sentence-transformers How to use tonnnnnnnnnnnnn/semantic_text_sim-v1 with sentence-transformers:
from sentence_transformers import SentenceTransformer
model = SentenceTransformer("tonnnnnnnnnnnnn/semantic_text_sim-v1")
sentences = [
"Koyomi Araragi is an aloof boy who holds a strange, supernatural secret which inadvertently leads him to others with similar stories. Gods, spirits and afflictions can be pesky things, taking important memories or causing unusual tendencies – a fact that Koyomi and others are unfortunately aware of. Using the help of an eccentric homeless man, Koyomi is able to help new friends he meets along the way with their own paranormal conundrums…",
"What is the name of the manga with the description \"For decades, the brave warriors of the Northern Heavenly Sect fought to keep the world safe from the evil Silent Night. But when the fourth generation leader is accused of colluding with the enemy, he is forced to disband the sect and commit suicide to protect his son, Moowon. With no family and allies left, Moowon leads a dreary life under close surveillance... until a surprise attack from the Silent Night gives him a chance to escape to the mountains. There, the young warrior trains to master the fighting techniques of his predecessors, before the mysterious disappearance of a loved one brings him back to the mainland. Follow Moowon as he embarks on a journey to avenge his father’s death and take down the villains who threaten to plunge the world into darkness!\"",
"What is a manga about a student with a disability who is ridiculed by the school's bullies until someone appears to stand by his side?",
"What manga has a protagonist named Kindaichi who travels to various places where a murder has taken place, typically involving ghosts, curses, myths and folklore of significant events from the distant past, and solves the mystery using ingenious deductions of curious clues and his cool magic?"
]
embeddings = model.encode(sentences)
similarities = model.similarity(embeddings, embeddings)
print(similarities.shape)
# [4, 4] 
