[Feature] Compile - tensorclass compatibility

[vmoens]

Stack from ghstack (oldest at bottom):

• [Feature] Compile - benchmarks #883
• -> [Feature] Compile - tensorclass compatibility #882

[github-actions]

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

Total Benchmarks: 133. Improved: $\large\color{#35bf28}4$. Worsened: $\large\color{#d91a1a}14$.

Expand to view detailed results

Name	Max	Mean	Ops	Ops on Repo HEAD	Change
test_plain_set_nested	41.1170μs	18.5890μs	53.7952 KOps/s	55.4261 KOps/s	$\color{#d91a1a}-2.94\%$
test_plain_set_stack_nested	47.0280μs	18.9038μs	52.8994 KOps/s	50.4211 KOps/s	$\color{#35bf28}+4.92\%$
test_plain_set_nested_inplace	67.2750μs	20.6969μs	48.3165 KOps/s	49.1865 KOps/s	$\color{#d91a1a}-1.77\%$
test_plain_set_stack_nested_inplace	72.6140μs	20.5495μs	48.6629 KOps/s	49.6198 KOps/s	$\color{#d91a1a}-1.93\%$
test_items	16.9610μs	2.6396μs	378.8460 KOps/s	392.0118 KOps/s	$\color{#d91a1a}-3.36\%$
test_items_nested	2.2345ms	0.3730ms	2.6809 KOps/s	2.6540 KOps/s	$\color{#35bf28}+1.02\%$
test_items_nested_locked	0.5305ms	0.3706ms	2.6981 KOps/s	2.7264 KOps/s	$\color{#d91a1a}-1.04\%$
test_items_nested_leaf	0.1603ms	85.7794μs	11.6578 KOps/s	11.5462 KOps/s	$\color{#35bf28}+0.97\%$
test_items_stack_nested	0.5487ms	0.3754ms	2.6640 KOps/s	2.7268 KOps/s	$\color{#d91a1a}-2.30\%$
test_items_stack_nested_leaf	0.1605ms	87.0092μs	11.4930 KOps/s	11.6208 KOps/s	$\color{#d91a1a}-1.10\%$
test_items_stack_nested_locked	0.8169ms	0.3790ms	2.6385 KOps/s	2.7395 KOps/s	$\color{#d91a1a}-3.68\%$
test_keys	0.1105ms	4.2346μs	236.1494 KOps/s	217.2154 KOps/s	$\textbf{\color{#35bf28}+8.72\%}$
test_keys_nested	0.2506ms	0.1457ms	6.8618 KOps/s	6.9305 KOps/s	$\color{#d91a1a}-0.99\%$
test_keys_nested_locked	0.6740ms	0.1511ms	6.6198 KOps/s	6.5846 KOps/s	$\color{#35bf28}+0.53\%$
test_keys_nested_leaf	0.2213ms	0.1233ms	8.1131 KOps/s	8.1110 KOps/s	$\color{#35bf28}+0.03\%$
test_keys_stack_nested	0.2339ms	0.1448ms	6.9062 KOps/s	6.8977 KOps/s	$\color{#35bf28}+0.12\%$
test_keys_stack_nested_leaf	0.2229ms	0.1232ms	8.1146 KOps/s	8.1304 KOps/s	$\color{#d91a1a}-0.19\%$
test_keys_stack_nested_locked	0.4618ms	0.1503ms	6.6517 KOps/s	6.6377 KOps/s	$\color{#35bf28}+0.21\%$
test_values	10.6675μs	1.1817μs	846.2389 KOps/s	888.8857 KOps/s	$\color{#d91a1a}-4.80\%$
test_values_nested	0.1070ms	49.7063μs	20.1182 KOps/s	20.2426 KOps/s	$\color{#d91a1a}-0.61\%$
test_values_nested_locked	0.1154ms	49.2208μs	20.3166 KOps/s	19.8622 KOps/s	$\color{#35bf28}+2.29\%$
test_values_nested_leaf	0.1155ms	44.7747μs	22.3341 KOps/s	22.4624 KOps/s	$\color{#d91a1a}-0.57\%$
test_values_stack_nested	0.1022ms	51.2425μs	19.5151 KOps/s	20.2883 KOps/s	$\color{#d91a1a}-3.81\%$
test_values_stack_nested_leaf	85.9000μs	44.3935μs	22.5258 KOps/s	22.4965 KOps/s	$\color{#35bf28}+0.13\%$
test_values_stack_nested_locked	98.8240μs	51.4953μs	19.4193 KOps/s	20.3545 KOps/s	$\color{#d91a1a}-4.59\%$
test_membership	26.9700μs	0.8916μs	1.1215 MOps/s	1.3998 MOps/s	$\textbf{\color{#d91a1a}-19.88\%}$
test_membership_nested	30.8380μs	2.6648μs	375.2613 KOps/s	363.4403 KOps/s	$\color{#35bf28}+3.25\%$
test_membership_nested_leaf	21.0890μs	2.7166μs	368.1136 KOps/s	366.9289 KOps/s	$\color{#35bf28}+0.32\%$
test_membership_stacked_nested	23.7640μs	2.6530μs	376.9310 KOps/s	363.5982 KOps/s	$\color{#35bf28}+3.67\%$
test_membership_stacked_nested_leaf	18.0840μs	2.6939μs	371.2123 KOps/s	368.7125 KOps/s	$\color{#35bf28}+0.68\%$
