id: 1
unit: def main()
file: scripts/scaling_benchmarks.py
start line: 0
end line: 0
size: 215 LOC
McCabe index: 63
number of parameters: 0

id: 2
unit: py::array build_mapping_impl()
file: src/nanotron/data/nemo_dataset/helpers.cpp
start line: 217
end line: 450
size: 164 LOC
McCabe index: 30
number of parameters: 9

id: 3
unit: def _forward_inference()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 150 LOC
McCabe index: 9
number of parameters: 8

id: 4
unit: py::array build_blocks_mapping_impl()
file: src/nanotron/data/nemo_dataset/helpers.cpp
start line: 483
end line: 698
size: 149 LOC
McCabe index: 29
number of parameters: 9

id: 5
unit: def forward()
file: src/nanotron/parallel/tensor_parallel/functional.py
start line: 0
end line: 0
size: 111 LOC
McCabe index: 15
number of parameters: 7

id: 6
unit: def create_nanotron_config()
file: slurm_launcher.py
start line: 0
end line: 0
size: 106 LOC
McCabe index: 5
number of parameters: 1

id: 7
unit: def parse_args()
file: slurm_launcher.py
start line: 0
end line: 0
size: 102 LOC
McCabe index: 1
number of parameters: 0

id: 8
unit: def backward()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 96 LOC
McCabe index: 24
number of parameters: 2

id: 9
unit: def data_generator()
file: src/nanotron/data/dataloader.py
start line: 0
end line: 0
size: 91 LOC
McCabe index: 14
number of parameters: 0

id: 10
unit: def _get_text()
file: src/nanotron/data/nemo_dataset/__init__.py
start line: 0
end line: 0
size: 84 LOC
McCabe index: 29
number of parameters: 2

id: 11
unit: def backward()
file: src/nanotron/parallel/tensor_parallel/functional.py
start line: 0
end line: 0
size: 76 LOC
McCabe index: 6
number of parameters: 2

id: 12
unit: def pre_training()
file: src/nanotron/trainer.py
start line: 0
end line: 0
size: 74 LOC
McCabe index: 12
number of parameters: 3

id: 13
unit: def __post_init__()
file: src/nanotron/config/config.py
start line: 0
end line: 0
size: 72 LOC
McCabe index: 24
number of parameters: 1

id: 14
unit: def _update_dataloader_based_on_training_stages()
file: src/nanotron/trainer.py
start line: 0
end line: 0
size: 72 LOC
McCabe index: 10
number of parameters: 3

id: 15
unit: def __call__()
file: src/nanotron/data/clm_collator.py
start line: 0
end line: 0
size: 68 LOC
McCabe index: 27
number of parameters: 3

id: 16
unit: def lr_scheduler_builder()
file: src/nanotron/helpers.py
start line: 0
end line: 0
size: 68 LOC
McCabe index: 22
number of parameters: 3

id: 17
unit: def __call__()
file: src/nanotron/data/clm_collator.py
start line: 0
end line: 0
size: 67 LOC
McCabe index: 33
number of parameters: 3

id: 18
unit: py::array build_sample_idx()
file: src/nanotron/data/nemo_dataset/helpers.cpp
start line: 100
end line: 198
size: 67 LOC
McCabe index: 6
number of parameters: 7

id: 19
unit: def forward()
file: src/nanotron/parallel/pipeline_parallel/block.py
start line: 0
end line: 0
size: 66 LOC
McCabe index: 22
number of parameters: 2

id: 20
unit: def save_weights()
file: src/nanotron/serialize/weights.py
start line: 0
end line: 0
size: 62 LOC
McCabe index: 13
number of parameters: 3

id: 21
unit: def init_model_randomly()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 61 LOC
McCabe index: 20
number of parameters: 2

id: 22
unit: def parse_ckpt_path()
file: src/nanotron/serialize/main.py
start line: 0
end line: 0
size: 61 LOC
McCabe index: 8
number of parameters: 2

id: 23
unit: def write_to_csv()
file: src/nanotron/helpers.py
start line: 0
end line: 0
size: 59 LOC
McCabe index: 17
number of parameters: 4

id: 24
unit: def _forward_packed()
file: src/nanotron/models/qwen.py
start line: 0
end line: 0
size: 59 LOC
McCabe index: 8
number of parameters: 5

id: 25
unit: def get_args()
file: tools/preprocess_data.py
start line: 0
end line: 0
size: 56 LOC
McCabe index: 3
number of parameters: 0

id: 26
unit: def _bwd_kernel()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 55 LOC
McCabe index: 1
number of parameters: 0

id: 27
unit: def set_logger_verbosity_format()
file: src/nanotron/logging/base.py
start line: 0
end line: 0
size: 55 LOC
McCabe index: 25
number of parameters: 2

id: 28
unit: def backward()
file: src/nanotron/parallel/tensor_parallel/functional.py
start line: 0
end line: 0
size: 52 LOC
McCabe index: 11
number of parameters: 2

id: 29
unit: def run()
file: scripts/log_lighteval_to_wandb.py
start line: 0
end line: 0
size: 49 LOC
McCabe index: 13
number of parameters: 1

id: 30
unit: def _partition_parameters()
file: src/nanotron/optim/zero.py
start line: 0
end line: 0
size: 48 LOC
McCabe index: 20
number of parameters: 1

id: 31
unit: def fp32_accum_hook()
file: src/nanotron/optim/gradient_accumulator.py
start line: 0
end line: 0
size: 48 LOC
McCabe index: 10
number of parameters: 2

id: 32
unit: def get_cpu_logitems()
file: src/nanotron/trainer.py
start line: 0
end line: 0
size: 48 LOC
McCabe index: 6
number of parameters: 0

id: 33
unit: def eval_single_checkpoint()
file: src/nanotron/eval/one_job_runner.py
start line: 0
end line: 0
size: 47 LOC
McCabe index: 11
number of parameters: 2

id: 34
unit: def new_build_nanoset_index()
file: src/nanotron/data/nanoset.py
start line: 0
end line: 0
size: 46 LOC
McCabe index: 6
number of parameters: 1

id: 35
unit: def _init_parallel_groups()
file: src/nanotron/parallel/context.py
start line: 0
end line: 0
size: 46 LOC
McCabe index: 15
number of parameters: 1

id: 36
unit: def main()
file: slurm_launcher.py
start line: 0
end line: 0
size: 45 LOC
McCabe index: 10
number of parameters: 0

id: 37
unit: def _fwd_kernel()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 44 LOC
McCabe index: 1
number of parameters: 0

id: 38
unit: def custom_load_state_dict()
file: src/nanotron/optim/base.py
start line: 0
end line: 0
size: 44 LOC
McCabe index: 20
number of parameters: 4

id: 39
unit: def save_checkpoint()
file: src/nanotron/trainer.py
start line: 0
end line: 0
size: 44 LOC
McCabe index: 4
number of parameters: 1

id: 40
unit: def __call__()
file: src/nanotron/generation/sampler.py
start line: 0
end line: 0
size: 43 LOC
McCabe index: 7
number of parameters: 2

id: 41
unit: def push_to_wandb()
file: src/nanotron/eval/upload_to_wandb.py
start line: 0
end line: 0
size: 41 LOC
McCabe index: 15
number of parameters: 6

id: 42
unit: def init_model_randomly()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 41 LOC
McCabe index: 9
number of parameters: 2

id: 43
unit: def init_model_randomly()
file: src/nanotron/models/qwen.py
start line: 0
end line: 0
size: 41 LOC
McCabe index: 9
number of parameters: 2

id: 44
unit: def save_experiment_configs()
file: scripts/scaling_benchmarks.py
start line: 0
end line: 0
size: 41 LOC
McCabe index: 6
number of parameters: 3

id: 45
unit: def get_warmup_percentages()
file: scripts/weka.py
start line: 0
end line: 0
size: 41 LOC
McCabe index: 9
number of parameters: 2

id: 46
unit: void build_blending_indices()
file: src/nanotron/data/nemo_dataset/helpers.cpp
start line: 35
end line: 97
size: 40 LOC
McCabe index: 7
number of parameters: 7

id: 47
unit: def _load_model_checkpoint()
file: src/nanotron/trainer.py
start line: 0
end line: 0
size: 40 LOC
McCabe index: 11
number of parameters: 2

id: 48
unit: def pre_save_checkpoint()
file: src/nanotron/trainer.py
start line: 0
end line: 0
size: 40 LOC
McCabe index: 10
number of parameters: 1

id: 49
unit: def _get_dataset_mix()
file: src/nanotron/data/processing.py
start line: 0
end line: 0
size: 39 LOC
McCabe index: 11
number of parameters: 3

id: 50
unit: def _bwd_kernel_one_col_block()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 36 LOC
McCabe index: 1
number of parameters: 0

id: 51
unit: def _recv_meta()
file: src/nanotron/parallel/pipeline_parallel/p2p.py
start line: 0
end line: 0
size: 36 LOC
McCabe index: 2
number of parameters: 3

id: 52
unit: def _update_cos_sin_cache()
file: src/nanotron/nn/rotary.py
start line: 0
end line: 0
size: 35 LOC
McCabe index: 17
number of parameters: 4

id: 53
unit: def compute_global_hidden_layer_metrics()
file: src/nanotron/metrics_logging.py
start line: 0
end line: 0
size: 34 LOC
McCabe index: 16
number of parameters: 2

id: 54
unit: def _build_sample_idx()
file: src/nanotron/data/nemo_dataset/__init__.py
start line: 0
end line: 0
size: 34 LOC
McCabe index: 11
number of parameters: 7

id: 55
unit: def __call__()
file: src/nanotron/generation/sampler.py
start line: 0
end line: 0
size: 34 LOC
McCabe index: 8
number of parameters: 2

id: 56
unit: def __post_init__()
file: src/nanotron/data/samplers.py
start line: 0
end line: 0
size: 33 LOC
McCabe index: 9
number of parameters: 1

id: 57
unit: def __post_init__()
file: src/nanotron/config/models_config.py
start line: 0
end line: 0
size: 33 LOC
McCabe index: 10
number of parameters: 1

id: 58
unit: def run_communication()
file: src/nanotron/parallel/pipeline_parallel/state.py
start line: 0
end line: 0
size: 33 LOC
McCabe index: 9
number of parameters: 2

id: 59
unit: def writer()
file: src/nanotron/data/nemo_dataset/indexed_dataset.py
start line: 0
end line: 0
size: 32 LOC
McCabe index: 2
number of parameters: 3

id: 60
unit: def __init__()
file: src/nanotron/data/nemo_dataset/indexed_dataset.py
start line: 0
end line: 0
size: 32 LOC
McCabe index: 2
number of parameters: 3

id: 61
unit: def _send_meta()
file: src/nanotron/parallel/pipeline_parallel/p2p.py
start line: 0
end line: 0
size: 32 LOC
McCabe index: 2
number of parameters: 4

id: 62
unit: def serialize()
file: src/nanotron/config/utils_config.py
start line: 0
end line: 0
size: 31 LOC
McCabe index: 16
number of parameters: 1

id: 63
unit: def sanity_check_dataloader()
file: src/nanotron/sanity_checks.py
start line: 0
end line: 0
size: 31 LOC
McCabe index: 7
number of parameters: 3

id: 64
unit: def process_sft()
file: src/nanotron/data/sft_processing.py
start line: 0
end line: 0
size: 30 LOC
McCabe index: 13
number of parameters: 3

id: 65
unit: def __call__()
file: src/nanotron/generation/sampler.py
start line: 0
end line: 0
size: 30 LOC
McCabe index: 9
number of parameters: 2

id: 66
unit: def __post_init__()
file: src/nanotron/config/config.py
start line: 0
end line: 0
size: 30 LOC
McCabe index: 15
number of parameters: 1

id: 67
unit: def backward()
file: src/nanotron/parallel/tensor_parallel/functional.py
start line: 0
end line: 0
size: 30 LOC
McCabe index: 7
number of parameters: 2

id: 68
unit: def _all_gather_params()
file: src/nanotron/optim/zero.py
start line: 0
end line: 0
size: 29 LOC
McCabe index: 11
number of parameters: 1

id: 69
unit: def checkpoint_method()
file: src/nanotron/utils.py
start line: 0
end line: 0
size: 29 LOC
McCabe index: 7
number of parameters: 1

id: 70
unit: def _forward_training()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 29 LOC
McCabe index: 2
number of parameters: 7

id: 71
unit: def __iter__()
file: src/nanotron/data/samplers.py
start line: 0
end line: 0
size: 28 LOC
McCabe index: 10
number of parameters: 1

id: 72
unit: def __init__()
file: src/nanotron/data/nemo_dataset/blendable_dataset.py
start line: 0
end line: 0
size: 28 LOC
McCabe index: 4
number of parameters: 5

id: 73
unit: def _build_dataset()
file: src/nanotron/data/nemo_dataset/__init__.py
start line: 0
end line: 0
size: 27 LOC
McCabe index: 2
number of parameters: 2

id: 74
unit: py::array build_mapping()
file: src/nanotron/data/nemo_dataset/helpers.cpp
start line: 453
end line: 480
size: 27 LOC
McCabe index: 4
number of parameters: 9

id: 75
unit: def forward()
file: src/nanotron/nn/rotary.py
start line: 0
end line: 0
size: 27 LOC
McCabe index: 16
number of parameters: 4

id: 76
unit: def apply_rotary_pos_emb()
file: src/nanotron/nn/rotary.py
start line: 0
end line: 0
size: 27 LOC
McCabe index: 9
number of parameters: 6

id: 77
unit: def __post_init__()
file: src/nanotron/config/config.py
start line: 0
end line: 0
size: 27 LOC
McCabe index: 5
number of parameters: 1

id: 78
unit: def init_model()
file: src/nanotron/trainer.py
start line: 0
end line: 0
size: 27 LOC
McCabe index: 4
number of parameters: 1

id: 79
unit: def update()
file: src/nanotron/s3_checkpoints/s3_mover.py
start line: 0
end line: 0
size: 27 LOC
McCabe index: 8
number of parameters: 1

id: 80
unit: def compute_tensor_stats()
file: src/nanotron/metrics_logging.py
start line: 0
end line: 0
size: 26 LOC
McCabe index: 2
number of parameters: 1

id: 81
unit: def all_ring_pass()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 26 LOC
McCabe index: 7
number of parameters: 5

id: 82
unit: def run_communication()
file: src/nanotron/parallel/pipeline_parallel/state.py
start line: 0
end line: 0
size: 26 LOC
McCabe index: 12
number of parameters: 2

id: 83
unit: def get_train_valid_test_split_()
file: src/nanotron/data/nemo_dataset/dataset_utils.py
start line: 0
end line: 0
size: 25 LOC
McCabe index: 10
number of parameters: 2

id: 84
unit: def unflatten_varlen_lse()
file: src/nanotron/nn/llama3_ring_attention.py
start line: 0
end line: 0
size: 25 LOC
McCabe index: 1
number of parameters: 3

id: 85
unit: def unflatten_varlen_lse()
file: src/nanotron/nn/ring_attention.py
start line: 0
end line: 0
size: 25 LOC
McCabe index: 1
number of parameters: 3

id: 86
unit: def basic_optimizer_builder()
file: src/nanotron/helpers.py
start line: 0
end line: 0
size: 25 LOC
McCabe index: 3
number of parameters: 1

id: 87
unit: def test_equal_dict()
file: src/nanotron/helpers.py
start line: 0
end line: 0
size: 25 LOC
McCabe index: 5
number of parameters: 3

id: 88
unit: def get_profiler()
file: src/nanotron/helpers.py
start line: 0
end line: 0
size: 25 LOC
McCabe index: 4
number of parameters: 1

