We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
We need #9002 to guarantee this. Script from https://github.com/dzhwinter/benchmark
Pass = 1, Training performance = 240.092875 imgs/s, Train accuracy = 0.099440, Test accuracy = 0.100400 Pass = 2, Training performance = 240.027211 imgs/s, Train accuracy = 0.099540, Test accuracy = 0.100000 Pass = 3, Training performance = 240.151129 imgs/s, Train accuracy = 0.099580, Test accuracy = 0.100100 Pass = 4, Training performance = 240.201239 imgs/s, Train accuracy = 0.099760, Test accuracy = 0.094900 Pass = 5, Training performance = 240.154561 imgs/s, Train accuracy = 0.098660, Test accuracy = 0.100000 Pass = 6, Training performance = 240.091699 imgs/s, Train accuracy = 0.099620, Test accuracy = 0.100000 Pass = 7, Training performance = 240.322419 imgs/s, Train accuracy = 0.097080, Test accuracy = 0.100000 Pass = 8, Training performance = 239.932700 imgs/s, Train accuracy = 0.098200, Test accuracy = 0.100300 Pass = 9, Training performance = 239.198880 imgs/s, Train accuracy = 0.099780, Test accuracy = 0.099800 Pass = 10, Training performance = 240.219263 imgs/s, Train accuracy = 0.098220, Test accuracy = 0.100000
The text was updated successfully, but these errors were encountered:
Locate the problem now.
Sorry, something went wrong.
[Tokenizer] Add BertTokenizerFast, support register new tokenizer (Pa…
cd22b0d
…ddlePaddle#9353) * Add , support register tokenizer, fix some typo * add more tests * CustomTokenizerFast2->CustomTokenizerFastWithoutSlow * lint
Superjomn
reyoung
dzhwinter
gongweibao
No branches or pull requests
We need #9002 to guarantee this.
Script from https://github.com/dzhwinter/benchmark
The text was updated successfully, but these errors were encountered: