explosion spaCy Help-model-advice Discussions
Help: Model Advice Discussions
Understanding a model's predictions and improving accuracy
-
You must be logged in to vote Do 3.4.0 English models use floret vectors?
docsDocumentation and website feat / vectorsFeature: Word vectors and similarity -
You must be logged in to vote Why do some entities perform better than others?
feat / nerFeature: Named Entity Recognizer perf / accuracyPerformance: accuracy feat / transformerFeature: Transformer -
You must be logged in to vote Splitting compound words into parts
third-partyThird-party packages and services feat / tokenizerFeature: Tokenizer -
You must be logged in to vote NER for alphanumeric tokens
feat / nerFeature: Named Entity Recognizer feat / tokenizerFeature: Tokenizer -
You must be logged in to vote Inaccurate tagging of csubjpass
lang / enEnglish language data and models feat / parserFeature: Dependency Parser perf / accuracyPerformance: accuracy -
You must be logged in to vote Train floret embedding iteratively
feat / vectorsFeature: Word vectors and similarity -
You must be logged in to vote Subordinate clauses incorrectly parsed by displacy
lang / enEnglish language data and models feat / parserFeature: Dependency Parser -
You must be logged in to vote Bad NER accuracy
feat / nerFeature: Named Entity Recognizer perf / accuracyPerformance: accuracy -
You must be logged in to vote Train a relation extraction model with spaCy
trainingTraining and updating models feat / relFeature: Relation Extractor -
You must be logged in to vote Custom NER training - understanding loss and learn rate schedule
🔮 thincspaCy's machine learning library Thinc trainingTraining and updating models -
You must be logged in to vote Updating NER trained model?
trainingTraining and updating models feat / nerFeature: Named Entity Recognizer -
You must be logged in to vote Inference speed between distilled and parent model is almost the same
perf / speedPerformance: speed feat / transformerFeature: Transformer -
You must be logged in to vote Using entity relation extraction to establish an entity hierarchy
feat / relFeature: Relation Extractor -
You must be logged in to vote CUDA out of memory while training Entity Linking
gpuUsing spaCy on GPU perf / memoryPerformance: memory use feat / nelFeature: Named Entity linking -
You must be logged in to vote Is Bloom Embedding also use for Chinese?
lang / zhChinese language data and models feat / vectorsFeature: Word vectors and similarity -
You must be logged in to vote Which hyperparameters are recommended for tuning on a custom SpaCy NER model?
feat / nerFeature: Named Entity Recognizer -
You must be logged in to vote Evaluation of German model on Universal Dependencies or Tiger
lang / deGerman language data and models -
You must be logged in to vote Not getting entity
feat / nerFeature: Named Entity Recognizer perf / accuracyPerformance: accuracy -
You must be logged in to vote How do I recommend improvements in spacy models NER extraction?
perf / accuracyPerformance: accuracy -
You must be logged in to vote What does it mean when the vector for a token is all 0's?
feat / vectorsFeature: Word vectors and similarity -
You must be logged in to vote Understanding the metrics displayed while training spacy components
trainingTraining and updating models -
You must be logged in to vote Additional training of pre-trained model does not work on BILUO
lang / ruRussian language data and models feat / nerFeature: Named Entity Recognizer -
You must be logged in to vote How to use a BiLSTM with a Transformer model?
feat / transformerFeature: Transformer -
You must be logged in to vote "The model you're using has no word vectors loaded"- Even though all my tokens do have a vector representation
feat / vectorsFeature: Word vectors and similarity -
You must be logged in to vote Precision on SpaCy pipeline and the possibility to use the same base models independantly
trainingTraining and updating models feat / taggerFeature: Part-of-speech tagger feat / parserFeature: Dependency Parser
