Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

Roberta SNLI and MNLI don't return results

See original GitHub issue

Frontend error:

react-dom.production.min.js:209 TypeError: Invalid attempt to destructure non-iterable instance.
In order to be iterable, non-array objects must have a [Symbol.iterator]() method.
    at nonIterableRest.js:2
    at o (slicedToArray.js:6)
    at Output (main.49d1a9e9.chunk.js:1)
    at qi (react-dom.production.min.js:153)
    at gs (react-dom.production.min.js:261)
    at cl (react-dom.production.min.js:246)
    at sl (react-dom.production.min.js:246)
    at Xs (react-dom.production.min.js:239)
    at react-dom.production.min.js:123
    at t.unstable_runWithPriority (scheduler.production.min.js:19)
es @ 2.8af9b7e1.chunk.js:2
main.49d1a9e9.chunk.js:1 TypeError: Invalid attempt to destructure non-iterable instance.
In order to be iterable, non-array objects must have a [Symbol.iterator]() method.
    at nonIterableRest.js:2
    at o (slicedToArray.js:6)
    at Output (main.49d1a9e9.chunk.js:1)
    at qi (react-dom.production.min.js:153)
    at gs (react-dom.production.min.js:261)
    at cl (react-dom.production.min.js:246)
    at sl (react-dom.production.min.js:246)
    at Xs (react-dom.production.min.js:239)
    at react-dom.production.min.js:123
    at t.unstable_runWithPriority (scheduler.production.min.js:19)
(anonymous) @ main.49d1a9e9.chunk.js:1
DevTools failed to load SourceMap: Could not parse content for https://demo.allennlp.org/static/js/main.49d1a9e9.chunk.js.map: Unexpected token < in JSON at position 0

I believe the backend is returning results:

{label: "contradiction", logits: [-4.040728569030762, 3.556131362915039, 1.1318637132644653],…}
label: "contradiction"
logits: [-4.040728569030762, 3.556131362915039, 1.1318637132644653]
probs: [0.00046097763697616756, 0.9182357788085938, 0.08130328357219696]
token_ids: [0, 1596, 390, 32, 26884, 552, 5, 8373, 4835, 1437, 12646, 6845, 4, 2, 2, 1596, 390, 32, 2828, 15, 10,…]
tokens: ["<s>", "ĠTwo", "Ġwomen", "Ġare", "Ġwandering", "Ġalong", "Ġthe", "Ġshore", "Ġdrinking", "Ġ", "iced",…]

Issue Analytics

State:
Created 3 years ago
Comments:9 (9 by maintainers)

Top GitHub Comments

1reaction

aimichalcommented, Jun 10, 2020

Alright, I think I got somewhere. PR coming up momentarily.

0reactions

aimichalcommented, Jun 10, 2020

Got it to repro locally by running docker-compose -f docker-compose.ui-only.yaml up --build (which uses the deployed models), and see a helpful stack trace like this:

Going to jump into the code now to see what’s what.

Top Results From Across the Web

Retrain transformer-based models in allennlp_models ...

I see BERT SRL checked here, but there's this issue that says that performance actually doesn't match, with a solution that seems to...

Natural Language Inference: An Overview | by Oleh Lokshyn

Natural Language Inference (NLI) is the task of determining whether the given “hypothesis” logically follows from the “premise”.

AllenNLP Models v2.5.0

Components provided: Dataset readers for several datasets, including SNLI and Quora Paraphrase. Reading Comprehension. Reading comprehension tasks involve ...

usc-isi/sbert-roberta-large-anli-mnli-snli - Hugging Face

We're on a journey to advance and democratize artificial intelligence through open source and open science.

MultiNLI Dataset - Papers With Code

The Multi-Genre Natural Language Inference (MultiNLI) dataset has 433K sentence pairs. Its size and mode of collection are modeled closely like SNLI.