MCPcopy Create free account

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

Endpoints41 in github.com/KellerJordan/modded-nanogpt

⨍ Functions 2,320◇ Types & classes 557↳ Endpoints 41
RouteXXT_kernel
triton.autotune( configs=_get_autotune_configs(), key=["M", "K", "a_stride_r", "a_stride_c", "c_stride
train_gpt_medium.py:None
RouteXXT_kernel
triton.autotune( configs=_get_autotune_configs(), key=["M", "K", "a_stride_r", "a_stride_c", "c_stride
records/track_1_short/2025-12-11_NorMuonOptimsAndFixes/profiler-example-traces/train_gpt-profiler-example.py:None
Routeba_plus_cAA_kernel
triton.autotune( configs=_get_autotune_configs(), key=["M", "a_stride_r", "a_stride_c", "c_stride_r",
train_gpt_medium.py:None
Routeba_plus_cAA_kernel
triton.autotune( configs=_get_autotune_configs(), key=["M", "a_stride_r", "a_stride_c", "c_stride_r",
records/track_1_short/2025-12-11_NorMuonOptimsAndFixes/profiler-example-traces/train_gpt-profiler-example.py:None
Routece_fwd_bwd
torch.library.custom_op("nanogpt::ce_fwd_bwd", mutates_args={"losses", "grad_input"})
triton_kernels.py:None
Routetest_adaptive_amortized_computation_frequency
unittest.skip( "RootInvShampooPreconditionerList does not support adaptive computation frequency."
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/shampoo_preconditioner_list_test.py:None
Routetest_convolution_like_dims
parametrize( "threshold, target_tensor_dimensionality, expected_new_tensor_shape", [
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/shampoo_utils_test.py:None
Routetest_defaults_when_env_not_set
patch.dict(os.environ, {}, clear=True)
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/utils_test.py:None
Routetest_distribute_buffer_sizes
parametrize( "blocked_params, group_size, load_balancing_config, expected_result", (
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/utils/tests/shampoo_utils_test.py:None
Routetest_flush_behavior_with_tensorboard
patch.object(loss_metrics_module, "SummaryWriter", return_value=MagicMock())
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/loss_metrics_test.py:None
Routetest_formatter_with_distributed_initialized
patch("distributed_shampoo.examples.utils.dist")
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/utils_test.py:None
Routetest_formatter_without_distributed_initialized
patch("distributed_shampoo.examples.utils.dist")
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/utils_test.py:None
Routetest_forward_pass_mismatched_input_size
parametrize( "model_height, model_width, input_height, input_width", [ (32, 32, 28
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/convnet_test.py:None
Routetest_global_metrics_distributed_behavior
patch.object(torch_distributed_module, "all_reduce")
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/loss_metrics_test.py:None
Routetest_global_metrics_logging
patch.object(loss_metrics_module, "logger")
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/loss_metrics_test.py:None
Routetest_global_metrics_logging_single_worker
patch.object(loss_metrics_module, "logger")
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/loss_metrics_test.py:None
Routetest_illegal_target_parameter_dimensionality
parametrize( "target_parameter_dimensionality, error_msg", [ (-1, "Must be >= 1.")
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/shampoo_types_test.py:None
Routetest_invalid_with_incorrect_hyperparameter_setting
parametrize( "incorrect_hyperparameter_setting, expected_error_msg", [ (
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/distributed_shampoo_test.py:None
Routetest_matrix_eigendecomposition_with_eigh
parametrize("perturb_before_computation", (False, True))
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/matrix_functions_test.py:None
Routetest_matrix_inverse_root
parametrize( "root_inv_config", ( EigenConfig(), # perturb_before_computation=Tru
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/matrix_functions_test.py:None
Routetest_multiple_updates_logging
patch.object(loss_metrics_module, "logger")
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/loss_metrics_test.py:None
Routetest_no_retry_double_precision_raise_exception
mock.patch.object( torch_linalg_module, "eigh", side_effect=RuntimeError("Mock Eigen Error") )
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/matrix_functions_test.py:None
Routetest_noop_hyperparameter_setting_warnings
parametrize( "noop_hyperparameter_setting, expected_warning_msgs", [ (
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/tests/distributed_shampoo_test.py:None
Routetest_perturbation_before_computation
parametrize("perturb_before_computation", (True, False))
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/matrix_functions_test.py:None
Routetest_reads_environment_variables
patch.dict(os.environ, {"LOCAL_RANK": "2", "RANK": "5", "WORLD_SIZE": "8"})
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/utils_test.py:None
Routetest_retry_double_precision_raise_exception
mock.patch.object( torch_linalg_module, "eigh", side_effect=RuntimeError("Mock Eigen Error") )
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/matrix_functions_test.py:None
Routetest_setup_distribution_cpu
patch.object(torch_distributed_module, "init_process_group")
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/utils_test.py:None
Routetest_setup_distribution_cuda
patch.object(torch_distributed_module, "init_process_group")
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/utils_test.py:None
Routetest_setup_per_rank_logging_configures_root_logger
patch("distributed_shampoo.examples.utils.dist")
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/utils_test.py:None
Routetest_setup_per_rank_logging_verbose_sets_debug
patch("distributed_shampoo.examples.utils.dist")
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/utils_test.py:None
Routetest_tensorboard_global_metrics
patch.object(loss_metrics_module, "SummaryWriter", return_value=MagicMock())
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/loss_metrics_test.py:None
Routetest_tensorboard_logging
patch.object(loss_metrics_module, "SummaryWriter", return_value=MagicMock())
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/loss_metrics_test.py:None
Routetest_test_perturbation_before_computation_with_disportionate_epsilon
parametrize("perturb_before_computation", (True, False))
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/preconditioner/tests/matrix_functions_test.py:None
Routetest_window_size_behavior
patch.object(loss_metrics_module, "logger")
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/loss_metrics_test.py:None
Routetest_wrap_model_calls_fully_shard
patch("distributed_shampoo.examples.parallelism.fully_shard")
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/parallelism_test.py:None
Routetest_wrap_model_gloo_backend
patch("distributed_shampoo.examples.parallelism.DDP")
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/parallelism_test.py:None
Routetest_wrap_model_nccl_backend
patch("distributed_shampoo.examples.parallelism.DDP")
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/parallelism_test.py:None
Routetest_wrap_model_passes_mesh
patch("distributed_shampoo.examples.parallelism.fully_shard")
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/parallelism_test.py:None
Routetest_wrap_model_uses_hybrid_shard_strategy
patch( "distributed_shampoo.examples.parallelism.compile_fsdp_parameter_metadata", return_valu
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/parallelism_test.py:None
Routetest_wrap_model_uses_orig_params
patch( "distributed_shampoo.examples.parallelism.compile_fsdp_parameter_metadata", return_valu
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/parallelism_test.py:None
Routetest_wrap_model_with_distributed_config
patch("distributed_shampoo.examples.parallelism.DDP")
records/track_3_optimization/results/20260513_shampoo_1_4_power/distributed_shampoo/examples/tests/parallelism_test.py:None