test_membership_nested_last	36.2770μs	3.9581μs	252.6476 KOps/s	249.5506 KOps/s	$\color{#35bf28}+1.24\%$
test_membership_nested_leaf_last	19.7160μs	3.9881μs	250.7467 KOps/s	249.7668 KOps/s	$\color{#35bf28}+0.39\%$
test_membership_stacked_nested_last	31.2680μs	4.5466μs	219.9431 KOps/s	251.7478 KOps/s	$\textbf{\color{#d91a1a}-12.63\%}$
test_membership_stacked_nested_leaf_last	52.2970μs	4.5843μs	218.1373 KOps/s	249.1957 KOps/s	$\textbf{\color{#d91a1a}-12.46\%}$
test_nested_getleaf	42.0590μs	10.9642μs	91.2061 KOps/s	92.3531 KOps/s	$\color{#d91a1a}-1.24\%$
test_nested_get	50.1330μs	10.4249μs	95.9239 KOps/s	98.5080 KOps/s	$\color{#d91a1a}-2.62\%$
test_stacked_getleaf	36.5580μs	10.8997μs	91.7457 KOps/s	93.9805 KOps/s	$\color{#d91a1a}-2.38\%$
test_stacked_get	35.0760μs	10.2582μs	97.4831 KOps/s	98.9593 KOps/s	$\color{#d91a1a}-1.49\%$
test_nested_getitemleaf	30.4870μs	11.3080μs	88.4330 KOps/s	88.7168 KOps/s	$\color{#d91a1a}-0.32\%$
test_nested_getitem	0.1193ms	10.5966μs	94.3696 KOps/s	96.8241 KOps/s	$\color{#d91a1a}-2.54\%$
test_stacked_getitemleaf	32.7000μs	11.3130μs	88.3941 KOps/s	88.8381 KOps/s	$\color{#d91a1a}-0.50\%$
test_stacked_getitem	40.2950μs	10.3639μs	96.4889 KOps/s	96.7847 KOps/s	$\color{#d91a1a}-0.31\%$
test_lock_nested	0.8743ms	0.4633ms	2.1584 KOps/s	2.1353 KOps/s	$\color{#35bf28}+1.08\%$
test_lock_stack_nested	0.9225ms	0.4314ms	2.3183 KOps/s	2.3052 KOps/s	$\color{#35bf28}+0.57\%$
test_unlock_nested	0.8646ms	0.3837ms	2.6061 KOps/s	2.2068 KOps/s	$\textbf{\color{#35bf28}+18.09\%}$
test_unlock_stack_nested	0.6887ms	0.3438ms	2.9087 KOps/s	2.8587 KOps/s	$\color{#35bf28}+1.75\%$
test_flatten_speed	0.6324ms	0.1051ms	9.5174 KOps/s	9.5568 KOps/s	$\color{#d91a1a}-0.41\%$
test_unflatten_speed	1.0108ms	0.4433ms	2.2556 KOps/s	2.2891 KOps/s	$\color{#d91a1a}-1.46\%$
test_common_ops	4.8669ms	0.8314ms	1.2028 KOps/s	1.2428 KOps/s	$\color{#d91a1a}-3.22\%$
test_creation	0.1032ms	2.4176μs	413.6366 KOps/s	432.9047 KOps/s	$\color{#d91a1a}-4.45\%$
test_creation_empty	45.8260μs	13.2240μs	75.6202 KOps/s	78.9336 KOps/s	$\color{#d91a1a}-4.20\%$
test_creation_nested_1	58.1780μs	16.4242μs	60.8857 KOps/s	62.5851 KOps/s	$\color{#d91a1a}-2.72\%$
test_creation_nested_2	67.8560μs	20.2307μs	49.4298 KOps/s	51.1970 KOps/s	$\color{#d91a1a}-3.45\%$
test_clone	0.1071ms	13.0327μs	76.7302 KOps/s	74.7032 KOps/s	$\color{#35bf28}+2.71\%$
test_getitem[int]	42.6290μs	11.7089μs	85.4050 KOps/s	85.6745 KOps/s	$\color{#d91a1a}-0.31\%$
test_getitem[slice_int]	59.6210μs	24.1382μs	41.4280 KOps/s	42.4525 KOps/s	$\color{#d91a1a}-2.41\%$
test_getitem[range]	0.2802ms	47.2438μs	21.1668 KOps/s	22.3256 KOps/s	$\textbf{\color{#d91a1a}-5.19\%}$
test_getitem[tuple]	50.0130μs	19.8379μs	50.4086 KOps/s	51.9597 KOps/s	$\color{#d91a1a}-2.99\%$
test_getitem[list]	0.3610ms	41.1971μs	24.2735 KOps/s	24.8605 KOps/s	$\color{#d91a1a}-2.36\%$
test_setitem_dim[int]	86.5710μs	36.6883μs	27.2567 KOps/s	28.6899 KOps/s	$\color{#d91a1a}-5.00\%$
test_setitem_dim[slice_int]	0.1262ms	64.0246μs	15.6190 KOps/s	15.8952 KOps/s	$\color{#d91a1a}-1.74\%$
test_setitem_dim[range]	0.1300ms	84.1439μs	11.8844 KOps/s	12.0544 KOps/s	$\color{#d91a1a}-1.41\%$
test_setitem_dim[tuple]	87.9340μs	52.1819μs	19.1637 KOps/s	19.3380 KOps/s	$\color{#d91a1a}-0.90\%$
test_setitem	0.1175ms	21.0265μs	47.5591 KOps/s	46.9194 KOps/s	$\color{#35bf28}+1.36\%$
test_set	0.1265ms	20.6264μs	48.4815 KOps/s	48.4159 KOps/s	$\color{#35bf28}+0.14\%$
