allennlp-demo
allennlp-demo copied to clipboard
Code for the AllenNLP demo.
This error is being throw ~500 times per day or so and should be investigated (and ideally resolved): ``` Traceback (most recent call last): File "server.py", line 20, in from...
I am trying to run the allennlp-demo right now and I just want to run the ui-only version of the website. I am able to build the docker correctly with...
one example is mnli refers to snli in training command: https://demo.allennlp.org/textual-entailment/roberta-mnli 
There are two pods running the NAQANet model, and they each consistently return different results. For example, pod allennlp-demo-api-naqanet-prod-597f89dc7f-7bwhw returns with answer.value = `game, players direct a nasce...`, while pod...
ive seen this on multiple models... but one example is: https://demo.allennlp.org/named-entity-recognition/fine-grained-ner 
is it only on this one task? https://demo.allennlp.org/constituency-parsing  i propose just cutting the line
MLM and ML do not have usage information.
### Task Reading Comprehension ### Model NAQANet ### Feature HotFlip Attack ### Issue Trying to run the HotFlip attack for the NAQANet model results in a blank white page. The...
I want to keep a list of features we're removing for performance reasons. I suspect a lot of these can come back when we start running the demo on GPUs....
AllenNLP has changed it's id scheme, so that pretrained models now include the task they were trained for in their id. For instance, `bidaf` is now `rc-bidaf`, where the `rc`...