MCPcopy Create free account

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

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

Methodtest_float64_resolution
Test that float64 is resolved correctly.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/resolvers_test.py:41
Methodtest_flush_behavior_with_tensorboard
Test flush behavior with tensorboard writer.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/loss_metrics_test.py:174
Methodtest_formatter_with_distributed_initialized
Test formatter includes rank when distributed is initialized.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/utils_test.py:221
Methodtest_formatter_without_distributed_initialized
Test formatter has default format when distributed is not initialized.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/utils_test.py:231
Methodtest_forward_pass_different_sizes
Test forward pass with different input sizes - indirectly validates _infer_conv_output_shape.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/convnet_test.py:35
Methodtest_forward_pass_mismatched_input_size
Test that forward pass fails with mismatched input size.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/convnet_test.py:54
Methodtest_fsdp_distributed_config_against_hsdp_distributed_config_bitwise_identical
( self, communicate_params: bool )
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_fsdp_distributor_test.py:219
Methodtest_fsdp_param_strategy_resolver_via_omegaconf
Test fsdp_param_strategy resolver works via OmegaConf interpolation.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/resolvers_test.py:114
Methodtest_fsdp_shampoo_against_default_shampoo
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_fsdp_distributor_test.py:156
Methodtest_fsdp_shampoo_block_index
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_fsdp_distributor_test.py:174
Methodtest_fully_shard_lossless_shampoo_block_index
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_fully_shard_lossless_distributor_test.py:258
Methodtest_fully_shard_shampoo_against_default_shampoo
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_fully_shard_distributor_test.py:162
Methodtest_fully_shard_shampoo_against_default_shampoo
( 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:228
Methodtest_fully_shard_shampoo_block_index
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_fully_shard_distributor_test.py:214
Methodtest_generate_pairwise_indices
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/shampoo_utils_test.py:226
Methodtest_generate_pairwise_indices_with_empty_list
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/shampoo_utils_test.py:233
Methodtest_get_device_mesh
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_dist_utils_test.py:47
Methodtest_get_dtype_size
(self, dtype: torch.dtype, size: int)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/shampoo_utils_test.py:221
Methodtest_global_metrics_distributed_behavior
Test global metrics behavior in distributed setting.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/loss_metrics_test.py:79
Methodtest_global_metrics_logging
Test global metrics logging behavior.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/loss_metrics_test.py:105
Methodtest_global_metrics_logging_single_worker
Test that global metrics are not logged for single worker.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/loss_metrics_test.py:127
Methodtest_gpa_vs_sgd_momentum
Test that Shampoo with GPA produces equivalent results to SGD with momentum. This test validates the theoretical equivalence between Generali
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/gpu_tests/iterate_averaging_test.py:74
Methodtest_hsdp_shampoo_against_default_shampoo
( self, num_trainers_per_group: int, communication_dtype: torch.dtype, communi
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_hsdp_distributor_test.py:143
Methodtest_hsdp_shampoo_block_index
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_hsdp_distributor_test.py:178
Methodtest_hybrid_shampoo_n_by_one_mesh_against_ddp_shampoo
Testing the correctness of hybrid shard Shampoo distributor of (n, 1) mesh by comparing it with DDP Shampoo. (n, 1) mesh is a special
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_hybrid_shard_distributor_test.py:258
Methodtest_hybrid_shampoo_n_by_one_mesh_against_default_shampoo
Testing the correctness of hybrid shard shampoo distributor of (n, 1) mesh by comparing it with default shampoo. (n, 1) mesh is a spe
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_hybrid_shard_distributor_test.py:207
Methodtest_hybrid_shard_distributed_config_against_fully_shard_distributed_config
Testing the correctness of hybrid shard shampoo distributor by comparing it with fully shard distributor.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_hybrid_shard_distributor_test.py:315
Methodtest_hybrid_shard_distributed_config_against_fully_shard_distributed_config_bitwise_identical
( self, communicate_params: bool )
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_fully_shard_distributor_test.py:181
Methodtest_hybrid_shard_lossless_shampoo_block_index
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_hybrid_shard_lossless_distributor_test.py:316
Methodtest_hybrid_shard_shampoo_against_default_shampoo
( self, num_trainers_per_group: int, communication_dtype: torch.dtype, communi
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_hybrid_shard_lossless_distributor_test.py:165
Methodtest_hybrid_shard_shampoo_against_default_shampoo
( self, num_trainers_per_group: int, communication_dtype: torch.dtype, communi
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_hybrid_shard_distributor_test.py:164
