Hanlei Zhang
Hanlei Zhang
First, the data need to be in the same format as in [Banking](https://github.com/thuiar/TEXTOIR/blob/main/data/banking/test.tsv). Second, you can just load the pre-trained model for inference (It should predict intent labels for each...
> Hi~ Thank you so much for your open-source platform. I want to run MSP method in intent detection. I run this commonds sh examples/run_ADB.sh and sh examples/run_MAP.sh. The platform...
> And I can not install all the dependencies by running pip install -r requirements.txt. The differences are: > 1. mkl_random==1.2.2 > 2. torch==1.7.1+cu110 > 3. Pillow==8.4.0 > 4. transformers==4.23.1...
Hello! I'm not entirely clear on your question. Since the newer versions of the library are built upon the older ones, the outcomes are typically consistent.
Q1) In this work, discovering new intents means finding new categories rather than new labels. It is because we do not know the true label of new intents in real-world...
Hi, thanks for your attention in this work! We have updated the details of data and basic usage of this code. More details about the framework will be updated in...
> Hi, the .csv file in the "results" folder shows 20 results, and the seeds are from 0 to 9 twice. Is the "acc" value in the paper the average...
Hi, the ood_detection_map is not used in this code. I have deleted it. Thanks for your mentioning this problem.
请问是按照源代码运行之后报的这个错吗？能给出具体一些的实验参数相关信息吗？
We have updated the raw data on the Google Drive, enjoy it!