[Performance] Minor efficiency improvements

[vmoens]

No description provided.

[github-actions]

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

Total Benchmarks: 126. Improved: $\large\color{#35bf28}4$. Worsened: $\large\color{#d91a1a}28$.

Expand to view detailed results

Name	Max	Mean	Ops	Ops on Repo HEAD	Change
test_plain_set_nested	53.2790μs	16.7948μs	59.5421 KOps/s	59.3768 KOps/s	$\color{#35bf28}+0.28\%$
test_plain_set_stack_nested	49.0590μs	16.8181μs	59.4598 KOps/s	59.8188 KOps/s	$\color{#d91a1a}-0.60\%$
test_plain_set_nested_inplace	46.6270μs	19.1037μs	52.3457 KOps/s	52.5112 KOps/s	$\color{#d91a1a}-0.32\%$
test_plain_set_stack_nested_inplace	59.3610μs	19.0214μs	52.5724 KOps/s	52.5418 KOps/s	$\color{#35bf28}+0.06\%$
test_items	34.4040μs	2.3238μs	430.3339 KOps/s	429.5628 KOps/s	$\color{#35bf28}+0.18\%$
test_items_nested	0.4760ms	0.2699ms	3.7055 KOps/s	3.7804 KOps/s	$\color{#d91a1a}-1.98\%$
test_items_nested_locked	0.3463ms	0.2700ms	3.7035 KOps/s	3.7906 KOps/s	$\color{#d91a1a}-2.30\%$
test_items_nested_leaf	0.3395ms	0.1654ms	6.0472 KOps/s	6.1905 KOps/s	$\color{#d91a1a}-2.31\%$
test_items_stack_nested	0.3429ms	0.2711ms	3.6884 KOps/s	3.8277 KOps/s	$\color{#d91a1a}-3.64\%$
test_items_stack_nested_leaf	0.3059ms	0.1663ms	6.0124 KOps/s	6.0949 KOps/s	$\color{#d91a1a}-1.35\%$
test_items_stack_nested_locked	0.3726ms	0.2712ms	3.6874 KOps/s	3.8388 KOps/s	$\color{#d91a1a}-3.94\%$
test_keys	29.3440μs	3.7209μs	268.7534 KOps/s	267.0487 KOps/s	$\color{#35bf28}+0.64\%$
test_keys_nested	0.8200ms	0.1439ms	6.9501 KOps/s	6.9868 KOps/s	$\color{#d91a1a}-0.53\%$
test_keys_nested_locked	0.2412ms	0.1464ms	6.8286 KOps/s	6.7330 KOps/s	$\color{#35bf28}+1.42\%$
test_keys_nested_leaf	37.2150ms	0.1300ms	7.6947 KOps/s	8.0764 KOps/s	$\color{#d91a1a}-4.73\%$
test_keys_stack_nested	0.2113ms	0.1437ms	6.9614 KOps/s	6.8573 KOps/s	$\color{#35bf28}+1.52\%$
test_keys_stack_nested_leaf	0.2240ms	0.1275ms	7.8420 KOps/s	7.7884 KOps/s	$\color{#35bf28}+0.69\%$
test_keys_stack_nested_locked	0.2172ms	0.1533ms	6.5247 KOps/s	6.7255 KOps/s	$\color{#d91a1a}-2.99\%$
test_values	6.8446μs	1.1218μs	891.4559 KOps/s	880.9146 KOps/s	$\color{#35bf28}+1.20\%$
test_values_nested	96.0890μs	50.8660μs	19.6595 KOps/s	20.2135 KOps/s	$\color{#d91a1a}-2.74\%$
test_values_nested_locked	95.1760μs	50.6384μs	19.7478 KOps/s	19.6699 KOps/s	$\color{#35bf28}+0.40\%$
test_values_nested_leaf	84.9880μs	45.7654μs	21.8506 KOps/s	22.0703 KOps/s	$\color{#d91a1a}-1.00\%$
test_values_stack_nested	0.1005ms	51.0894μs	19.5735 KOps/s	19.9743 KOps/s	$\color{#d91a1a}-2.01\%$