test_set_shared	1.6311ms	0.1673ms	5.9786 KOps/s	5.9238 KOps/s	$\color{#35bf28}+0.93\%$
test_update	0.1879ms	24.4710μs	40.8648 KOps/s	41.2466 KOps/s	$\color{#d91a1a}-0.93\%$
test_update_nested	0.1455ms	33.3237μs	30.0087 KOps/s	29.8121 KOps/s	$\color{#35bf28}+0.66\%$
test_update__nested	0.1106ms	25.1612μs	39.7437 KOps/s	39.5094 KOps/s	$\color{#35bf28}+0.59\%$
test_set_nested	0.1531ms	22.7857μs	43.8873 KOps/s	43.8483 KOps/s	$\color{#35bf28}+0.09\%$
test_set_nested_new	0.1343ms	27.5675μs	36.2747 KOps/s	36.6207 KOps/s	$\color{#d91a1a}-0.95\%$
test_select	0.1674ms	43.4993μs	22.9889 KOps/s	23.4945 KOps/s	$\color{#d91a1a}-2.15\%$
test_select_nested	0.1213ms	60.8845μs	16.4246 KOps/s	16.4609 KOps/s	$\color{#d91a1a}-0.22\%$
test_exclude_nested	0.1812ms	80.6000μs	12.4070 KOps/s	12.4216 KOps/s	$\color{#d91a1a}-0.12\%$
test_empty[True]	0.4670ms	0.3469ms	2.8827 KOps/s	2.9137 KOps/s	$\color{#d91a1a}-1.06\%$
test_empty[False]	7.4538μs	1.2636μs	791.4072 KOps/s	797.4495 KOps/s	$\color{#d91a1a}-0.76\%$
test_unbind_speed	0.5084ms	0.2823ms	3.5425 KOps/s	3.5615 KOps/s	$\color{#d91a1a}-0.53\%$
test_unbind_speed_stack0	0.4246ms	0.2717ms	3.6807 KOps/s	3.5543 KOps/s	$\color{#35bf28}+3.56\%$
test_unbind_speed_stack1	79.2189ms	0.7640ms	1.3090 KOps/s	1.2839 KOps/s	$\color{#35bf28}+1.95\%$
test_split	76.2580ms	1.6504ms	605.9219 Ops/s	670.2860 Ops/s	$\textbf{\color{#d91a1a}-9.60\%}$
test_chunk	77.6195ms	1.6599ms	602.4580 Ops/s	619.5921 Ops/s	$\color{#d91a1a}-2.77\%$
test_creation[device0]	0.2082ms	92.7146μs	10.7858 KOps/s	10.4681 KOps/s	$\color{#35bf28}+3.03\%$
test_creation_from_tensor	4.0686ms	95.6449μs	10.4553 KOps/s	10.2115 KOps/s	$\color{#35bf28}+2.39\%$
test_add_one[memmap_tensor0]	0.1795ms	5.5595μs	179.8716 KOps/s	182.8270 KOps/s	$\color{#d91a1a}-1.62\%$
test_contiguous[memmap_tensor0]	21.7510μs	0.6325μs	1.5810 MOps/s	1.5928 MOps/s	$\color{#d91a1a}-0.75\%$
test_stack[memmap_tensor0]	44.0820μs	3.6928μs	270.8004 KOps/s	276.1779 KOps/s	$\color{#d91a1a}-1.95\%$
test_memmaptd_index	1.0625ms	0.2645ms	3.7806 KOps/s	3.8825 KOps/s	$\color{#d91a1a}-2.62\%$
test_memmaptd_index_astensor	0.5856ms	0.3350ms	2.9847 KOps/s	3.0108 KOps/s	$\color{#d91a1a}-0.87\%$
test_memmaptd_index_op	0.9146ms	0.6499ms	1.5386 KOps/s	1.5548 KOps/s	$\color{#d91a1a}-1.04\%$
test_serialize_model	0.1275s	0.1223s	8.1759 Ops/s	7.1484 Ops/s	$\textbf{\color{#35bf28}+14.37\%}$
test_serialize_model_pickle	0.4453s	0.3901s	2.5633 Ops/s	2.4871 Ops/s	$\color{#35bf28}+3.06\%$
test_serialize_weights	0.1967s	0.1341s	7.4597 Ops/s	8.0480 Ops/s	$\textbf{\color{#d91a1a}-7.31\%}$
test_serialize_weights_returnearly	0.1862s	0.1699s	5.8855 Ops/s	5.6251 Ops/s	$\color{#35bf28}+4.63\%$
test_serialize_weights_pickle	0.4779s	0.4128s	2.4225 Ops/s	2.3751 Ops/s	$\color{#35bf28}+2.00\%$
test_serialize_weights_filesystem	0.1475s	0.1435s	6.9699 Ops/s	7.0671 Ops/s	$\color{#d91a1a}-1.38\%$
test_serialize_model_filesystem	0.1537s	0.1507s	6.6370 Ops/s	6.5676 Ops/s	$\color{#35bf28}+1.06\%$
test_reshape_pytree	95.1870μs	25.4823μs	39.2429 KOps/s	38.5837 KOps/s	$\color{#35bf28}+1.71\%$
test_reshape_td	0.1244ms	34.3210μs	29.1366 KOps/s	28.7214 KOps/s	$\color{#35bf28}+1.45\%$
test_view_pytree	76.5020μs	25.6813μs	38.9388 KOps/s	38.3590 KOps/s	$\color{#35bf28}+1.51\%$
test_view_td	91.1700μs	39.1584μs	25.5373 KOps/s	24.6672 KOps/s	$\color{#35bf28}+3.53\%$
test_unbind_pytree	0.1011ms	29.6094μs	33.7730 KOps/s	33.8963 KOps/s	$\color{#d91a1a}-0.36\%$
