MCPcopy Create free account

hub / github.com/KellerJordan/modded-nanogpt / functions

Functions2,320 in github.com/KellerJordan/modded-nanogpt

Method_expected_numel_list
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/adagrad_preconditioner_list_test.py:118
Method_expected_numel_list
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/sign_descent_preconditioner_list_test.py:81
Function_get_autotune_configs
()
train_gpt_medium.py:121
Function_get_autotune_configs
()
records/track_1_short/2025-12-11_NorMuonOptimsAndFixes/profiler-example-traces/train_gpt-profiler-example.py:125
Method_get_composable_block_id_rank
For lossless shampoo distributor, it's unnecessary to include rank in block id.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/_shampoo_hybrid_shard_lossless_distributor.py:121
Method_get_default_amortized_computation_config
()
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/shampoo_types.py:324
Method_get_default_amortized_computation_config
()
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/shampoo_types.py:390
Method_get_default_amortized_computation_config
()
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/shampoo_types.py:458
Method_get_default_cost_model
()
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/shampoo_types.py:954
Method_get_default_load_balancing_config
()
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/shampoo_types.py:1042
Method_get_default_rank_deficient_stability_config
()
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/matrix_functions_types.py:100
Method_get_default_scale_by_dims_fn
()
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/matrix_functions_types.py:265
Method_get_field_dict
Creates a dictionary containing shallow copies of this dataclass's fields, excluding specific fields. This method overrides the pare
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/shampoo_preconditioner_list.py:1015
Method_get_inverse_roots_from_override
( self, preconditioned_dims_selector: tuple[bool, ...] )
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/shampoo_preconditioner_list.py:1583
Method_get_inverse_roots_from_override
( self, preconditioned_dims_selector: tuple[bool, ...] )
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/shampoo_preconditioner_list.py:1680
Method_has_lambda_recursively
Recursively check if an object contains lambda functions.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributed_shampoo.py:1579
Method_instantiate_block_list
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/sgd_preconditioner_list_test.py:23
Method_instantiate_block_list
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/shampoo_preconditioner_list_test.py:205
Method_instantiate_block_list
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/spectral_descent_preconditioner_list_test.py:40
Method_instantiate_block_list
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/adagrad_preconditioner_list_test.py:28
Method_instantiate_block_list
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/sign_descent_preconditioner_list_test.py:34
Method_iterate_averaging_config
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/distributed_shampoo_test.py:1110
Method_iterate_averaging_config
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/distributed_shampoo_test.py:1371
Method_iterate_averaging_config
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/distributed_shampoo_test.py:1387
Method_load
(self)
train_gpt_medium.py:1256
Method_load
(self)
train_gpt_medium.py:1308
Method_load
(self)
records/track_1_short/2025-12-11_NorMuonOptimsAndFixes/profiler-example-traces/train_gpt-profiler-example.py:1170
Method_load
(self)
records/track_1_short/2025-12-11_NorMuonOptimsAndFixes/profiler-example-traces/train_gpt-profiler-example.py:1222
Method_merge_and_block_gradients
Split, merge, and block gradients. Returns: local_masked_blocked_grads (tuple[Tensor, ...]): Local gradients with grad not None.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/shampoo_fsdp_distributor.py:108
Method_merge_and_block_parameters
Split, merge, and block parameters.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/shampoo_fsdp_distributor.py:63
Method_optim_factory
( parameters: ParamsT, optim_cls: type[torch.optim.Optimizer], **kwargs: Any, )
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/gpu_tests/shampoo_grafting_test.py:51
Method_optim_factory
( parameters: ParamsT, optim_cls: type[torch.optim.Optimizer], **kwargs: Any, )
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/gpu_tests/iterate_averaging_test.py:62
Method_optim_factory
( parameters: ParamsT, optim_cls: type[torch.optim.Optimizer], **kwargs: Any, )
