SikuBERT-verb-couplet-simplified-0.1

This is a fine-tuned SikuBERT model for token-level verb classification in Classical Chinese couplets.
It classifies each character as either verb (v) or non-verb (nv).

---

Usage

from transformers import pipeline

classifier = pipeline(
    "token-classification",
    model="qhchina/SikuBERT-verb-couplet-simplified-0.1",
)

text = "无边落木萧萧下，不尽长江滚滚来"
outputs = classifier(text)

for token in outputs:
    label = 1 if token["entity"] == "v" else 0
    print(f"{token['word']} -> {label}")

# 无 -> 0
# 边 -> 0
# 落 -> 0
# 木 -> 0
# 萧 -> 0
# 萧 -> 0
# 下 -> 1
# 不 -> 0
# 尽 -> 0
# 长 -> 0
# 江 -> 0
# 滚 -> 0
# 滚 -> 0
# 来 -> 1