datalee
datalee
did you solve the problem?
> Updates please, we have issues with too many threaded connection to postgresql using the app. postgresql also has this issue?
在 export_saved_model.py 中更新 export_path = ‘../my_image_classifier/1’为 export_path = ‘../my_image_classifier/2’。 TensorFlow Serving 会自动检测出 my_image_classifier 目录下模型的新版本,并在服务器中更新它。 ===》没有类似操作吗?求作者解惑,谢谢
淡了淡了,这个不管了
@vi3k6i5 also,there is some issues with chinese from file: `processor.add_keyword_from_file('D:/keywords.txt')` > --------------------------------------------------------------------------- UnicodeDecodeError Traceback (most recent call last) in () ----> 1 processor.add_keyword_from_file('D:/keywords.txt') D:\Program Files\Python35\lib\site-packages\flashtext\keyword.py in add_keyword_from_file(self, keyword_file) 313 raise...
yes,i know must be encoding issue, but i don't find add_keyword_from_file has the parameter of setting it.so...... add_keyword_from_file() got an unexpected keyword argument 'encoding'
这个只是针对图像类的模型吗? NLP类的能用不?
> @datalee 可以,目前已经有UIE。 PaddleNLP的所有部署之后都会由FastDeploy来支持哈 有对比过和onnx之间的性能差异吗?
> 必须要把同义词 加到 分词库里 , 男人,男孩,男子,男性,男孩子,男同胞 要加到ik的分词库里 未必吧,你定义好tokenizer不就行了吗