Methodtest_hybrid_shard_shampoo_block_index
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_hybrid_shard_distributor_test.py:386
Methodtest_hybrid_shard_shampoo_config_against_fully_shard_shampoo_config
Testing the correctness of hybrid shard shampoo distributor by comparing it with fully shard distributor.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_hybrid_shard_lossless_distributor_test.py:226
Methodtest_illegal_beta2
( self, cls: type[RMSpropPreconditionerConfig], beta2: float )
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/shampoo_types_test.py:71
Methodtest_illegal_epsilon
( self, cls: type[AdaGradPreconditionerConfig], epsilon: float )
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/shampoo_types_test.py:52
Methodtest_illegal_eval_interp_coeff
(self, eval_interp_coeff: float)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/shampoo_types_test.py:324
Methodtest_illegal_ignored_basis_change_dims
( self, cls: type[EigenvalueCorrectedShampooPreconditionerConfig] )
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/shampoo_types_test.py:180
Methodtest_illegal_inverse_exponent_override
( self, cls: type[ClassicShampooPreconditionerConfig] )
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/shampoo_types_test.py:115
Methodtest_illegal_inverse_exponent_override
( self, cls: type[EigenvalueCorrectedShampooPreconditionerConfig] )
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/shampoo_types_test.py:222
Methodtest_illegal_num_tolerated_failed_amortized_computations
( self, cls: type[BaseShampooPreconditionerConfig] )
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/shampoo_types_test.py:91
Methodtest_illegal_rank_deficient_stability_config
( self, cls: type[EigenvalueCorrectedShampooPreconditionerConfig] )
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/shampoo_types_test.py:258
Methodtest_illegal_target_parameter_dimensionality
( self, cls: type[DistributedConfig], target_parameter_dimensionality: int, er
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/shampoo_types_test.py:368
Methodtest_illegal_tensor_shard_size
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/tests/shampoo_fsdp_distributor_test.py:46
Methodtest_illegal_tensor_shard_size
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/tests/shampoo_hsdp_distributor_test.py:44
Methodtest_illegal_tolerance
( self, cls: type[EigendecompositionConfig], tolerance: float )
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/matrix_functions_types_test.py:37
Methodtest_illegal_tolerance
(self, cls: type[EigenConfig])
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/matrix_functions_types_test.py:57
Methodtest_illegal_train_interp_coeff
( self, cls: type[IterateAveragingConfig], train_interp_coeff: float )
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/shampoo_types_test.py:308
Methodtest_incompatible_replicated_group_size_and_num_trainers_per_group
( self, )
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_hybrid_shard_distributor_test.py:492
Methodtest_incompatible_replicated_group_size_and_num_trainers_per_group
( self, )
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_hsdp_distributor_test.py:324
Methodtest_independent_weight_decay
Test INDEPENDENT weight decay matches AdamW with scaled weight_decay. Independent weight decay applies params *= (1 - (lr / peak_lr) * weight
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/gpu_tests/weight_decay_test.py:215
Methodtest_init_assign_full_parameters
Test that FullyShardLosslessDistributor initializes correctly and blocked params match original tensor size.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_fully_shard_lossless_distributor_test.py:140
Methodtest_init_float_from_dequantized
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/shampoo_quantization_test.py:38
Methodtest_init_from_quantized_tensors
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/shampoo_quantization_test.py:156
Methodtest_initialization
(self, bias: bool, seed: int)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/shampoo_model_utils_test.py:111
Methodtest_inplace_dequantize_quantize
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/shampoo_quantization_test.py:230
Methodtest_instantiate_adam_optimizer
Test instantiating Adam optimizer.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/utils_test.py:150
Methodtest_instantiate_sgd_optimizer
Test instantiating SGD optimizer.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/utils_test.py:132
Methodtest_invalid_batch_size
Test that batched raises ValueError for batch size < 1.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/commons_test.py:50
Methodtest_invalid_computation_dtype
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/shampoo_quantization_test.py:130
Methodtest_invalid_distributed_config
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/distributed_shampoo_test.py:237
Methodtest_invalid_dtype_empty_raises_error
Test that empty dtype string raises ValueError.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/resolvers_test.py:58
Methodtest_invalid_dtype_float_raises_error
Test that 'float' dtype string raises ValueError.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/resolvers_test.py:52
Methodtest_invalid_dtype_int32_raises_error
Test that int32 dtype string raises ValueError.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/resolvers_test.py:46
Methodtest_invalid_eigendecomposition_config
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/matrix_functions_test.py:843
Methodtest_invalid_init
(self, abstract_cls: type[AbstractDataclass])
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/abstract_dataclass_test.py:30
