MCPcopy Create free account

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

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

MethodsetUp
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/tests/shampoo_distributor_test.py:199
MethodsetUp
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_ddp_distributor_test.py:903
MethodsetUp
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/preconditioner_list_test_utils.py:52
MethodsetUp
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/shampoo_preconditioner_list_test.py:214
MethodsetUp
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/adagrad_preconditioner_list_test.py:45
MethodsetUp
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/distributed_shampoo_test.py:61
MethodsetUp
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/distributed_shampoo_test.py:254
MethodsetUp
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/distributed_shampoo_test.py:328
MethodsetUp
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/distributed_shampoo_test.py:1116
MethodsetUp
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/distributed_shampoo_test.py:1603
MethodsetUp
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/loss_metrics_test.py:21
Functionsetup_context
(ctx: torch.autograd.function.FunctionCtx, inputs, output)
train_gpt_medium.py:109
Functionsetup_context
(ctx: torch.autograd.function.FunctionCtx, inputs, output)
records/track_1_short/2025-12-11_NorMuonOptimsAndFixes/profiler-example-traces/train_gpt-profiler-example.py:113
Functionsetup_context_t
(ctx: torch.autograd.function.FunctionCtx, inputs, output)
train_gpt.py:148
Methodshould_assign_param_idx
(i: int)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/_shampoo_hybrid_shard_lossless_distributor.py:75
Methodshould_assign_param_idx
(i: int)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/_shampoo_fully_shard_lossless_distributor.py:69
Methodstart_transition
(self, freeze_count=40)
train_gpt_medium.py:1581
Functionstea_extrapolate
(iterates, regularization=1e-6, damping=1.0)
records/track_3_optimization/results/20260520_rre_extrapolation_pr300_2925/train_gpt_simple_rre_pr300_2925.py:845
Methodstep
(self)
train_gpt_medium.py:782
Methodstep
(self)
records/track_1_short/2025-12-11_NorMuonOptimsAndFixes/profiler-example-traces/train_gpt-profiler-example.py:775
Methodstep
(self)
records/track_3_optimization/results/20260611_tailema_2720_submission/train_gpt_tailema_2720.py:641
Methodstep
(self)
records/track_3_optimization/results/20260611_tailema_2720_submission/train_gpt_tailema_2720.py:746
Methodstep
(self)
records/track_3_optimization/results/20260519_dynmuon/train_gpt_simple_dynmuon.py:329
Methodstep
(self)
records/track_3_optimization/results/20260609_soap_f1_auxb2_clean/train_gpt_clean_SOTA.py:492
Methodstep
(self)
records/track_3_optimization/results/20260609_soap_f1_auxb2_clean/train_gpt_clean_SOTA.py:597
Methodstep
(self)
records/track_3_optimization/results/20260529_tail_phase_readout_2850/train_gpt_tail_phase_readout_2850.py:1008
Methodstep
(self)
records/track_3_optimization/results/20260619_cwd_rowfloor_tailema/train_gpt_cwd_SOTA.py:536
Methodstep
(self)
records/track_3_optimization/results/20260619_cwd_rowfloor_tailema/train_gpt_cwd_SOTA.py:641
Methodstep
(self)
records/track_3_optimization/results/20260525_aurora_ema_ref/train_gpt_simple_aurora_ema_ref.py:561
Methodstep
(self)
records/track_3_optimization/results/20260611_tailema_2730_submission/train_gpt_tailema_2730.py:640
Methodstep
(self)
records/track_3_optimization/results/20260611_tailema_2730_submission/train_gpt_tailema_2730.py:745
Methodstep
(self)
records/track_3_optimization/results/20260611_tailema_2730_submission/ablation/combo_2740/train_gpt_combo_2740.py:623
Methodstep
(self)
records/track_3_optimization/results/20260611_tailema_2730_submission/ablation/combo_2740/train_gpt_combo_2740.py:728
Functionsvd_orthogonalization
Compute the orthogonalization of a matrix using SVD. Args: A (Tensor): The input matrix. scale_by_nuclear_norm (bool)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/matrix_functions.py:912
MethodtearDown
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_ddp_distributor_test.py:908
Methodtest_adagrad_eigenvalue_correction
( self, weight_decay: float, device: torch.device, start_preconditioning_step:
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/gpu_tests/shampoo_eigenvalue_correction_test.py:77
Methodtest_adagrad_grafting
( self, weight_decay: float, device: torch.device, preconditioner_config: Prec
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/gpu_tests/shampoo_grafting_test.py:73
Methodtest_adam_eigenvalue_correction
( self, weight_decay: float, device: torch.device, start_preconditioning_step:
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/gpu_tests/shampoo_eigenvalue_correction_test.py:127
Methodtest_adam_grafting
( self, weight_decay: float, device: torch.device, preconditioner_config: Prec