test_unbind_td	0.3591ms	41.2604μs	24.2363 KOps/s	24.1274 KOps/s	$\color{#35bf28}+0.45\%$
test_split_pytree	80.4100μs	29.7283μs	33.6380 KOps/s	33.8553 KOps/s	$\color{#d91a1a}-0.64\%$
test_split_td	0.5098ms	42.1016μs	23.7521 KOps/s	24.1775 KOps/s	$\color{#d91a1a}-1.76\%$
test_add_pytree	78.5760μs	35.3420μs	28.2950 KOps/s	28.4040 KOps/s	$\color{#d91a1a}-0.38\%$
test_add_td	0.1335ms	60.2769μs	16.5901 KOps/s	16.7226 KOps/s	$\color{#d91a1a}-0.79\%$
test_distributed	0.2660ms	0.1295ms	7.7208 KOps/s	7.4305 KOps/s	$\color{#35bf28}+3.91\%$
test_tdmodule	85.8800μs	17.8089μs	56.1516 KOps/s	57.2739 KOps/s	$\color{#d91a1a}-1.96\%$
test_tdmodule_dispatch	69.1790μs	38.2876μs	26.1181 KOps/s	27.8117 KOps/s	$\textbf{\color{#d91a1a}-6.09\%}$
test_tdseq	49.0720μs	19.9053μs	50.2378 KOps/s	51.2352 KOps/s	$\color{#d91a1a}-1.95\%$
test_tdseq_dispatch	71.5230μs	42.3573μs	23.6087 KOps/s	24.6007 KOps/s	$\color{#d91a1a}-4.03\%$
test_instantiation_functorch	2.0638ms	1.3264ms	753.9160 Ops/s	737.8022 Ops/s	$\color{#35bf28}+2.18\%$
test_instantiation_td	2.4848ms	1.0271ms	973.6399 Ops/s	887.9638 Ops/s	$\textbf{\color{#35bf28}+9.65\%}$
test_exec_functorch	0.3898ms	0.1718ms	5.8203 KOps/s	6.1583 KOps/s	$\textbf{\color{#d91a1a}-5.49\%}$
test_exec_functional_call	0.2894ms	0.1479ms	6.7600 KOps/s	6.4803 KOps/s	$\color{#35bf28}+4.32\%$
test_exec_td	0.2791ms	0.1531ms	6.5320 KOps/s	6.7685 KOps/s	$\color{#d91a1a}-3.49\%$
test_exec_td_decorator	0.6850ms	0.2356ms	4.2454 KOps/s	4.2881 KOps/s	$\color{#d91a1a}-1.00\%$
test_vmap_mlp_speed[True-True]	0.7499ms	0.5006ms	1.9977 KOps/s	2.0141 KOps/s	$\color{#d91a1a}-0.81\%$
test_vmap_mlp_speed[True-False]	0.7471ms	0.4962ms	2.0155 KOps/s	2.0369 KOps/s	$\color{#d91a1a}-1.05\%$
test_vmap_mlp_speed[False-True]	0.6883ms	0.3987ms	2.5082 KOps/s	2.4655 KOps/s	$\color{#35bf28}+1.73\%$
test_vmap_mlp_speed[False-False]	0.6996ms	0.4007ms	2.4956 KOps/s	2.4697 KOps/s	$\color{#35bf28}+1.05\%$
test_vmap_mlp_speed_decorator[True-True]	1.2531ms	0.5851ms	1.7092 KOps/s	1.7126 KOps/s	$\color{#d91a1a}-0.20\%$
test_vmap_mlp_speed_decorator[True-False]	0.7661ms	0.5817ms	1.7191 KOps/s	1.7084 KOps/s	$\color{#35bf28}+0.62\%$
test_vmap_mlp_speed_decorator[False-True]	0.7909ms	0.4775ms	2.0941 KOps/s	2.0883 KOps/s	$\color{#35bf28}+0.28\%$
test_vmap_mlp_speed_decorator[False-False]	0.7379ms	0.4742ms	2.1090 KOps/s	2.0743 KOps/s	$\color{#35bf28}+1.68\%$
test_to_module_speed[True]	1.9603ms	1.8133ms	551.4772 Ops/s	546.3937 Ops/s	$\color{#35bf28}+0.93\%$
test_to_module_speed[False]	2.3258ms	1.7822ms	561.1191 Ops/s	550.6449 Ops/s	$\color{#35bf28}+1.90\%$
test_tc_init	0.1134ms	45.7457μs	21.8600 KOps/s	25.8977 KOps/s	$\textbf{\color{#d91a1a}-15.59\%}$
test_tc_init_nested	0.1699ms	91.1193μs	10.9746 KOps/s	12.8109 KOps/s	$\textbf{\color{#d91a1a}-14.33\%}$
test_tc_first_layer_tensor	58.7290μs	9.2043μs	108.6445 KOps/s	120.8621 KOps/s	$\textbf{\color{#d91a1a}-10.11\%}$
test_tc_first_layer_nontensor	32.8110μs	9.1637μs	109.1260 KOps/s	120.4470 KOps/s	$\textbf{\color{#d91a1a}-9.40\%}$
test_tc_second_layer_tensor	41.1960μs	2.8368μs	352.5154 KOps/s	391.6604 KOps/s	$\textbf{\color{#d91a1a}-9.99\%}$
test_tc_second_layer_nontensor	51.9470μs	10.2739μs	97.3342 KOps/s	106.2061 KOps/s	$\textbf{\color{#d91a1a}-8.35\%}$

[github-actions]

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

Total Benchmarks: 141. Improved: $\large\color{#35bf28}13$. Worsened: $\large\color{#d91a1a}8$.