Methodtest_invalid_orthogonalization_config
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/matrix_functions_test.py:949
Methodtest_invalid_preconditioner_config
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/distributed_shampoo_test.py:66
Methodtest_invalid_quantized_data_type
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/shampoo_quantization_test.py:87
Methodtest_invalid_quantized_data_type
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/shampoo_quantization_test.py:110
Methodtest_invalid_quantized_data_type
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/shampoo_quantization_test.py:280
Methodtest_invalid_rank_deficient_stability_config
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/matrix_functions_test.py:196
Methodtest_invalid_strategy_raises_error
Test that invalid strategy strings raise ValueError.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/resolvers_test.py:71
Methodtest_invalid_weight_decay_type_raises_error
Test that invalid weight decay type strings raise ValueError.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/resolvers_test.py:89
Methodtest_invalid_with_incorrect_hyperparameter_setting
( self, incorrect_hyperparameter_setting: dict[str, Any], expected_error_msg: str )
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/distributed_shampoo_test.py:141
Methodtest_inverse_exponent_override
For this example, we modify the one given above such that the inverse_exponent_override = {0: 1.0, 1: 1.0, 2: 1.0}. This effe
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/shampoo_preconditioner_list_test.py:1013
Methodtest_inverse_exponent_override
For this example, we modify the one given above such that the inverse_exponent_override = {0: 1.0, 1: 1.0, 2: 1.0}. This effectively
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/shampoo_preconditioner_list_test.py:1324
Methodtest_iterate_averaging_vs_gpa_adamw
Test that Shampoo with iterate averaging + Adam grafting produces equivalent results to GPA-AdamW. This test validates the theoretical equiva
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/gpu_tests/iterate_averaging_test.py:141
Methodtest_l2_resolution
Test that L2 weight decay type is resolved correctly.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/resolvers_test.py:84
Methodtest_l2_weight_decay
Test L2 weight decay matches PyTorch Adam with L2 regularization. L2 weight decay adds weight_decay * params to the gradient before p
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/gpu_tests/weight_decay_test.py:82
Methodtest_len
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/shampoo_quantization_test.py:170
Methodtest_linear_forward_backward
( self, in_features: int, out_features: int, bias: bool, seed: int )
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/shampoo_model_utils_test.py:100
Methodtest_load_checkpoint_returns_early_if_no_checkpoint_dir
Test that load_checkpoint returns early if checkpoint_dir is None.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/utils_test.py:169
Methodtest_load_checkpoint_returns_early_if_not_shampoo
Test that load_checkpoint returns early for non-Shampoo optimizers.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/utils_test.py:177
Methodtest_load_state_dict
Test that load_state_dict() loads the correct state dict by comparing optimizer.state_dict() and the reference state dict. No
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/distributed_shampoo_test.py:398
Methodtest_load_state_dict
Test that load_state_dict() correctly restores weight_buffer.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/distributed_shampoo_test.py:1335
Methodtest_load_state_dict_preserves_mode
Test that load_state_dict() preserves the caller's train/eval mode.
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/distributed_shampoo_test.py:1506
Methodtest_load_state_dict_with_non_matching_objects
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/optimizer_modules_test.py:111
Methodtest_load_state_dict_with_non_tensor_objects
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/optimizer_modules_test.py:76
Methodtest_load_state_dict_without_non_tensor_objects
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/optimizer_modules_test.py:97
Methodtest_local_block_info_list
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/tests/shampoo_distributor_test.py:109
Methodtest_local_block_info_list
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/tests/shampoo_distributor_test.py:285
Methodtest_local_block_info_list
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/distributor_test_utils.py:131
Methodtest_local_block_info_list
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_ddp_distributor_test.py:885
Methodtest_local_block_info_list
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_fully_shard_distributor_test.py:372
Methodtest_local_block_info_list
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_hybrid_shard_lossless_distributor_test.py:493
Methodtest_local_block_info_list
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_hybrid_shard_distributor_test.py:677
Methodtest_local_block_info_list
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/shampoo_fully_shard_lossless_distributor_test.py:415
Methodtest_local_blocked_params
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/tests/shampoo_distributor_test.py:97
Methodtest_local_blocked_params
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/tests/shampoo_distributor_test.py:272
Methodtest_local_blocked_params
(self)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/distributor/gpu_tests/distributor_test_utils.py:113
← previousnext →2,001–2,100 of 2,320, ranked by callers