search

yuanzhoulvpi2017 / quick_sentence_transformers

sentence-transformers to onnx 让sbert模型推理效率更快
MIT License
162 stars 30 forks source link

无法指定GPU #7

Open easonforai opened 8 months ago

easonforai commented 8 months ago

InferSentenceTransformer(model_name_or_path=model_path, device="cuda:1", onnx_model_name=onnx_model_name) 这里指定第二块卡时，运行时还是都用在第一块卡上