5794μs	17.3673 KOps/s	17.3779 KOps/s	$\color{#d91a1a}-0.06\%$
test_add_pytree	96.0190μs	43.6047μs	22.9333 KOps/s	22.2177 KOps/s	$\color{#35bf28}+3.22\%$
test_add_td	0.1695ms	85.8646μs	11.6462 KOps/s	10.8455 KOps/s	$\textbf{\color{#35bf28}+7.38\%}$
test_compile_add_one_nested[tensordict-compile]	0.1290ms	58.6062μs	17.0630 KOps/s	17.1953 KOps/s	$\color{#d91a1a}-0.77\%$
test_compile_add_one_nested[tensordict-eager]	0.3960ms	0.1980ms	5.0516 KOps/s	5.0462 KOps/s	$\color{#35bf28}+0.11\%$
test_compile_add_one_nested[pytree-compile]	0.1286ms	57.1068μs	17.5110 KOps/s	17.6727 KOps/s	$\color{#d91a1a}-0.92\%$
test_compile_add_one_nested[pytree-eager]	0.3144ms	0.1379ms	7.2503 KOps/s	7.1254 KOps/s	$\color{#35bf28}+1.75\%$
test_compile_copy_nested[tensordict-compile]	71.3940μs	23.2920μs	42.9332 KOps/s	44.2836 KOps/s	$\color{#d91a1a}-3.05\%$
test_compile_copy_nested[tensordict-eager]	0.1478ms	74.6866μs	13.3893 KOps/s	13.0550 KOps/s	$\color{#35bf28}+2.56\%$
test_compile_copy_nested[pytree-compile]	0.1349ms	75.2887μs	13.2822 KOps/s	13.0446 KOps/s	$\color{#35bf28}+1.82\%$
test_compile_copy_nested[pytree-eager]	0.1358ms	68.3377μs	14.6332 KOps/s	14.0820 KOps/s	$\color{#35bf28}+3.91\%$
test_compile_add_one_flat[tensordict-compile]	0.3044ms	0.1818ms	5.4993 KOps/s	5.4429 KOps/s	$\color{#35bf28}+1.04\%$
test_compile_add_one_flat[tensordict-eager]	1.4672ms	0.2422ms	4.1287 KOps/s	4.0823 KOps/s	$\color{#35bf28}+1.14\%$
test_compile_add_one_flat[tensorclass-compile]	0.1117ms	49.3149μs	20.2779 KOps/s	20.8496 KOps/s	$\color{#d91a1a}-2.74\%$
test_compile_add_one_flat[tensorclass-eager]	0.4650ms	79.6544μs	12.5542 KOps/s	11.8601 KOps/s	$\textbf{\color{#35bf28}+5.85\%}$
test_compile_add_one_flat[pytree-compile]	0.2799ms	0.1744ms	5.7324 KOps/s	5.6395 KOps/s	$\color{#35bf28}+1.65\%$
test_compile_add_one_flat[pytree-eager]	0.5176ms	0.2821ms	3.5449 KOps/s	3.4597 KOps/s	$\color{#35bf28}+2.46\%$
test_compile_add_self_flat[tensordict-eager]	0.4822ms	0.2753ms	3.6324 KOps/s	3.6002 KOps/s	$\color{#35bf28}+0.89\%$
test_compile_add_self_flat[tensordict-compile]	0.5089ms	0.1872ms	5.3427 KOps/s	5.4754 KOps/s	$\color{#d91a1a}-2.42\%$
test_compile_add_self_flat[tensorclass-eager]	0.2020ms	75.3398μs	13.2732 KOps/s	13.1895 KOps/s	$\color{#35bf28}+0.63\%$
test_compile_add_self_flat[tensorclass-compile]	0.1388ms	48.5738μs	20.5872 KOps/s	20.6688 KOps/s	$\color{#d91a1a}-0.39\%$
test_compile_add_self_flat[pytree-eager]	0.3808ms	0.2274ms	4.3983 KOps/s	4.3105 KOps/s	$\color{#35bf28}+2.04\%$
test_compile_add_self_flat[pytree-compile]	0.3794ms	0.1715ms	5.8296 KOps/s	5.6198 KOps/s	$\color{#35bf28}+3.73\%$
test_compile_copy_flat[tensordict-compile]	0.2446ms	0.1092ms	9.1608 KOps/s	9.0070 KOps/s	$\color{#35bf28}+1.71\%$
test_compile_copy_flat[tensordict-eager]	0.1738ms	78.5355μs	12.7331 KOps/s	12.4012 KOps/s	$\color{#35bf28}+2.68\%$
test_compile_copy_flat[pytree-compile]	0.1378ms	76.0635μs	13.1469 KOps/s	12.8520 KOps/s	$\color{#35bf28}+2.29\%$
test_compile_copy_flat[pytree-eager]	0.1681ms	66.8078μs	14.9683 KOps/s	14.1643 KOps/s	$\textbf{\color{#35bf28}+5.68\%}$
test_compile_assign_and_add[tensordict-compile]	0.2714ms	0.1927ms	5.1905 KOps/s	5.2148 KOps/s	$\color{#d91a1a}-0.47\%$
test_compile_assign_and_add[tensordict-eager]	2.8873ms	1.6751ms	596.9635 Ops/s	571.9744 Ops/s	$\color{#35bf28}+4.37\%$
test_compile_assign_and_add[pytree-compile]	0.2892ms	0.1916ms	5.2182 KOps/s	5.1871 KOps/s	$\color{#35bf28}+0.60\%$
test_compile_assign_and_add[pytree-eager]	1.3372ms	1.0770ms	928.4779 Ops/s	904.9282 Ops/s	$\color{#35bf28}+2.60\%$
test_compile_assign_and_add_stack[compile]	0.5232ms	0.4143ms	2.4136 KOps/s	2.2727 KOps/s	$\textbf{\color{#35bf28}+6.20\%}$
test_compile_assign_and_add_stack[eager]	5.1599ms	4.1068ms	243.4977 Ops/s	234.1584 Ops/s	$\color{#35bf28}+3.99\%$
test_compile_indexing[tensor-tensordict-compile]	78.6870μs	33.5657μs	29.7923 KOps/s	28.8566 KOps/s	$\color{#35bf28}+3.24\%$
test_compile_indexing[tensor-tensordict-eager]	0.7245ms	48.2340μs	20.7323 KOps/s	20.2371 KOps/s	$\color{#35bf28}+2.45\%$
test_compile_indexing[tensor-tensorclass-compile]	84.8280μs	29.2525μs	34.1851 KOps/s	32.9953 KOps/s	$\color{#35bf28}+3.61\%$
