SPARQA
SPARQA copied to clipboard
nju-websoft
oracle_grounded_graph_graphq/result/ 缺失
您好，在执行 code/running/freebase/pipeline_cwq.py 的时候，设置module = "2.1"，报错：
[Errno 2] No such file or directory: 'SPARQA/dataset_graphquestions/oracle_grounded_graph_graphq/result/'
请问应该如何解决呢？谢谢～
您运行的文件是跑cwq的，报的是graphq，可能是配置地方不一致造成的
---原始邮件--- 发件人: "GentleMing_X"<[email protected]> 发送时间: 2020年10月20日(周二) 中午11:57 收件人: "nju-websoft/SPARQA"<[email protected]>; 抄送: "Subscribed"<[email protected]>; 主题: [nju-websoft/SPARQA] oracle_grounded_graph_graphq/result/ 缺失 (#7)
您好，在执行 code/running/freebase/pipeline_cwq.py 的时候，设置module = "2.1"，报错：
[Errno 2] No such file or directory: 'SPARQA/dataset_graphquestions/oracle_grounded_graph_graphq/result/'
请问应该如何解决呢？谢谢～
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or unsubscribe.
您好，在执行code/running/freebase/pipeline_graphq.py，设置module = "2.1"时：
No such file or directory: 'SPARQA/dataset_graphquestions/2019.05.13_test_answers'
请问应该如何解决呢？谢谢～
给您带来不便，抱歉，这个文件应该是已经不用了，晚上我再检查一下，答复您
---原始邮件--- 发件人: "GentleMing_X"<[email protected]> 发送时间: 2020年11月10日(周二) 下午4:03 收件人: "nju-websoft/SPARQA"<[email protected]>; 抄送: "Comment"<[email protected]>;"Yawei Sun"<[email protected]>; 主题: Re: [nju-websoft/SPARQA] oracle_grounded_graph_graphq/result/ 缺失 (#7)
您好，在执行code/running/freebase/pipeline_graphq.py，设置module = "2.1"时：
No such file or directory: 'SPARQA/dataset_graphquestions/2019.05.13_test_answers'
请问应该如何解决呢？谢谢～
— You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe.
您好，在执行 code/running/freebase/pipeline_cwq.py 的时候，设置module = "2.3_word_match"，又报错：
Exception has occurred: ModuleNotFoundError
No module named 'grounding.ranking.path_match_nn'
File "SPARQA/code/grounding/ranking/path_match_word_level/path_match_interface.py", line 30, in set_model_data
self.model = torch.load(model_file,map_location=lambda storage, location: storage)
请问应该如何解决呢？谢谢～
您好, 抱歉README.md我可能没讲清楚, 说明如下: 您跑CWQ问句集的话, globals_args.py里面: q_mode='cwq', kb_mode='kb_freebase_latest', 您跑GraphQ问句集的话, q_mode='graphq', kb_mode='kb_freebase_en_2013'. ------------------ Original ------------------ From: "GentleMing_X"; Date: 2020年11月16日(星期一) 上午10:31 To: "nju-websoft/SPARQA"; Cc: "Yawei Sun"; "Comment"; Subject: Re: [nju-websoft/SPARQA] oracle_grounded_graph_graphq/result/ 缺失 (#7)
您好，在执行 code/running/freebase/pipeline_cwq.py 的时候，设置module = "2.3_word_match"，又报错：
[Errno 2] No such file or directory: "SPARQA/kb_freebase_en_2013/relortype_level_words.json"
请问应该如何解决呢？谢谢～
— You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe.
relortype_level_words是我的问题，刚发出来就删除了，现在是No module named 'grounding.ranking.path_match_nn'的问题~
您好, 抱歉, 这个问题是我上周更改的文件夹命名跟model_file保存的旧的文件名不一致造成的, 您要是想跑我提供的models, 把文件夹名字path_match_word_level改回原来的path_match_nn试试
------------------ Original ------------------ From: "GentleMing_X"; Date: 2020年11月16日(星期一) 中午11:12 To: "nju-websoft/SPARQA"; Cc: "Yawei Sun"; "Comment"; Subject: Re: [nju-websoft/SPARQA] oracle_grounded_graph_graphq/result/ 缺失 (#7)
您好，在执行 code/running/freebase/pipeline_cwq.py 的时候，设置module = "2.3_word_match"，又报错：
Exception has occurred: ModuleNotFoundError No module named 'grounding.ranking.path_match_nn' File "SPARQA/code/grounding/ranking/path_match_word_level/path_match_interface.py", line 30, in set_model_data self.model = torch.load(model_file,map_location=lambda storage, location: storage)
请问应该如何解决呢？谢谢～
— You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe.
要不您别改了, 改的地方有点多, 我现在改回原来的, 上传一下, 您更新一下grounding/ranking下面的文件 ------------------ Original ------------------ From: "GentleMing_X"; Date: 2020年11月16日(星期一) 下午5:20 To: "nju-websoft/SPARQA"; Cc: "Yawei Sun"; "Comment"; Subject: Re: [nju-websoft/SPARQA] oracle_grounded_graph_graphq/result/ 缺失 (#7)
relortype_level_words是我的问题，刚发出来就删除了，现在是No module named 'grounding.ranking.path_match_nn'的问题~
— You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe.
running文件夹, 您也更新一下, 不是, 可能是有些尝试过代码,我没删除, 闲了, 可以再精简优化一下.
感谢您的关注
------------------ Original ------------------ From: "GentleMing_X"; Date: 2020年11月16日(星期一) 下午5:34 To: "nju-websoft/SPARQA"; Cc: "Yawei Sun"; "Comment"; Subject: Re: [nju-websoft/SPARQA] oracle_grounded_graph_graphq/result/ 缺失 (#7)
好，感谢感谢，这个工作代码量真的好大啊
— You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe.
您好，跑graphquestion时，pipeline_graphq.py设置module = "2.2"，运行过程中run_grounded_graph_generation_by_structure_transformation函数中的grounded_graph_forest变量一直是[]，即generate_candidates_by_2_1_grounded_graph_interface函数的返回值oracle_graphs一直是[]，想问下这是正常现象吗？
您好，不正常，您或许要检查一下您本地的dataset_name里面的GraphqFileName类下的oracle_graphs_file是否更改为您本地的路径 ------------------ Original ------------------ From: "GentleMing_X"; Date: 2020年11月16日(星期一) 晚上11:01 To: "nju-websoft/SPARQA"; Cc: "Yawei Sun"; "Comment"; Subject: Re: [nju-websoft/SPARQA] oracle_grounded_graph_graphq/result/ 缺失 (#7)
您好，跑graphquestion时，pipeline_graphq.py设置module = "2.2"，运行过程中run_grounded_graph_generation_by_structure_transformation函数中的grounded_graph_forest变量一直是[]，即generate_candidates_by_2_1_grounded_graph_interface函数的返回值oracle_graphs一直是[]，想问下这是正常现象吗？
— You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe.
你好，在跑2.2的时候oracle_grounded_graph_graphq/result/ 的缺失导致grounded_graph_forest，为空请问如何解决？本地路径导向的是什么文件？
您好，您需要下载README.md提示的Oracle Grounded Graph，它解压以后就是result里面的文件，也可能我没讲清楚，抱歉。
下载并生成，有点繁琐，为帮助理解，针对数据集提供2.2样例输出，具体参见dataset_graphquestions/output_graphq和dataset_cwq_1_1/output_cwq下面的2.2打头的文件夹