bootleg/configs/standard/train.yaml at master · HazyResearch/bootleg · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
emmental:
  lr: 2e-5
  n_steps: 5
  evaluation_freq: 6
  warmup_percentage: 0.1
  lr_scheduler: linear
  log_path: bootleg_logs
  checkpointing: true
  checkpoint_all: true
  checkpoint_freq: 1
  counter_unit: batch
  online_eval: false
  clear_intermediate_checkpoints: false
  checkpoint_metric: NED/Bootleg/dev/final_loss/acc_boot:max
  l2: 0.01
  grad_clip: 1.0
  gradient_accumulation_steps: 4
  fp16: true
run_config:
  log_level: DEBUG
  eval_batch_size: 32
  dataloader_threads: 4
  dataset_threads: 30
  spawn_method: fork
train_config:
  batch_size: 32
model_config:
  hidden_size: 200
  normalize: true
  temperature: 0.01
data_config:
  data_dir: /lfs/raiders8/0/lorr1/output
  train_in_candidates: true
  data_prep_dir: prep
  use_entity_desc: true
  context_mask_perc: 0.0
  entity_type_data:
    use_entity_types: true
    type_symbols_dir: type_mappings/wiki
  entity_kg_data:
    use_entity_kg: true
    kg_symbols_dir: kg_mappings
  entity_dir: /lfs/raiders8/0/lorr1/output/entity_db
  max_seq_len: 128
  max_seq_window_len: 64
  max_ent_len: 128
  overwrite_preprocessed_data: false
  dev_dataset:
    file: dev.jsonl
    use_weak_label: true
  test_dataset:
    file: test.jsonl
    use_weak_label: true
  train_dataset:
    file: train.jsonl
    use_weak_label: true
  word_embedding:
    cache_dir: logs/pretrained_bert_models
    bert_model: bert-base-chinese
    context_layers: 6
    entity_layers: 6