test_compile_indexing[tensor-tensorclass-eager]	83.6360μs	27.9640μs	35.7603 KOps/s	33.6872 KOps/s	$\textbf{\color{#35bf28}+6.15\%}$
test_compile_indexing[tensor-pytree-compile]	87.9840μs	29.3508μs	34.0706 KOps/s	32.9735 KOps/s	$\color{#35bf28}+3.33\%$
test_compile_indexing[tensor-pytree-eager]	0.1415ms	28.0931μs	35.5959 KOps/s	34.1323 KOps/s	$\color{#35bf28}+4.29\%$
test_compile_indexing[slice-tensordict-compile]	0.1400ms	71.7819μs	13.9311 KOps/s	13.5626 KOps/s	$\color{#35bf28}+2.72\%$
test_compile_indexing[slice-tensordict-eager]	0.5580ms	27.1962μs	36.7699 KOps/s	36.1797 KOps/s	$\color{#35bf28}+1.63\%$
test_compile_indexing[slice-tensorclass-compile]	0.1233ms	67.9277μs	14.7215 KOps/s	14.6971 KOps/s	$\color{#35bf28}+0.17\%$
test_compile_indexing[slice-tensorclass-eager]	66.0930μs	23.1179μs	43.2566 KOps/s	43.1661 KOps/s	$\color{#35bf28}+0.21\%$
test_compile_indexing[slice-pytree-compile]	0.1355ms	68.2881μs	14.6438 KOps/s	14.5204 KOps/s	$\color{#35bf28}+0.85\%$
test_compile_indexing[slice-pytree-eager]	0.1816ms	23.1755μs	43.1491 KOps/s	42.9198 KOps/s	$\color{#35bf28}+0.53\%$
test_compile_indexing[int-tensordict-compile]	0.1372ms	71.9380μs	13.9009 KOps/s	13.7729 KOps/s	$\color{#35bf28}+0.93\%$
test_compile_indexing[int-tensordict-eager]	0.7859ms	27.0446μs	36.9759 KOps/s	36.7677 KOps/s	$\color{#35bf28}+0.57\%$
test_compile_indexing[int-tensorclass-compile]	0.1920ms	67.9597μs	14.7146 KOps/s	14.6084 KOps/s	$\color{#35bf28}+0.73\%$
test_compile_indexing[int-tensorclass-eager]	92.5050μs	22.5085μs	44.4276 KOps/s	43.6118 KOps/s	$\color{#35bf28}+1.87\%$
test_compile_indexing[int-pytree-compile]	0.1443ms	67.3176μs	14.8550 KOps/s	14.6330 KOps/s	$\color{#35bf28}+1.52\%$
test_compile_indexing[int-pytree-eager]	56.3350μs	22.6959μs	44.0609 KOps/s	43.1405 KOps/s	$\color{#35bf28}+2.13\%$
test_mod_add[eager]	70.6820μs	26.7297μs	37.4116 KOps/s	35.9577 KOps/s	$\color{#35bf28}+4.04\%$
test_mod_add[compile]	85.1690μs	37.6413μs	26.5665 KOps/s	26.0622 KOps/s	$\color{#35bf28}+1.94\%$
test_mod_add[compile-overhead]	96.7810μs	37.6755μs	26.5425 KOps/s	26.6547 KOps/s	$\color{#d91a1a}-0.42\%$
test_mod_wrap[eager]	0.3179ms	0.2081ms	4.8062 KOps/s	4.7678 KOps/s	$\color{#35bf28}+0.80\%$
test_mod_wrap[compile]	0.3563ms	0.2306ms	4.3368 KOps/s	4.3110 KOps/s	$\color{#35bf28}+0.60\%$
test_mod_wrap[compile-overhead]	0.3827ms	0.2296ms	4.3552 KOps/s	4.2691 KOps/s	$\color{#35bf28}+2.02\%$
test_mod_wrap_and_backward[eager]	18.3251ms	11.6997ms	85.4722 Ops/s	82.4218 Ops/s	$\color{#35bf28}+3.70\%$
test_mod_wrap_and_backward[compile]	13.7414ms	11.6520ms	85.8219 Ops/s	84.7389 Ops/s	$\color{#35bf28}+1.28\%$
test_mod_wrap_and_backward[compile-overhead]	13.2989ms	11.8545ms	84.3560 Ops/s	75.0334 Ops/s	$\textbf{\color{#35bf28}+12.42\%}$
test_seq_add[eager]	0.2197ms	94.0548μs	10.6321 KOps/s	10.1531 KOps/s	$\color{#35bf28}+4.72\%$
test_seq_add[compile]	0.1195ms	63.8330μs	15.6659 KOps/s	15.6793 KOps/s	$\color{#d91a1a}-0.09\%$
test_seq_add[compile-overhead]	0.1209ms	63.5418μs	15.7377 KOps/s	15.7536 KOps/s	$\color{#d91a1a}-0.10\%$
test_seq_wrap[eager]	0.7301ms	0.3915ms	2.5541 KOps/s	2.4466 KOps/s	$\color{#35bf28}+4.40\%$
test_seq_wrap[compile]	0.4909ms	0.2710ms	3.6894 KOps/s	3.6267 KOps/s	$\color{#35bf28}+1.73\%$
test_seq_wrap[compile-overhead]	0.3662ms	0.2676ms	3.7369 KOps/s	3.5771 KOps/s	$\color{#35bf28}+4.47\%$
test_func_call_runtime[False-eager]	0.6723ms	0.5175ms	1.9322 KOps/s	1.8526 KOps/s	$\color{#35bf28}+4.30\%$
test_func_call_runtime[False-compile]	0.6339ms	0.4960ms	2.0159 KOps/s	1.9879 KOps/s	$\color{#35bf28}+1.41\%$
test_func_call_runtime[False-compile-overhead]	1.0460ms	0.4986ms	2.0057 KOps/s	1.9705 KOps/s	$\color{#35bf28}+1.78\%$
test_func_call_runtime[True-eager]	0.9430ms	0.7296ms	1.3705 KOps/s	1.3189 KOps/s	$\color{#35bf28}+3.91\%$
test_func_call_runtime[True-compile]	0.6001ms	0.5039ms	1.9846 KOps/s	1.9034 KOps/s	$\color{#35bf28}+4.27\%$
test_func_call_runtime[True-compile-overhead]	0.7738ms	0.5063ms	1.9752 KOps/s	1.9179 KOps/s	$\color{#35bf28}+2.99\%$
test_func_call_cm_runtime[False-eager]	0.7254ms	0.5149ms	1.9422 KOps/s	1.9017 KOps/s	$\color{#35bf28}+2.13\%$
test_func_call_cm_runtime[False-compile]	0.9302ms	0.4958ms	2.0169 KOps/s	1.9779 KOps/s	$\color{#35bf28}+1.97\%$
