hwaking

[email protected]

EWELL Hangzhou, China machine learning; nlp

Results 8 comments of


                                            hwaking

Add deprecations for tests written for `nose` [SQUASH]

邮件已收到！

[tosa] Add legalization for TF/TFL conv3d

邮件已收到！

Support language which need tokenizer (Chinese, Japanese .etc)

@eromoe I am doing Chinese EMR information extraction ， can i use iepy to do entity relationship extraction ？

一句话中有多对实体关系的情况您怎么考虑呢

前后窗口指的是训练语句长度，您的这个模型长度设置的应该是70。

Add cosine_simirarity_v1 which is more faster than the original cosine_similarity method.

@vowelparrot

vllm hangs when reinitializing ray

I just tried using 32 cores and it solved my problem.

预训练细节（fim）

请问下预训练脚本以及数据样式能share 下么？@lightdf @pkuzqh

deepseek coder能够在base模型基础上继续与训练吗？

@guoday 能提供下预训练的脚步和数据样例么，谢谢