zhanghaoie
zhanghaoie
> Hi, 请问你使用light seq对unilm加速了吗？效果怎么样啊？
sh tools/build_and_run_unittests.sh $PWD -DWITH_GPU=ON 按照readme运行的这个指令，难道这个不是支持gpu的吗？
按照gpu docker的方式编译的，应该没什么问题的。。。
我训练了一个中文模型， 貌似nsp模型更倾向于给重复率较高的句子高分
thank you very much for your reply! Is it possible to convert the pretrained model saved with fluid.io.save_params to .pdparams format?
ok, thank you very much for your reply
Hi，is it possible to load .pdparams model for fine-tuning with develop branch?
yeah, i have found paddlenlp releases a plato-mini model and want to use it for fine-tuning.
> OK, I will update some convert tool later. Hi, how about the convert tool.
> Hi @jbkyang-nvi, ok I will try and let you know as soon as possible. Thanks for the support! Hi, @GabrieleFerrario I have run into the same issue. Could you...