test_values_stack_nested_leaf	93.1330μs	45.5138μs	21.9714 KOps/s	22.3772 KOps/s	$\color{#d91a1a}-1.81\%$
test_values_stack_nested_locked	0.1183ms	51.7904μs	19.3086 KOps/s	19.5553 KOps/s	$\color{#d91a1a}-1.26\%$
test_membership	28.9730μs	1.3292μs	752.3521 KOps/s	769.7910 KOps/s	$\color{#d91a1a}-2.27\%$
test_membership_nested	27.7020μs	3.3711μs	296.6392 KOps/s	291.9265 KOps/s	$\color{#35bf28}+1.61\%$
test_membership_nested_leaf	34.6040μs	3.4143μs	292.8817 KOps/s	289.8442 KOps/s	$\color{#35bf28}+1.05\%$
test_membership_stacked_nested	26.9800μs	3.4044μs	293.7408 KOps/s	241.0081 KOps/s	$\textbf{\color{#35bf28}+21.88\%}$
test_membership_stacked_nested_leaf	26.1180μs	3.3952μs	294.5316 KOps/s	283.1872 KOps/s	$\color{#35bf28}+4.01\%$
test_membership_nested_last	33.5620μs	4.1875μs	238.8080 KOps/s	233.0439 KOps/s	$\color{#35bf28}+2.47\%$
test_membership_nested_leaf_last	36.0870μs	4.2018μs	237.9932 KOps/s	235.0371 KOps/s	$\color{#35bf28}+1.26\%$
test_membership_stacked_nested_last	23.5830μs	4.1911μs	238.6009 KOps/s	240.3064 KOps/s	$\color{#d91a1a}-0.71\%$
test_membership_stacked_nested_leaf_last	32.7110μs	4.1781μs	239.3413 KOps/s	237.6010 KOps/s	$\color{#35bf28}+0.73\%$
test_nested_getleaf	47.2370μs	10.5358μs	94.9145 KOps/s	97.3747 KOps/s	$\color{#d91a1a}-2.53\%$
test_nested_get	41.0060μs	9.8264μs	101.7662 KOps/s	100.5304 KOps/s	$\color{#35bf28}+1.23\%$
test_stacked_getleaf	51.7540μs	10.9491μs	91.3314 KOps/s	94.1009 KOps/s	$\color{#d91a1a}-2.94\%$
test_stacked_get	27.3410μs	9.9294μs	100.7115 KOps/s	103.9720 KOps/s	$\color{#d91a1a}-3.14\%$
test_nested_getitemleaf	35.4360μs	10.9399μs	91.4085 KOps/s	91.2843 KOps/s	$\color{#35bf28}+0.14\%$
test_nested_getitem	30.7370μs	10.4969μs	95.2663 KOps/s	96.6390 KOps/s	$\color{#d91a1a}-1.42\%$
test_stacked_getitemleaf	30.7770μs	10.9587μs	91.2518 KOps/s	91.8325 KOps/s	$\color{#d91a1a}-0.63\%$
test_stacked_getitem	45.3540μs	10.3934μs	96.2151 KOps/s	96.8220 KOps/s	$\color{#d91a1a}-0.63\%$
test_lock_nested	0.7058ms	0.3246ms	3.0808 KOps/s	3.0284 KOps/s	$\color{#35bf28}+1.73\%$
test_lock_stack_nested	0.4135ms	0.2957ms	3.3815 KOps/s	3.3492 KOps/s	$\color{#35bf28}+0.96\%$
test_unlock_nested	74.9622ms	0.4062ms	2.4621 KOps/s	2.4357 KOps/s	$\color{#35bf28}+1.08\%$
test_unlock_stack_nested	0.5010ms	0.3047ms	3.2814 KOps/s	3.2523 KOps/s	$\color{#35bf28}+0.89\%$
test_flatten_speed	0.5951ms	0.2832ms	3.5315 KOps/s	3.6646 KOps/s	$\color{#d91a1a}-3.63\%$
test_unflatten_speed	0.6101ms	0.4025ms	2.4846 KOps/s	2.5987 KOps/s	$\color{#d91a1a}-4.39\%$
test_common_ops	4.9187ms	0.7067ms	1.4150 KOps/s	1.4967 KOps/s	$\textbf{\color{#d91a1a}-5.46\%}$