test_func_call_cm_runtime[False-compile-overhead]	0.9188ms	0.4955ms	2.0182 KOps/s	1.9753 KOps/s	$\color{#35bf28}+2.17\%$
test_func_call_cm_runtime[True-eager]	1.2164ms	0.8741ms	1.1441 KOps/s	1.0963 KOps/s	$\color{#35bf28}+4.35\%$
test_func_call_cm_runtime[True-compile]	1.1599ms	0.7291ms	1.3715 KOps/s	1.3197 KOps/s	$\color{#35bf28}+3.93\%$
test_func_call_cm_runtime[True-compile-overhead]	1.1813ms	0.7312ms	1.3677 KOps/s	1.3255 KOps/s	$\color{#35bf28}+3.18\%$
test_vmap_func_call_cm_runtime[eager]	2.3936ms	1.8723ms	534.0915 Ops/s	512.1015 Ops/s	$\color{#35bf28}+4.29\%$
test_vmap_func_call_cm_runtime[compile]	3.4336ms	1.9577ms	510.7938 Ops/s	499.6178 Ops/s	$\color{#35bf28}+2.24\%$
test_vmap_func_call_cm_runtime[compile-overhead]	3.3404ms	1.9524ms	512.1930 Ops/s	492.1656 Ops/s	$\color{#35bf28}+4.07\%$
test_distributed	0.3029ms	0.1266ms	7.9017 KOps/s	7.5913 KOps/s	$\color{#35bf28}+4.09\%$
test_tdmodule	98.5340μs	19.2525μs	51.9414 KOps/s	48.6060 KOps/s	$\textbf{\color{#35bf28}+6.86\%}$
test_tdmodule_dispatch	65.8230μs	38.4040μs	26.0389 KOps/s	25.1138 KOps/s	$\color{#35bf28}+3.68\%$
test_tdseq	49.0020μs	22.3377μs	44.7673 KOps/s	43.3337 KOps/s	$\color{#35bf28}+3.31\%$
test_tdseq_dispatch	74.6900μs	44.3183μs	22.5640 KOps/s	21.9582 KOps/s	$\color{#35bf28}+2.76\%$
test_instantiation_functorch	2.1361ms	1.5350ms	651.4693 Ops/s	582.6844 Ops/s	$\textbf{\color{#35bf28}+11.80\%}$
test_exec_functorch	0.3295ms	0.1824ms	5.4823 KOps/s	5.3086 KOps/s	$\color{#35bf28}+3.27\%$
test_exec_functional_call	0.3120ms	0.1716ms	5.8288 KOps/s	5.8008 KOps/s	$\color{#35bf28}+0.48\%$
test_exec_td_decorator	0.4782ms	0.2328ms	4.2962 KOps/s	4.2328 KOps/s	$\color{#35bf28}+1.50\%$
test_vmap_mlp_speed_decorator[True-True]	1.0297ms	0.6635ms	1.5070 KOps/s	1.4996 KOps/s	$\color{#35bf28}+0.49\%$
test_vmap_mlp_speed_decorator[True-False]	0.8457ms	0.6333ms	1.5791 KOps/s	1.5058 KOps/s	$\color{#35bf28}+4.87\%$
test_vmap_mlp_speed_decorator[False-True]	0.8092ms	0.5194ms	1.9253 KOps/s	1.8408 KOps/s	$\color{#35bf28}+4.59\%$
test_vmap_mlp_speed_decorator[False-False]	0.7022ms	0.5158ms	1.9389 KOps/s	1.7222 KOps/s	$\textbf{\color{#35bf28}+12.59\%}$
test_to_module_speed[True]	2.2988ms	1.4231ms	702.6701 Ops/s	685.8409 Ops/s	$\color{#35bf28}+2.45\%$
test_to_module_speed[False]	1.7128ms	1.3816ms	723.8200 Ops/s	714.2990 Ops/s	$\color{#35bf28}+1.33\%$
test_tc_init	0.1249ms	47.4474μs	21.0760 KOps/s	20.4833 KOps/s	$\color{#35bf28}+2.89\%$
test_tc_init_nested	0.2054ms	94.8297μs	10.5452 KOps/s	10.2663 KOps/s	$\color{#35bf28}+2.72\%$
test_tc_first_layer_tensor	24.0050μs	1.6030μs	623.8207 KOps/s	647.5391 KOps/s	$\color{#d91a1a}-3.66\%$
test_tc_first_layer_nontensor	24.2950μs	4.7150μs	212.0886 KOps/s	201.3532 KOps/s	$\textbf{\color{#35bf28}+5.33\%}$
test_tc_second_layer_tensor	28.6940μs	2.9248μs	341.9068 KOps/s	340.9035 KOps/s	$\color{#35bf28}+0.29\%$
test_tc_second_layer_nontensor	44.5550μs	5.9719μs	167.4504 KOps/s	162.6014 KOps/s	$\color{#35bf28}+2.98\%$
test_unbind	0.5180s	13.6787ms	73.1064 Ops/s	132.9157 Ops/s	$\textbf{\color{#d91a1a}-45.00\%}$
test_full_like	13.8388ms	8.6218ms	115.9851 Ops/s	74.1177 Ops/s	$\textbf{\color{#35bf28}+56.49\%}$
test_zeros_like	4.8817ms	3.0767ms	325.0257 Ops/s	118.9574 Ops/s	$\textbf{\color{#35bf28}+173.23\%}$
test_ones_like	12.6680ms	6.3519ms	157.4337 Ops/s	111.7366 Ops/s	$\textbf{\color{#35bf28}+40.90\%}$
test_clone	16.2253ms	9.0088ms	111.0024 Ops/s	90.1455 Ops/s	$\textbf{\color{#35bf28}+23.14\%}$
test_squeeze	74.0480μs	12.3087μs	81.2432 KOps/s	81.8361 KOps/s	$\color{#d91a1a}-0.72\%$
test_unsqueeze	0.1681ms	91.5289μs	10.9255 KOps/s	11.1825 KOps/s	$\color{#d91a1a}-2.30\%$
test_split	0.5531ms	0.1927ms	5.1889 KOps/s	5.2204 KOps/s	$\color{#d91a1a}-0.60\%$
test_permute	0.4814ms	0.2144ms	4.6631 KOps/s	4.6142 KOps/s	$\color{#35bf28}+1.06\%$
test_stack	35.1199ms	27.2849ms	36.6503 Ops/s	36.4029 Ops/s	$\color{#35bf28}+0.68\%$
test_cat	31.0045ms	26.8414ms	37.2559 Ops/s	36.7647 Ops/s	$\color{#35bf28}+1.34\%$

[github-actions]

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

Total Benchmarks: 218. Improved: $\large\color{#35bf28}14$. Worsened: $\large\color{#d91a1a}5$.

