File size: 4,619 Bytes
3d36724 | 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 | """
HuggingFaceμμ λͺ¨λΈμ λ‘λνμ¬ μΆλ‘ νλ μμ
Usage:
from inference_example import extract_sentences
results = extract_sentences("μΌμ±μ μμ μ€μ μ΄ μμ₯ μμμ μννλ€. ...")
"""
import re
from typing import List, Dict
import torch
from transformers import AutoTokenizer
from model import (
DocumentEncoderConfig,
DocumentEncoderForExtractiveSummarization,
IDX_TO_ROLE,
)
def split_into_sentences(text: str) -> List[str]:
sentences = re.split(r"(?<=[.!?])\s+", text.strip())
return [s.strip() for s in sentences if s.strip()]
def extract_sentences(
text: str,
model_name_or_path: str = "./", # λ‘컬 λλ HuggingFace repo ID
top_k: int = 3,
threshold: float = 0.5,
device: str = None,
) -> Dict:
"""
ν
μ€νΈμμ λνλ¬Έμ₯μ μΆμΆνκ³ μν μ λΆλ₯ν©λλ€.
Args:
text: μ
λ ₯ ν
μ€νΈ (κΈμ΅ 리ν¬νΈ λ±)
model_name_or_path: λͺ¨λΈ κ²½λ‘ λλ HuggingFace repo ID
top_k: μΆμΆν μ΅λ λ¬Έμ₯ μ
threshold: λνλ¬Έμ₯ νλ¨ μκ³κ°
device: cuda λλ cpu
Returns:
dict with 'sentences', 'all_scores', 'all_roles', 'selected'
"""
device = device or ("cuda" if torch.cuda.is_available() else "cpu")
config = DocumentEncoderConfig.from_pretrained(model_name_or_path)
model = DocumentEncoderForExtractiveSummarization.from_pretrained(
model_name_or_path, config=config
)
model = model.to(device)
model.eval()
tokenizer = AutoTokenizer.from_pretrained(model_name_or_path)
sentences = split_into_sentences(text)
if not sentences:
return {"sentences": [], "all_scores": [], "all_roles": [], "selected": []}
max_sentences = config.max_sentences
max_length = config.max_length
padded = sentences[:max_sentences]
num_real = len(padded)
while len(padded) < max_sentences:
padded.append("")
all_input_ids, all_attention_mask = [], []
for s in padded:
if s:
enc = tokenizer(s, max_length=max_length, padding="max_length", truncation=True, return_tensors="pt")
else:
enc = {
"input_ids": torch.zeros(1, max_length, dtype=torch.long),
"attention_mask": torch.zeros(1, max_length, dtype=torch.long),
}
all_input_ids.append(enc["input_ids"])
all_attention_mask.append(enc["attention_mask"])
input_ids = torch.cat(all_input_ids, dim=0).unsqueeze(0).to(device)
attention_mask = torch.cat(all_attention_mask, dim=0).unsqueeze(0).to(device)
document_mask = torch.zeros(1, max_sentences, device=device)
document_mask[0, :num_real] = 1
with torch.no_grad():
scores, role_logits = model(input_ids, attention_mask, document_mask)
scores_list = scores[0, :num_real].tolist()
role_indices = role_logits[0, :num_real].argmax(dim=-1).tolist()
roles_list = [IDX_TO_ROLE[idx] for idx in role_indices]
selected = []
for i, (sent, score, role) in enumerate(zip(sentences, scores_list, roles_list)):
if score >= threshold:
selected.append({"index": i, "sentence": sent, "score": score, "role": role})
selected.sort(key=lambda x: x["score"], reverse=True)
selected = selected[:top_k]
selected.sort(key=lambda x: x["index"])
return {
"sentences": sentences,
"all_scores": scores_list,
"all_roles": roles_list,
"selected": selected,
}
if __name__ == "__main__":
text = """
μΌμ±μ μμ 2024λ
4λΆκΈ° μ€μ μ΄ μμ₯ μμμ μννλ€.
λ©λͺ¨λ¦¬ λ°λ체 κ°κ²© μμΉμΌλ‘ μμ
μ΄μ΅μ΄ μ λΆκΈ° λλΉ 30% μ¦κ°νλ€.
HBM3E μμ°μ΄ 본격νλλ©΄μ AI λ°λ체 μμ₯ μ μ μ¨μ΄ νλλ μ λ§μ΄λ€.
λ€λ§, μ€κ΅ μμ₯μ λΆνμ€μ±μ΄ μ¬μ ν 리μ€ν¬ μμΈμΌλ‘ μμ©νκ³ μλ€.
νμ¬λ μ¬ν΄ μ€λΉ ν¬μλ₯Ό 20% νλν κ³νμ΄λ€.
"""
result = extract_sentences(text, model_name_or_path="./")
print("=" * 60)
print("μ 체 λ¬Έμ₯ λΆμ:")
for i, (s, sc, r) in enumerate(zip(result["sentences"], result["all_scores"], result["all_roles"])):
marker = "*" if sc >= 0.5 else " "
print(f" {marker} {i+1}. [{sc:.4f}] [{r:10s}] {s}")
print(f"\nμ νλ λνλ¬Έμ₯:")
for item in result["selected"]:
print(f" - [{item['score']:.4f}] [{item['role']:10s}] {item['sentence']}")
|