test_creation	16.3200μs	1.7963μs	556.7074 KOps/s	570.5859 KOps/s	$\color{#d91a1a}-2.43\%$
test_creation_empty	48.0990μs	10.9198μs	91.5770 KOps/s	105.6021 KOps/s	$\textbf{\color{#d91a1a}-13.28\%}$
test_creation_nested_1	36.2770μs	13.4935μs	74.1098 KOps/s	80.7965 KOps/s	$\textbf{\color{#d91a1a}-8.28\%}$
test_creation_nested_2	44.8630μs	16.5979μs	60.2487 KOps/s	63.7997 KOps/s	$\textbf{\color{#d91a1a}-5.57\%}$
test_clone	59.5810μs	12.7096μs	78.6806 KOps/s	76.9423 KOps/s	$\color{#35bf28}+2.26\%$
test_getitem[int]	37.7200μs	10.6934μs	93.5153 KOps/s	90.8638 KOps/s	$\color{#35bf28}+2.92\%$
test_getitem[slice_int]	60.3920μs	22.0614μs	45.3280 KOps/s	44.9791 KOps/s	$\color{#35bf28}+0.78\%$
test_getitem[range]	0.1532ms	43.4105μs	23.0359 KOps/s	24.1646 KOps/s	$\color{#d91a1a}-4.67\%$
test_getitem[tuple]	49.3120μs	18.0811μs	55.3065 KOps/s	53.9962 KOps/s	$\color{#35bf28}+2.43\%$
test_getitem[list]	0.1613ms	37.4906μs	26.6734 KOps/s	27.8949 KOps/s	$\color{#d91a1a}-4.38\%$
test_setitem_dim[int]	67.1940μs	36.2458μs	27.5894 KOps/s	31.1813 KOps/s	$\textbf{\color{#d91a1a}-11.52\%}$
test_setitem_dim[slice_int]	0.1030ms	63.1862μs	15.8262 KOps/s	17.4055 KOps/s	$\textbf{\color{#d91a1a}-9.07\%}$
test_setitem_dim[range]	0.1147ms	83.2869μs	12.0067 KOps/s	12.6599 KOps/s	$\textbf{\color{#d91a1a}-5.16\%}$
test_setitem_dim[tuple]	87.0320μs	51.5092μs	19.4140 KOps/s	20.3265 KOps/s	$\color{#d91a1a}-4.49\%$
test_setitem	73.4970μs	20.0347μs	49.9133 KOps/s	53.5022 KOps/s	$\textbf{\color{#d91a1a}-6.71\%}$
test_set	70.9820μs	19.4956μs	51.2935 KOps/s	55.0577 KOps/s	$\textbf{\color{#d91a1a}-6.84\%}$
test_set_shared	3.4358ms	0.1388ms	7.2039 KOps/s	7.1631 KOps/s	$\color{#35bf28}+0.57\%$
test_update	87.7030μs	22.8901μs	43.6869 KOps/s	46.3136 KOps/s	$\textbf{\color{#d91a1a}-5.67\%}$
test_update_nested	84.5670μs	29.7726μs	33.5879 KOps/s	35.6086 KOps/s	$\textbf{\color{#d91a1a}-5.67\%}$
test_set_nested	83.8750μs	21.5288μs	46.4494 KOps/s	49.7746 KOps/s	$\textbf{\color{#d91a1a}-6.68\%}$
test_set_nested_new	60.9830μs	24.8120μs	40.3031 KOps/s	42.3334 KOps/s	$\color{#d91a1a}-4.80\%$
test_select	86.7010μs	39.1428μs	25.5475 KOps/s	26.8230 KOps/s	$\color{#d91a1a}-4.76\%$
test_select_nested	0.1586ms	58.1795μs	17.1882 KOps/s	17.4385 KOps/s	$\color{#d91a1a}-1.44\%$
test_exclude_nested	0.2672ms	0.1168ms	8.5590 KOps/s	8.8690 KOps/s	$\color{#d91a1a}-3.49\%$
test_empty[True]	0.7846ms	0.4040ms	2.4754 KOps/s	2.4846 KOps/s	$\color{#d91a1a}-0.37\%$
test_empty[False]	5.0374μs	1.0205μs	979.9268 KOps/s	1.0069 MOps/s	$\color{#d91a1a}-2.68\%$