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/gpu_tests/shampoo_eigenvalue_correction_test.py:51
Method_optim_factory
( parameters: ParamsT, optim_cls: type[torch.optim.Optimizer], **kwargs: Any, )
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/gpu_tests/weight_decay_test.py:60
Function_patched_trace_structured
(name, metadata_fn, **kwargs)
records/track_2_medium/2025-11-12_BlockMaskRedundantOp/train_gpt_medium.py:392
Method_per_group_step_impl
( self, state_lists: dict[str, Any], step: torch.Tensor, lr: torch.Tensor,
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributed_shampoo.py:1314
Method_post_load_state_dict_hook
Perform post-load operations after checkpoint loading. This hook performs two operations: 1. Refreshes assigned params in lossless di
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributed_shampoo.py:1635
Method_post_state_dict_hook
Process the state dictionary after it's created by state_dict(). This hook extracts the actual content from each parameter state using the
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributed_shampoo.py:1562
Method_pre_load_state_dict_hook
Save the current train mode for each parameter group before loading state dict. This allows the post-load hook to restore the original mode a
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributed_shampoo.py:1617
Method_preconditioner_config
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/distributed_shampoo_test.py:322
Method_preconditioner_config
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/distributed_shampoo_test.py:497
Method_preconditioner_config
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/distributed_shampoo_test.py:660
Method_preconditioner_config
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/distributed_shampoo_test.py:831
Method_preconditioner_config
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/distributed_shampoo_test.py:1012
Method_preconditioner_config
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/distributed_shampoo_test.py:1018
Method_preconditioner_config
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/distributed_shampoo_test.py:1024
Method_preconditioner_config
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/distributed_shampoo_test.py:1062
Method_preconditioner_config
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/distributed_shampoo_test.py:1113
Method_preconditioner_list_factory
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/shampoo_preconditioner_list_test.py:1107
Method_preconditioner_list_factory
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/shampoo_preconditioner_list_test.py:1135
Method_preconditioner_list_factory
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/shampoo_preconditioner_list_test.py:1158
Method_preconditioner_list_factory
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/shampoo_preconditioner_list_test.py:1420
Method_preconditioner_list_factory
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/shampoo_preconditioner_list_test.py:1444
Method_ref_state_dict
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/distributed_shampoo_test.py:326
Method_ref_state_dict
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/distributed_shampoo_test.py:422
Method_ref_state_dict
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/distributed_shampoo_test.py:501
Method_ref_state_dict
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/distributed_shampoo_test.py:664
Method_ref_state_dict
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/distributed_shampoo_test.py:835
Method_ref_state_dict
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/distributed_shampoo_test.py:1144
Method_scan
(self)
train_gpt.py:1545
Method_shampoo_optim_factory
( shampoo_pt2_compile_config: ShampooPT2CompileConfig | None, precondition_frequency: int,
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/gpu_tests/shampoo_pt2_test.py:38
Method_sync_gradient
(self, param)
train_gpt_medium.py:749
Method_sync_gradient
(self, param)
records/track_1_short/2025-12-11_NorMuonOptimsAndFixes/profiler-example-traces/train_gpt-profiler-example.py:761
Function_transpose_add_kernel
( src_ptr, dst_ptr, M, N, src_stride_m, src_stride_n, dst_stride_0, dst_stride_1, BLOCK_M:
triton_kernels.py:636
Function_transpose_copy_kernel
( src_ptr, dst_ptr, M, N, src_stride_m, src_stride_n, dst_stride_0, dst_stride_1, BLOCK_M:
triton_kernels.py:574
Methodall_gather_into_tensor
()
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/shampoo_hsdp_distributor.py:245
Methodall_gather_into_tensor
()
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/shampoo_hybrid_shard_distributor.py:267
Functionanderson_extrapolate
(iterates, regularization=1e-6, damping=1.0, beta=1.0)
