-
We have to do a fair amount of text preprocessing of our data before feeding it into tensorflow. Since the text manipulation abilities of tensorflow and tensorflow tranform are still relatively immatu…
-
Hey guys, I am trying to add some customized new tokens in a Code related model `deepseek-ai/deepseek-coder-6.7b-instruct`. However the model **completely loses performance** after using `tokenizer.ad…
-
Hello
The `aac-metrics-download` command fails because it can't download Stanford CoreNLP from the official link. It seems like the link to CoreNLP zip file is no longer valid, and the version used…
-
Running the following code leads to a `google.protobuf.message.DecodeError: Error parsing message`.
This admittedly is a very weird document. A chinese(?) text snuck into my English pipeline and c…
-
Hi there,
I'm trying to use epic in an Apache Spark Streaming environment but I'm experiencing some difficulty loading the models. I'm not really sure whether this is an Epic issue, a Breeze issue, a…
-
When I try to install I got errors as follow:
I have JDK 1.8 and Ptyhon 2.7 installed.
Is there something wrong with my configuration ?
`
====Executing command 'npm install stanford-corenlp --save '=…
-
I'd like to play around with this UI running locally, but my laptop doesn’t have a GUI. But there are public APIs to run models (e.g. https://replicate.com/stability-ai/stable-diffusion), and I’d be w…
-
Hello. I am interested in the evaluation of AI systems and LLMs both in general and specifically with respect to reading comprehension, story comprehension (e.g., [NarrativeQA](https://crfm.stanford.e…
-
Specifying the MULTI_RELATION as optional seems to cause an infinite loop of printing matches. I am using Tregex 4.2.0 downloaded from [https://nlp.stanford.edu/software/tregex.shtml#Download](https:/…
-
The code contains bugs and cannot produce reasonable results. For example, the code tries to sort all; training instances in a batch by length, but the roots are NOT reordered accordingly [see here](h…