Expand to view detailed results

Name	Max	Mean	Ops	Ops on Repo HEAD	Change
test_plain_set_nested	0.1481ms	18.0352μs	55.4471 KOps/s	56.0120 KOps/s	$\color{#d91a1a}-1.01\%$
test_plain_set_stack_nested	38.5210μs	17.9399μs	55.7418 KOps/s	56.2071 KOps/s	$\color{#d91a1a}-0.83\%$
test_plain_set_nested_inplace	47.0510μs	19.2175μs	52.0360 KOps/s	52.2696 KOps/s	$\color{#d91a1a}-0.45\%$
test_plain_set_stack_nested_inplace	61.6610μs	19.0721μs	52.4326 KOps/s	52.5799 KOps/s	$\color{#d91a1a}-0.28\%$
test_items	22.5700μs	2.8824μs	346.9361 KOps/s	345.2369 KOps/s	$\color{#35bf28}+0.49\%$
test_items_nested	0.4969ms	0.3374ms	2.9643 KOps/s	2.9884 KOps/s	$\color{#d91a1a}-0.81\%$
test_items_nested_locked	0.3675ms	0.3396ms	2.9445 KOps/s	2.9785 KOps/s	$\color{#d91a1a}-1.14\%$
test_items_nested_leaf	85.9520μs	62.9998μs	15.8731 KOps/s	16.0185 KOps/s	$\color{#d91a1a}-0.91\%$
test_items_stack_nested	0.3721ms	0.3411ms	2.9317 KOps/s	2.9422 KOps/s	$\color{#d91a1a}-0.36\%$
test_items_stack_nested_leaf	86.6210μs	63.5325μs	15.7400 KOps/s	15.7040 KOps/s	$\color{#35bf28}+0.23\%$
test_items_stack_nested_locked	0.4368ms	0.3415ms	2.9284 KOps/s	2.9613 KOps/s	$\color{#d91a1a}-1.11\%$
test_keys	30.8000μs	3.4327μs	291.3153 KOps/s	289.5678 KOps/s	$\color{#35bf28}+0.60\%$
test_keys_nested	0.1229ms	71.2072μs	14.0435 KOps/s	14.1398 KOps/s	$\color{#d91a1a}-0.68\%$
test_keys_nested_locked	0.7876ms	76.5246μs	13.0677 KOps/s	13.0194 KOps/s	$\color{#35bf28}+0.37\%$
test_keys_nested_leaf	97.2320μs	61.6266μs	16.2268 KOps/s	16.3216 KOps/s	$\color{#d91a1a}-0.58\%$
test_keys_stack_nested	0.1111ms	71.1836μs	14.0482 KOps/s	14.1016 KOps/s	$\color{#d91a1a}-0.38\%$
test_keys_stack_nested_leaf	95.1820μs	62.9837μs	15.8771 KOps/s	15.8853 KOps/s	$\color{#d91a1a}-0.05\%$
test_keys_stack_nested_locked	0.1054ms	77.4681μs	12.9085 KOps/s	12.8721 KOps/s	$\color{#35bf28}+0.28\%$
test_values	4.8050μs	0.8369μs	1.1949 MOps/s	1.1830 MOps/s	$\color{#35bf28}+1.01\%$
test_values_nested	74.6510μs	48.4863μs	20.6244 KOps/s	20.5079 KOps/s	$\color{#35bf28}+0.57\%$
test_values_nested_locked	74.0710μs	50.0881μs	19.9648 KOps/s	19.8581 KOps/s	$\color{#35bf28}+0.54\%$
test_values_nested_leaf	67.9920μs	42.8792μs	23.3214 KOps/s	23.4114 KOps/s	$\color{#d91a1a}-0.38\%$
test_values_stack_nested	85.1720μs	49.6950μs	20.1228 KOps/s	20.1938 KOps/s	$\color{#d91a1a}-0.35\%$
test_values_stack_nested_leaf	71.4820μs	43.7713μs	22.8460 KOps/s	23.0268 KOps/s	$\color{#d91a1a}-0.79\%$
test_values_stack_nested_locked	80.2810μs	51.8904μs	19.2714 KOps/s	19.5605 KOps/s	$\color{#d91a1a}-1.48\%$
test_membership	1.5356μs	0.5066μs	1.9740 MOps/s	1.9716 MOps/s	$\color{#35bf28}+0.12\%$
test_membership_nested	13.3500μs	1.8705μs	534.6106 KOps/s	548.0333 KOps/s	$\color{#d91a1a}-2.45\%$
test_membership_nested_leaf	14.6405μs	1.8397μs	543.5701 KOps/s	557.6154 KOps/s	$\color{#d91a1a}-2.52\%$
test_membership_stacked_nested	33.2710μs	1.8903μs	529.0027 KOps/s	534.2995 KOps/s	$\color{#d91a1a}-0.99\%$
test_membership_stacked_nested_leaf	25.0100μs	1.9309μs	517.9049 KOps/s	534.7840 KOps/s	$\color{#d91a1a}-3.16\%$
test_membership_nested_last	38.7310μs	2.9437μs	339.7118 KOps/s	343.5265 KOps/s	$\color{#d91a1a}-1.11\%$
test_membership_nested_leaf_last	36.2610μs	2.9618μs	337.6271 KOps/s	344.6008 KOps/s	$\color{#d91a1a}-2.02\%$
test_membership_stacked_nested_last	26.4400μs	2.9099μs	343.6511 KOps/s	346.2904 KOps/s	$\color{#d91a1a}-0.76\%$
test_membership_stacked_nested_leaf_last	33.2110μs	2.9296μs	341.3441 KOps/s	344.7707 KOps/s	$\color{#d91a1a}-0.99\%$
test_nested_getleaf	32.7800μs	6.1453μs	162.7272 KOps/s	165.5011 KOps/s	$\color{#d91a1a}-1.68\%$
test_nested_get	34.0700μs	5.7354μs	174.3544 KOps/s	174.1047 KOps/s	$\color{#35bf28}+0.14\%$
test_stacked_getleaf	32.9110μs	6.0763μs	164.5725 KOps/s	165.3511 KOps/s	$\color{#d91a1a}-0.47\%$
test_stacked_get	35.0510μs	5.6410μs	177.2726 KOps/s	176.6379 KOps/s	$\color{#35bf28}+0.36\%$
test_nested_getitemleaf	40.2210μs	6.1116μs	163.6242 KOps/s	164.4866 KOps/s	$\color{#d91a1a}-0.52\%$
test_nested_getitem	32.1100μs	5.7612μs	173.5752 KOps/s	174.3752 KOps/s	$\color{#d91a1a}-0.46\%$
test_stacked_getitemleaf	31.1600μs	6.0795μs	164.4863 KOps/s	164.1013 KOps/s	$\color{#35bf28}+0.23\%$
test_stacked_getitem	35.9810μs	5.7058μs	175.2593 KOps/s	177.5636 KOps/s	$\color{#d91a1a}-1.30\%$
test_lock_nested	7.6557ms	0.4264ms	2.3453 KOps/s	2.3772 KOps/s	$\color{#d91a1a}-1.34\%$
test_lock_stack_nested	0.4219ms	0.3887ms	2.5726 KOps/s	2.5879 KOps/s	$\color{#d91a1a}-0.59\%$
test_unlock_nested	0.7798ms	0.3600ms	2.7778 KOps/s	2.8043 KOps/s	$\color{#d91a1a}-0.94\%$