id: 89
unit: def distributed_wait_for_completion()
file: src/nanotron/s3_checkpoints/s3_mover.py
start line: 0
end line: 0
size: 25 LOC
McCabe index: 12
number of parameters: 2

id: 90
unit: def collect_parameter_metrics()
file: src/nanotron/metrics_logging.py
start line: 0
end line: 0
size: 24 LOC
McCabe index: 13
number of parameters: 2

id: 91
unit: def prepare_sft_dataset()
file: src/nanotron/data/sft_processing.py
start line: 0
end line: 0
size: 24 LOC
McCabe index: 11
number of parameters: 5

id: 92
unit: def test_index_blending()
file: src/nanotron/data/nemo_dataset/blendable_dataset.py
start line: 0
end line: 0
size: 24 LOC
McCabe index: 3
number of parameters: 1

id: 93
unit: def backward()
file: src/nanotron/nn/llama3_ring_attention.py
start line: 0
end line: 0
size: 24 LOC
McCabe index: 1
number of parameters: 3

id: 94
unit: def end()
file: src/nanotron/logging/timers.py
start line: 0
end line: 0
size: 24 LOC
McCabe index: 11
number of parameters: 1

id: 95
unit: def flush()
file: src/nanotron/parallel/pipeline_parallel/p2p.py
start line: 0
end line: 0
size: 24 LOC
McCabe index: 5
number of parameters: 1

id: 96
unit: def main()
file: tools/preprocess_data.py
start line: 0
end line: 0
size: 24 LOC
McCabe index: 2
number of parameters: 1

id: 97
unit: py::array build_blocks_mapping()
file: src/nanotron/data/nemo_dataset/helpers.cpp
start line: 700
end line: 723
size: 23 LOC
McCabe index: 4
number of parameters: 9

id: 98
unit: def sync_gradients_across_dp()
file: src/nanotron/optim/gradient_accumulator.py
start line: 0
end line: 0
size: 23 LOC
McCabe index: 7
number of parameters: 4

id: 99
unit: def __call__()
file: src/nanotron/generation/sampler.py
start line: 0
end line: 0
size: 23 LOC
McCabe index: 7
number of parameters: 2

id: 100
unit: def isend_tensors()
file: src/nanotron/parallel/pipeline_parallel/p2p.py
start line: 0
end line: 0
size: 23 LOC
McCabe index: 5
number of parameters: 4

id: 101
unit: def init_rotary_embeddings()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 23 LOC
McCabe index: 5
number of parameters: 1

id: 102
unit: def _post_uploading()
file: src/nanotron/s3_checkpoints/s3_mover.py
start line: 0
end line: 0
size: 23 LOC
McCabe index: 9
number of parameters: 1

id: 103
unit: def __iter__()
file: src/nanotron/data/samplers.py
start line: 0
end line: 0
size: 22 LOC
McCabe index: 9
number of parameters: 1

id: 104
unit: def build_dataset()
file: src/nanotron/data/nemo_dataset/__init__.py
start line: 0
end line: 0
size: 21 LOC
McCabe index: 3
number of parameters: 2

id: 105
unit: def elapsed()
file: src/nanotron/logging/timers.py
start line: 0
end line: 0
size: 21 LOC
McCabe index: 8
number of parameters: 1

id: 106
unit: def initialize_torch_distributed()
file: src/nanotron/distributed.py
start line: 0
end line: 0
size: 21 LOC
McCabe index: 3
number of parameters: 0

id: 107
unit: def llama3_flash_attn_varlen_backward()
file: src/nanotron/nn/llama3_ring_attention.py
start line: 0
end line: 0
size: 20 LOC
McCabe index: 1
number of parameters: 18

id: 108
unit: def flatten_varlen_lse()
file: src/nanotron/nn/llama3_ring_attention.py
start line: 0
end line: 0
size: 20 LOC
McCabe index: 1
number of parameters: 2

id: 109
unit: def backward()
file: src/nanotron/nn/ring_attention.py
start line: 0
end line: 0
size: 20 LOC
McCabe index: 1
number of parameters: 3

id: 110
unit: def flatten_varlen_lse()
file: src/nanotron/nn/ring_attention.py
start line: 0
end line: 0
size: 20 LOC
McCabe index: 1
number of parameters: 2

id: 111
unit: def print_config_details()
file: src/nanotron/config/config.py
start line: 0
end line: 0
size: 20 LOC
McCabe index: 1
number of parameters: 1

id: 112
unit: def from_metadata()
file: src/nanotron/parallel/pipeline_parallel/p2p.py
start line: 0
end line: 0
size: 20 LOC
McCabe index: 1
number of parameters: 3

id: 113
unit: def forward()
file: src/nanotron/nn/llama3_ring_attention.py
start line: 0
end line: 0
size: 19 LOC
McCabe index: 1
number of parameters: 0

id: 114
unit: def all_gather_variable_dim()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 19 LOC
McCabe index: 3
number of parameters: 3

id: 115
unit: def start()
file: src/nanotron/logging/timers.py
start line: 0
end line: 0
size: 19 LOC
McCabe index: 8
number of parameters: 1

id: 116
unit: def post_init()
file: src/nanotron/trainer.py
start line: 0
end line: 0
size: 19 LOC
McCabe index: 6
number of parameters: 1

id: 117
unit: def forward()
file: src/nanotron/parallel/tensor_parallel/nn.py
start line: 0
end line: 0
size: 19 LOC
McCabe index: 7
number of parameters: 2

id: 118
unit: def get_flops_per_sec()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 19 LOC
McCabe index: 2
number of parameters: 4

id: 119
unit: def get_flops_per_sec()
file: src/nanotron/models/qwen.py
start line: 0
end line: 0
size: 19 LOC
McCabe index: 3
number of parameters: 4

id: 120
unit: def _post_downloading()
file: src/nanotron/s3_checkpoints/s3_mover.py
start line: 0
end line: 0
size: 19 LOC
McCabe index: 10
number of parameters: 1

id: 121
unit: def _post_removing()
file: src/nanotron/s3_checkpoints/s3_mover.py
start line: 0
end line: 0
size: 19 LOC
McCabe index: 6
number of parameters: 1

id: 122
unit: def __init__()
file: src/nanotron/fp8/linear.py
start line: 0
end line: 0
size: 19 LOC
McCabe index: 5
number of parameters: 5

id: 123
unit: def llama3_flash_attn_varlen_func()
file: src/nanotron/nn/llama3_ring_attention.py
start line: 0
end line: 0
size: 18 LOC
McCabe index: 1
number of parameters: 14

id: 124
unit: def log_libraries_versions()
file: src/nanotron/logging/base.py
start line: 0
end line: 0
size: 18 LOC
McCabe index: 2
number of parameters: 1

id: 125
unit: def log_all()
file: src/nanotron/logging/timers.py
start line: 0
end line: 0
size: 18 LOC
McCabe index: 10
number of parameters: 4

id: 126
unit: def forward()
file: src/nanotron/parallel/tensor_parallel/distributed_differentiable_primitives.py
start line: 0
end line: 0
size: 18 LOC
McCabe index: 3
number of parameters: 3

id: 127
unit: def forward()
file: src/nanotron/parallel/tensor_parallel/distributed_differentiable_primitives.py
start line: 0
end line: 0
size: 18 LOC
McCabe index: 3
number of parameters: 3

id: 128
unit: def tie_custom_params()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 18 LOC
McCabe index: 5
number of parameters: 1

id: 129
unit: def __iter__()
file: src/nanotron/data/samplers.py
start line: 0
end line: 0
size: 17 LOC
McCabe index: 8
number of parameters: 1

id: 130
unit: def llama3_flash_attn_varlen_forward()
file: src/nanotron/nn/llama3_ring_attention.py
start line: 0
end line: 0
size: 17 LOC
McCabe index: 1
number of parameters: 15

id: 131
unit: def llama3_flash_attn_varlen_kvpacked_func()
file: src/nanotron/nn/llama3_ring_attention.py
start line: 0
end line: 0
size: 17 LOC
McCabe index: 1
number of parameters: 13

id: 132
unit: def flash_attn_forward()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 17 LOC
McCabe index: 1
number of parameters: 0

id: 133
unit: def flash_attn_backward()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 17 LOC
McCabe index: 1
number of parameters: 0

id: 134
unit: def get_logger()
file: src/nanotron/logging/base.py
start line: 0
end line: 0
size: 17 LOC
McCabe index: 8
number of parameters: 2

id: 135
unit: def get_position_ids()
file: src/nanotron/generation/decode.py
start line: 0
end line: 0
size: 17 LOC
McCabe index: 9
number of parameters: 2

id: 136
unit: def cos_sin()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 17 LOC
McCabe index: 3
number of parameters: 5

id: 137
unit: def __init__()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 17 LOC
McCabe index: 4
number of parameters: 4

id: 138
unit: def forward()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 17 LOC
McCabe index: 3
number of parameters: 2

id: 139
unit: def _start_removing()
file: src/nanotron/s3_checkpoints/s3_mover.py
start line: 0
end line: 0
size: 17 LOC
McCabe index: 8
number of parameters: 1

id: 140
unit: def __init__()
file: src/nanotron/scaling/parametrization.py
start line: 0
end line: 0
size: 17 LOC
McCabe index: 3
number of parameters: 2

id: 141
unit: def create_config()
file: scripts/scaling_benchmarks.py
start line: 0
end line: 0
size: 17 LOC
McCabe index: 1
number of parameters: 0

id: 142
unit: def __init__()
file: src/nanotron/data/tokenized_bytes.py
start line: 0
end line: 0
size: 16 LOC
McCabe index: 1
number of parameters: 16

id: 143
unit: def get_tb_dataloader()
file: src/nanotron/data/tokenized_bytes.py
start line: 0
end line: 0
size: 16 LOC
McCabe index: 1
number of parameters: 0

id: 144
unit: def __getitem__()
file: src/nanotron/data/nemo_dataset/indexed_dataset.py
start line: 0
end line: 0
size: 16 LOC
McCabe index: 4
number of parameters: 2

id: 145
unit: def llama3_flash_attn_varlen_qkvpacked_func()
file: src/nanotron/nn/llama3_ring_attention.py
start line: 0
end line: 0
size: 16 LOC
McCabe index: 1
number of parameters: 12

id: 146
unit: def ring_flash_attn_varlen_backward()
file: src/nanotron/nn/ring_attention.py
start line: 0
end line: 0
size: 16 LOC
McCabe index: 1
number of parameters: 14

id: 147
unit: def ring_flash_attn_varlen_func()
file: src/nanotron/nn/ring_attention.py
start line: 0
end line: 0
size: 16 LOC
McCabe index: 1
number of parameters: 11

id: 148
unit: def _bwd_preprocess_do_o_dot()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 16 LOC
McCabe index: 1
number of parameters: 0

id: 149
unit: def log()
file: src/nanotron/logging/timers.py
start line: 0
end line: 0
size: 16 LOC
McCabe index: 8
number of parameters: 5

id: 150
unit: def get_all_comps()
file: src/nanotron/helpers.py
start line: 0
end line: 0
size: 16 LOC
McCabe index: 6
number of parameters: 1

id: 151
unit: def _accumulate_grad()
file: src/nanotron/optim/gradient_accumulator.py
start line: 0
end line: 0
size: 16 LOC
McCabe index: 2
number of parameters: 3

id: 152
unit: def __post_init__()
file: src/nanotron/config/lighteval_config.py
start line: 0
end line: 0
size: 16 LOC
McCabe index: 6
number of parameters: 1

id: 153
unit: def check_buffers_empty()
file: src/nanotron/parallel/pipeline_parallel/state.py
start line: 0
end line: 0
size: 16 LOC
McCabe index: 1
number of parameters: 1

id: 154
unit: def get_flops_per_sec()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 16 LOC
McCabe index: 2
number of parameters: 4

id: 155
unit: def _core_forward()
file: src/nanotron/models/qwen.py
start line: 0
end line: 0
size: 16 LOC
McCabe index: 6
number of parameters: 2

id: 156
unit: def _start_downloading()
file: src/nanotron/s3_checkpoints/s3_mover.py
start line: 0
end line: 0
size: 16 LOC
McCabe index: 4
number of parameters: 2

id: 157
unit: def build_nanoset_index()
file: src/nanotron/data/nanoset.py
start line: 0
end line: 0
size: 15 LOC
McCabe index: 5
number of parameters: 1

id: 158
unit: def get_train_dataloader()
file: src/nanotron/data/dataloader.py
start line: 0
end line: 0
size: 15 LOC
McCabe index: 1
number of parameters: 0

id: 159
unit: def _build_shuffle_idx()
file: src/nanotron/data/nemo_dataset/__init__.py
start line: 0
end line: 0
size: 15 LOC
McCabe index: 3
number of parameters: 3

id: 160
unit: def forward()
file: src/nanotron/nn/ring_attention.py
start line: 0
end line: 0
size: 15 LOC
McCabe index: 1
number of parameters: 0

id: 161
unit: def ring_flash_attn_cuda()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 15 LOC
McCabe index: 1
number of parameters: 0

id: 162
unit: def set_formatter()
file: src/nanotron/logging/base.py
start line: 0
end line: 0
size: 15 LOC
McCabe index: 6
number of parameters: 1

id: 163
unit: def load_state_dict()
file: src/nanotron/optim/named_optimizer.py
start line: 0
end line: 0
size: 15 LOC
McCabe index: 3
number of parameters: 4

id: 164
unit: def _set_grad()
file: src/nanotron/optim/zero.py
start line: 0
end line: 0
size: 15 LOC
McCabe index: 3
number of parameters: 2

id: 165
unit: def __post_init__()
file: src/nanotron/config/config.py
start line: 0
end line: 0
size: 15 LOC
McCabe index: 6
number of parameters: 1

id: 166
unit: def create_empty_storage()
file: src/nanotron/parallel/pipeline_parallel/p2p.py
start line: 0
end line: 0
size: 15 LOC
McCabe index: 3
number of parameters: 2

id: 167
unit: def _send_recv_second_metadata()
file: src/nanotron/parallel/pipeline_parallel/p2p.py
start line: 0
end line: 0
size: 15 LOC
McCabe index: 5
number of parameters: 2

id: 168
unit: def _send_recv_data()
file: src/nanotron/parallel/pipeline_parallel/p2p.py
start line: 0
end line: 0
size: 15 LOC
McCabe index: 4
number of parameters: 2

id: 169
unit: def __init__()
file: src/nanotron/parallel/tensor_parallel/nn.py
start line: 0
end line: 0
size: 15 LOC
McCabe index: 1
number of parameters: 0

id: 170
unit: def create_new_group()
file: src/nanotron/parallel/context.py
start line: 0
end line: 0
size: 15 LOC
McCabe index: 4
number of parameters: 2

id: 171
unit: def get_sliced_parameter()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 15 LOC
McCabe index: 3
number of parameters: 2

id: 172
unit: def _acquire_lock()
file: src/nanotron/s3_checkpoints/s3_mover.py
start line: 0
end line: 0
size: 15 LOC
McCabe index: 4
number of parameters: 2

id: 173
unit: def save_meta()
file: src/nanotron/serialize/metadata.py
start line: 0
end line: 0
size: 15 LOC
McCabe index: 2
number of parameters: 3

id: 174
unit: def build_dataset()
file: src/nanotron/data/tokenized_bytes.py
start line: 0
end line: 0
size: 14 LOC
McCabe index: 1
number of parameters: 0

id: 175
unit: def get_indexed_dataset()
file: src/nanotron/data/nemo_dataset/__init__.py
start line: 0
end line: 0
size: 14 LOC
McCabe index: 1
number of parameters: 2

id: 176
unit: def _core_forward()
file: src/nanotron/nn/moe.py
start line: 0
end line: 0
size: 14 LOC
McCabe index: 4
number of parameters: 2

