Closed FanY1999 closed 3 years ago
Hi, we are indeed using the V2 data, and the call to data preprocessing can be found here in main.py line 174. The evaluation code is general and not specific to V1 or V2 data. When you run setup.sh, the data will be downloaded to Span-ASTE/aste/data, and you can replace the files with your own data as long as the format is the same. Hope this is helpful!
Hi, we are indeed using the V2 data, and the call to data preprocessing can be found here in main.py line 174. The evaluation code is general and not specific to V1 or V2 data. When you run setup.sh, the data will be downloaded to Span-ASTE/aste/data, and you can replace the files with your own data as long as the format is the same. Hope this is helpful!
thanks for your reply! I didn’t look at the files carefully before,Now I have reproduced it successfully
On the one hand,Do you have any preprocessing of the dataset? Since I didn't see the data folder in this repo. Can I just drop the ASTE-Dataset-V2 into the corresponding folder without any preprocess?
On the other hand,I can't find your code for processing the ASTE-Dataset-V2 data set, the relevant code seems to be processing the V1 data set. (evaluation.py->class TagReader->def read_inst->(line 159))