test_unbind_speed	0.5116ms	0.2383ms	4.1959 KOps/s	4.1918 KOps/s	$\color{#35bf28}+0.10\%$
test_unbind_speed_stack0	0.3919ms	0.2385ms	4.1926 KOps/s	4.1322 KOps/s	$\color{#35bf28}+1.46\%$
test_unbind_speed_stack1	0.1211s	0.6744ms	1.4827 KOps/s	1.4973 KOps/s	$\color{#d91a1a}-0.97\%$
test_split	0.1052s	1.5880ms	629.7209 Ops/s	629.3406 Ops/s	$\color{#35bf28}+0.06\%$
test_chunk	1.5112ms	1.3812ms	723.9841 Ops/s	714.4765 Ops/s	$\color{#35bf28}+1.33\%$
test_creation[device0]	3.3523ms	0.1024ms	9.7633 KOps/s	9.8519 KOps/s	$\color{#d91a1a}-0.90\%$
test_creation_from_tensor	0.1711ms	80.5514μs	12.4144 KOps/s	12.2989 KOps/s	$\color{#35bf28}+0.94\%$
test_add_one[memmap_tensor0]	92.6520μs	5.2961μs	188.8179 KOps/s	188.1702 KOps/s	$\color{#35bf28}+0.34\%$
test_contiguous[memmap_tensor0]	10.8910μs	0.6134μs	1.6302 MOps/s	1.6020 MOps/s	$\color{#35bf28}+1.76\%$
test_stack[memmap_tensor0]	21.0890μs	3.4342μs	291.1913 KOps/s	269.0497 KOps/s	$\textbf{\color{#35bf28}+8.23\%}$
test_memmaptd_index	1.1414ms	0.2298ms	4.3517 KOps/s	4.1371 KOps/s	$\textbf{\color{#35bf28}+5.19\%}$
test_memmaptd_index_astensor	0.6284ms	0.2922ms	3.4218 KOps/s	3.3254 KOps/s	$\color{#35bf28}+2.90\%$
test_memmaptd_index_op	1.1774ms	0.5923ms	1.6884 KOps/s	1.7039 KOps/s	$\color{#d91a1a}-0.91\%$
test_serialize_model	0.2081s	0.1115s	8.9663 Ops/s	8.8171 Ops/s	$\color{#35bf28}+1.69\%$
test_serialize_model_pickle	0.4524s	0.3753s	2.6643 Ops/s	2.6335 Ops/s	$\color{#35bf28}+1.17\%$
test_serialize_weights	0.1085s	0.1003s	9.9728 Ops/s	10.2583 Ops/s	$\color{#d91a1a}-2.78\%$
test_serialize_weights_returnearly	0.2344s	0.1314s	7.6086 Ops/s	7.2721 Ops/s	$\color{#35bf28}+4.63\%$
test_serialize_weights_pickle	1.0109s	0.5860s	1.7065 Ops/s	2.4168 Ops/s	$\textbf{\color{#d91a1a}-29.39\%}$
test_serialize_weights_filesystem	0.1019s	91.5194ms	10.9266 Ops/s	10.6056 Ops/s	$\color{#35bf28}+3.03\%$
test_serialize_model_filesystem	95.2376ms	89.9942ms	11.1118 Ops/s	10.7953 Ops/s	$\color{#35bf28}+2.93\%$
test_reshape_pytree	45.2840μs	20.6860μs	48.3420 KOps/s	50.3734 KOps/s	$\color{#d91a1a}-4.03\%$
test_reshape_td	81.8830μs	30.8450μs	32.4202 KOps/s	33.2837 KOps/s	$\color{#d91a1a}-2.59\%$
test_view_pytree	67.8360μs	20.6098μs	48.5206 KOps/s	50.1535 KOps/s	$\color{#d91a1a}-3.26\%$
test_view_td	0.1202s	58.1472μs	17.1977 KOps/s	17.9954 KOps/s	$\color{#d91a1a}-4.43\%$
test_unbind_pytree	71.5730μs	24.2345μs	41.2635 KOps/s	44.7255 KOps/s	$\textbf{\color{#d91a1a}-7.74\%}$
test_unbind_td	0.4184ms	35.6790μs	28.0277 KOps/s	26.5691 KOps/s	$\textbf{\color{#35bf28}+5.49\%}$
