Ecosyste.ms: Issues

An open API service for providing issue and pull request metadata for open source projects.

GitHub / tomaarsen/SpanMarkerNER issues and pull requests

#73 - Use getattr instead of __getattribute__ to get tokenizer args

Pull Request - State: closed - Opened by tomaarsen about 1 month ago

#72 - Tag revisions of models on HuggingFace

Issue - State: open - Opened by sanmai-NL 2 months ago

#71 - Issue with CUDA 12.x Inference, Performance completely obliterated

Issue - State: open - Opened by gilljon 2 months ago - 4 comments

#69 - `span_marker` 1.6.0: Add missing `sklearn` (`scikit-learn`) dependency

Issue - State: open - Opened by sanmai-NL 3 months ago - 7 comments

#67 - AttributeError: 'SpanMarkerConfig' object has no attribute 'label2id'

Issue - State: open - Opened by jdb110 3 months ago - 1 comment

#66 - Inference threshold

Issue - State: open - Opened by Definelymes 5 months ago

#65 - handle removed DatasetFilter class

Pull Request - State: closed - Opened by logan-markewich 7 months ago - 4 comments

#63 - No f1, precision and recall scores during training

Issue - State: open - Opened by adap7 7 months ago

#62 - Urgent : Example to use custom dataset from csv file

Issue - State: open - Opened by vatsaldin 8 months ago

#61 - `mode="Strict"` for seqeval?

Issue - State: open - Opened by tan-js 8 months ago

#59 - Stacking of embeddings

Issue - State: open - Opened by ganga7445 10 months ago

#58 - UnicodeDecodeError when import span_marker

Issue - State: open - Opened by p-dre 11 months ago - 1 comment

#57 - Stop enabling progress bars when they are disabled

Pull Request - State: open - Opened by vangheem 11 months ago

#56 - deberta-v3 encoder error

Issue - State: open - Opened by eek 11 months ago - 1 comment

#55 - is <start> <end> ever used?

Issue - State: closed - Opened by 2533245542 12 months ago - 1 comment

#54 - implementing the relation identification part?

Issue - State: open - Opened by 2533245542 12 months ago

#53 - Spacy Integration - "detect an empty sentence"

Issue - State: open - Opened by mrw-dev about 1 year ago - 5 comments

#52 - Possible to load your own trained models with internet disabled?

Issue - State: open - Opened by justin-donnelly about 1 year ago - 4 comments

#51 - SpanMaker not working on custom dataset

Issue - State: open - Opened by andysingal about 1 year ago - 1 comment

#50 - Add `num_proc=4` to Dataset maps in Trainer

Pull Request - State: open - Opened by tomaarsen about 1 year ago

#49 - Bert-based models crash

Issue - State: open - Opened by lambdavi about 1 year ago - 4 comments

#48 - Evaluation Metrics with Nervalute

Issue - State: closed - Opened by lambdavi about 1 year ago - 1 comment

#47 - num_proc not specified in .map functions

Issue - State: open - Opened by sandstromviktor about 1 year ago - 3 comments

#46 - Warning update for evaluation done without document-level context

Pull Request - State: closed - Opened by jayant-yadav about 1 year ago - 4 comments

#44 - Error loading SpanMarkerTokenizer

Issue - State: closed - Opened by polodealvarado over 1 year ago - 2 comments

#43 - onnx_implementation added

Pull Request - State: open - Opened by polodealvarado over 1 year ago - 28 comments

#42 - inference time cpu vs gpu

Issue - State: open - Opened by ganga7445 over 1 year ago - 3 comments

#41 - Using spaCy `nlp.pipe` now processes texts sentence-wise, just like for `nlp(...)`.

Pull Request - State: closed - Opened by tomaarsen over 1 year ago - 2 comments

#40 - Improve the error when passing an empty sentence to the tokenizer

Pull Request - State: closed - Opened by tomaarsen over 1 year ago

#37 - spaCy_integration `.pipe()` does not behave as expected

Issue - State: closed - Opened by q-jackboylan over 1 year ago - 1 comment
Labels: enhancement

#36 - add spanmarker to spacy in poetry plugins in pyproject.toml

Pull Request - State: closed - Opened by davidberenstein1957 over 1 year ago - 2 comments

#35 - should return same no. of list as of inputs

Issue - State: closed - Opened by mallapraveen over 1 year ago - 6 comments

#34 - Add pipeline support for SpanMarker

Pull Request - State: closed - Opened by tomaarsen over 1 year ago
Labels: enhancement