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/gpu_tests/shampoo_grafting_test.py:110
Methodtest_adamw_eigenvalue_correction
( self, weight_decay: float, device: torch.device, start_preconditioning_step:
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/gpu_tests/shampoo_eigenvalue_correction_test.py:181
Methodtest_adamw_grafting
( self, weight_decay: float, device: torch.device, preconditioner_config: Prec
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/gpu_tests/shampoo_grafting_test.py:147
Methodtest_adaptive_amortized_computation_frequency
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/shampoo_preconditioner_list_test.py:133
Methodtest_adaptive_amortized_computation_frequency
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/shampoo_preconditioner_list_test.py:1113
Methodtest_adaptive_amortized_computation_frequency
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/shampoo_preconditioner_list_test.py:1426
Methodtest_all_non_abstract_subclasses
( self, cls: object, expected_subclasses: list[object] )
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/commons_test.py:121
Methodtest_all_ranks_with_no_grads
(self, communicate_params: bool)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_ddp_distributor_test.py:673
Methodtest_all_ranks_with_no_grads
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_fully_shard_distributor_test.py:139
Methodtest_all_ranks_with_no_grads
( self, communicate_params: bool, model_linear_layers_dims: tuple[int, ...], m
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_hybrid_shard_lossless_distributor_test.py:278
Methodtest_all_ranks_with_no_grads
(self, communicate_params: bool)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_hybrid_shard_distributor_test.py:356
Methodtest_all_ranks_with_no_grads
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_fsdp_distributor_test.py:133
Methodtest_all_ranks_with_no_grads
( self, model_linear_layers_dims: tuple[int, ...], model_dead_layers_dims: tuple[int,
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_fully_shard_lossless_distributor_test.py:187
Methodtest_all_ranks_with_no_grads
(self, communicate_params: bool)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_hsdp_distributor_test.py:238
Methodtest_amortized_computation_failure_tolerance
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/shampoo_preconditioner_list_test.py:386
Methodtest_amortized_computation_internal_failure
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/shampoo_preconditioner_list_test.py:351
Methodtest_batch_size_one
Test batching with batch size of 1.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/commons_test.py:43
Methodtest_bfloat16_resolution
Test that bfloat16 is resolved correctly.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/resolvers_test.py:36
Methodtest_can_run_with_empty_local_params
A variant of test_losses() that tests the case where ranks can have empty local blocked parameters.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_ddp_distributor_test.py:221
Methodtest_check_tensor_for_non_two_dim_matrix
(self, A: Tensor)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/matrix_functions_test.py:66
Methodtest_check_tensor_for_rectangle_matrix
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/matrix_functions_test.py:79
Methodtest_check_tensor_for_rectangle_matrix
Override the test from Check2DTensorTest to test non-square matrix.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/matrix_functions_test.py:92
Methodtest_check_tensor_for_square_matrix
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/matrix_functions_test.py:74
Methodtest_checkpoint_preemption_bitwise_equivalence
This test validates that when using fully_shard/hybrid_shard lossless distributor, resuming training after preemption produces **bitw
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_checkpoint_test.py:330
Methodtest_compile_fsdp_parameter_metadata
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_fsdp_utils_test.py:59
Methodtest_compile_fsdp_parameter_metadata_with_no_flat_param
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_fsdp_utils_test.py:109
Methodtest_compress
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/shampoo_quantization_test.py:173
Methodtest_compress_list
( self, selector: tuple[bool, ...], compressed_tuple: tuple[int,] )
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/shampoo_utils_test.py:198
Methodtest_compress_list_with_different_size
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/shampoo_utils_test.py:205
Methodtest_compress_preconditioner_list
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/preconditioner_list_test_utils.py:113
Methodtest_convolution_like_dims
( self, threshold: int, target_tensor_dimensionality: int, expected_new_tensor
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/shampoo_utils_test.py:130
Methodtest_corrected_weight_decay