id: 177
unit: def forward()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 14 LOC
McCabe index: 1
number of parameters: 0

id: 178
unit: def _init_model_instance()
file: src/nanotron/trainer.py
start line: 0
end line: 0
size: 14 LOC
McCabe index: 1
number of parameters: 1

id: 179
unit: def get_pp_rank_of()
file: src/nanotron/parallel/pipeline_parallel/utils.py
start line: 0
end line: 0
size: 14 LOC
McCabe index: 6
number of parameters: 2

id: 180
unit: def attach_pipeline_state_to_model()
file: src/nanotron/parallel/pipeline_parallel/context_manager.py
start line: 0
end line: 0
size: 14 LOC
McCabe index: 4
number of parameters: 2

id: 181
unit: def to_first_metadata()
file: src/nanotron/parallel/pipeline_parallel/p2p.py
start line: 0
end line: 0
size: 14 LOC
McCabe index: 2
number of parameters: 2

id: 182
unit: def backward()
file: src/nanotron/parallel/tensor_parallel/functional.py
start line: 0
end line: 0
size: 14 LOC
McCabe index: 3
number of parameters: 3

id: 183
unit: def create_pg_for_tied_weights()
file: src/nanotron/parallel/tied_parameters.py
start line: 0
end line: 0
size: 14 LOC
McCabe index: 6
number of parameters: 2

id: 184
unit: def __init__()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 14 LOC
McCabe index: 1
number of parameters: 3

id: 185
unit: def named_modules_in_pp_rank()
file: src/nanotron/models/base.py
start line: 0
end line: 0
size: 14 LOC
McCabe index: 5
number of parameters: 1

id: 186
unit: def _dispatch_tokens()
file: src/nanotron/models/qwen.py
start line: 0
end line: 0
size: 14 LOC
McCabe index: 5
number of parameters: 4

id: 187
unit: def backward()
file: src/nanotron/fp8/linear.py
start line: 0
end line: 0
size: 14 LOC
McCabe index: 2
number of parameters: 3

id: 188
unit: def load_meta()
file: src/nanotron/serialize/metadata.py
start line: 0
end line: 0
size: 14 LOC
McCabe index: 1
number of parameters: 2

id: 189
unit: def get_checkpoint_version()
file: src/nanotron/serialize/weights.py
start line: 0
end line: 0
size: 14 LOC
McCabe index: 2
number of parameters: 3

id: 190
unit: def get_args()
file: scripts/log_lighteval_to_wandb.py
start line: 0
end line: 0
size: 14 LOC
McCabe index: 1
number of parameters: 0

id: 191
unit: def __init__()
file: src/nanotron/data/tokenized_bytes.py
start line: 0
end line: 0
size: 13 LOC
McCabe index: 1
number of parameters: 0

id: 192
unit: def get_tb_datasets()
file: src/nanotron/data/tokenized_bytes.py
start line: 0
end line: 0
size: 13 LOC
McCabe index: 1
number of parameters: 0

id: 193
unit: def _stream_file()
file: src/nanotron/data/s3_utils.py
start line: 0
end line: 0
size: 13 LOC
McCabe index: 4
number of parameters: 4

id: 194
unit: def print_nanoset_info()
file: src/nanotron/data/nanoset.py
start line: 0
end line: 0
size: 13 LOC
McCabe index: 2
number of parameters: 1

id: 195
unit: def __init__()
file: src/nanotron/data/nemo_dataset/__init__.py
start line: 0
end line: 0
size: 13 LOC
McCabe index: 1
number of parameters: 0

id: 196
unit: def build_nanoset_dataloader()
file: src/nanotron/data/dataloader_builder.py
start line: 0
end line: 0
size: 13 LOC
McCabe index: 1
number of parameters: 0

id: 197
unit: def ring_flash_attn_varlen_forward()
file: src/nanotron/nn/ring_attention.py
start line: 0
end line: 0
size: 13 LOC
McCabe index: 1
number of parameters: 11

id: 198
unit: def flex_attention_forward()
file: src/nanotron/nn/attention.py
start line: 0
end line: 0
size: 13 LOC
McCabe index: 1
number of parameters: 0

id: 199
unit: def zero_grad()
file: src/nanotron/optim/zero.py
start line: 0
end line: 0
size: 13 LOC
McCabe index: 9
number of parameters: 1

id: 200
unit: def __new__()
file: src/nanotron/optim/zero.py
start line: 0
end line: 0
size: 13 LOC
McCabe index: 1
number of parameters: 4

id: 201
unit: def forward()
file: src/nanotron/parallel/tensor_parallel/functional.py
start line: 0
end line: 0
size: 13 LOC
McCabe index: 3
number of parameters: 6

id: 202
unit: def unshape()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 13 LOC
McCabe index: 2
number of parameters: 1

id: 203
unit: def get_named_params_with_correct_tied()
file: src/nanotron/models/base.py
start line: 0
end line: 0
size: 13 LOC
McCabe index: 3
number of parameters: 1

id: 204
unit: def __init__()
file: src/nanotron/s3_checkpoints/s3_mover.py
start line: 0
end line: 0
size: 13 LOC
McCabe index: 1
number of parameters: 0

id: 205
unit: def __new__()
file: src/nanotron/fp8/tensor.py
start line: 0
end line: 0
size: 13 LOC
McCabe index: 3
number of parameters: 3

id: 206
unit: def __eq__()
file: src/nanotron/random.py
start line: 0
end line: 0
size: 13 LOC
McCabe index: 9
number of parameters: 2

id: 207
unit: def branch_random_state()
file: src/nanotron/random.py
start line: 0
end line: 0
size: 13 LOC
McCabe index: 2
number of parameters: 3

id: 208
unit: def _parametrize_mup_weight()
file: src/nanotron/scaling/parametrization.py
start line: 0
end line: 0
size: 13 LOC
McCabe index: 4
number of parameters: 3

id: 209
unit: def save()
file: src/nanotron/serialize/main.py
start line: 0
end line: 0
size: 13 LOC
McCabe index: 1
number of parameters: 0

id: 210
unit: def _get_next_from_stream()
file: src/nanotron/data/tokenized_bytes.py
start line: 0
end line: 0
size: 12 LOC
McCabe index: 4
number of parameters: 1

id: 211
unit: def __len__()
file: src/nanotron/data/samplers.py
start line: 0
end line: 0
size: 12 LOC
McCabe index: 4
number of parameters: 1

id: 212
unit: def __len__()
file: src/nanotron/data/samplers.py
start line: 0
end line: 0
size: 12 LOC
McCabe index: 4
number of parameters: 1

id: 213
unit: def __len__()
file: src/nanotron/data/samplers.py
start line: 0
end line: 0
size: 12 LOC
McCabe index: 4
number of parameters: 1

id: 214
unit: def group_texts()
file: src/nanotron/data/processing.py
start line: 0
end line: 0
size: 12 LOC
McCabe index: 6
number of parameters: 2

id: 215
unit: def _build_index_mappings()
file: src/nanotron/data/nemo_dataset/__init__.py
start line: 0
end line: 0
size: 12 LOC
McCabe index: 1
number of parameters: 0

id: 216
unit: def permute()
file: src/nanotron/data/nemo_dataset/__init__.py
start line: 0
end line: 0
size: 12 LOC
McCabe index: 1
number of parameters: 0

id: 217
unit: inline int32_t get_target_sample_len()
file: src/nanotron/data/nemo_dataset/helpers.cpp
start line: 201
end line: 213
size: 12 LOC
McCabe index: 3
number of parameters: 3

id: 218
unit: def _bwd_store_dk_dv()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 12 LOC
McCabe index: 1
number of parameters: 0

id: 219
unit: def set_verbosity()
file: src/nanotron/logging/base.py
start line: 0
end line: 0
size: 12 LOC
McCabe index: 7
number of parameters: 1

id: 220
unit: def human_format()
file: src/nanotron/logging/base.py
start line: 0
end line: 0
size: 12 LOC
McCabe index: 5
number of parameters: 3

id: 221
unit: def grad_optimizer_builder()
file: src/nanotron/helpers.py
start line: 0
end line: 0
size: 12 LOC
McCabe index: 1
number of parameters: 1

id: 222
unit: def __init__()
file: src/nanotron/optim/inherit_from_other_optimizer.py
start line: 0
end line: 0
size: 12 LOC
McCabe index: 4
number of parameters: 4

id: 223
unit: def attach_store()
file: src/nanotron/generation/generate_store.py
start line: 0
end line: 0
size: 12 LOC
McCabe index: 4
number of parameters: 2

id: 224
unit: def _calculate_model_params()
file: src/nanotron/config/config.py
start line: 0
end line: 0
size: 12 LOC
McCabe index: 2
number of parameters: 1

id: 225
unit: def __post_init__()
file: src/nanotron/config/models_config.py
start line: 0
end line: 0
size: 12 LOC
McCabe index: 6
number of parameters: 1

id: 226
unit: def __new__()
file: src/nanotron/parallel/parameters.py
start line: 0
end line: 0
size: 12 LOC
McCabe index: 2
number of parameters: 3

id: 227
unit: def __init__()
file: src/nanotron/parallel/tensor_parallel/nn.py
start line: 0
end line: 0
size: 12 LOC
McCabe index: 1
number of parameters: 0

id: 228
unit: def ddp_trigger_sync_in_bwd()
file: src/nanotron/parallel/data_parallel/utils.py
start line: 0
end line: 0
size: 12 LOC
McCabe index: 1
number of parameters: 1

id: 229
unit: def forward()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 12 LOC
McCabe index: 3
number of parameters: 3

id: 230
unit: def log_modules()
file: src/nanotron/models/base.py
start line: 0
end line: 0
size: 12 LOC
McCabe index: 3
number of parameters: 4

id: 231
unit: def assert_fail_except_rank_with()
file: src/nanotron/sanity_checks.py
start line: 0
end line: 0
size: 12 LOC
McCabe index: 5
number of parameters: 3

id: 232
unit: def state_dict_to_device()
file: src/nanotron/serialize/optimizer.py
start line: 0
end line: 0
size: 12 LOC
McCabe index: 3
number of parameters: 2

id: 233
unit: def update_checkpoint()
file: scripts/fix_checkpoint_bad_naming.py
start line: 0
end line: 0
size: 12 LOC
McCabe index: 5
number of parameters: 1

id: 234
unit: def check_params()
file: scripts/scaling_benchmarks.py
start line: 0
end line: 0
size: 12 LOC
McCabe index: 3
number of parameters: 1

id: 235
unit: def warmup_datasets()
file: scripts/weka.py
start line: 0
end line: 0
size: 12 LOC
McCabe index: 5
number of parameters: 2

id: 236
unit: def get_attribute_by_path()
file: src/nanotron/metrics_logging.py
start line: 0
end line: 0
size: 11 LOC
McCabe index: 5
number of parameters: 2

id: 237
unit: def collect_embeddings_metrics()
file: src/nanotron/metrics_logging.py
start line: 0
end line: 0
size: 11 LOC
McCabe index: 4
number of parameters: 2

id: 238
unit: def __init__()
file: src/nanotron/data/nanoset.py
start line: 0
end line: 0
size: 11 LOC
McCabe index: 1
number of parameters: 0

id: 239
unit: def _build_doc_idx()
file: src/nanotron/data/nemo_dataset/__init__.py
start line: 0
end line: 0
size: 11 LOC
McCabe index: 3
number of parameters: 4

id: 240
unit: def __init__()
file: src/nanotron/nn/llama3_ring_attention.py
start line: 0
end line: 0
size: 11 LOC
McCabe index: 2
number of parameters: 2

id: 241
unit: def __init__()
file: src/nanotron/nn/ring_attention.py
start line: 0
end line: 0
size: 11 LOC
McCabe index: 2
number of parameters: 2

id: 242
unit: def flash_attention_forward()
file: src/nanotron/nn/attention.py
start line: 0
end line: 0
size: 11 LOC
McCabe index: 1
number of parameters: 0

id: 243
unit: def get_attention_mask()
file: src/nanotron/nn/attention.py
start line: 0
end line: 0
size: 11 LOC
McCabe index: 3
number of parameters: 2

id: 244
unit: def _get_default_logging_level()
file: src/nanotron/logging/base.py
start line: 0
end line: 0
size: 11 LOC
McCabe index: 3
number of parameters: 0

id: 245
unit: def log_memory()
file: src/nanotron/logging/base.py
start line: 0
end line: 0
size: 11 LOC
McCabe index: 1
number of parameters: 2

id: 246
unit: def _vocab_size_with_padding()
file: src/nanotron/helpers.py
start line: 0
end line: 0
size: 11 LOC
McCabe index: 2
number of parameters: 3

id: 247
unit: def decode_text()
file: src/nanotron/generation/decode.py
start line: 0
end line: 0
size: 11 LOC
McCabe index: 1
number of parameters: 0

id: 248
unit: def __post_init__()
file: src/nanotron/config/parallelism_config.py
start line: 0
end line: 0
size: 11 LOC
McCabe index: 6
number of parameters: 1

id: 249
unit: def post_save_checkpoint()
file: src/nanotron/trainer.py
start line: 0
end line: 0
size: 11 LOC
McCabe index: 7
number of parameters: 1

id: 250
unit: def view_as_contiguous()
file: src/nanotron/parallel/pipeline_parallel/p2p.py
start line: 0
end line: 0
size: 11 LOC
McCabe index: 1
number of parameters: 1

id: 251
unit: def __init__()
file: src/nanotron/parallel/tensor_parallel/nn.py
start line: 0
end line: 0
size: 11 LOC
McCabe index: 1
number of parameters: 0

id: 252
unit: def _mark_all_parameters_in_module_as_sharded()
file: src/nanotron/parallel/tensor_parallel/nn.py
start line: 0
end line: 0
size: 11 LOC
McCabe index: 3
number of parameters: 2

id: 253
unit: def init_rotary_embeddings()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 11 LOC
McCabe index: 1
number of parameters: 1

id: 254
unit: def init_rotary_embeddings()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 11 LOC
McCabe index: 5
number of parameters: 1

id: 255
unit: def __init__()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 11 LOC
McCabe index: 1
number of parameters: 0

id: 256
unit: def get_state_as_int()
file: src/nanotron/s3_checkpoints/s3_mover.py
start line: 0
end line: 0
size: 11 LOC
McCabe index: 5
number of parameters: 1

id: 257
unit: def convert_torch_dtype_to_te_dtype()
file: src/nanotron/fp8/tensor.py
start line: 0
end line: 0
size: 11 LOC
McCabe index: 1
number of parameters: 1

id: 258
unit: def __getitem__()
file: src/nanotron/data/tokenized_bytes.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 10
number of parameters: 2

id: 259
unit: def get_sampler()
file: src/nanotron/data/samplers.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 1
number of parameters: 0

id: 260
unit: def __init__()
file: src/nanotron/data/samplers.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 1
number of parameters: 0

id: 261
unit: def dummy_infinite_data_generator()
file: src/nanotron/data/dataloader.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 1
number of parameters: 0

id: 262
unit: def _do_init()
file: src/nanotron/data/nemo_dataset/indexed_dataset.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 2
number of parameters: 3

id: 263
unit: def build_dataset()
file: src/nanotron/data/nemo_dataset/__init__.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 1
number of parameters: 0

id: 264
unit: def build_train_valid_test_datasets()
file: src/nanotron/data/nemo_dataset/__init__.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 1
number of parameters: 0

id: 265
unit: def _build_train_valid_test_datasets()
file: src/nanotron/data/nemo_dataset/__init__.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 1
number of parameters: 0

id: 266
unit: def flatten_kernel()
file: src/nanotron/nn/llama3_ring_attention.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 1
number of parameters: 0

id: 267
unit: def unflatten_kernel()
file: src/nanotron/nn/llama3_ring_attention.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 1
number of parameters: 0