test_unlock_stack_nested	0.3621ms	0.3261ms	3.0668 KOps/s	3.0869 KOps/s	$\color{#d91a1a}-0.65\%$
test_flatten_speed	0.1564ms	77.7880μs	12.8555 KOps/s	13.0576 KOps/s	$\color{#d91a1a}-1.55\%$
test_unflatten_speed	0.4111ms	0.3251ms	3.0762 KOps/s	3.1157 KOps/s	$\color{#d91a1a}-1.27\%$
test_common_ops	1.6757ms	1.2923ms	773.8041 Ops/s	770.1550 Ops/s	$\color{#35bf28}+0.47\%$
test_creation	29.0010μs	1.4484μs	690.4227 KOps/s	685.5652 KOps/s	$\color{#35bf28}+0.71\%$
test_creation_empty	63.8210μs	17.8292μs	56.0879 KOps/s	55.7126 KOps/s	$\color{#35bf28}+0.67\%$
test_creation_nested_1	55.2910μs	19.4460μs	51.4245 KOps/s	50.9733 KOps/s	$\color{#35bf28}+0.89\%$
test_creation_nested_2	50.2110μs	22.3847μs	44.6733 KOps/s	43.9264 KOps/s	$\color{#35bf28}+1.70\%$
test_clone	66.6510μs	28.7896μs	34.7348 KOps/s	34.8615 KOps/s	$\color{#d91a1a}-0.36\%$
test_getitem[int]	1.3639ms	15.6425μs	63.9286 KOps/s	66.3906 KOps/s	$\color{#d91a1a}-3.71\%$
test_getitem[slice_int]	0.1204ms	26.6492μs	37.5245 KOps/s	38.5328 KOps/s	$\color{#d91a1a}-2.62\%$
test_getitem[range]	0.1583ms	0.1082ms	9.2463 KOps/s	9.3098 KOps/s	$\color{#d91a1a}-0.68\%$
test_getitem[tuple]	0.1260ms	23.4704μs	42.6068 KOps/s	45.0243 KOps/s	$\textbf{\color{#d91a1a}-5.37\%}$
test_getitem[list]	0.1940ms	98.6454μs	10.1373 KOps/s	10.2572 KOps/s	$\color{#d91a1a}-1.17\%$
test_setitem_dim[int]	66.5020μs	43.7993μs	22.8314 KOps/s	22.5867 KOps/s	$\color{#35bf28}+1.08\%$
test_setitem_dim[slice_int]	94.2420μs	66.2949μs	15.0841 KOps/s	15.0286 KOps/s	$\color{#35bf28}+0.37\%$
test_setitem_dim[range]	0.1873ms	0.1265ms	7.9050 KOps/s	7.9499 KOps/s	$\color{#d91a1a}-0.56\%$
test_setitem_dim[tuple]	90.5620μs	59.8409μs	16.7110 KOps/s	16.8128 KOps/s	$\color{#d91a1a}-0.61\%$
test_setitem	74.2120μs	43.2232μs	23.1357 KOps/s	23.2663 KOps/s	$\color{#d91a1a}-0.56\%$
test_set	81.1010μs	41.3635μs	24.1759 KOps/s	23.8252 KOps/s	$\color{#35bf28}+1.47\%$
test_set_shared	0.3656ms	53.2160μs	18.7913 KOps/s	18.4497 KOps/s	$\color{#35bf28}+1.85\%$
test_update	94.6120μs	52.3516μs	19.1016 KOps/s	19.2711 KOps/s	$\color{#d91a1a}-0.88\%$
test_update_nested	94.5610μs	59.4605μs	16.8179 KOps/s	16.8241 KOps/s	$\color{#d91a1a}-0.04\%$
test_update__nested	0.1957ms	63.5399μs	15.7381 KOps/s	14.8851 KOps/s	$\textbf{\color{#35bf28}+5.73\%}$
test_set_nested	94.4020μs	45.9516μs	21.7621 KOps/s	22.2602 KOps/s	$\color{#d91a1a}-2.24\%$
test_set_nested_new	92.0520μs	48.8504μs	20.4706 KOps/s	20.9657 KOps/s	$\color{#d91a1a}-2.36\%$
test_select	97.9920μs	61.4597μs	16.2708 KOps/s	16.5933 KOps/s	$\color{#d91a1a}-1.94\%$
test_select_nested	69.7210μs	41.5868μs	24.0461 KOps/s	23.9394 KOps/s	$\color{#35bf28}+0.45\%$
test_exclude_nested	85.4520μs	57.4668μs	17.4014 KOps/s	17.2400 KOps/s	$\color{#35bf28}+0.94\%$
test_empty[True]	0.2880ms	0.2612ms	3.8283 KOps/s	3.8948 KOps/s	$\color{#d91a1a}-1.71\%$
test_empty[False]	6.6261μs	0.7383μs	1.3545 MOps/s	1.3479 MOps/s	$\color{#35bf28}+0.49\%$
test_to	63.8810μs	26.3096μs	38.0090 KOps/s	38.7987 KOps/s	$\color{#d91a1a}-2.04\%$
test_to_nonblocking	53.4410μs	24.9690μs	40.0496 KOps/s	41.0036 KOps/s	$\color{#d91a1a}-2.33\%$
test_unbind_speed	1.5734ms	0.2717ms	3.6811 KOps/s	3.6459 KOps/s	$\color{#35bf28}+0.97\%$
test_unbind_speed_stack0	0.3191ms	0.2743ms	3.6450 KOps/s	3.6568 KOps/s	$\color{#d91a1a}-0.32\%$
test_unbind_speed_stack1	93.2680ms	0.7126ms	1.4032 KOps/s	1.3981 KOps/s	$\color{#35bf28}+0.37\%$
test_split	95.0083ms	2.0875ms	479.0488 Ops/s	469.5180 Ops/s	$\color{#35bf28}+2.03\%$
test_chunk	95.2293ms	2.1098ms	473.9778 Ops/s	468.0460 Ops/s	$\color{#35bf28}+1.27\%$
test_creation[device0]	0.3066ms	0.1264ms	7.9087 KOps/s	7.9615 KOps/s	$\color{#d91a1a}-0.66\%$
test_creation_from_tensor	0.3511ms	0.1276ms	7.8385 KOps/s	7.8619 KOps/s	$\color{#d91a1a}-0.30\%$
test_add_one[memmap_tensor0]	0.1474ms	8.8321μs	113.2240 KOps/s	112.8768 KOps/s	$\color{#35bf28}+0.31\%$
test_contiguous[memmap_tensor0]	29.6900μs	2.0792μs	480.9534 KOps/s	446.1647 KOps/s	$\textbf{\color{#35bf28}+7.80\%}$
test_stack[memmap_tensor0]	30.9610μs	6.3821μs	156.6885 KOps/s	154.5334 KOps/s	$\color{#35bf28}+1.39\%$
test_memmaptd_index	1.1309ms	0.4248ms	2.3543 KOps/s	2.4536 KOps/s	$\color{#d91a1a}-4.05\%$
test_memmaptd_index_astensor	0.7840ms	0.4979ms	2.0084 KOps/s	2.0661 KOps/s	$\color{#d91a1a}-2.79\%$
test_memmaptd_index_op	1.4813ms	1.0732ms	931.8274 Ops/s	945.6589 Ops/s	$\color{#d91a1a}-1.46\%$
test_serialize_model	0.1305s	0.1297s	7.7129 Ops/s	7.6844 Ops/s	$\color{#35bf28}+0.37\%$