Test CORRECTED weight decay matches AdamW when lr == peak_lr. Corrected weight decay applies params *= (1 - lr * (lr / peak_lr) * weight_deca
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/gpu_tests/weight_decay_test.py:166
Methodtest_custom_scale_fn
(self, cls: type[SignDescentPreconditionerConfig])
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/shampoo_types_test.py:289
Methodtest_ddp_shampoo_checkpoint
This test is intended to make sure that Shampoo `state_dict` and `load_state_dict` are compatible with PyTorch's API, including `torc
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_checkpoint_test.py:129
Methodtest_decoupled_resolution
Test that DECOUPLED weight decay type is resolved correctly.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/resolvers_test.py:79
Methodtest_decoupled_weight_decay
Test DECOUPLED weight decay matches PyTorch AdamW. Decoupled weight decay applies params *= (1 - lr * weight_decay) independent of th
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/gpu_tests/weight_decay_test.py:124
Methodtest_default_scale_fn
(self, cls: type[SignDescentPreconditionerConfig])
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/shampoo_types_test.py:282
Methodtest_default_strategy_resolution
Test that DEFAULT strategy is resolved correctly.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/resolvers_test.py:66
Methodtest_defaults_when_env_not_set
Test that get_distributed_env returns defaults when env vars not set.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/utils_test.py:53
Methodtest_dequantize_quantize
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/shampoo_quantization_test.py:47
Methodtest_dequantize_quantize
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/shampoo_quantization_test.py:196
Methodtest_dequantize_with_same_quantize_dtype_and_computation_dtype
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/shampoo_quantization_test.py:180
Methodtest_dict_zip_iterator_different_lengths_raises_error
Test that ValueError is raised when iterators have different lengths.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/dict_zip_iterator_test.py:54
Methodtest_dict_zip_iterator_empty_iterators
Test that empty iterators raise StopIteration immediately.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/dict_zip_iterator_test.py:80
Methodtest_dict_zip_iterator_same_length
( self, data: dict[str, Iterator[object]], expected_results: list[dict[str, object]],
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/dict_zip_iterator_test.py:41
Methodtest_dict_zip_iterator_single_element
Test with single element iterators.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/dict_zip_iterator_test.py:92
Methodtest_dims_list
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/preconditioner_list_test_utils.py:79
Methodtest_disable_eager_nan_check
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/tests/shampoo_distributor_test.py:180
Methodtest_dist_is_initialized
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_hybrid_shard_distributor_test.py:465
Methodtest_dist_is_initialized
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_hsdp_distributor_test.py:298
Methodtest_distribute_buffer_sizes
( self, blocked_params: tuple[torch.Tensor, ...], group_size: int, load_balanc
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/shampoo_utils_test.py:362
Methodtest_edge_case_zero_window_size
Test behavior with edge case of zero window size.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/loss_metrics_test.py:219
Methodtest_eigen_root_identity
(self, n: int, root: int)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/matrix_functions_test.py:421
Methodtest_eigen_root_nonfull_rank
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/matrix_functions_test.py:473
Methodtest_eigen_root_tridiagonal
( self, n: int, root: int, alpha: float, beta: float )
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/matrix_functions_test.py:433
Methodtest_empty_dims
(self, threshold: int)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/shampoo_utils_test.py:88
Methodtest_empty_iterable
Test batching an empty iterable.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/commons_test.py:37
Methodtest_enable_eager_nan_check
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/tests/shampoo_distributor_test.py:162
Methodtest_extra_repr
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/shampoo_model_utils_test.py:134
Methodtest_extract_state_dict_content
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/shampoo_state_dict_utils_test.py:43
Methodtest_float16_resolution
Test that float16 is resolved correctly.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/resolvers_test.py:31
Methodtest_float32_resolution
Test that float32 is resolved correctly.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/resolvers_test.py:26
← previousnext →1,901–2,000 of 2,320, ranked by callers