id: 268
unit: def send_recv()
file: src/nanotron/nn/llama3_ring_attention.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 2
number of parameters: 3

id: 269
unit: def flatten_kernel()
file: src/nanotron/nn/ring_attention.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 1
number of parameters: 0

id: 270
unit: def unflatten_kernel()
file: src/nanotron/nn/ring_attention.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 1
number of parameters: 0

id: 271
unit: def send_recv()
file: src/nanotron/nn/ring_attention.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 2
number of parameters: 3

id: 272
unit: def __init__()
file: src/nanotron/nn/moe.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 1
number of parameters: 3

id: 273
unit: def send_and_receive_()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 3
number of parameters: 5

id: 274
unit: def forward()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 3
number of parameters: 4

id: 275
unit: def sdpa_attention_forward()
file: src/nanotron/nn/attention.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 1
number of parameters: 0

id: 276
unit: def get_tbi_logs()
file: src/nanotron/logging/logmixin.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 9
number of parameters: 2

id: 277
unit: def log_rank()
file: src/nanotron/logging/base.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 1
number of parameters: 0

id: 278
unit: def decode_tokenized()
file: src/nanotron/generation/decode.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 1
number of parameters: 0

id: 279
unit: def find_free_port()
file: src/nanotron/utils.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 3
number of parameters: 2

id: 280
unit: def check_buffers_empty()
file: src/nanotron/parallel/pipeline_parallel/state.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 1
number of parameters: 1

id: 281
unit: def get_sort_key()
file: src/nanotron/parallel/pipeline_parallel/block.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 2
number of parameters: 1

id: 282
unit: def add_send()
file: src/nanotron/parallel/pipeline_parallel/p2p.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 1
number of parameters: 4

id: 283
unit: def forward()
file: src/nanotron/parallel/tensor_parallel/nn.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 1
number of parameters: 2

id: 284
unit: def _mark_all_parameters_in_module_as_tied()
file: src/nanotron/parallel/tensor_parallel/nn.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 4
number of parameters: 1

id: 285
unit: def __init__()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 2
number of parameters: 4

id: 286
unit: def get_block_compute_costs()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 1
number of parameters: 1

id: 287
unit: def get_block_compute_costs()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 2
number of parameters: 1

id: 288
unit: def get_flops()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 1
number of parameters: 0

id: 289
unit: def forward()
file: src/nanotron/models/qwen.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 1
number of parameters: 0

id: 290
unit: def __init__()
file: src/nanotron/models/qwen.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 2
number of parameters: 4

id: 291
unit: def get_block_compute_costs()
file: src/nanotron/models/qwen.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 1
number of parameters: 1

id: 292
unit: def _popen()
file: src/nanotron/s3_checkpoints/s3_mover.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 2
number of parameters: 2

id: 293
unit: def get_current_stdout()
file: src/nanotron/s3_checkpoints/s3_mover.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 5
number of parameters: 1

id: 294
unit: def _parametrize_grouped_mlp()
file: src/nanotron/scaling/parametrization.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 4
number of parameters: 3

id: 295
unit: def _get_mup_lr()
file: src/nanotron/scaling/parametrization.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 3
number of parameters: 3

id: 296
unit: def __post_init__()
file: src/nanotron/serialize/metadata.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 6
number of parameters: 1

id: 297
unit: def get_args()
file: run_train.py
start line: 0
end line: 0
size: 10 LOC
McCabe index: 1
number of parameters: 0

id: 298
unit: def compute_kurtosis()
file: src/nanotron/metrics_logging.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 2
number of parameters: 1

id: 299
unit: def __init__()
file: src/nanotron/data/tokenized_bytes.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 1
number of parameters: 0

id: 300
unit: def __init__()
file: src/nanotron/data/nemo_dataset/blendable_dataset.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 1
number of parameters: 0

id: 301
unit: def make_indexed_dataset()
file: src/nanotron/data/nemo_dataset/indexed_dataset.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 3
number of parameters: 2

id: 302
unit: def print_split_stats()
file: src/nanotron/data/nemo_dataset/__init__.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 1
number of parameters: 2

id: 303
unit: def rotate_half()
file: src/nanotron/nn/rotary.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 2
number of parameters: 2

id: 304
unit: def __init__()
file: src/nanotron/nn/rotary.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 1
number of parameters: 0

id: 305
unit: def split_by_rank()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 3
number of parameters: 1

id: 306
unit: def lengths_to_offsets()
file: src/nanotron/nn/flex_attention.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 2
number of parameters: 2

id: 307
unit: def generate_doc_mask_mod()
file: src/nanotron/nn/flex_attention.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 1
number of parameters: 2

id: 308
unit: def _create_timer_decorator()
file: src/nanotron/logging/timers.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 1
number of parameters: 5

id: 309
unit: def create_table_log()
file: src/nanotron/helpers.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 1
number of parameters: 0

id: 310
unit: def create_table_output()
file: src/nanotron/helpers.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 4
number of parameters: 2

id: 311
unit: def generator()
file: src/nanotron/generation/decode.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 3
number of parameters: 0

id: 312
unit: def generator()
file: src/nanotron/generation/decode.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 3
number of parameters: 0

id: 313
unit: def get_parameter_and_parent_module()
file: src/nanotron/utils.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 3
number of parameters: 2

id: 314
unit: def cast_pipeline_engine_to_str()
file: src/nanotron/config/utils_config.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 3
number of parameters: 1

id: 315
unit: def backward()
file: src/nanotron/parallel/pipeline_parallel/functional.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 3
number of parameters: 2

id: 316
unit: def _send_first_metadata_p2p_op()
file: src/nanotron/parallel/pipeline_parallel/p2p.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 1
number of parameters: 4

id: 317
unit: def _recv_first_metadata_p2p_op()
file: src/nanotron/parallel/pipeline_parallel/p2p.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 1
number of parameters: 3

id: 318
unit: def _send_second_metadata_p2p_op()
file: src/nanotron/parallel/pipeline_parallel/p2p.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 1
number of parameters: 4

id: 319
unit: def forward()
file: src/nanotron/parallel/tensor_parallel/nn.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 1
number of parameters: 2

id: 320
unit: def __init__()
file: src/nanotron/parallel/tensor_parallel/nn.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 1
number of parameters: 0

id: 321
unit: def forward()
file: src/nanotron/parallel/tensor_parallel/nn.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 3
number of parameters: 2

id: 322
unit: def backward()
file: src/nanotron/parallel/tensor_parallel/functional.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 1
number of parameters: 2

id: 323
unit: def initial_sync()
file: src/nanotron/parallel/utils.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 3
number of parameters: 2

id: 324
unit: def __init__()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 1
number of parameters: 4

id: 325
unit: def pad_to_right()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 5
number of parameters: 3

id: 326
unit: def get_flops()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 1
number of parameters: 0

id: 327
unit: def pad_to_right()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 5
number of parameters: 3

id: 328
unit: def forward()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 2
number of parameters: 0

id: 329
unit: def __init__()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 2
number of parameters: 4

id: 330
unit: def __init__()
file: src/nanotron/models/base.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 2
number of parameters: 3

id: 331
unit: def ignore_init_on_device_and_dtype()
file: src/nanotron/models/base.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 3
number of parameters: 0

id: 332
unit: def get_flops()
file: src/nanotron/models/qwen.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 1
number of parameters: 0

id: 333
unit: def _reset_state()
file: src/nanotron/s3_checkpoints/s3_mover.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 2
number of parameters: 1

id: 334
unit: def start_uploading()
file: src/nanotron/s3_checkpoints/s3_mover.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 4
number of parameters: 2

id: 335
unit: def _compute_scaling_factor()
file: src/nanotron/scaling/parametrization.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 4
number of parameters: 1

id: 336
unit: def __init__()
file: src/nanotron/scaling/parametrization.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 1
number of parameters: 2

id: 337
unit: def process_type()
file: src/nanotron/serialize/metadata.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 4
number of parameters: 4

id: 338
unit: def convert_to_string()
file: src/nanotron/serialize/optimizer.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 7
number of parameters: 1

id: 339
unit: def read_checkpoint_version_from_shard_file()
file: src/nanotron/serialize/weights.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 2
number of parameters: 1

id: 340
unit: def generate_slurm_script()
file: scripts/scaling_benchmarks.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 1
number of parameters: 0

id: 341
unit: def parse_input_paths()
file: scripts/weka.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 8
number of parameters: 2

id: 342
unit: def generate_model_config()
file: slurm_launcher.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 1
number of parameters: 0

id: 343
unit: def create_slurm_script()
file: slurm_launcher.py
start line: 0
end line: 0
size: 9 LOC
McCabe index: 1
number of parameters: 0

id: 344
unit: def __init__()
file: src/nanotron/metrics_logging.py
start line: 0
end line: 0
size: 8 LOC
McCabe index: 2
number of parameters: 2

id: 345
unit: def _get_new_stream()
file: src/nanotron/data/tokenized_bytes.py
start line: 0
end line: 0
size: 8 LOC
McCabe index: 2
number of parameters: 2

id: 346
unit: def get_consumption_stats()
file: src/nanotron/data/nemo_dataset/blendable_dataset.py
start line: 0
end line: 0
size: 8 LOC
McCabe index: 2
number of parameters: 1

id: 347
unit: def _num_epochs()
file: src/nanotron/data/nemo_dataset/__init__.py
start line: 0
end line: 0
size: 8 LOC
McCabe index: 5
number of parameters: 4

id: 348
unit: def __init__()
file: src/nanotron/nn/rotary.py
start line: 0
end line: 0
size: 8 LOC
McCabe index: 1
number of parameters: 0

id: 349
unit: def _get_default_args()
file: src/nanotron/nn/llama3_ring_attention.py
start line: 0
end line: 0
size: 8 LOC
McCabe index: 3
number of parameters: 1

id: 350
unit: def unflatten_varlen_lse()
file: src/nanotron/nn/llama3_ring_attention.py
start line: 0
end line: 0
size: 8 LOC
McCabe index: 2
number of parameters: 3

id: 351
unit: def _get_default_args()
file: src/nanotron/nn/ring_attention.py
start line: 0
end line: 0
size: 8 LOC
McCabe index: 3
number of parameters: 1

id: 352
unit: def unflatten_varlen_lse()
file: src/nanotron/nn/ring_attention.py
start line: 0
end line: 0
size: 8 LOC
McCabe index: 2
number of parameters: 3

id: 353
unit: def reset()
file: src/nanotron/logging/timers.py
start line: 0
end line: 0
size: 8 LOC
McCabe index: 1
number of parameters: 1

id: 354
unit: def total_time()
file: src/nanotron/logging/timers.py
start line: 0
end line: 0
size: 8 LOC
McCabe index: 3
number of parameters: 1

id: 355
unit: def init_random_states()
file: src/nanotron/helpers.py
start line: 0
end line: 0
size: 8 LOC
McCabe index: 3
number of parameters: 2

id: 356
unit: def log_throughput()
file: src/nanotron/helpers.py
start line: 0
end line: 0
size: 8 LOC
McCabe index: 1
number of parameters: 0

id: 357
unit: def state_dict()
file: src/nanotron/optim/named_optimizer.py
start line: 0
end line: 0
size: 8 LOC
McCabe index: 3
number of parameters: 1

id: 358
unit: def __init__()
file: src/nanotron/optim/zero.py
start line: 0
end line: 0
size: 8 LOC
McCabe index: 1
number of parameters: 4

id: 359
unit: def __post_init__()
file: src/nanotron/config/config.py
start line: 0
end line: 0
size: 8 LOC
McCabe index: 4
number of parameters: 1

id: 360
unit: def __post_init__()
file: src/nanotron/config/models_config.py
start line: 0
end line: 0
size: 8 LOC
McCabe index: 4
number of parameters: 1

id: 361
unit: def get_input_output_pp_ranks()
file: src/nanotron/parallel/pipeline_parallel/utils.py
start line: 0
end line: 0
size: 8 LOC
McCabe index: 2
number of parameters: 1

id: 362
unit: def _send_data_p2p_op()
file: src/nanotron/parallel/pipeline_parallel/p2p.py
start line: 0
end line: 0
size: 8 LOC
McCabe index: 1
number of parameters: 4

id: 363
unit: def _set_metadata()
file: src/nanotron/parallel/parameters.py
start line: 0
end line: 0
size: 8 LOC
McCabe index: 2
number of parameters: 3

id: 364
unit: def column_linear()
file: src/nanotron/parallel/tensor_parallel/functional.py
start line: 0
end line: 0
size: 8 LOC
McCabe index: 1
number of parameters: 0

id: 365
unit: def __init__()
file: src/nanotron/parallel/context.py
start line: 0
end line: 0
size: 8 LOC
McCabe index: 1
number of parameters: 0

id: 366
unit: def __init__()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 8 LOC
McCabe index: 2
number of parameters: 4

id: 367
unit: def forward()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 8 LOC
McCabe index: 2
number of parameters: 4

id: 368
unit: def check_model_has_grad()
file: src/nanotron/models/base.py
start line: 0
end line: 0
size: 8 LOC
McCabe index: 3
number of parameters: 2

id: 369
unit: def _combine_expert_outputs()
file: src/nanotron/models/qwen.py
start line: 0
end line: 0
size: 8 LOC
McCabe index: 4
number of parameters: 4

id: 370
unit: def check_optim_state_in_sync()
file: src/nanotron/sanity_checks.py
start line: 0
end line: 0
size: 8 LOC
McCabe index: 4
number of parameters: 2

id: 371
unit: def _inner()
file: src/nanotron/fp8/tensor.py
start line: 0
end line: 0
size: 8 LOC
McCabe index: 1
number of parameters: 4

id: 372
unit: def __new__()
file: src/nanotron/fp8/parameter.py
start line: 0
end line: 0
size: 8 LOC
McCabe index: 2
number of parameters: 4

id: 373
unit: def set_random_state()
file: src/nanotron/random.py
start line: 0
end line: 0
size: 8 LOC
McCabe index: 2
number of parameters: 1

id: 374
unit: def _parametrize_row_linear()
file: src/nanotron/scaling/parametrization.py
start line: 0
end line: 0
size: 8 LOC
McCabe index: 3
number of parameters: 3

id: 375
unit: def __init__()
file: src/nanotron/scaling/parametrization.py
start line: 0
end line: 0
size: 8 LOC
McCabe index: 1
number of parameters: 4

id: 376
unit: def load_random_states()
file: src/nanotron/serialize/random.py
start line: 0
end line: 0
size: 8 LOC
McCabe index: 2
number of parameters: 2

id: 377
unit: def estimate_num_params()
file: scripts/scaling_benchmarks.py
start line: 0
end line: 0
size: 8 LOC
McCabe index: 1
number of parameters: 0

id: 378
unit: def get_dataloader_from_data_stage()
file: run_train.py
start line: 0
end line: 0
size: 8 LOC
McCabe index: 1
number of parameters: 0

id: 379
unit: def _format_paths()
file: src/nanotron/metrics_logging.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 4
number of parameters: 3

id: 380
unit: def clm_process()
file: src/nanotron/data/processing.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 1
number of parameters: 0

id: 381
unit: def compile_helper()
file: src/nanotron/data/nemo_dataset/dataset_utils.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 2
number of parameters: 0

id: 382
unit: def update_consumption_metrics()
file: src/nanotron/data/nemo_dataset/blendable_dataset.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 4
number of parameters: 4

id: 383
unit: def get_ds_sample_idx()
file: src/nanotron/data/nemo_dataset/blendable_dataset.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 1
number of parameters: 2

id: 384
unit: def get()
file: src/nanotron/data/nemo_dataset/indexed_dataset.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 2
number of parameters: 4