Expand to view detailed results

Name	Max	Mean	Ops	Ops on Repo HEAD	Change
test_plain_set_nested	78.4920μs	12.5118μs	79.9245 KOps/s	75.7568 KOps/s	$\textbf{\color{#35bf28}+5.50\%}$
test_plain_set_stack_nested	0.1035ms	12.6947μs	78.7730 KOps/s	76.0204 KOps/s	$\color{#35bf28}+3.62\%$
test_plain_set_nested_inplace	42.1010μs	13.6621μs	73.1953 KOps/s	70.3745 KOps/s	$\color{#35bf28}+4.01\%$
test_plain_set_stack_nested_inplace	31.1900μs	13.6561μs	73.2273 KOps/s	70.5577 KOps/s	$\color{#35bf28}+3.78\%$
test_items	21.7810μs	4.7603μs	210.0729 KOps/s	209.3041 KOps/s	$\color{#35bf28}+0.37\%$
test_items_nested	0.5054ms	0.4024ms	2.4848 KOps/s	2.5246 KOps/s	$\color{#d91a1a}-1.57\%$
test_items_nested_locked	0.5935ms	0.4034ms	2.4787 KOps/s	2.4946 KOps/s	$\color{#d91a1a}-0.64\%$
test_items_nested_leaf	0.2345ms	85.8378μs	11.6499 KOps/s	11.5173 KOps/s	$\color{#35bf28}+1.15\%$
test_items_stack_nested	0.4519ms	0.4013ms	2.4917 KOps/s	2.5135 KOps/s	$\color{#d91a1a}-0.86\%$
test_items_stack_nested_leaf	0.2714ms	86.2013μs	11.6008 KOps/s	11.4087 KOps/s	$\color{#35bf28}+1.68\%$
test_items_stack_nested_locked	0.4715ms	0.4030ms	2.4813 KOps/s	2.4902 KOps/s	$\color{#d91a1a}-0.36\%$
test_keys	17.4410μs	4.3696μs	228.8539 KOps/s	227.8169 KOps/s	$\color{#35bf28}+0.46\%$
test_keys_nested	0.1047ms	67.3765μs	14.8420 KOps/s	14.4676 KOps/s	$\color{#35bf28}+2.59\%$
test_keys_nested_locked	0.7625ms	73.1832μs	13.6643 KOps/s	13.4555 KOps/s	$\color{#35bf28}+1.55\%$
test_keys_nested_leaf	87.1720μs	57.5700μs	17.3702 KOps/s	17.3422 KOps/s	$\color{#35bf28}+0.16\%$
test_keys_stack_nested	0.1009ms	67.8143μs	14.7462 KOps/s	14.8907 KOps/s	$\color{#d91a1a}-0.97\%$
test_keys_stack_nested_leaf	95.5110μs	56.5338μs	17.6885 KOps/s	17.1243 KOps/s	$\color{#35bf28}+3.29\%$
test_keys_stack_nested_locked	0.1269ms	71.6087μs	13.9648 KOps/s	13.5747 KOps/s	$\color{#35bf28}+2.87\%$
test_values	7.4733μs	1.7767μs	562.8457 KOps/s	560.1742 KOps/s	$\color{#35bf28}+0.48\%$
test_values_nested	54.3910μs	33.8827μs	29.5136 KOps/s	28.7828 KOps/s	$\color{#35bf28}+2.54\%$
test_values_nested_locked	70.5320μs	35.7615μs	27.9630 KOps/s	27.1477 KOps/s	$\color{#35bf28}+3.00\%$
test_values_nested_leaf	0.1370ms	30.0789μs	33.2459 KOps/s	32.1363 KOps/s	$\color{#35bf28}+3.45\%$
test_values_stack_nested	0.1301ms	34.0329μs	29.3833 KOps/s	28.1075 KOps/s	$\color{#35bf28}+4.54\%$
test_values_stack_nested_leaf	51.9510μs	30.4465μs	32.8445 KOps/s	31.6257 KOps/s	$\color{#35bf28}+3.85\%$
test_values_stack_nested_locked	0.1626ms	36.0327μs	27.7526 KOps/s	26.8645 KOps/s	$\color{#35bf28}+3.31\%$
test_membership	1.8350μs	0.5373μs	1.8611 MOps/s	1.8816 MOps/s	$\color{#d91a1a}-1.09\%$
test_membership_nested	17.0200μs	2.0761μs	481.6732 KOps/s	484.0788 KOps/s	$\color{#d91a1a}-0.50\%$
test_membership_nested_leaf	9.5950μs	2.0149μs	496.3148 KOps/s	498.0365 KOps/s	$\color{#d91a1a}-0.35\%$
test_membership_stacked_nested	17.9500μs	2.0907μs	478.2996 KOps/s	477.3000 KOps/s	$\color{#35bf28}+0.21\%$
test_membership_stacked_nested_leaf	16.0300μs	2.0880μs	478.9353 KOps/s	483.8790 KOps/s	$\color{#d91a1a}-1.02\%$
test_membership_nested_last	17.1000μs	3.0155μs	331.6151 KOps/s	334.0297 KOps/s	$\color{#d91a1a}-0.72\%$
test_membership_nested_leaf_last	0.2009ms	2.9681μs	336.9135 KOps/s	331.8967 KOps/s	$\color{#35bf28}+1.51\%$
test_membership_stacked_nested_last	72.0410μs	3.4183μs	292.5401 KOps/s	291.0557 KOps/s	$\color{#35bf28}+0.51\%$
test_membership_stacked_nested_leaf_last	98.9220μs	3.4236μs	292.0919 KOps/s	292.5499 KOps/s	$\color{#d91a1a}-0.16\%$