records/track_3_optimization/results/20260520_rre_extrapolation_pr300_2925/train_gpt_simple_rre_pr300_2925.py:788
Methodassert_state_dict_close
( actual: dict[Any, Any], expected: dict[Any, Any], path: str
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_ddp_distributor_test.py:642
Functionba_plus_cAA_kernel
( A_ptr, C_ptr, M, a_stride_b, a_stride_r, a_stride_c, c_stride_b, c_stride_r, c_stride_c,
train_gpt_medium.py:264
Functionba_plus_cAA_kernel
( A_ptr, C_ptr, M, a_stride_b, a_stride_r, a_stride_c, c_stride_b, c_stride_r, c_stride_c,
triton_kernels.py:282
Functionba_plus_cAA_kernel
( A_ptr, C_ptr, M, a_stride_b, a_stride_r, a_stride_c, c_stride_b, c_stride_r, c_stride_c,
records/track_1_short/2025-12-11_NorMuonOptimsAndFixes/profiler-example-traces/train_gpt-profiler-example.py:268
Functionbackward
(ctx, grad_out: Tensor, *_)
train_gpt_medium.py:101
Functionbackward
(ctx, grad_out: Tensor, *_)
records/track_1_short/2025-12-11_NorMuonOptimsAndFixes/profiler-example-traces/train_gpt-profiler-example.py:105
Methodbackward
(ctx, grad_output)
triton_kernels.py:558
Functionbackward_t
(ctx, grad_out: Tensor, *_)
train_gpt.py:140
Methodblock_info_equality
( a: BlockInfo, b: BlockInfo, msg: str | None = None )
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/tests/shampoo_distributor_test.py:110
Methodblock_info_equality
( a: BlockInfo, b: BlockInfo, msg: str | None = None )
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/tests/shampoo_distributor_test.py:289
Methodblock_within_tensor_shard_recovery
( block_within_tensor_shard: Tensor, dimension: int, block_start_idx: int,
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/shampoo_hsdp_distributor.py:726
Methodblock_within_tensor_shard_recovery
( block_within_tensor_shard: Tensor, dimension: int, block_start_idx: int,
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/shampoo_fsdp_distributor.py:277
Methodbuild_bm
(window_size_blocks: Tensor)
records/track_2_medium/2025-11-12_BlockMaskRedundantOp/train_gpt_medium.py:261
Methodcheck_tensor_func
Helper function decorated with _check_square_matrix for testing.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/matrix_functions_test.py:88
Methodclosure
()
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/distributed_shampoo_test.py:280
Methodcompress_preconditioner_list
( self, local_grad_selector: tuple[bool, ...] )
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/shampoo_preconditioner_list.py:1307
Methodcompress_preconditioner_list
( self, local_grad_selector: tuple[bool, ...] )
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/spectral_descent_preconditioner_list.py:65
Methodcompress_preconditioner_list
( self, local_grad_selector: tuple[bool, ...] )
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/sgd_preconditioner_list.py:39
Methodcompress_preconditioner_list
( self, local_grad_selector: tuple[bool, ...] )
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/sign_descent_preconditioner_list.py:52
Methodcompress_preconditioner_list
( self, local_grad_selector: tuple[bool, ...] )
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/adagrad_preconditioner_list.py:165
Methodcost
(self, tensor: torch.Tensor)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/load_balancing_utils.py:55
Methodcost
(self, tensor: torch.Tensor)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/load_balancing_utils.py:73
Methodcreate_doc_swc_block_mask
(sliding_window_num_blocks)
records/track_2_medium/2024-12-31_Target350M/train_gpt.py:271
Methoddense_to_ordered
(dense_mask)
records/track_2_medium/2024-12-31_Target350M/train_gpt.py:266
Methoddense_to_ordered
(dense_blockmask: Tensor)
records/track_2_medium/2025-11-12_BlockMaskRedundantOp/train_gpt_medium.py:241
Methoddequantized_value
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/shampoo_quantization.py:228
Methoddims_list
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/preconditioner_list.py:91
Methoddocument_causal
(b, h, q_idx, kv_idx)
records/track_2_medium/2024-12-31_Target350M/train_gpt.py:261
Methoddocument_causal
(b, h, q_idx, kv_idx)
records/track_2_medium/2025-11-12_BlockMaskRedundantOp/train_gpt_medium.py:236
Methoddocument_causal_mask
(b, h, q_idx, kv_idx)
records/track_1_short/2024-12-04_ValueEmbed/train_gpt2.py:255
Functioneigh_eigenvalue_decomposition
Compute the eigendecomposition of a symmetric matrix using torch.linalg.eigh. Args: A (Tensor): The input symmetric matrix.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/matrix_functions.py:706
Methodempty_tensor_list
(sizes: tuple[int, ...])
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/shampoo_utils_test.py:289
← previousnext →1,501–1,600 of 2,320, ranked by callers