Task-level Distributionally Robust Optimization for Large Language Model-based Dense Retrieval
Paper • 2408.10613 • Published
s0-baseline-Qwen1.5-0.5B
tDRO: Task-level Distributionally Robust Optimization for Large Language Model-based Dense Retrieval. Guangyuan Ma, Yongliang Ma, Xing Wu, Zhenpeng Su, Ming Zhou and Songlin Hu.
This is a fine-tuned baseline retriever with uniform sampling weights of tdro-llm/finetune_data. This model is also used as a reference model in tDRO.
Model tree for tdro-llm/s0-baseline-Qwen1.5-0.5B
Base model
Qwen/Qwen1.5-0.5B
from sentence_transformers import SentenceTransformer model = SentenceTransformer("tdro-llm/s0-baseline-Qwen1.5-0.5B") sentences = [ "That is a happy person", "That is a happy dog", "That is a very happy person", "Today is a sunny day" ] embeddings = model.encode(sentences) similarities = model.similarity(embeddings, embeddings) print(similarities.shape) # [4, 4]