id: 385
unit: def merge_file_()
file: src/nanotron/data/nemo_dataset/indexed_dataset.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 2
number of parameters: 2

id: 386
unit: def wait()
file: src/nanotron/nn/llama3_ring_attention.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 3
number of parameters: 1

id: 387
unit: def wait()
file: src/nanotron/nn/ring_attention.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 3
number of parameters: 1

id: 388
unit: def __init__()
file: src/nanotron/nn/activations.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 2
number of parameters: 1

id: 389
unit: def document_mask_direct()
file: src/nanotron/nn/flex_attention.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 3
number of parameters: 4

id: 390
unit: def sliding_window_causal()
file: src/nanotron/nn/flex_attention.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 2
number of parameters: 4

id: 391
unit: def get_block_mask_from_type()
file: src/nanotron/nn/flex_attention.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 1
number of parameters: 0

id: 392
unit: def __init__()
file: src/nanotron/nn/layer_norm.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 1
number of parameters: 5

id: 393
unit: def clear_all_tbi_logs()
file: src/nanotron/logging/logmixin.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 4
number of parameters: 1

id: 394
unit: def add_scalars_from_list()
file: src/nanotron/logging/base.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 2
number of parameters: 3

id: 395
unit: def set_ranks_logging_level()
file: src/nanotron/logging/base.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 4
number of parameters: 2

id: 396
unit: def _process_value_according_to_param_policy()
file: src/nanotron/optim/base.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 1
number of parameters: 0

id: 397
unit: def no_sync()
file: src/nanotron/optim/gradient_accumulator.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 1
number of parameters: 1

id: 398
unit: def zero_grad()
file: src/nanotron/optim/gradient_accumulator.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 5
number of parameters: 1

id: 399
unit: def micro_batcher()
file: src/nanotron/generation/decode.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 1
number of parameters: 0

id: 400
unit: def get_local_store()
file: src/nanotron/generation/generate_store.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 3
number of parameters: 1

id: 401
unit: def main_rank_first()
file: src/nanotron/utils.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 3
number of parameters: 1

id: 402
unit: def local_ranks_zero_first()
file: src/nanotron/utils.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 3
number of parameters: 1

id: 403
unit: def __post_init__()
file: src/nanotron/config/config.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 3
number of parameters: 1

id: 404
unit: def save_as_yaml()
file: src/nanotron/config/config.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 2
number of parameters: 3

id: 405
unit: def cast_str_to_pipeline_engine()
file: src/nanotron/config/utils_config.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 3
number of parameters: 1

id: 406
unit: def __post_init__()
file: src/nanotron/config/models_config.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 3
number of parameters: 1

id: 407
unit: def validation_step()
file: src/nanotron/trainer.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 2
number of parameters: 4

id: 408
unit: def train_batch_iter()
file: src/nanotron/parallel/pipeline_parallel/engine.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 1
number of parameters: 0

id: 409
unit: def train_batch_iter()
file: src/nanotron/parallel/pipeline_parallel/engine.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 1
number of parameters: 0

id: 410
unit: def train_batch_iter()
file: src/nanotron/parallel/pipeline_parallel/engine.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 1
number of parameters: 0

id: 411
unit: def forward()
file: src/nanotron/parallel/pipeline_parallel/functional.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 1
number of parameters: 0

id: 412
unit: def send_to_pipeline_state_buffer()
file: src/nanotron/parallel/pipeline_parallel/functional.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 2
number of parameters: 4

id: 413
unit: def __init__()
file: src/nanotron/parallel/pipeline_parallel/block.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 1
number of parameters: 0

id: 414
unit: def reshape()
file: src/nanotron/parallel/pipeline_parallel/p2p.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 4
number of parameters: 2

id: 415
unit: def forward()
file: src/nanotron/parallel/tensor_parallel/functional.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 1
number of parameters: 0

id: 416
unit: def row_linear()
file: src/nanotron/parallel/tensor_parallel/functional.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 1
number of parameters: 0

id: 417
unit: def get()
file: src/nanotron/parallel/utils.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 3
number of parameters: 4

id: 418
unit: def get_global_rank()
file: src/nanotron/parallel/context.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 1
number of parameters: 0

id: 419
unit: def forward()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 1
number of parameters: 13

id: 420
unit: def forward()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 1
number of parameters: 13

id: 421
unit: def __init__()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 1
number of parameters: 0

id: 422
unit: def register_empty_parameter()
file: src/nanotron/models/base.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 3
number of parameters: 3

id: 423
unit: def register_empty_buffer()
file: src/nanotron/models/base.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 3
number of parameters: 4

id: 424
unit: def __init__()
file: src/nanotron/models/qwen.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 1
number of parameters: 0

id: 425
unit: def __init__()
file: src/nanotron/models/qwen.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 1
number of parameters: 0

id: 426
unit: def start_downloading()
file: src/nanotron/s3_checkpoints/s3_mover.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 2
number of parameters: 1

id: 427
unit: def update_output()
file: src/nanotron/distributed.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 3
number of parameters: 0

id: 428
unit: def forward()
file: src/nanotron/fp8/linear.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 8
number of parameters: 3

id: 429
unit: def get_current_random_state()
file: src/nanotron/random.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 2
number of parameters: 0

id: 430
unit: def from_str_dict()
file: src/nanotron/serialize/metadata.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 1
number of parameters: 3

id: 431
unit: def load_optimizer()
file: src/nanotron/serialize/optimizer.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 1
number of parameters: 9

id: 432
unit: def get_checkpoint_paths_list()
file: src/nanotron/serialize/weights.py
start line: 0
end line: 0
size: 7 LOC
McCabe index: 1
number of parameters: 0

id: 433
unit: def _get_s3_object()
file: src/nanotron/data/s3_utils.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 2

id: 434
unit: def deallocate_indexed_dataset_memory()
file: src/nanotron/data/nemo_dataset/indexed_dataset.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 3
number of parameters: 1

id: 435
unit: def __init__()
file: src/nanotron/data/nemo_dataset/indexed_dataset.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 3

id: 436
unit: def update_out_and_lse()
file: src/nanotron/nn/llama3_ring_attention.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 437
unit: def flatten_varlen_lse()
file: src/nanotron/nn/llama3_ring_attention.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 2
number of parameters: 2

id: 438
unit: def send_recv_kv()
file: src/nanotron/nn/llama3_ring_attention.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 439
unit: def update_out_and_lse()
file: src/nanotron/nn/ring_attention.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 440
unit: def flatten_varlen_lse()
file: src/nanotron/nn/ring_attention.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 2
number of parameters: 2

id: 441
unit: def send_recv_kv()
file: src/nanotron/nn/ring_attention.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 442
unit: def __init__()
file: src/nanotron/nn/activations.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 2
number of parameters: 2

id: 443
unit: def __init__()
file: src/nanotron/nn/activations.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 2
number of parameters: 3

id: 444
unit: def __init__()
file: src/nanotron/nn/activations.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 2
number of parameters: 1

id: 445
unit: def __init__()
file: src/nanotron/nn/moe.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 446
unit: def forward()
file: src/nanotron/nn/moe.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 3
number of parameters: 2

id: 447
unit: def circular_rank_left()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 3
number of parameters: 4

id: 448
unit: def circular_rank_right()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 3
number of parameters: 4

id: 449
unit: def ring_pass()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 450
unit: def all_gather_same_dim()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 2
number of parameters: 1

id: 451
unit: def is_flash_attn_greater_or_equal_2_10()
file: src/nanotron/nn/attention.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 2
number of parameters: 0

id: 452
unit: def score_mod()
file: src/nanotron/nn/flex_attention.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 3
number of parameters: 5

id: 453
unit: def document_mask_from_positions()
file: src/nanotron/nn/flex_attention.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 4

id: 454
unit: def _offsets_to_doc_ids_tensor()
file: src/nanotron/nn/flex_attention.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 1

id: 455
unit: def forward()
file: src/nanotron/nn/layer_norm.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 2

id: 456
unit: def filter()
file: src/nanotron/logging/base.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 5
number of parameters: 2

id: 457
unit: def add_scalar()
file: src/nanotron/logging/base.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 3
number of parameters: 5

id: 458
unit: def __call__()
file: src/nanotron/logging/timers.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 7

id: 459
unit: def get_custom_weight_decay_for_named_parameters()
file: src/nanotron/helpers.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 460
unit: def inherit_from()
file: src/nanotron/optim/inherit_from_other_optimizer.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 3
number of parameters: 2

id: 461
unit: def clip_grad_norm()
file: src/nanotron/optim/clip_grads.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 462
unit: def __torch_dispatch__()
file: src/nanotron/optim/zero.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 2
number of parameters: 4

id: 463
unit: def merge_dp_shard_in_zero1_optimizer()
file: src/nanotron/optim/zero.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 464
unit: def micro_splitter()
file: src/nanotron/generation/decode.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 465
unit: def get_config_from_file()
file: src/nanotron/config/config.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 466
unit: def __init__()
file: src/nanotron/trainer.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 467
unit: def train()
file: src/nanotron/trainer.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 468
unit: def train_step_logs()
file: src/nanotron/trainer.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 469
unit: def forward()
file: src/nanotron/parallel/pipeline_parallel/engine.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 470
unit: def forward()
file: src/nanotron/parallel/pipeline_parallel/functional.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 471
unit: def backward()
file: src/nanotron/parallel/pipeline_parallel/functional.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 2

id: 472
unit: def recv_from_pipeline_state_buffer()
file: src/nanotron/parallel/pipeline_parallel/functional.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 2
number of parameters: 3

id: 473
unit: def _reset()
file: src/nanotron/parallel/pipeline_parallel/p2p.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 1

id: 474
unit: def add_recv()
file: src/nanotron/parallel/pipeline_parallel/p2p.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 3

id: 475
unit: def _send_recv_first_metadata()
file: src/nanotron/parallel/pipeline_parallel/p2p.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 4
number of parameters: 1

id: 476
unit: def mark_as_tied()
file: src/nanotron/parallel/parameters.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 477
unit: def sanity_check()
file: src/nanotron/parallel/parameters.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 3
number of parameters: 1

id: 478
unit: def forward()
file: src/nanotron/parallel/tensor_parallel/functional.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 2
number of parameters: 5

id: 479
unit: def sharded_cross_entropy()
file: src/nanotron/parallel/tensor_parallel/functional.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 480
unit: def create_tied_parameter()
file: src/nanotron/parallel/tied_parameters.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 481
unit: def sync_gradients_across_dp()
file: src/nanotron/parallel/data_parallel/utils.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 482
unit: def __init__()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 4

id: 483
unit: def apply_rotary_pos_emb()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 6

id: 484
unit: def __init__()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 485
unit: def __init__()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 486
unit: def __init__()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 487
unit: def forward()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 488
unit: def reset_parameters()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 3
number of parameters: 1

id: 489
unit: def __init__()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 490
unit: def __init__()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 491
unit: def __init__()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 492
unit: def __init__()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 493
unit: def forward()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 494
unit: def build_model()
file: src/nanotron/models/base.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 6

id: 495
unit: def patch_tensor_constructor()
file: src/nanotron/models/base.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 1

id: 496
unit: def __init__()
file: src/nanotron/models/qwen.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 497
unit: def __init__()
file: src/nanotron/models/qwen.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 498
unit: def forward()
file: src/nanotron/models/qwen.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 3
number of parameters: 2

id: 499
unit: def __init__()
file: src/nanotron/models/qwen.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 500
unit: def forward()
file: src/nanotron/models/qwen.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 3
number of parameters: 2

id: 501
unit: def __init__()
file: src/nanotron/models/qwen.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 502
unit: def forward()
file: src/nanotron/models/qwen.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 503
unit: def before_tbi_sanity_checks()
file: src/nanotron/sanity_checks.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 504
unit: def before_optim_step_sanity_checks()
file: src/nanotron/sanity_checks.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 505
unit: def reduce_scatter_tensor()
file: src/nanotron/distributed.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 506
unit: def reduce_scatter_coalesced()
file: src/nanotron/distributed.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 507
unit: def convert_tensor_from_fp8()
file: src/nanotron/fp8/tensor.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 3

id: 508
unit: def is_fp8_available()
file: src/nanotron/fp8/utils.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 3
number of parameters: 0

id: 509
unit: def fp8_matmul_kernel()
file: src/nanotron/fp8/kernel.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 510
unit: def set_random_seed()
file: src/nanotron/random.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 2
number of parameters: 1

id: 511
unit: def _parametrize_column_linear()
file: src/nanotron/scaling/parametrization.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 3
number of parameters: 3

id: 512
unit: def _parametrize_layer_norm()
file: src/nanotron/scaling/parametrization.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 3
number of parameters: 3

id: 513
unit: def _parametrize_layer_norm()
file: src/nanotron/scaling/parametrization.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 3
number of parameters: 3

id: 514
unit: def consumed_tokens_per_dataset_folder_total()
file: src/nanotron/serialize/metadata.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 3
number of parameters: 1

id: 515
unit: def to_str_dict()
file: src/nanotron/serialize/metadata.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 1

id: 516
unit: def get_path()
file: src/nanotron/serialize/utils.py
start line: 0
end line: 0
size: 6 LOC
McCabe index: 1
number of parameters: 0

id: 517
unit: def __getitem__()
file: src/nanotron/data/tokenized_bytes.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 4
number of parameters: 2

id: 518
unit: def normalize()
file: src/nanotron/data/utils.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 1

id: 519
unit: def count_dataset_indexes()
file: src/nanotron/data/utils.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 2
number of parameters: 2

id: 520
unit: def __getitem__()
file: src/nanotron/data/nanoset.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 2

id: 521
unit: def get_dataloader_worker_init()
file: src/nanotron/data/dataloader.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 1

id: 522
unit: def __best_fitting_dtype()
file: src/nanotron/data/nemo_dataset/indexed_dataset.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 3
number of parameters: 1

id: 523
unit: def code()
file: src/nanotron/data/nemo_dataset/indexed_dataset.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 3
number of parameters: 1

id: 524
unit: def __del__()
file: src/nanotron/data/nemo_dataset/indexed_dataset.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 2
number of parameters: 1

id: 525
unit: def exists()
file: src/nanotron/data/nemo_dataset/indexed_dataset.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 2
number of parameters: 1

id: 526
unit: def __init__()
file: src/nanotron/data/nemo_dataset/indexed_dataset.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 3

id: 527
unit: def get_default_args()
file: src/nanotron/nn/llama3_ring_attention.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 2
number of parameters: 1

id: 528
unit: def _update_out_and_lse()
file: src/nanotron/nn/llama3_ring_attention.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 529
unit: def get_default_args()
file: src/nanotron/nn/ring_attention.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 2
number of parameters: 1

id: 530
unit: def _update_out_and_lse()
file: src/nanotron/nn/ring_attention.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 531
unit: def get_activation()
file: src/nanotron/nn/activations.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 2
number of parameters: 1

id: 532
unit: def routing()
file: src/nanotron/nn/moe.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 2

id: 533
unit: def get_attention_mod_from_type()
file: src/nanotron/nn/flex_attention.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 534
unit: def validate_attention_args()
file: src/nanotron/nn/flex_attention.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 535
unit: def emit()
file: src/nanotron/logging/base.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 2
number of parameters: 2

id: 536
unit: def __new__()
file: src/nanotron/logging/timers.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 2
number of parameters: 1

id: 537
unit: def __init__()
file: src/nanotron/eval/one_job_runner.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 3

id: 538
unit: def run_slurm_one_job()
file: src/nanotron/eval/one_job_runner.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 539
unit: def get_custom_lr_for_named_parameters()
file: src/nanotron/helpers.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 540
unit: def init_optimizer_and_grad_accumulator()
file: src/nanotron/helpers.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 541
unit: def state_dict_additional_keys()
file: src/nanotron/optim/inherit_from_other_optimizer.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 2
number of parameters: 1

