-
I intend to add my model to train on wmt task. But I found ParlAi can only build one dictionary, and only support English tokenize.
Could you tell me how can I do?
-
Hello everyone:
I saw this framework support SimpleQuestion WebQuestion, but haven't some examples about KB-QA, so can I use this framework for KB-QA task? thanks for your attention.
-
I was able to create a sqlite db earlier, as described in the retriever Readme as a test, but after deleting the initial test db, I tried again a several hours later and got an UnicodeDecodeError: utf…
-
Добрый день!
Прошу уточнить, каким способом можно указать собственный dataset (json?/tar.gz?) для обучения модели.
Насколько я понимаю, `squad_dataset_reader` загружает и делит dataset на train/…
-
python scripts/reader/interactive.py
09/20/2018 12:46:08 PM: [ CUDA enabled (GPU -1) ]
09/20/2018 12:46:08 PM: [ Initializing model... ]
09/20/2018 12:46:08 PM: [ Loading model /home/vraj/DrQA/data…
-
```
python scripts/interactive.py --tokenizer spacy
11/01/2018 12:15:23 PM: [ Running on CPU only. ]
11/01/2018 12:15:23 PM: [ Initializing pipeline... ]
11/01/2018 12:15:23 PM: [ Initializing doc…
-
Hi I just installed ParlAI, and when trying to run the interactive.py example, it is missing the "crawl-300d-2M.vec" data file. Also, the .opt file refer to specific folders such as "/private/home/edi…
-
здравствуйте
при запуске этого файла **predict** ошибка выдают **ValueError: unsupported pickle protocol: 3**
вы можете объяснить что за файлы "model/vocab.pickle" , " model/weights.pickle" и ка…
-
Hello, thanks for releasing the code. I was playing with your pretrained retriever using `scripts/retriever/interactive.py`, and just wondering if I can also retrieve the text (content) of the Wikiped…
-
__async__ is a reserved word in Python 3.7 so cuda is deprecating the function parameter __async__ in favor of __non_blocking__.
https://github.com/pytorch/pytorch/blob/4d28b65fb8978f1c34765d56f18c…