test_serialize_model_pickle	1.3774s	1.2186s	0.8206 Ops/s	0.8230 Ops/s	$\color{#d91a1a}-0.29\%$
test_serialize_weights	0.2232s	0.1425s	7.0177 Ops/s	6.9837 Ops/s	$\color{#35bf28}+0.49\%$
test_serialize_weights_returnearly	0.2209s	55.9398ms	17.8764 Ops/s	17.7111 Ops/s	$\color{#35bf28}+0.93\%$
test_serialize_weights_pickle	1.3898s	1.2213s	0.8188 Ops/s	0.8213 Ops/s	$\color{#d91a1a}-0.30\%$
test_reshape_pytree	71.0610μs	34.9648μs	28.6002 KOps/s	29.0496 KOps/s	$\color{#d91a1a}-1.55\%$
test_reshape_td	71.3610μs	40.5373μs	24.6686 KOps/s	24.5762 KOps/s	$\color{#35bf28}+0.38\%$
test_view_pytree	63.5410μs	34.7931μs	28.7414 KOps/s	29.7911 KOps/s	$\color{#d91a1a}-3.52\%$
test_view_td	84.0510μs	46.6653μs	21.4292 KOps/s	22.5641 KOps/s	$\textbf{\color{#d91a1a}-5.03\%}$
test_unbind_pytree	59.5210μs	33.6475μs	29.7199 KOps/s	29.0042 KOps/s	$\color{#35bf28}+2.47\%$
test_unbind_td	0.6643ms	42.6029μs	23.4726 KOps/s	22.9168 KOps/s	$\color{#35bf28}+2.43\%$
test_split_pytree	75.4220μs	45.2372μs	22.1057 KOps/s	21.5993 KOps/s	$\color{#35bf28}+2.34\%$
test_split_td	0.1748ms	53.7848μs	18.5926 KOps/s	17.6705 KOps/s	$\textbf{\color{#35bf28}+5.22\%}$
test_add_pytree	87.0920μs	54.9620μs	18.1944 KOps/s	16.6733 KOps/s	$\textbf{\color{#35bf28}+9.12\%}$
test_add_td	0.2310ms	95.8459μs	10.4334 KOps/s	9.8267 KOps/s	$\textbf{\color{#35bf28}+6.17\%}$
test_compile_add_one_nested[tensordict-compile]	0.3203ms	0.1576ms	6.3437 KOps/s	6.1657 KOps/s	$\color{#35bf28}+2.89\%$
test_compile_add_one_nested[tensordict-eager]	0.2903ms	0.1603ms	6.2384 KOps/s	6.3181 KOps/s	$\color{#d91a1a}-1.26\%$
test_compile_add_one_nested[pytree-compile]	0.2143ms	0.1560ms	6.4091 KOps/s	6.3472 KOps/s	$\color{#35bf28}+0.98\%$
test_compile_add_one_nested[pytree-eager]	0.2804ms	0.1831ms	5.4605 KOps/s	5.4447 KOps/s	$\color{#35bf28}+0.29\%$
test_compile_copy_nested[tensordict-compile]	56.1010μs	21.1438μs	47.2951 KOps/s	45.7675 KOps/s	$\color{#35bf28}+3.34\%$
test_compile_copy_nested[tensordict-eager]	0.1056ms	47.0824μs	21.2394 KOps/s	21.2115 KOps/s	$\color{#35bf28}+0.13\%$
test_compile_copy_nested[pytree-compile]	0.3112ms	64.3962μs	15.5289 KOps/s	15.4803 KOps/s	$\color{#35bf28}+0.31\%$
test_compile_copy_nested[pytree-eager]	92.3210μs	49.8423μs	20.0633 KOps/s	20.1295 KOps/s	$\color{#d91a1a}-0.33\%$
test_compile_add_one_flat[tensordict-compile]	0.3645ms	0.3107ms	3.2187 KOps/s	3.1819 KOps/s	$\color{#35bf28}+1.16\%$
test_compile_add_one_flat[tensordict-eager]	0.3132ms	0.2294ms	4.3584 KOps/s	4.3423 KOps/s	$\color{#35bf28}+0.37\%$
test_compile_add_one_flat[tensorclass-compile]	0.1942ms	0.1256ms	7.9615 KOps/s	7.8559 KOps/s	$\color{#35bf28}+1.34\%$
test_compile_add_one_flat[tensorclass-eager]	0.1262ms	64.1341μs	15.5923 KOps/s	15.5553 KOps/s	$\color{#35bf28}+0.24\%$
test_compile_add_one_flat[pytree-compile]	0.3670ms	0.3205ms	3.1201 KOps/s	3.1406 KOps/s	$\color{#d91a1a}-0.65\%$
test_compile_add_one_flat[pytree-eager]	0.7032ms	0.6115ms	1.6354 KOps/s	1.6192 KOps/s	$\color{#35bf28}+1.00\%$
test_compile_add_self_flat[tensordict-eager]	0.4067ms	0.2796ms	3.5765 KOps/s	3.5700 KOps/s	$\color{#35bf28}+0.18\%$
test_compile_add_self_flat[tensordict-compile]	0.4662ms	0.3149ms	3.1760 KOps/s	3.1863 KOps/s	$\color{#d91a1a}-0.32\%$
test_compile_add_self_flat[tensorclass-eager]	0.1650ms	81.2969μs	12.3006 KOps/s	12.5918 KOps/s	$\color{#d91a1a}-2.31\%$
test_compile_add_self_flat[tensorclass-compile]	0.1945ms	0.1312ms	7.6191 KOps/s	7.5804 KOps/s	$\color{#35bf28}+0.51\%$
test_compile_add_self_flat[pytree-eager]	0.6565ms	0.5365ms	1.8639 KOps/s	1.8975 KOps/s	$\color{#d91a1a}-1.77\%$
test_compile_add_self_flat[pytree-compile]	0.3847ms	0.3189ms	3.1353 KOps/s	3.1103 KOps/s	$\color{#35bf28}+0.81\%$
test_compile_copy_flat[tensordict-compile]	55.6210μs	19.1763μs	52.1478 KOps/s	52.9529 KOps/s	$\color{#d91a1a}-1.52\%$
test_compile_copy_flat[tensordict-eager]	86.9120μs	38.5554μs	25.9367 KOps/s	26.0905 KOps/s	$\color{#d91a1a}-0.59\%$
test_compile_copy_flat[pytree-compile]	0.1147ms	70.0492μs	14.2757 KOps/s	14.3784 KOps/s	$\color{#d91a1a}-0.71\%$
test_compile_copy_flat[pytree-eager]	90.5920μs	51.3442μs	19.4764 KOps/s	19.4955 KOps/s	$\color{#d91a1a}-0.10\%$
test_compile_assign_and_add[tensordict-compile]	2.3974ms	0.8367ms	1.1952 KOps/s	1.1339 KOps/s	$\textbf{\color{#35bf28}+5.41\%}$
test_compile_assign_and_add[tensordict-eager]	3.6554ms	3.2978ms	303.2296 Ops/s	312.6941 Ops/s	$\color{#d91a1a}-3.03\%$
test_compile_assign_and_add[pytree-compile]	2.4501ms	0.8409ms	1.1892 KOps/s	1.1199 KOps/s	$\textbf{\color{#35bf28}+6.19\%}$