id: 542
unit: def get_base_optimizer()
file: src/nanotron/optim/inherit_from_other_optimizer.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 2
number of parameters: 1

id: 543
unit: def __init__()
file: src/nanotron/optim/optimizer_from_gradient_accumulator.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 544
unit: def state_dict()
file: src/nanotron/optim/optimizer_from_gradient_accumulator.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 1

id: 545
unit: def __init__()
file: src/nanotron/optim/zero.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 546
unit: def _get_grad()
file: src/nanotron/optim/zero.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 2
number of parameters: 1

id: 547
unit: def find_optim_index_from_param_name()
file: src/nanotron/optim/zero.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 2
number of parameters: 4

id: 548
unit: def backward()
file: src/nanotron/optim/gradient_accumulator.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 2
number of parameters: 2

id: 549
unit: def step()
file: src/nanotron/optim/gradient_accumulator.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 2
number of parameters: 1

id: 550
unit: def load_state_dict()
file: src/nanotron/optim/gradient_accumulator.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 2
number of parameters: 3

id: 551
unit: def chunks()
file: src/nanotron/generation/decode.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 2
number of parameters: 2

id: 552
unit: def all_gather_batches()
file: src/nanotron/generation/sampler.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 3

id: 553
unit: def get_untyped_storage()
file: src/nanotron/utils.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 2
number of parameters: 1

id: 554
unit: def tensor_from_untyped_storage()
file: src/nanotron/utils.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 2

id: 555
unit: def __post_init__()
file: src/nanotron/config/config.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 3
number of parameters: 1

id: 556
unit: def __post_init__()
file: src/nanotron/config/config.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 3
number of parameters: 1

id: 557
unit: def __post_init__()
file: src/nanotron/config/config.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 3
number of parameters: 1

id: 558
unit: def __post_init__()
file: src/nanotron/config/config.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 3
number of parameters: 1

id: 559
unit: def __post_init__()
file: src/nanotron/config/config.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 3
number of parameters: 1

id: 560
unit: def __post_init__()
file: src/nanotron/config/lighteval_config.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 3
number of parameters: 1

id: 561
unit: def cast_str_to_torch_dtype()
file: src/nanotron/config/utils_config.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 2
number of parameters: 1

id: 562
unit: def get_size()
file: src/nanotron/trainer.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 3
number of parameters: 1

id: 563
unit: def _mark_tied_parameters()
file: src/nanotron/trainer.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 564
unit: def _get_bwd_context()
file: src/nanotron/parallel/pipeline_parallel/engine.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 565
unit: def validate_batch_iter()
file: src/nanotron/parallel/pipeline_parallel/engine.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 566
unit: def forward()
file: src/nanotron/parallel/pipeline_parallel/functional.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 5

id: 567
unit: def build_and_set_rank()
file: src/nanotron/parallel/pipeline_parallel/block.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 2
number of parameters: 2

id: 568
unit: def __init__()
file: src/nanotron/parallel/pipeline_parallel/p2p.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 3

id: 569
unit: def recv_tensors()
file: src/nanotron/parallel/pipeline_parallel/p2p.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 2
number of parameters: 4

id: 570
unit: def from_str()
file: src/nanotron/parallel/parameters.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 2

id: 571
unit: def mark_as_sharded()
file: src/nanotron/parallel/parameters.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 572
unit: def forward()
file: src/nanotron/parallel/tensor_parallel/functional.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 4

id: 573
unit: def forward()
file: src/nanotron/parallel/tensor_parallel/distributed_differentiable_primitives.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 2
number of parameters: 3

id: 574
unit: def create_sharded_parameter()
file: src/nanotron/parallel/sharded_parameters.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 575
unit: def mark_all_parameters_in_module_as_sharded()
file: src/nanotron/parallel/sharded_parameters.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 3
number of parameters: 3

id: 576
unit: def tie_parameters()
file: src/nanotron/parallel/tied_parameters.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 577
unit: def destroy()
file: src/nanotron/parallel/context.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 2
number of parameters: 1

id: 578
unit: def __init__()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 579
unit: def __init__()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 580
unit: def forward()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 581
unit: def forward()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 582
unit: def get_embeddings_lm_head_tied_names()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 2
number of parameters: 1

id: 583
unit: def __init__()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 584
unit: def forward()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 2

id: 585
unit: def forward()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 586
unit: def get_embeddings_lm_head_tied_names()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 1

id: 587
unit: def to()
file: src/nanotron/models/base.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 4
number of parameters: 3

id: 588
unit: def forward()
file: src/nanotron/models/qwen.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 589
unit: def _compute_router_probabilities()
file: src/nanotron/models/qwen.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 2

id: 590
unit: def _core_forward()
file: src/nanotron/models/qwen.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 591
unit: def _checkpointed_forward()
file: src/nanotron/models/qwen.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 592
unit: def forward()
file: src/nanotron/models/qwen.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 593
unit: def __init__()
file: src/nanotron/models/qwen.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 594
unit: def forward()
file: src/nanotron/models/qwen.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 595
unit: def forward()
file: src/nanotron/models/qwen.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 596
unit: def get_embeddings_lm_head_tied_names()
file: src/nanotron/models/qwen.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 2
number of parameters: 1

id: 597
unit: def _release_lock()
file: src/nanotron/s3_checkpoints/s3_mover.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 4
number of parameters: 1

id: 598
unit: def assert_tensor_synced_across_pg()
file: src/nanotron/sanity_checks.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 599
unit: def after_tbi_sanity_checks()
file: src/nanotron/sanity_checks.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 600
unit: def after_optim_step_sanity_checks()
file: src/nanotron/sanity_checks.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 601
unit: def update_output()
file: src/nanotron/distributed.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 2
number of parameters: 0

id: 602
unit: def all_reduce_coalesced()
file: src/nanotron/distributed.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 603
unit: def all_gather_coalesced()
file: src/nanotron/distributed.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 604
unit: def get_global_rank()
file: src/nanotron/distributed.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 2
number of parameters: 2

id: 605
unit: def get_rank()
file: src/nanotron/distributed.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 2
number of parameters: 1

id: 606
unit: def check_type()
file: src/nanotron/random.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 3
number of parameters: 2

id: 607
unit: def __setitem__()
file: src/nanotron/random.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 2
number of parameters: 3

id: 608
unit: def _parametrize_router()
file: src/nanotron/scaling/parametrization.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 3
number of parameters: 3

id: 609
unit: def to_dict()
file: src/nanotron/serialize/metadata.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 2
number of parameters: 4

id: 610
unit: def optimizer_filename()
file: src/nanotron/serialize/optimizer.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 2
number of parameters: 2

id: 611
unit: def lr_scheduler_filename()
file: src/nanotron/serialize/optimizer.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 2
number of parameters: 2

id: 612
unit: def save_lr_scheduler()
file: src/nanotron/serialize/optimizer.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 613
unit: def load_lr_scheduler()
file: src/nanotron/serialize/optimizer.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 614
unit: def extract_tp_pp_rank_from_shard_path()
file: src/nanotron/serialize/utils.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 1

id: 615
unit: def merge_and_shard_tp_tensors()
file: src/nanotron/serialize/utils.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 616
unit: def load_sharded_param_latest()
file: src/nanotron/serialize/weights.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 617
unit: def load_weights()
file: src/nanotron/serialize/weights.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 618
unit: def main()
file: scripts/fix_checkpoint_bad_naming.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 0

id: 619
unit: def launch_slurm_job()
file: slurm_launcher.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 1
number of parameters: 2

id: 620
unit: def tail_output_file()
file: slurm_launcher.py
start line: 0
end line: 0
size: 5 LOC
McCabe index: 2
number of parameters: 1

id: 621
unit: def _get_s3_path_components()
file: src/nanotron/data/s3_utils.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 2

id: 622
unit: def __iter__()
file: src/nanotron/data/samplers.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 3
number of parameters: 1

id: 623
unit: def __getitem__()
file: src/nanotron/data/samplers.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 2
number of parameters: 2

id: 624
unit: def get_start_end_idx()
file: src/nanotron/data/samplers.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 1

id: 625
unit: def _tokenize_and_group_texts()
file: src/nanotron/data/processing.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 3
number of parameters: 1

id: 626
unit: def get_datasets()
file: src/nanotron/data/processing.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 0

id: 627
unit: def sanity_check_dataloader()
file: src/nanotron/data/dataloader.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 0

id: 628
unit: def __getitem__()
file: src/nanotron/data/nemo_dataset/blendable_dataset.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 2
number of parameters: 2

id: 629
unit: def _warmup_mmap_file()
file: src/nanotron/data/nemo_dataset/indexed_dataset.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 2
number of parameters: 1

id: 630
unit: def __del__()
file: src/nanotron/data/nemo_dataset/indexed_dataset.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 2
number of parameters: 1

id: 631
unit: def add_item()
file: src/nanotron/data/nemo_dataset/indexed_dataset.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 2

id: 632
unit: def finalize()
file: src/nanotron/data/nemo_dataset/indexed_dataset.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 2

id: 633
unit: def commit()
file: src/nanotron/nn/llama3_ring_attention.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 2
number of parameters: 1

id: 634
unit: def wait()
file: src/nanotron/nn/llama3_ring_attention.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 2
number of parameters: 1

id: 635
unit: def commit()
file: src/nanotron/nn/ring_attention.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 2
number of parameters: 1

id: 636
unit: def wait()
file: src/nanotron/nn/ring_attention.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 2
number of parameters: 1

id: 637
unit: def forward()
file: src/nanotron/nn/activations.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 2

id: 638
unit: def __getitem__()
file: src/nanotron/nn/activations.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 2
number of parameters: 2

id: 639
unit: def forward()
file: src/nanotron/nn/moe.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 0

id: 640
unit: def _dispatch_tokens()
file: src/nanotron/nn/moe.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 0

id: 641
unit: def pad_dim_to()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 2
number of parameters: 3

id: 642
unit: def gather_sizes()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 3

id: 643
unit: def backward()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 3

id: 644
unit: def create_document_mask_func()
file: src/nanotron/nn/flex_attention.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 0

id: 645
unit: def create_attention_mask()
file: src/nanotron/nn/flex_attention.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 0

id: 646
unit: def doc_causal_mask()
file: src/nanotron/nn/flex_attention.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 4

id: 647
unit: def sliding_window_causal_mask_func()
file: src/nanotron/nn/flex_attention.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 5

id: 648
unit: def __init__()
file: src/nanotron/nn/layer_norm.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 3

id: 649
unit: def _configure_library_root_logger()
file: src/nanotron/logging/base.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 0

id: 650
unit: def average_time()
file: src/nanotron/logging/timers.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 2
number of parameters: 1

id: 651
unit: def items()
file: src/nanotron/logging/timers.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 2
number of parameters: 1

id: 652
unit: def normalize_s3_path()
file: src/nanotron/eval/one_job_runner.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 1

id: 653
unit: def get_formatted_value()
file: src/nanotron/helpers.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 2
number of parameters: 1

id: 654
unit: def __init__()
file: src/nanotron/optim/named_optimizer.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 0

id: 655
unit: def step()
file: src/nanotron/optim/optimizer_from_gradient_accumulator.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 3

id: 656
unit: def load_state_dict()
file: src/nanotron/optim/optimizer_from_gradient_accumulator.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 4

id: 657
unit: def step()
file: src/nanotron/optim/zero.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 4
number of parameters: 3

id: 658
unit: def __init__()
file: src/nanotron/optim/gradient_accumulator.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 0

id: 659
unit: def assign_param_offsets()
file: src/nanotron/optim/gradient_accumulator.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 3
number of parameters: 6

id: 660
unit: def __call__()
file: src/nanotron/utils.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 2
number of parameters: 3

id: 661
unit: def post_train_step()
file: src/nanotron/trainer.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 2
number of parameters: 1

id: 662
unit: def _init_model()
file: src/nanotron/trainer.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 0

id: 663
unit: def register_send_activation()
file: src/nanotron/parallel/pipeline_parallel/state.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 3
number of parameters: 4

id: 664
unit: def register_recv_activation()
file: src/nanotron/parallel/pipeline_parallel/state.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 3
number of parameters: 3

id: 665
unit: def _get_fwd_context()
file: src/nanotron/parallel/pipeline_parallel/engine.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 2
number of parameters: 1

id: 666
unit: def backward()
file: src/nanotron/parallel/pipeline_parallel/functional.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 2

id: 667
unit: def send_tensors()
file: src/nanotron/parallel/pipeline_parallel/p2p.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 2
number of parameters: 4

id: 668
unit: def __str__()
file: src/nanotron/parallel/parameters.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 1

id: 669
unit: def get_full_name_from_model()
file: src/nanotron/parallel/parameters.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 2
number of parameters: 2

id: 670
unit: def get_tied_info()
file: src/nanotron/parallel/parameters.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 1

id: 671
unit: def is_tied()
file: src/nanotron/parallel/parameters.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 1

id: 672
unit: def get_sharded_info()
file: src/nanotron/parallel/parameters.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 1

id: 673
unit: def is_sharded()
file: src/nanotron/parallel/parameters.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 1

id: 674
unit: def backward()
file: src/nanotron/parallel/tensor_parallel/distributed_differentiable_primitives.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 2

id: 675
unit: def create_sharded_parameter_from_config()
file: src/nanotron/parallel/sharded_parameters.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 0

id: 676
unit: def sync_tied_weights_gradients()
file: src/nanotron/parallel/tied_parameters.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 0

id: 677
unit: def set_device()
file: src/nanotron/parallel/context.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 1

id: 678
unit: def forward()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 0

id: 679
unit: def forward()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 0

id: 680
unit: def rotate_half()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 2

id: 681
unit: def forward()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 2

id: 682
unit: def forward()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 0

id: 683
unit: def _core_forward()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 0

id: 684
unit: def _checkpointed_forward()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 0

id: 685
unit: def forward()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 0

id: 686
unit: def forward()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 0

id: 687
unit: def forward_with_hidden_states()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 0

id: 688
unit: def forward()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 0

id: 689
unit: def forward()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 5

id: 690
unit: def dropout_add()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 4

id: 691
unit: def forward()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 0

id: 692
unit: def forward()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 3
number of parameters: 3

id: 693
unit: def forward()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 0

id: 694
unit: def before_tbi_sanity_checks()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 3
number of parameters: 1

id: 695
unit: def should_ignore_init_on_device_and_dtype()
file: src/nanotron/models/base.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 2
number of parameters: 0

id: 696
unit: def forward()
file: src/nanotron/models/qwen.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 3

id: 697
unit: def forward()
file: src/nanotron/models/qwen.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 0

id: 698
unit: def _warning()
file: src/nanotron/s3_checkpoints/s3_mover.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 2
number of parameters: 2

id: 699
unit: def _info()
file: src/nanotron/s3_checkpoints/s3_mover.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 2
number of parameters: 2

id: 700
unit: def wait_for_completion()
file: src/nanotron/s3_checkpoints/s3_mover.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 2
number of parameters: 1

id: 701
unit: def get_filesystem_and_path()
file: src/nanotron/s3_checkpoints/fsspec.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 2

id: 702
unit: def convert_tensor_to_fp8()
file: src/nanotron/fp8/tensor.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 2

id: 703
unit: def __init__()
file: src/nanotron/random.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 3
number of parameters: 2

id: 704
unit: def __eq__()
file: src/nanotron/random.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 2
number of parameters: 2

id: 705
unit: def parametrize()
file: src/nanotron/scaling/parametrization.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 2
number of parameters: 3

id: 706
unit: def _parametrize_embedding()
file: src/nanotron/scaling/parametrization.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 2
number of parameters: 3

id: 707
unit: def _parametrize_embedding()
file: src/nanotron/scaling/parametrization.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 2
number of parameters: 3