test_split_pytree	69.3490μs	23.8176μs	41.9857 KOps/s	43.5307 KOps/s	$\color{#d91a1a}-3.55\%$
test_split_td	0.1480ms	38.9018μs	25.7058 KOps/s	26.1930 KOps/s	$\color{#d91a1a}-1.86\%$
test_add_pytree	66.8140μs	29.5203μs	33.8750 KOps/s	35.5674 KOps/s	$\color{#d91a1a}-4.76\%$
test_add_td	0.1644ms	56.0561μs	17.8393 KOps/s	19.6850 KOps/s	$\textbf{\color{#d91a1a}-9.38\%}$
test_distributed	0.3782ms	98.5473μs	10.1474 KOps/s	10.3650 KOps/s	$\color{#d91a1a}-2.10\%$
test_tdmodule	68.4770μs	18.1236μs	55.1766 KOps/s	58.1555 KOps/s	$\textbf{\color{#d91a1a}-5.12\%}$
test_tdmodule_dispatch	60.7220μs	34.1705μs	29.2650 KOps/s	31.7089 KOps/s	$\textbf{\color{#d91a1a}-7.71\%}$
test_tdseq	49.4720μs	20.8243μs	48.0209 KOps/s	52.0778 KOps/s	$\textbf{\color{#d91a1a}-7.79\%}$
test_tdseq_dispatch	72.6850μs	39.7685μs	25.1455 KOps/s	26.4804 KOps/s	$\textbf{\color{#d91a1a}-5.04\%}$
test_instantiation_functorch	2.0346ms	1.3031ms	767.4175 Ops/s	790.6928 Ops/s	$\color{#d91a1a}-2.94\%$
test_instantiation_td	6.3793ms	0.9908ms	1.0093 KOps/s	1.0459 KOps/s	$\color{#d91a1a}-3.50\%$
test_exec_functorch	0.2480ms	0.1574ms	6.3537 KOps/s	6.4934 KOps/s	$\color{#d91a1a}-2.15\%$
test_exec_functional_call	0.2908ms	0.1496ms	6.6836 KOps/s	7.2649 KOps/s	$\textbf{\color{#d91a1a}-8.00\%}$
test_exec_td	0.3435ms	0.1415ms	7.0648 KOps/s	7.3531 KOps/s	$\color{#d91a1a}-3.92\%$
test_exec_td_decorator	0.5375ms	0.1941ms	5.1531 KOps/s	5.4406 KOps/s	$\textbf{\color{#d91a1a}-5.28\%}$
test_vmap_mlp_speed[True-True]	0.6527ms	0.4758ms	2.1018 KOps/s	2.2168 KOps/s	$\textbf{\color{#d91a1a}-5.19\%}$
test_vmap_mlp_speed[True-False]	0.8998ms	0.4759ms	2.1012 KOps/s	2.2206 KOps/s	$\textbf{\color{#d91a1a}-5.38\%}$
test_vmap_mlp_speed[False-True]	2.8308ms	0.3911ms	2.5568 KOps/s	2.7174 KOps/s	$\textbf{\color{#d91a1a}-5.91\%}$
test_vmap_mlp_speed[False-False]	0.6258ms	0.3854ms	2.5944 KOps/s	2.7411 KOps/s	$\textbf{\color{#d91a1a}-5.35\%}$
test_vmap_mlp_speed_decorator[True-True]	0.9554ms	0.4963ms	2.0147 KOps/s	2.1279 KOps/s	$\textbf{\color{#d91a1a}-5.32\%}$
test_vmap_mlp_speed_decorator[True-False]	0.7598ms	0.4954ms	2.0187 KOps/s	2.1263 KOps/s	$\textbf{\color{#d91a1a}-5.06\%}$
test_vmap_mlp_speed_decorator[False-True]	0.5445ms	0.4004ms	2.4977 KOps/s	2.6137 KOps/s	$\color{#d91a1a}-4.44\%$
test_vmap_mlp_speed_decorator[False-False]	0.7839ms	0.4021ms	2.4871 KOps/s	2.6271 KOps/s	$\textbf{\color{#d91a1a}-5.33\%}$
test_to_module_speed[True]	2.1659ms	1.3673ms	731.3583 Ops/s	756.6089 Ops/s	$\color{#d91a1a}-3.34\%$
test_to_module_speed[False]	1.8075ms	1.3624ms	733.9827 Ops/s	746.4541 Ops/s	$\color{#d91a1a}-1.67\%$