test_compile_assign_and_add[pytree-eager]	3.4766ms	3.2177ms	310.7825 Ops/s	315.8217 Ops/s	$\color{#d91a1a}-1.60\%$
test_compile_indexing[tensor-tensordict-compile]	0.2900ms	0.1201ms	8.3288 KOps/s	8.0653 KOps/s	$\color{#35bf28}+3.27\%$
test_compile_indexing[tensor-tensordict-eager]	0.2254ms	60.6529μs	16.4873 KOps/s	15.3506 KOps/s	$\textbf{\color{#35bf28}+7.40\%}$
test_compile_indexing[tensor-tensorclass-compile]	0.1707ms	0.1196ms	8.3620 KOps/s	8.3832 KOps/s	$\color{#d91a1a}-0.25\%$
test_compile_indexing[tensor-tensorclass-eager]	0.2107ms	44.4288μs	22.5079 KOps/s	22.0900 KOps/s	$\color{#35bf28}+1.89\%$
test_compile_indexing[tensor-pytree-compile]	0.2484ms	0.1207ms	8.2873 KOps/s	8.3561 KOps/s	$\color{#d91a1a}-0.82\%$
test_compile_indexing[tensor-pytree-eager]	0.1800ms	45.1731μs	22.1370 KOps/s	22.0677 KOps/s	$\color{#35bf28}+0.31\%$
test_compile_indexing[slice-tensordict-compile]	0.2086ms	0.1493ms	6.6990 KOps/s	6.6385 KOps/s	$\color{#35bf28}+0.91\%$
test_compile_indexing[slice-tensordict-eager]	0.1568ms	24.2689μs	41.2050 KOps/s	40.9120 KOps/s	$\color{#35bf28}+0.72\%$
test_compile_indexing[slice-tensorclass-compile]	0.2004ms	0.1401ms	7.1398 KOps/s	7.2378 KOps/s	$\color{#d91a1a}-1.35\%$
test_compile_indexing[slice-tensorclass-eager]	60.8410μs	21.3058μs	46.9356 KOps/s	51.0325 KOps/s	$\textbf{\color{#d91a1a}-8.03\%}$
test_compile_indexing[slice-pytree-compile]	0.1992ms	0.1454ms	6.8767 KOps/s	7.2018 KOps/s	$\color{#d91a1a}-4.51\%$
test_compile_indexing[slice-pytree-eager]	60.3110μs	20.8463μs	47.9702 KOps/s	51.7000 KOps/s	$\textbf{\color{#d91a1a}-7.21\%}$
test_compile_indexing[int-tensordict-compile]	0.2397ms	0.1449ms	6.9022 KOps/s	6.9607 KOps/s	$\color{#d91a1a}-0.84\%$
test_compile_indexing[int-tensordict-eager]	0.4889ms	23.5764μs	42.4152 KOps/s	41.2442 KOps/s	$\color{#35bf28}+2.84\%$
test_compile_indexing[int-tensorclass-compile]	0.2065ms	0.1394ms	7.1719 KOps/s	7.1801 KOps/s	$\color{#d91a1a}-0.12\%$
test_compile_indexing[int-tensorclass-eager]	65.8410μs	19.7345μs	50.6726 KOps/s	49.7924 KOps/s	$\color{#35bf28}+1.77\%$
test_compile_indexing[int-pytree-compile]	0.1934ms	0.1394ms	7.1736 KOps/s	7.1890 KOps/s	$\color{#d91a1a}-0.21\%$
test_compile_indexing[int-pytree-eager]	98.4820μs	19.9576μs	50.1062 KOps/s	50.6246 KOps/s	$\color{#d91a1a}-1.02\%$
test_mod_add[eager]	74.5510μs	33.1648μs	30.1525 KOps/s	29.6786 KOps/s	$\color{#35bf28}+1.60\%$
test_mod_add[compile]	0.1943ms	82.3029μs	12.1502 KOps/s	12.6656 KOps/s	$\color{#d91a1a}-4.07\%$
test_mod_add[compile-overhead]	0.2974ms	0.1493ms	6.6962 KOps/s	6.0980 KOps/s	$\textbf{\color{#35bf28}+9.81\%}$
test_mod_wrap[eager]	0.3222ms	0.2458ms	4.0689 KOps/s	3.9105 KOps/s	$\color{#35bf28}+4.05\%$
test_mod_wrap[compile]	1.3769ms	0.2910ms	3.4366 KOps/s	3.3769 KOps/s	$\color{#35bf28}+1.77\%$
test_mod_wrap[compile-overhead]	7.6847ms	4.1189ms	242.7832 Ops/s	247.2398 Ops/s	$\color{#d91a1a}-1.80\%$
test_mod_wrap_and_backward[eager]	1.4687ms	1.3339ms	749.6822 Ops/s	692.7018 Ops/s	$\textbf{\color{#35bf28}+8.23\%}$
test_mod_wrap_and_backward[compile]	1.5726ms	1.3083ms	764.3560 Ops/s	701.2890 Ops/s	$\textbf{\color{#35bf28}+8.99\%}$
test_mod_wrap_and_backward[compile-overhead]	1.3298ms	0.8879ms	1.1262 KOps/s	992.3240 Ops/s	$\textbf{\color{#35bf28}+13.49\%}$
test_seq_add[eager]	0.1466ms	97.1126μs	10.2973 KOps/s	9.8690 KOps/s	$\color{#35bf28}+4.34\%$
test_seq_add[compile]	0.1326ms	90.6671μs	11.0294 KOps/s	11.1978 KOps/s	$\color{#d91a1a}-1.50\%$
test_seq_add[compile-overhead]	0.1655ms	0.1223ms	8.1771 KOps/s	8.1551 KOps/s	$\color{#35bf28}+0.27\%$
test_seq_wrap[eager]	0.5148ms	0.3949ms	2.5322 KOps/s	2.5373 KOps/s	$\color{#d91a1a}-0.20\%$
test_seq_wrap[compile]	0.3548ms	0.3065ms	3.2631 KOps/s	3.1169 KOps/s	$\color{#35bf28}+4.69\%$
test_seq_wrap[compile-overhead]	0.2693ms	0.2221ms	4.5026 KOps/s	4.6016 KOps/s	$\color{#d91a1a}-2.15\%$
test_func_call_runtime[False-eager]	0.7841ms	0.7194ms	1.3900 KOps/s	1.2983 KOps/s	$\textbf{\color{#35bf28}+7.07\%}$
test_func_call_runtime[False-compile]	0.9784ms	0.7862ms	1.2719 KOps/s	1.2786 KOps/s	$\color{#d91a1a}-0.52\%$
test_func_call_runtime[False-compile-overhead]	0.4893ms	0.3556ms	2.8121 KOps/s	2.8095 KOps/s	$\color{#35bf28}+0.09\%$
test_func_call_runtime[True-eager]	0.9905ms	0.8845ms	1.1306 KOps/s	1.1066 KOps/s	$\color{#35bf28}+2.17\%$
test_func_call_runtime[True-compile]	0.8582ms	0.7973ms	1.2543 KOps/s	1.2349 KOps/s	$\color{#35bf28}+1.57\%$
test_func_call_runtime[True-compile-overhead]	0.4438ms	0.3770ms	2.6524 KOps/s	2.6812 KOps/s	$\color{#d91a1a}-1.08\%$
test_func_call_cm_runtime[False-eager]	0.9115ms	0.7172ms	1.3943 KOps/s	1.3577 KOps/s	$\color{#35bf28}+2.70\%$
