buckhx / gobert Goto Github PK

View Code? Open in Web Editor NEW

99.0 99.0 18.0 108 KB

Go Bindings for BERT NLP Models

License: MIT License

Go 73.77% Makefile 3.24% Python 22.43% Dockerfile 0.51% Shell 0.05%

bert embeddings go golang machine-learning natural-language-processing natural-language-understanding nlp tensorflow

gobert's People

Contributors

Stargazers

Watchers

Forkers

wyh454412805 dhlee347 go-dockly aicelerity-py kabychow sunhailin-leo fulquan wujiadw404 nazandr ruoxuwang ihauk brunowang wuyanxin dahvlhldle hzylyq cocainecong

gobert's Issues

loadModel

Hello,

Thanks for the repo.

I successfully export the model and try to run the "embedding" function or "raw-model" function in the Example folder. However, it returned error message: " Could not find SavedModel .pb or .pbtxt at supplied export directory". Actually, the .pb file is there, and vocab.txt can be found.

Could you give me any hints to solve this issue, please? Many thanks,

`make model` not working

running make model has this error:

make: *** No rule to make target 'export_image', needed by 'model'.  Stop.

Is export_image target missing? (cannot see it in Makefile)

cause oom

Full.Tokenize may cause OOM(out of memery), such as: text="suvZR∨"

finBERT

May I ask is this module still maintained? Just wondering if there's a way to implement finBERT model? Tnx

tokenizer goes into deadlock when some special cases

The tokenizer goes into deadlock when some special cases, and the error occurs when a wordpiece cannot be found in vocabulary and the '#' key in the vocabulary.

Recommend Projects