test_nested_getleaf	33.2410μs	8.0239μs	124.6283 KOps/s	123.9154 KOps/s	$\color{#35bf28}+0.58\%$
test_nested_get	19.5210μs	7.5773μs	131.9738 KOps/s	131.9285 KOps/s	$\color{#35bf28}+0.03\%$
test_stacked_getleaf	33.8000μs	8.0396μs	124.3845 KOps/s	123.8000 KOps/s	$\color{#35bf28}+0.47\%$
test_stacked_get	22.1310μs	7.5415μs	132.5992 KOps/s	132.5021 KOps/s	$\color{#35bf28}+0.07\%$
test_nested_getitemleaf	22.8100μs	8.1945μs	122.0336 KOps/s	122.1271 KOps/s	$\color{#d91a1a}-0.08\%$
test_nested_getitem	21.6100μs	7.6999μs	129.8716 KOps/s	130.0097 KOps/s	$\color{#d91a1a}-0.11\%$
test_stacked_getitemleaf	32.6010μs	8.2287μs	121.5251 KOps/s	122.3144 KOps/s	$\color{#d91a1a}-0.65\%$
test_stacked_getitem	23.0610μs	7.6868μs	130.0930 KOps/s	130.3581 KOps/s	$\color{#d91a1a}-0.20\%$
test_lock_nested	9.6803ms	0.4293ms	2.3294 KOps/s	2.3519 KOps/s	$\color{#d91a1a}-0.96\%$
test_lock_stack_nested	0.4621ms	0.3890ms	2.5707 KOps/s	2.5536 KOps/s	$\color{#35bf28}+0.67\%$
test_unlock_nested	0.8300ms	0.3428ms	2.9173 KOps/s	2.8960 KOps/s	$\color{#35bf28}+0.73\%$
test_unlock_stack_nested	0.3727ms	0.3105ms	3.2208 KOps/s	3.2075 KOps/s	$\color{#35bf28}+0.42\%$
test_flatten_speed	0.3933ms	0.1057ms	9.4618 KOps/s	9.4876 KOps/s	$\color{#d91a1a}-0.27\%$
test_unflatten_speed	0.5006ms	0.2921ms	3.4229 KOps/s	3.4340 KOps/s	$\color{#d91a1a}-0.32\%$
test_common_ops	0.9497ms	0.5658ms	1.7674 KOps/s	1.7029 KOps/s	$\color{#35bf28}+3.79\%$
test_creation	35.6610μs	1.9506μs	512.6646 KOps/s	538.7516 KOps/s	$\color{#d91a1a}-4.84\%$
test_creation_empty	29.2600μs	8.7541μs	114.2317 KOps/s	102.0831 KOps/s	$\textbf{\color{#35bf28}+11.90\%}$
test_creation_nested_1	30.9100μs	10.6353μs	94.0269 KOps/s	84.8531 KOps/s	$\textbf{\color{#35bf28}+10.81\%}$
test_creation_nested_2	28.4600μs	13.0172μs	76.8217 KOps/s	70.9614 KOps/s	$\textbf{\color{#35bf28}+8.26\%}$
test_clone	91.4620μs	11.0417μs	90.5658 KOps/s	88.6636 KOps/s	$\color{#35bf28}+2.15\%$
test_getitem[int]	25.5300μs	10.2578μs	97.4864 KOps/s	97.3175 KOps/s	$\color{#35bf28}+0.17\%$
test_getitem[slice_int]	0.1213ms	20.2695μs	49.3351 KOps/s	49.4720 KOps/s	$\color{#d91a1a}-0.28\%$
test_getitem[range]	0.2435ms	37.8344μs	26.4310 KOps/s	26.5985 KOps/s	$\color{#d91a1a}-0.63\%$
test_getitem[tuple]	36.4500μs	17.7440μs	56.3571 KOps/s	57.0150 KOps/s	$\color{#d91a1a}-1.15\%$
test_getitem[list]	0.2655ms	32.3529μs	30.9091 KOps/s	30.7658 KOps/s	$\color{#35bf28}+0.47\%$
test_setitem_dim[int]	42.1310μs	23.6833μs	42.2239 KOps/s	38.2694 KOps/s	$\textbf{\color{#35bf28}+10.33\%}$
test_setitem_dim[slice_int]	73.6410μs	45.1344μs	22.1560 KOps/s	20.8670 KOps/s	$\textbf{\color{#35bf28}+6.18\%}$
test_setitem_dim[range]	0.1061ms	62.0848μs	16.1070 KOps/s	15.5197 KOps/s	$\color{#35bf28}+3.78\%$
test_setitem_dim[tuple]	0.1467ms	39.2671μs	25.4666 KOps/s	24.5486 KOps/s	$\color{#35bf28}+3.74\%$
test_setitem	90.1010μs	15.7266μs	63.5866 KOps/s	61.2210 KOps/s	$\color{#35bf28}+3.86\%$
test_set	99.3320μs	15.1360μs	66.0676 KOps/s	63.8574 KOps/s	$\color{#35bf28}+3.46\%$
test_set_shared	3.0625ms	0.1014ms	9.8581 KOps/s	10.2343 KOps/s	$\color{#d91a1a}-3.68\%$
test_update	98.2620μs	17.4419μs	57.3331 KOps/s	52.0450 KOps/s	$\textbf{\color{#35bf28}+10.16\%}$
test_update_nested	0.1090ms	23.1195μs	43.2535 KOps/s	41.7386 KOps/s	$\color{#35bf28}+3.63\%$
test_update__nested	97.7520μs	20.4971μs	48.7874 KOps/s	46.7755 KOps/s	$\color{#35bf28}+4.30\%$
test_set_nested	0.1086ms	15.9571μs	62.6682 KOps/s	59.8245 KOps/s	$\color{#35bf28}+4.75\%$
test_set_nested_new	99.1820μs	18.8219μs	53.1297 KOps/s	51.7153 KOps/s	$\color{#35bf28}+2.73\%$