#33 - Confusing error thrown when tokens is empty

Issue - State: closed - Opened by tomaarsen over 1 year ago
Labels: enhancement

#32 - Hugging Face Space URL not working for FewNERD fine-tuned model

Issue - State: closed - Opened by alvarobartt over 1 year ago - 2 comments

#31 - Add support for BILO label scheme

Pull Request - State: closed - Opened by tomaarsen over 1 year ago - 8 comments

#30 - Cannot train BILOU scheme with no singletons

Issue - State: closed - Opened by david-waterworth over 1 year ago - 1 comment

#29 - Tests: Reintroduce Windows tests

Pull Request - State: closed - Opened by tomaarsen over 1 year ago

#28 - Heavily improve automatic model card generation + Patch XLM-R

Pull Request - State: closed - Opened by tomaarsen over 1 year ago - 13 comments
Labels: enhancement

#27 - Use `dataclasses.replace` due to immutability of TrainingArguments

Pull Request - State: closed - Opened by tomaarsen over 1 year ago
Labels: bug

#26 - SpanMarker with ONNX models

Issue - State: open - Opened by Ulipenitz over 1 year ago - 11 comments

#25 - Integrate Entity Ruler with Span Marker model

Issue - State: closed - Opened by ralucaginga over 1 year ago - 1 comment

#24 - Downgrade dependencies to try and fix CI breaking on Windows

Pull Request - State: closed - Opened by tomaarsen over 1 year ago - 1 comment

#23 - Note: (XLM-)RoBERTa-based SpanMarker models require text preprocessing

Issue - State: closed - Opened by tomaarsen over 1 year ago - 3 comments

#22 - Choose class-candidates during inference

Issue - State: open - Opened by Ulipenitz over 1 year ago

#21 - How to make this work for overlapping entities?

Issue - State: closed - Opened by seekingpeace over 1 year ago - 2 comments
Labels: question

#20 - Unexpectedly (bad) predictions?

Issue - State: closed - Opened by mbauwens over 1 year ago - 5 comments

#19 - Prevent re-adding contextual information when training with document-level context

Issue - State: open - Opened by tomaarsen over 1 year ago
Labels: enhancement

#18 - Different results on ontonotes

Issue - State: closed - Opened by YeDeming over 1 year ago - 7 comments

#17 - spaCy integration ignores old entities

Issue - State: closed - Opened by davidberenstein1957 over 1 year ago - 2 comments
Labels: enhancement

#16 - added `.pipe()` method to spaCy integration

Pull Request - State: closed - Opened by davidberenstein1957 over 1 year ago - 10 comments

#15 - spaCy integration has no `.pipe()` method, hence will fallback to individual `.call()`

Issue - State: closed - Opened by davidberenstein1957 over 1 year ago - 1 comment
Labels: enhancement

#14 - Pretrained tokenizer

Issue - State: closed - Opened by darebfh over 1 year ago - 2 comments

#13 - Issue with adding span_marker pipe within spaCy

Issue - State: closed - Opened by Rao-Sanjayy over 1 year ago - 3 comments
Labels: bug

#11 - Add `trained_with_document_context` to SpanMarkerConfig

Pull Request - State: closed - Opened by tomaarsen over 1 year ago

#10 - Improve speed of sample spreading - much faster inference

Pull Request - State: closed - Opened by tomaarsen over 1 year ago

#9 - Add support for document-level context for training, evaluating & inference

Pull Request - State: closed - Opened by tomaarsen over 1 year ago - 1 comment
Labels: enhancement

#8 - Can use with BetterTransformer?

Issue - State: closed - Opened by cceyda over 1 year ago - 3 comments

#7 - Entity type

Issue - State: closed - Opened by similang almost 2 years ago - 2 comments

#6 - MobileBert

Issue - State: closed - Opened by ande4485 almost 2 years ago - 2 comments

#5 - amazing works but minimum token ?

Issue - State: closed - Opened by noctetemp almost 2 years ago - 4 comments

#4 - [Fix] Ensure models are on eval mode when predicting

Pull Request - State: closed - Opened by tomaarsen almost 2 years ago
Labels: enhancement

#3 - [Docs] Mention thesis in docs & README

Pull Request - State: closed - Opened by tomaarsen almost 2 years ago
Labels: documentation

#2 - `Trainer.preprocess_dataset` - `KeyError` with `datasets<2.6.0`

Issue - State: closed - Opened by pdhall99 almost 2 years ago - 3 comments
Labels: bug

#1 - Change position_id calculation to fix last token being ignored.

Pull Request - State: closed - Opened by cceyda almost 2 years ago - 11 comments
Labels: bug