File size: 2,392 Bytes
52f70ed |
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 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 |
seed: 1337
deterministic: false
gpu: 0
name: null
debug: false
load_model: null
distillation: false
data:
dataset_path: explainable_medical_coding/datasets/mdace_inpatient_icd9.py
target_columns:
- diagnosis_codes
- procedure_codes
max_length: 6000
dataloader:
max_batch_size: 8
batch_size: 8
num_workers: 0
drop_last: false
pin_memory: false
batch_sampler:
name: BySequenceLengthSampler
configs:
bucket_boundaries:
- 400
- 600
- 800
- 1000
- 1200
- 1400
- 1600
- 1800
- 2000
- 2200
- 2600
- 3000
- 3400
- 4000
- 5000
model:
name: PLMICD
autoregressive: false
configs:
model_path: models/roberta-base-pm-m3-voc-hf
chunk_size: 128
cross_attention: true
loss: binary_cross_entropy
lambda_1: 0.0
scale: 1
mask_input: false
trainer:
name: Trainer
epochs: 20
validate_on_training_data: true
print_metrics: false
use_amp: true
threshold_tuning: true
clip_grad_norm: 1
clip_value: 10
optimizer:
name: AdamW
configs:
lr: 5.0e-05
weight_decay: 0
lr_scheduler:
name: linear
configs:
warmup: 0.1
metrics:
- name: F1Score
configs:
average: micro
- name: F1Score
configs:
average: macro
- name: Recall
configs:
average: micro
- name: Recall
configs:
average: macro
- name: Precision
configs:
average: micro
- name: Precision
configs:
average: macro
- name: FPR
configs:
average: micro
- name: FPR
configs:
average: macro
- name: ExactMatchRatio
configs: {}
- name: Precision_K
configs:
k: 5
- name: Precision_K
configs:
k: 8
- name: Precision_K
configs:
k: 15
- name: Recall_K
configs:
k: 5
- name: Recall_K
configs:
k: 10
- name: Recall_K
configs:
k: 15
- name: MeanAveragePrecision
configs: {}
- name: PrecisionAtRecall
configs: {}
- name: AUC
configs:
average: micro
- name: AUC
configs:
average: macro
- name: LossMetric
configs: {}
callbacks:
- name: WandbCallback
configs:
project: explainable-medical-coding
entity: null
- name: SaveBestModelCallback
configs:
split: validation
target: all
metric: map
- name: EarlyStoppingCallback
configs:
split: validation
target: all
metric: map
patience: 3
loss:
name: kl_attention_loss
configs:
lambda_1: 2.5
|