Method__init__(self, optimizer, start_lr,
warmup_iter, total_iters,
decay_style, last_iter
megatron/learning_rates.py:26
Method__init__(self, input_in_fp16, upper_triang_mask_fusion,
general_mask_fusion, mask_func, softmax_in_f
megatron/model/fused_softmax.py:86
Method__init__(self, attention_mask_func, init_method,
output_layer_init_method, layer_number,
megatron/model/transformer.py:118
Method__init__(self, attention_mask_func, init_method,
output_layer_init_method, layer_number,
megatron/model/transformer.py:429
Method__init__(self, max_decode_len, bos_id, eos_id, beam_size=5, alpha=0.6, topk_evidence=-1)
megatron/model/search_strategy.py:125
Method__init__(self, max_decode_len, bos_id, eos_id, sample=False, topk_evidence=-1)
megatron/model/search_strategy.py:182
Method__init__(self, mpu_vocab_size, hidden_size, init_method,
layernorm_epsilon, parallel_output)
megatron/model/bert_model.py:72
Method__init__(self, sampler, batch_size, drop_last, rank=-1,
world_size=2, wrap_last=False, interleave=Fal
megatron/data/samplers.py:95
Method__init__(self, task_name, dataset_name, datapath, tokenizer, max_seq_length)
megatron/data/orqa_wiki_dataset.py:143
Method__init__(self, input_size, output_size, bias=True, gather_output=True,
init_method=init.xavier_normal
megatron/mpu/layers.py:194
Method__init__(self, task_name, dataset_name, datapaths,
tokenizer, max_seq_length, decoder_seq_length)
tasks/openqa/e2eqa/train_data_utils.py:106
Method__init__(self, task_name, dataset_name, datapaths, tokenizer, max_seq_length, evaluate=False)
tasks/openqa/dense_retriever/train_data_utils.py:133
Method__init__(self, name, datapaths, tokenizer, max_seq_length, evaluate=False)
tasks/openqa/dense_retriever/train_data_utils.py:228