id: 708
unit: def get_lr()
file: src/nanotron/scaling/parametrization.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 2
number of parameters: 3

id: 709
unit: def save_optimizer()
file: src/nanotron/serialize/optimizer.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 0

id: 710
unit: def read_checkpoint_version_from_meta()
file: src/nanotron/serialize/weights.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 2

id: 711
unit: def save_random_states()
file: src/nanotron/serialize/random.py
start line: 0
end line: 0
size: 4 LOC
McCabe index: 1
number of parameters: 0

id: 712
unit: def collect_all_metrics()
file: src/nanotron/metrics_logging.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 0

id: 713
unit: def __init__()
file: src/nanotron/data/samplers.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 4

id: 714
unit: def __getitem__()
file: src/nanotron/data/nemo_dataset/blendable_dataset.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 2

id: 715
unit: def __getitem__()
file: src/nanotron/data/nemo_dataset/__init__.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 2

id: 716
unit: def __init__()
file: src/nanotron/nn/llama3_ring_attention.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 2

id: 717
unit: def all_gather()
file: src/nanotron/nn/llama3_ring_attention.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 3

id: 718
unit: def __init__()
file: src/nanotron/nn/ring_attention.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 2

id: 719
unit: def all_gather()
file: src/nanotron/nn/ring_attention.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 3

id: 720
unit: def __init__()
file: src/nanotron/nn/activations.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 1

id: 721
unit: def forward()
file: src/nanotron/nn/activations.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 4

id: 722
unit: def forward()
file: src/nanotron/nn/moe.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 2

id: 723
unit: def _combine_expert_outputs()
file: src/nanotron/nn/moe.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 4

id: 724
unit: def __init__()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 3

id: 725
unit: def get_ring_flash_attn_cuda()
file: src/nanotron/nn/attention.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 0

id: 726
unit: def create_softcapped_causal_score_mod()
file: src/nanotron/nn/flex_attention.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 0

id: 727
unit: def create_block_mask_cached()
file: src/nanotron/nn/flex_attention.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 6

id: 728
unit: def __init__()
file: src/nanotron/logging/logmixin.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 3

id: 729
unit: def _reset_library_root_logger()
file: src/nanotron/logging/base.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 0

id: 730
unit: def __enter__()
file: src/nanotron/logging/timers.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 1

id: 731
unit: def __exit__()
file: src/nanotron/logging/timers.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 4

id: 732
unit: def reset_all()
file: src/nanotron/logging/timers.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 2
number of parameters: 1

id: 733
unit: def reset()
file: src/nanotron/logging/timers.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 2
number of parameters: 2

id: 734
unit: def merge_named_param_groups()
file: src/nanotron/helpers.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 0

id: 735
unit: def is_last_stage()
file: src/nanotron/helpers.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 0

id: 736
unit: def zero_grad()
file: src/nanotron/optim/optimizer_from_gradient_accumulator.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 1

id: 737
unit: def get_sliced_flat_tensor()
file: src/nanotron/optim/zero.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 3

id: 738
unit: def get_sliced_tensor()
file: src/nanotron/optim/zero.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 3

id: 739
unit: def get_numel_of_unsharded_dp_param()
file: src/nanotron/optim/zero.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 3
number of parameters: 1

id: 740
unit: def assign_shard_to_buffer()
file: src/nanotron/optim/zero.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 3

id: 741
unit: def get_fp32_accum_hook()
file: src/nanotron/optim/gradient_accumulator.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 0

id: 742
unit: def flush()
file: src/nanotron/generation/generate_store.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 2
number of parameters: 1

id: 743
unit: def _attach_store()
file: src/nanotron/generation/generate_store.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 2

id: 744
unit: def __init__()
file: src/nanotron/utils.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 2

id: 745
unit: def __enter__()
file: src/nanotron/utils.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 2
number of parameters: 1

id: 746
unit: def __post_init__()
file: src/nanotron/config/config.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 2
number of parameters: 1

id: 747
unit: def __post_init__()
file: src/nanotron/config/config.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 2
number of parameters: 1

id: 748
unit: def __post_init__()
file: src/nanotron/config/config.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 2
number of parameters: 1

id: 749
unit: def __post_init__()
file: src/nanotron/config/config.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 2
number of parameters: 1

id: 750
unit: def create_empty()
file: src/nanotron/config/config.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 2
number of parameters: 1

id: 751
unit: def get_yaml()
file: src/nanotron/config/config.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 1

id: 752
unit: def load_from_yaml()
file: src/nanotron/config/config.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 2

id: 753
unit: def __post_init__()
file: src/nanotron/config/lighteval_config.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 2
number of parameters: 1

id: 754
unit: def post_training()
file: src/nanotron/trainer.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 2
number of parameters: 1

id: 755
unit: def __init__()
file: src/nanotron/parallel/pipeline_parallel/engine.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 1

id: 756
unit: def get_min_max_rank()
file: src/nanotron/parallel/pipeline_parallel/block.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 3
number of parameters: 1

id: 757
unit: def __init__()
file: src/nanotron/parallel/pipeline_parallel/p2p.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 2

id: 758
unit: def forward()
file: src/nanotron/parallel/tensor_parallel/distributed_differentiable_primitives.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 3

id: 759
unit: def backward()
file: src/nanotron/parallel/tensor_parallel/distributed_differentiable_primitives.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 2

id: 760
unit: def backward()
file: src/nanotron/parallel/tensor_parallel/distributed_differentiable_primitives.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 2

id: 761
unit: def get_local_ranks()
file: src/nanotron/parallel/context.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 3
number of parameters: 2

id: 762
unit: def __init__()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 2

id: 763
unit: def forward()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 2

id: 764
unit: def __init__()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 2

id: 765
unit: def __init__()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 3

id: 766
unit: def rotate_half()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 1

id: 767
unit: def shape()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 0

id: 768
unit: def __init__()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 2

id: 769
unit: def __new__()
file: src/nanotron/models/base.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 3

id: 770
unit: def init_on_device_and_dtype()
file: src/nanotron/models/base.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 1

id: 771
unit: def __init__()
file: src/nanotron/models/qwen.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 2

id: 772
unit: def __init__()
file: src/nanotron/models/qwen.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 3

id: 773
unit: def is_previous_save_finished()
file: src/nanotron/s3_checkpoints/s3_mover.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 1

id: 774
unit: def fs_open()
file: src/nanotron/s3_checkpoints/fsspec.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 0

id: 775
unit: def fs_copy()
file: src/nanotron/s3_checkpoints/fsspec.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 0

id: 776
unit: def get_synced_random_state()
file: src/nanotron/random.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 0

id: 777
unit: def __init__()
file: src/nanotron/scaling/parametrization.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 1
number of parameters: 4

id: 778
unit: def __post_init__()
file: src/nanotron/serialize/metadata.py
start line: 0
end line: 0
size: 3 LOC
McCabe index: 2
number of parameters: 1

id: 779
unit: def compute_tensor_norm()
file: src/nanotron/metrics_logging.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 780
unit: def compute_zero_fraction()
file: src/nanotron/metrics_logging.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 781
unit: def __len__()
file: src/nanotron/data/tokenized_bytes.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 782
unit: def __len__()
file: src/nanotron/data/tokenized_bytes.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 2
number of parameters: 1

id: 783
unit: def _get_s3_file_list()
file: src/nanotron/data/s3_utils.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 0

id: 784
unit: def pack_sft_sequences()
file: src/nanotron/data/sft_processing.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 4
number of parameters: 3

id: 785
unit: def total_length()
file: src/nanotron/data/samplers.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 786
unit: def __len__()
file: src/nanotron/data/samplers.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 787
unit: def __init__()
file: src/nanotron/data/samplers.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 788
unit: def __len__()
file: src/nanotron/data/samplers.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 789
unit: def __iter__()
file: src/nanotron/data/samplers.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 790
unit: def __len__()
file: src/nanotron/data/nanoset.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 791
unit: def build_nanoset_index_helper()
file: src/nanotron/data/nanoset.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 0

id: 792
unit: def set_tensor_pointers()
file: src/nanotron/data/dataloader.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 0

id: 793
unit: def get_datasets_weights_and_num_samples()
file: src/nanotron/data/nemo_dataset/dataset_utils.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 0

id: 794
unit: def __len__()
file: src/nanotron/data/nemo_dataset/blendable_dataset.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 795
unit: def __len__()
file: src/nanotron/data/nemo_dataset/blendable_dataset.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 796
unit: def make_builder()
file: src/nanotron/data/nemo_dataset/indexed_dataset.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 6

id: 797
unit: def index_file_path()
file: src/nanotron/data/nemo_dataset/indexed_dataset.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 798
unit: def data_file_path()
file: src/nanotron/data/nemo_dataset/indexed_dataset.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 799
unit: def dtype()
file: src/nanotron/data/nemo_dataset/indexed_dataset.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 800
unit: def sizes()
file: src/nanotron/data/nemo_dataset/indexed_dataset.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 801
unit: def doc_idx()
file: src/nanotron/data/nemo_dataset/indexed_dataset.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 802
unit: def __getitem__()
file: src/nanotron/data/nemo_dataset/indexed_dataset.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 803
unit: def __len__()
file: src/nanotron/data/nemo_dataset/indexed_dataset.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 804
unit: def __getstate__()
file: src/nanotron/data/nemo_dataset/indexed_dataset.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 805
unit: def __len__()
file: src/nanotron/data/nemo_dataset/indexed_dataset.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 806
unit: def sizes()
file: src/nanotron/data/nemo_dataset/indexed_dataset.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 807
unit: def doc_idx()
file: src/nanotron/data/nemo_dataset/indexed_dataset.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 808
unit: def get_doc_idx()
file: src/nanotron/data/nemo_dataset/indexed_dataset.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 809
unit: def set_doc_idx()
file: src/nanotron/data/nemo_dataset/indexed_dataset.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 810
unit: def supports_prefetch()
file: src/nanotron/data/nemo_dataset/indexed_dataset.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 811
unit: def deallocate_indexed_dataset_memory()
file: src/nanotron/data/nemo_dataset/indexed_dataset.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 812
unit: def end_document()
file: src/nanotron/data/nemo_dataset/indexed_dataset.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 813
unit: def __len__()
file: src/nanotron/data/nemo_dataset/__init__.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 814
unit: def _num_tokens()
file: src/nanotron/data/nemo_dataset/__init__.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 815
unit: def llama3_flash_attn_prepare_cu_seqlens()
file: src/nanotron/nn/llama3_ring_attention.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 0

id: 816
unit: def forward()
file: src/nanotron/nn/activations.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 817
unit: def forward()
file: src/nanotron/nn/activations.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 818
unit: def _gelu_python()
file: src/nanotron/nn/activations.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 819
unit: def forward()
file: src/nanotron/nn/activations.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 820
unit: def forward()
file: src/nanotron/nn/activations.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 821
unit: def forward()
file: src/nanotron/nn/activations.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 822
unit: def forward()
file: src/nanotron/nn/activations.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 823
unit: def forward()
file: src/nanotron/nn/activations.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 824
unit: def forward()
file: src/nanotron/nn/activations.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 825
unit: def _mish_python()
file: src/nanotron/nn/activations.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 826
unit: def forward()
file: src/nanotron/nn/activations.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 827
unit: def forward()
file: src/nanotron/nn/activations.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 828
unit: def __init__()
file: src/nanotron/nn/moe.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 0

id: 829
unit: def gating()
file: src/nanotron/nn/moe.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 830
unit: def _checkpointed_forward()
file: src/nanotron/nn/moe.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 831
unit: def exists()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 832
unit: def default()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 2
number of parameters: 2

id: 833
unit: def divisible_by()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 834
unit: def exists()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 835
unit: def default()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 2
number of parameters: 2

id: 836
unit: def cast_tuple()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 2
number of parameters: 2

id: 837
unit: def circular_index_left()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 3

id: 838
unit: def circular_index_right()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 3

id: 839
unit: def null_ring_pass()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 5

id: 840
unit: def exists()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 841
unit: def default()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 2
number of parameters: 2

id: 842
unit: def divisible_by()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 843
unit: def get_rank()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 2
number of parameters: 0

id: 844
unit: def get_world_size()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 2
number of parameters: 0

id: 845
unit: def is_distributed()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 2
number of parameters: 0

id: 846
unit: def has_only_one_value()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 847
unit: def forward()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 3

id: 848
unit: def exists()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 849
unit: def default()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 2
number of parameters: 2

id: 850
unit: def is_contiguous()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 851
unit: def grid()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 852
unit: def init_to_zero()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 853
unit: def grid()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 854
unit: def grid()
file: src/nanotron/nn/ring_attention_lucidrain.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 2
number of parameters: 1

id: 855
unit: def is_torch_flex_attn_available()
file: src/nanotron/nn/attention.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 2
number of parameters: 0

id: 856
unit: def causal_mask_func()
file: src/nanotron/nn/flex_attention.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 4

id: 857
unit: def forward()
file: src/nanotron/nn/layer_norm.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 0

id: 858
unit: def reset_parameters()
file: src/nanotron/nn/layer_norm.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 859
unit: def forward()
file: src/nanotron/nn/layer_norm.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 0

id: 860
unit: def extra_repr()
file: src/nanotron/nn/layer_norm.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 861
unit: def tbi_logger()
file: src/nanotron/logging/logmixin.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 4
number of parameters: 3

id: 862
unit: def _get_internal_logs()
file: src/nanotron/logging/logmixin.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 863
unit: def _clear_internal_logs()
file: src/nanotron/logging/logmixin.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 864
unit: def get_library_root_logger()
file: src/nanotron/logging/base.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 0

id: 865
unit: def get_verbosity()
file: src/nanotron/logging/base.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 0

id: 866
unit: def warn_once()
file: src/nanotron/logging/base.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 0

id: 867
unit: def enable()
file: src/nanotron/logging/timers.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 868
unit: def disable()
file: src/nanotron/logging/timers.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 869
unit: def is_enabled()
file: src/nanotron/logging/timers.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 870
unit: def test_all_pair_to_pair()
file: src/nanotron/helpers.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 0

id: 871
unit: def compute_remain_train_steps_of_a_data_stage_from_ckp()
file: src/nanotron/helpers.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 0

id: 872
unit: def is_resume_from_training()
file: src/nanotron/helpers.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 0

id: 873
unit: def get_consumed_train_samples_of_a_data_stage_from_ckp()
file: src/nanotron/helpers.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 0

id: 874
unit: def __getstate__()
file: src/nanotron/optim/inherit_from_other_optimizer.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 875
unit: def __setstate__()
file: src/nanotron/optim/inherit_from_other_optimizer.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 876
unit: def __repr__()
file: src/nanotron/optim/inherit_from_other_optimizer.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 877
unit: def zero_grad()
file: src/nanotron/optim/inherit_from_other_optimizer.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 878
unit: def state_dict()
file: src/nanotron/optim/inherit_from_other_optimizer.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 879
unit: def load_state_dict()
file: src/nanotron/optim/inherit_from_other_optimizer.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 4

id: 880
unit: def step()
file: src/nanotron/optim/inherit_from_other_optimizer.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 3

id: 881
unit: def param_groups()
file: src/nanotron/optim/inherit_from_other_optimizer.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 882
unit: def __getstate__()
file: src/nanotron/optim/base.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 883
unit: def __setstate__()
file: src/nanotron/optim/base.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 884
unit: def __repr__()
file: src/nanotron/optim/base.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 885
unit: def zero_grad()
file: src/nanotron/optim/base.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 886
unit: def state_dict_additional_keys()
file: src/nanotron/optim/base.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 887
unit: def state_dict()
file: src/nanotron/optim/base.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 888
unit: def load_state_dict()
file: src/nanotron/optim/base.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 4