test_func_call_cm_runtime[False-compile]	0.8523ms	0.7712ms	1.2967 KOps/s	1.2774 KOps/s	$\color{#35bf28}+1.51\%$
test_func_call_cm_runtime[False-compile-overhead]	0.4057ms	0.3574ms	2.7981 KOps/s	2.8008 KOps/s	$\color{#d91a1a}-0.09\%$
test_func_call_cm_runtime[True-eager]	1.0688ms	0.9862ms	1.0140 KOps/s	988.8272 Ops/s	$\color{#35bf28}+2.54\%$
test_func_call_cm_runtime[True-compile]	0.9036ms	0.8253ms	1.2116 KOps/s	1.2024 KOps/s	$\color{#35bf28}+0.77\%$
test_func_call_cm_runtime[True-compile-overhead]	0.4475ms	0.4015ms	2.4905 KOps/s	2.4796 KOps/s	$\color{#35bf28}+0.44\%$
test_vmap_func_call_cm_runtime[eager]	2.5650ms	2.1025ms	475.6268 Ops/s	469.5910 Ops/s	$\color{#35bf28}+1.29\%$
test_vmap_func_call_cm_runtime[compile]	0.9261ms	0.8369ms	1.1949 KOps/s	1.1813 KOps/s	$\color{#35bf28}+1.15\%$
test_vmap_func_call_cm_runtime[compile-overhead]	0.4584ms	0.4075ms	2.4538 KOps/s	2.4692 KOps/s	$\color{#d91a1a}-0.62\%$
test_distributed	2.3598ms	0.1720ms	5.8129 KOps/s	8.7610 KOps/s	$\textbf{\color{#d91a1a}-33.65\%}$
test_tdmodule	36.1510μs	15.8888μs	62.9375 KOps/s	60.7466 KOps/s	$\color{#35bf28}+3.61\%$
test_tdmodule_dispatch	60.3910μs	31.6683μs	31.5773 KOps/s	31.9786 KOps/s	$\color{#d91a1a}-1.25\%$
test_tdseq	49.9900μs	16.8165μs	59.4653 KOps/s	59.1002 KOps/s	$\color{#35bf28}+0.62\%$
test_tdseq_dispatch	57.1210μs	33.9812μs	29.4280 KOps/s	29.1847 KOps/s	$\color{#35bf28}+0.83\%$
test_instantiation_functorch	1.9780ms	1.8246ms	548.0543 Ops/s	549.0712 Ops/s	$\color{#d91a1a}-0.19\%$
test_exec_functorch	0.2646ms	0.2034ms	4.9169 KOps/s	4.7975 KOps/s	$\color{#35bf28}+2.49\%$
test_exec_functional_call	0.2757ms	0.2082ms	4.8039 KOps/s	4.8082 KOps/s	$\color{#d91a1a}-0.09\%$
test_exec_td_decorator	0.4564ms	0.2630ms	3.8017 KOps/s	3.8531 KOps/s	$\color{#d91a1a}-1.34\%$
test_vmap_mlp_speed_decorator[True-True]	0.8333ms	0.6892ms	1.4510 KOps/s	1.4452 KOps/s	$\color{#35bf28}+0.40\%$
test_vmap_mlp_speed_decorator[True-False]	0.8160ms	0.6864ms	1.4569 KOps/s	1.4476 KOps/s	$\color{#35bf28}+0.65\%$
test_vmap_mlp_speed_decorator[False-True]	0.7385ms	0.6070ms	1.6474 KOps/s	1.6597 KOps/s	$\color{#d91a1a}-0.74\%$
test_vmap_mlp_speed_decorator[False-False]	0.7040ms	0.6078ms	1.6454 KOps/s	1.6408 KOps/s	$\color{#35bf28}+0.28\%$
test_vmap_transformer_speed_decorator[True-True]	20.6510ms	19.8679ms	50.3324 Ops/s	50.7021 Ops/s	$\color{#d91a1a}-0.73\%$
test_vmap_transformer_speed_decorator[True-False]	20.5037ms	19.7164ms	50.7191 Ops/s	50.2595 Ops/s	$\color{#35bf28}+0.91\%$
test_vmap_transformer_speed_decorator[False-True]	20.5028ms	19.8369ms	50.4111 Ops/s	51.0525 Ops/s	$\color{#d91a1a}-1.26\%$
test_vmap_transformer_speed_decorator[False-False]	20.9611ms	19.7388ms	50.6616 Ops/s	50.8175 Ops/s	$\color{#d91a1a}-0.31\%$
test_to_module_speed[True]	1.4442ms	0.9878ms	1.0123 KOps/s	1.0095 KOps/s	$\color{#35bf28}+0.28\%$
test_to_module_speed[False]	1.4271ms	0.9685ms	1.0326 KOps/s	1.0460 KOps/s	$\color{#d91a1a}-1.29\%$
test_tc_init	76.1620μs	37.6332μs	26.5723 KOps/s	26.1944 KOps/s	$\color{#35bf28}+1.44\%$
test_tc_init_nested	0.2014ms	74.0647μs	13.5017 KOps/s	12.9097 KOps/s	$\color{#35bf28}+4.59\%$
test_tc_first_layer_tensor	4.9300μs	0.6617μs	1.5112 MOps/s	1.5082 MOps/s	$\color{#35bf28}+0.20\%$
test_tc_first_layer_nontensor	42.6310μs	2.1702μs	460.7816 KOps/s	459.7175 KOps/s	$\color{#35bf28}+0.23\%$
test_tc_second_layer_tensor	9.6477μs	1.3201μs	757.4973 KOps/s	742.2117 KOps/s	$\color{#35bf28}+2.06\%$
test_tc_second_layer_nontensor	24.1110μs	2.8740μs	347.9504 KOps/s	348.5403 KOps/s	$\color{#d91a1a}-0.17\%$
test_unbind	0.1941s	9.5224ms	105.0161 Ops/s	92.7657 Ops/s	$\textbf{\color{#35bf28}+13.21\%}$
test_full_like	0.6638ms	0.5744ms	1.7411 KOps/s	1.7450 KOps/s	$\color{#d91a1a}-0.22\%$
test_zeros_like	0.2628ms	0.1980ms	5.0517 KOps/s	5.0463 KOps/s	$\color{#35bf28}+0.11\%$
test_ones_like	0.2332ms	0.1978ms	5.0553 KOps/s	5.0501 KOps/s	$\color{#35bf28}+0.10\%$
test_clone	0.4430ms	0.4147ms	2.4113 KOps/s	2.4101 KOps/s	$\color{#35bf28}+0.05\%$
test_squeeze	44.7600μs	9.5882μs	104.2947 KOps/s	106.6937 KOps/s	$\color{#d91a1a}-2.25\%$
test_unsqueeze	0.2169ms	73.8036μs	13.5495 KOps/s	13.8298 KOps/s	$\color{#d91a1a}-2.03\%$
test_split	0.4195ms	0.1513ms	6.6097 KOps/s	6.5958 KOps/s	$\color{#35bf28}+0.21\%$
test_permute	0.2893ms	0.1799ms	5.5586 KOps/s	5.6511 KOps/s	$\color{#d91a1a}-1.64\%$
test_stack	1.2568ms	0.8294ms	1.2057 KOps/s	1.1538 KOps/s	$\color{#35bf28}+4.50\%$
test_cat	1.2624ms	1.2313ms	812.1598 Ops/s	812.0162 Ops/s	$\color{#35bf28}+0.02\%$