test_select	0.1159ms	31.4979μs	31.7481 KOps/s	30.1584 KOps/s	$\textbf{\color{#35bf28}+5.27\%}$
test_select_nested	92.2120μs	52.7568μs	18.9549 KOps/s	19.0010 KOps/s	$\color{#d91a1a}-0.24\%$
test_exclude_nested	0.1429ms	72.4181μs	13.8087 KOps/s	13.7605 KOps/s	$\color{#35bf28}+0.35\%$
test_empty[True]	0.4063ms	0.2985ms	3.3505 KOps/s	3.3351 KOps/s	$\color{#35bf28}+0.46\%$
test_empty[False]	16.9113μs	0.9355μs	1.0690 MOps/s	1.0822 MOps/s	$\color{#d91a1a}-1.23\%$
test_to	89.1230μs	59.1288μs	16.9122 KOps/s	17.0539 KOps/s	$\color{#d91a1a}-0.83\%$
test_to_nonblocking	0.1840ms	36.8469μs	27.1393 KOps/s	26.4599 KOps/s	$\color{#35bf28}+2.57\%$
test_unbind_speed	0.3283ms	0.2638ms	3.7906 KOps/s	3.7153 KOps/s	$\color{#35bf28}+2.03\%$
test_unbind_speed_stack0	0.3610ms	0.2634ms	3.7966 KOps/s	3.7670 KOps/s	$\color{#35bf28}+0.79\%$
test_unbind_speed_stack1	94.3900ms	0.8021ms	1.2467 KOps/s	1.2368 KOps/s	$\color{#35bf28}+0.80\%$
test_split	93.1856ms	1.5563ms	642.5634 Ops/s	627.1538 Ops/s	$\color{#35bf28}+2.46\%$
test_chunk	1.5399ms	1.4148ms	706.8164 Ops/s	691.2008 Ops/s	$\color{#35bf28}+2.26\%$
test_creation[device0]	0.1896ms	55.4836μs	18.0233 KOps/s	17.2059 KOps/s	$\color{#35bf28}+4.75\%$
test_creation_from_tensor	0.1948ms	55.2455μs	18.1010 KOps/s	18.2142 KOps/s	$\color{#d91a1a}-0.62\%$
test_add_one[memmap_tensor0]	97.6820μs	6.8050μs	146.9501 KOps/s	145.9794 KOps/s	$\color{#35bf28}+0.66\%$
test_contiguous[memmap_tensor0]	11.7800μs	0.6052μs	1.6524 MOps/s	1.7039 MOps/s	$\color{#d91a1a}-3.02\%$
test_stack[memmap_tensor0]	32.2910μs	4.3144μs	231.7793 KOps/s	233.3979 KOps/s	$\color{#d91a1a}-0.69\%$
test_memmaptd_index	1.0943ms	0.2548ms	3.9250 KOps/s	3.7821 KOps/s	$\color{#35bf28}+3.78\%$
test_memmaptd_index_astensor	0.6243ms	0.3188ms	3.1366 KOps/s	3.0705 KOps/s	$\color{#35bf28}+2.15\%$
test_memmaptd_index_op	94.1474ms	0.6546ms	1.5277 KOps/s	1.6012 KOps/s	$\color{#d91a1a}-4.59\%$
test_serialize_model	94.2990ms	90.2210ms	11.0839 Ops/s	10.3109 Ops/s	$\textbf{\color{#35bf28}+7.50\%}$
test_serialize_model_pickle	1.3480s	1.2351s	0.8096 Ops/s	0.7186 Ops/s	$\textbf{\color{#35bf28}+12.66\%}$
test_serialize_weights	0.1864s	99.6104ms	10.0391 Ops/s	10.7579 Ops/s	$\textbf{\color{#d91a1a}-6.68\%}$
test_serialize_weights_returnearly	0.2958s	79.5005ms	12.5785 Ops/s	13.8962 Ops/s	$\textbf{\color{#d91a1a}-9.48\%}$
test_serialize_weights_pickle	1.3523s	1.2487s	0.8008 Ops/s	0.8010 Ops/s	$\color{#d91a1a}-0.02\%$
test_reshape_pytree	0.2389ms	25.3708μs	39.4154 KOps/s	39.2822 KOps/s	$\color{#35bf28}+0.34\%$
test_reshape_td	99.1620μs	30.7905μs	32.4775 KOps/s	32.8166 KOps/s	$\color{#d91a1a}-1.03\%$
test_view_pytree	0.1437ms	25.0124μs	39.9802 KOps/s	39.3006 KOps/s	$\color{#35bf28}+1.73\%$
test_view_td	0.2455ms	37.9981μs	26.3171 KOps/s	27.3142 KOps/s	$\color{#d91a1a}-3.65\%$
test_unbind_pytree	0.1770ms	32.2356μs	31.0216 KOps/s	32.3281 KOps/s	$\color{#d91a1a}-4.04\%$
test_unbind_td	0.6119ms	41.9748μs	23.8238 KOps/s	25.1636 KOps/s	$\textbf{\color{#d91a1a}-5.32\%}$
test_split_pytree	63.5910μs	35.4142μs	28.2373 KOps/s	29.3353 KOps/s	$\color{#d91a1a}-3.74\%$
test_split_td	0.2516ms	39.8755μs	25.0781 KOps/s	26.6280 KOps/s	$\textbf{\color{#d91a1a}-5.82\%}$
test_add_pytree	0.1660ms	37.1620μs	26.9092 KOps/s	26.4577 KOps/s	$\color{#35bf28}+1.71\%$
test_add_td	0.2472ms	46.9774μs	21.2868 KOps/s	20.1175 KOps/s	$\textbf{\color{#35bf28}+5.81\%}$
test_distributed	0.2681ms	74.0026μs	13.5130 KOps/s	14.6156 KOps/s	$\textbf{\color{#d91a1a}-7.54\%}$
test_tdmodule	59.6210μs	14.2998μs	69.9310 KOps/s	66.7691 KOps/s	$\color{#35bf28}+4.74\%$