id: 889
unit: def step()
file: src/nanotron/optim/base.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 3

id: 890
unit: def inherit_from()
file: src/nanotron/optim/base.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 891
unit: def state_dict_additional_keys()
file: src/nanotron/optim/optimizer_from_gradient_accumulator.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 892
unit: def _del_grad()
file: src/nanotron/optim/zero.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 893
unit: def data_ptr()
file: src/nanotron/optim/zero.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 894
unit: def extract_parallel_ranks_from_shard_path()
file: src/nanotron/optim/zero.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 0

id: 895
unit: def __init__()
file: src/nanotron/optim/gradient_accumulator.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 3

id: 896
unit: def backward()
file: src/nanotron/optim/gradient_accumulator.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 897
unit: def step()
file: src/nanotron/optim/gradient_accumulator.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 898
unit: def sync_gradients_across_dp()
file: src/nanotron/optim/gradient_accumulator.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 4

id: 899
unit: def zero_grad()
file: src/nanotron/optim/gradient_accumulator.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 900
unit: def get_parameter_for_optimizer()
file: src/nanotron/optim/gradient_accumulator.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 901
unit: def get_grad_buffer()
file: src/nanotron/optim/gradient_accumulator.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 902
unit: def state_dict()
file: src/nanotron/optim/gradient_accumulator.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 903
unit: def load_state_dict()
file: src/nanotron/optim/gradient_accumulator.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 904
unit: def build_grad_buffers()
file: src/nanotron/optim/gradient_accumulator.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 0

id: 905
unit: def get_parameter_for_optimizer()
file: src/nanotron/optim/gradient_accumulator.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 906
unit: def get_grad_buffer()
file: src/nanotron/optim/gradient_accumulator.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 907
unit: def state_dict()
file: src/nanotron/optim/gradient_accumulator.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 2
number of parameters: 1

id: 908
unit: def __post_init__()
file: src/nanotron/optim/gradient_accumulator.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 909
unit: def broadcast_tensors()
file: src/nanotron/generation/decode.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 0

id: 910
unit: def __call__()
file: src/nanotron/generation/sampler.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 911
unit: def __init__()
file: src/nanotron/generation/generate_store.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 912
unit: def _detach_store()
file: src/nanotron/generation/generate_store.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 913
unit: def __exit__()
file: src/nanotron/utils.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 3

id: 914
unit: def __repr__()
file: src/nanotron/utils.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 915
unit: def global_batch_size()
file: src/nanotron/config/config.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 916
unit: def global_batch_size_in_tokens()
file: src/nanotron/config/config.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 917
unit: def as_dict()
file: src/nanotron/config/config.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 918
unit: def get_config_from_dict()
file: src/nanotron/config/config.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 0

id: 919
unit: def __post_init__()
file: src/nanotron/config/lighteval_config.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 920
unit: def __post_init__()
file: src/nanotron/config/lighteval_config.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 921
unit: def __post_init__()
file: src/nanotron/config/models_config.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 922
unit: def is_using_mup()
file: src/nanotron/config/models_config.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 923
unit: def is_using_mup()
file: src/nanotron/config/models_config.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 924
unit: def is_moe_model()
file: src/nanotron/config/models_config.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 2
number of parameters: 1

id: 925
unit: def n_embed()
file: src/nanotron/config/models_config.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 926
unit: def n_head()
file: src/nanotron/config/models_config.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 927
unit: def n_layer()
file: src/nanotron/config/models_config.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 928
unit: def n_positions()
file: src/nanotron/config/models_config.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 929
unit: def n_inner()
file: src/nanotron/config/models_config.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 930
unit: def pre_init()
file: src/nanotron/trainer.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 931
unit: def training_step()
file: src/nanotron/trainer.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 0

id: 932
unit: def setup_log_writers()
file: src/nanotron/trainer.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 0

id: 933
unit: def mark_tied_parameters()
file: src/nanotron/trainer.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 0

id: 934
unit: def mark_unsharded_params_as_tied_across_tp()
file: src/nanotron/trainer.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 0

id: 935
unit: def mark_unsharded_params_as_tied_across_expert()
file: src/nanotron/trainer.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 0

id: 936
unit: def __call__()
file: src/nanotron/parallel/pipeline_parallel/state.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 937
unit: def __call__()
file: src/nanotron/parallel/pipeline_parallel/state.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 938
unit: def __call__()
file: src/nanotron/parallel/pipeline_parallel/state.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 939
unit: def __call__()
file: src/nanotron/parallel/pipeline_parallel/state.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 940
unit: def register_activation_requiring_backward()
file: src/nanotron/parallel/pipeline_parallel/state.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 941
unit: def register_send_activation()
file: src/nanotron/parallel/pipeline_parallel/state.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 4

id: 942
unit: def register_recv_activation()
file: src/nanotron/parallel/pipeline_parallel/state.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 3

id: 943
unit: def register_send_grad()
file: src/nanotron/parallel/pipeline_parallel/state.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 4

id: 944
unit: def register_recv_grad()
file: src/nanotron/parallel/pipeline_parallel/state.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 3

id: 945
unit: def run_communication()
file: src/nanotron/parallel/pipeline_parallel/state.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 946
unit: def new_micro_batch_forward()
file: src/nanotron/parallel/pipeline_parallel/state.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 947
unit: def pop_last_activations_requiring_backward()
file: src/nanotron/parallel/pipeline_parallel/state.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 948
unit: def register_activation_requiring_backward()
file: src/nanotron/parallel/pipeline_parallel/state.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 949
unit: def register_send_activation()
file: src/nanotron/parallel/pipeline_parallel/state.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 4

id: 950
unit: def register_recv_activation()
file: src/nanotron/parallel/pipeline_parallel/state.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 3

id: 951
unit: def register_send_grad()
file: src/nanotron/parallel/pipeline_parallel/state.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 4

id: 952
unit: def register_recv_grad()
file: src/nanotron/parallel/pipeline_parallel/state.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 3

id: 953
unit: def new_micro_batch_forward()
file: src/nanotron/parallel/pipeline_parallel/state.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 954
unit: def pop_last_activations_requiring_backward()
file: src/nanotron/parallel/pipeline_parallel/state.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 955
unit: def register_activation_requiring_backward()
file: src/nanotron/parallel/pipeline_parallel/state.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 956
unit: def register_send_grad()
file: src/nanotron/parallel/pipeline_parallel/state.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 4

id: 957
unit: def register_recv_grad()
file: src/nanotron/parallel/pipeline_parallel/state.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 3

id: 958
unit: def new_micro_batch_forward()
file: src/nanotron/parallel/pipeline_parallel/state.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 959
unit: def pop_last_activations_requiring_backward()
file: src/nanotron/parallel/pipeline_parallel/state.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 960
unit: def backward()
file: src/nanotron/parallel/pipeline_parallel/engine.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 0

id: 961
unit: def __str__()
file: src/nanotron/parallel/pipeline_parallel/engine.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 962
unit: def __format__()
file: src/nanotron/parallel/pipeline_parallel/engine.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 963
unit: def __init__()
file: src/nanotron/parallel/pipeline_parallel/engine.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 964
unit: def __str__()
file: src/nanotron/parallel/pipeline_parallel/engine.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 965
unit: def __init__()
file: src/nanotron/parallel/pipeline_parallel/engine.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 966
unit: def __str__()
file: src/nanotron/parallel/pipeline_parallel/engine.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 967
unit: def extra_repr()
file: src/nanotron/parallel/pipeline_parallel/block.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 2
number of parameters: 1

id: 968
unit: def set_pipeline_state()
file: src/nanotron/parallel/pipeline_parallel/block.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 969
unit: def to_second_metadata()
file: src/nanotron/parallel/pipeline_parallel/p2p.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 970
unit: def _recv_second_metadata_p2p_op()
file: src/nanotron/parallel/pipeline_parallel/p2p.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 0

id: 971
unit: def _recv_data_p2p_op()
file: src/nanotron/parallel/pipeline_parallel/p2p.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 0

id: 972
unit: def irecv_tensors()
file: src/nanotron/parallel/pipeline_parallel/p2p.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 0

id: 973
unit: def __str__()
file: src/nanotron/parallel/pipeline_parallel/p2p.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 974
unit: def slice_to_str()
file: src/nanotron/parallel/parameters.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 3
number of parameters: 1

id: 975
unit: def str_to_slice()
file: src/nanotron/parallel/parameters.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 3
number of parameters: 1

id: 976
unit: def tuple_to_str()
file: src/nanotron/parallel/parameters.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 3
number of parameters: 2

id: 977
unit: def tuple_from_str()
file: src/nanotron/parallel/parameters.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 978
unit: def get_full_name_from_module_id_to_prefix()
file: src/nanotron/parallel/parameters.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 3

id: 979
unit: def is_tp_sharded()
file: src/nanotron/parallel/parameters.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 980
unit: def is_expert_sharded()
file: src/nanotron/parallel/parameters.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 981
unit: def is_dp_sharded()
file: src/nanotron/parallel/parameters.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 982
unit: def extra_repr()
file: src/nanotron/parallel/tensor_parallel/nn.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 983
unit: def extra_repr()
file: src/nanotron/parallel/tensor_parallel/nn.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 984
unit: def extra_repr()
file: src/nanotron/parallel/tensor_parallel/nn.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 985
unit: def __format__()
file: src/nanotron/parallel/tensor_parallel/enum.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 986
unit: def __str__()
file: src/nanotron/parallel/tensor_parallel/enum.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 987
unit: def backward()
file: src/nanotron/parallel/tensor_parallel/distributed_differentiable_primitives.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 988
unit: def differentiable_identity()
file: src/nanotron/parallel/tensor_parallel/distributed_differentiable_primitives.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 989
unit: def differentiable_all_reduce_sum()
file: src/nanotron/parallel/tensor_parallel/distributed_differentiable_primitives.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 990
unit: def differentiable_all_gather()
file: src/nanotron/parallel/tensor_parallel/distributed_differentiable_primitives.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 991
unit: def differentiable_reduce_scatter_sum()
file: src/nanotron/parallel/tensor_parallel/distributed_differentiable_primitives.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 992
unit: def __init__()
file: src/nanotron/parallel/utils.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 993
unit: def get_tied_id_to_param()
file: src/nanotron/parallel/tied_parameters.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 0

id: 994
unit: def masked_mean()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 3

id: 995
unit: def get_block_compute_costs()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 996
unit: def get_flops_per_sec()
file: src/nanotron/models/llama.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 4

id: 997
unit: def backward()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 0

id: 998
unit: def dropout_add_fused_train()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 3

id: 999
unit: def masked_mean()
file: src/nanotron/models/starcoder2.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 3

id: 1000
unit: def init_model_randomly()
file: src/nanotron/models/base.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 1001
unit: def tie_custom_params()
file: src/nanotron/models/base.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 1002
unit: def get_embeddings_lm_head_tied_names()
file: src/nanotron/models/base.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 1003
unit: def before_tbi_sanity_checks()
file: src/nanotron/models/base.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 1004
unit: def after_tbi_sanity_checks()
file: src/nanotron/models/base.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 1005
unit: def before_optim_step_sanity_checks()
file: src/nanotron/models/base.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 1006
unit: def after_optim_step_sanity_checks()
file: src/nanotron/models/base.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 1007
unit: def detach()
file: src/nanotron/models/base.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 3

id: 1008
unit: def type()
file: src/nanotron/models/base.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 3

id: 1009
unit: def float()
file: src/nanotron/models/base.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 3

id: 1010
unit: def double()
file: src/nanotron/models/base.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 3

id: 1011
unit: def half()
file: src/nanotron/models/base.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 3

id: 1012
unit: def long()
file: src/nanotron/models/base.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 3

id: 1013
unit: def int()
file: src/nanotron/models/base.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 3

id: 1014
unit: def short()
file: src/nanotron/models/base.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 3

id: 1015
unit: def char()
file: src/nanotron/models/base.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 3

id: 1016
unit: def byte()
file: src/nanotron/models/base.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 3

id: 1017
unit: def bool()
file: src/nanotron/models/base.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 3

id: 1018
unit: def bfloat16()
file: src/nanotron/models/base.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 3

id: 1019
unit: def _checkpointed_forward()
file: src/nanotron/models/qwen.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 1020
unit: def masked_mean()
file: src/nanotron/models/qwen.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 3

id: 1021
unit: def get_block_compute_costs()
file: src/nanotron/models/qwen.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 1022
unit: def get_flops_per_sec()
file: src/nanotron/models/qwen.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 4

id: 1023
unit: def __init__()
file: src/nanotron/s3_checkpoints/s3_mover.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 3

id: 1024
unit: def poll()
file: src/nanotron/s3_checkpoints/s3_mover.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 1025
unit: def communicate()
file: src/nanotron/s3_checkpoints/s3_mover.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 1026
unit: def _start_uploading()
file: src/nanotron/s3_checkpoints/s3_mover.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 0

id: 1027
unit: def check_path_is_local()
file: src/nanotron/s3_checkpoints/fsspec.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 1028
unit: def new_group()
file: src/nanotron/distributed.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 0

id: 1029
unit: def all_gather_into_tensor()
file: src/nanotron/distributed.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 0

id: 1030
unit: def get_global_ranks()
file: src/nanotron/distributed.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 2
number of parameters: 1

id: 1031
unit: def __repr__()
file: src/nanotron/fp8/tensor.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 1032
unit: def update_scaling_factor()
file: src/nanotron/fp8/tensor.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 0

id: 1033
unit: def data()
file: src/nanotron/fp8/parameter.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 1034
unit: def data()
file: src/nanotron/fp8/parameter.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 1035
unit: def fp8_meta()
file: src/nanotron/fp8/parameter.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 1036
unit: def __repr__()
file: src/nanotron/fp8/parameter.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 1037
unit: def te_dtype()
file: src/nanotron/fp8/meta.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 1038
unit: def __post_init__()
file: src/nanotron/fp8/meta.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 2
number of parameters: 1

id: 1039
unit: def fp8_max()
file: src/nanotron/fp8/meta.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 1040
unit: def inverse_scale()
file: src/nanotron/fp8/meta.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 1041
unit: def __repr__()
file: src/nanotron/fp8/meta.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 1042
unit: def forward()
file: src/nanotron/fp8/linear.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 0

id: 1043
unit: def __getitem__()
file: src/nanotron/random.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 1044
unit: def __iter__()
file: src/nanotron/random.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 1045
unit: def __len__()
file: src/nanotron/random.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 1046
unit: def __delitem__()
file: src/nanotron/random.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 1047
unit: def __init__()
file: src/nanotron/scaling/parametrization.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 2

id: 1048
unit: def _compute_spectral_std()
file: src/nanotron/scaling/parametrization.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 3

id: 1049
unit: def get_lr()
file: src/nanotron/scaling/parametrization.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 3

id: 1050
unit: def get_lr()
file: src/nanotron/scaling/parametrization.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 3

id: 1051
unit: def _get_global_lr()
file: src/nanotron/scaling/parametrization.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 3

id: 1052
unit: def sanity_consumed_train_samples()
file: src/nanotron/serialize/metadata.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 1053
unit: def consumed_tokens_all_datasets()
file: src/nanotron/serialize/metadata.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 1054
unit: def current_stage()
file: src/nanotron/serialize/metadata.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 1

id: 1055
unit: def to_list()
file: src/nanotron/serialize/metadata.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 2
number of parameters: 5

id: 1056
unit: def get_checkpoint_state_metadata()
file: src/nanotron/serialize/optimizer.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 3

id: 1057
unit: def get_exp_tp_pp_rank_and_size_from()
file: src/nanotron/serialize/utils.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 0

id: 1058
unit: def get_dataloader()
file: run_train.py
start line: 0
end line: 0
size: 2 LOC
McCabe index: 1
number of parameters: 0