chokkan / crfsuite

CRFsuite: a fast implementation of Conditional Random Fields (CRFs)
http://www.chokkan.org/software/crfsuite/
Other
648 stars 208 forks source link

how to continue training of existing crf model #93

Open yanghanxy opened 7 years ago

yanghanxy commented 7 years ago

Is there a way to continue training a trained crf model? For example, I trained a crf model using training data Corpus_A, and save the model as file crf.model. Some days later, I get some extra training data as Corpus_B. Can I just load the trained crf.model and continue training with Corpus_B?

guybas commented 5 years ago

yanghanxy -- I have the same question as yours above. Did you find an answer by any chance?

deepindeed2022 commented 2 years ago

I have the same question, and i try add model_filename as input, but it don't work. Anyone have solution? crf = sklearn_crfsuite.CRF( algorithm='lbfgs', c1=0.1, c2=0.1, max_iterations=MAX_ITER_TIME, all_possible_transitions=True, verbose=verbose, model_filename=init_model )