test_tdmodule_dispatch	45.5110μs	29.6399μs	33.7383 KOps/s	35.3126 KOps/s	$\color{#d91a1a}-4.46\%$
test_tdseq	0.1008ms	15.3914μs	64.9715 KOps/s	65.2193 KOps/s	$\color{#d91a1a}-0.38\%$
test_tdseq_dispatch	56.8810μs	32.2853μs	30.9739 KOps/s	31.6974 KOps/s	$\color{#d91a1a}-2.28\%$
test_instantiation_functorch	1.6262ms	1.3770ms	726.2266 Ops/s	721.3382 Ops/s	$\color{#35bf28}+0.68\%$
test_instantiation_td	1.4452ms	0.9739ms	1.0268 KOps/s	914.9997 Ops/s	$\textbf{\color{#35bf28}+12.22\%}$
test_exec_functorch	0.2546ms	0.1444ms	6.9269 KOps/s	6.7972 KOps/s	$\color{#35bf28}+1.91\%$
test_exec_functional_call	0.3243ms	0.1291ms	7.7443 KOps/s	7.5294 KOps/s	$\color{#35bf28}+2.85\%$
test_exec_td	0.1567ms	0.1258ms	7.9484 KOps/s	7.5727 KOps/s	$\color{#35bf28}+4.96\%$
test_exec_td_decorator	0.6016ms	0.1965ms	5.0883 KOps/s	4.9476 KOps/s	$\color{#35bf28}+2.84\%$
test_vmap_mlp_speed[True-True]	0.7708ms	0.5672ms	1.7631 KOps/s	1.7371 KOps/s	$\color{#35bf28}+1.50\%$
test_vmap_mlp_speed[True-False]	0.8013ms	0.5661ms	1.7664 KOps/s	1.7653 KOps/s	$\color{#35bf28}+0.06\%$
test_vmap_mlp_speed[False-True]	0.7103ms	0.5149ms	1.9422 KOps/s	1.9887 KOps/s	$\color{#d91a1a}-2.34\%$
test_vmap_mlp_speed[False-False]	0.7012ms	0.4990ms	2.0039 KOps/s	1.9940 KOps/s	$\color{#35bf28}+0.50\%$
test_vmap_mlp_speed_decorator[True-True]	1.1047ms	0.6408ms	1.5605 KOps/s	1.5450 KOps/s	$\color{#35bf28}+1.00\%$
test_vmap_mlp_speed_decorator[True-False]	0.8552ms	0.6410ms	1.5602 KOps/s	1.5576 KOps/s	$\color{#35bf28}+0.16\%$
test_vmap_mlp_speed_decorator[False-True]	0.7536ms	0.5583ms	1.7912 KOps/s	1.6886 KOps/s	$\textbf{\color{#35bf28}+6.08\%}$
test_vmap_mlp_speed_decorator[False-False]	0.7910ms	0.5615ms	1.7808 KOps/s	1.7233 KOps/s	$\color{#35bf28}+3.34\%$
test_vmap_transformer_speed[True-True]	8.0670ms	7.5948ms	131.6694 Ops/s	129.3257 Ops/s	$\color{#35bf28}+1.81\%$
test_vmap_transformer_speed[True-False]	7.7453ms	7.4968ms	133.3897 Ops/s	129.5591 Ops/s	$\color{#35bf28}+2.96\%$
test_vmap_transformer_speed[False-True]	7.6446ms	7.4405ms	134.4004 Ops/s	130.3973 Ops/s	$\color{#35bf28}+3.07\%$
test_vmap_transformer_speed[False-False]	8.0195ms	7.5994ms	131.5900 Ops/s	130.5478 Ops/s	$\color{#35bf28}+0.80\%$
test_vmap_transformer_speed_decorator[True-True]	18.9022ms	18.5444ms	53.9245 Ops/s	52.3234 Ops/s	$\color{#35bf28}+3.06\%$
test_vmap_transformer_speed_decorator[True-False]	19.0490ms	18.6123ms	53.7280 Ops/s	52.4073 Ops/s	$\color{#35bf28}+2.52\%$
test_vmap_transformer_speed_decorator[False-True]	18.6452ms	18.3531ms	54.4867 Ops/s	52.9137 Ops/s	$\color{#35bf28}+2.97\%$
test_vmap_transformer_speed_decorator[False-False]	19.0367ms	18.3893ms	54.3796 Ops/s	52.9993 Ops/s	$\color{#35bf28}+2.60\%$
test_to_module_speed[True]	1.6022ms	1.4842ms	673.7843 Ops/s	655.7122 Ops/s	$\color{#35bf28}+2.76\%$
test_to_module_speed[False]	1.5901ms	1.4612ms	684.3782 Ops/s	662.0518 Ops/s	$\color{#35bf28}+3.37\%$
test_tc_init	54.0410μs	35.9928μs	27.7833 KOps/s	28.1650 KOps/s	$\color{#d91a1a}-1.36\%$
test_tc_init_nested	0.1028ms	71.4880μs	13.9884 KOps/s	14.2762 KOps/s	$\color{#d91a1a}-2.02\%$
test_tc_first_layer_tensor	18.8800μs	3.9748μs	251.5840 KOps/s	277.3163 KOps/s	$\textbf{\color{#d91a1a}-9.28\%}$
test_tc_first_layer_nontensor	0.1664ms	4.0099μs	249.3857 KOps/s	273.6621 KOps/s	$\textbf{\color{#d91a1a}-8.87\%}$
test_tc_second_layer_tensor	45.2413μs	1.2895μs	775.5228 KOps/s	811.1156 KOps/s	$\color{#d91a1a}-4.39\%$
test_tc_second_layer_nontensor	21.0410μs	4.5790μs	218.3893 KOps/s	240.6797 KOps/s	$\textbf{\color{#d91a1a}-9.26\%}$