test eatcosmos
  • Joined on Jun 09, 2022
Loading Heatmap…

eatcosmos pushed to master at eatcosmos/gpt2-wikitext-2

1 month ago

eatcosmos pushed to master at eatcosmos/gpt2-wikitext-2

1 month ago

eatcosmos pushed to master at eatcosmos/gpt2-wikitext-2

1 month ago

eatcosmos pushed to master at eatcosmos/gpt2-wikitext-2

1 month ago

eatcosmos created repository eatcosmos/gpt2-wikitext-2

1 month ago

eatcosmos created repository eatcosmos/Infinity

1 month ago

eatcosmos pushed to master at eatcosmos/CUDA_Programming

2 months ago

eatcosmos created repository eatcosmos/CUDA_Programming

2 months ago

eatcosmos opened issue OpenIOSSG/OpenI_Cloudbrain_Example#2

c2net_context = prepare() 这行报错

3 months ago

eatcosmos opened issue zeizei/OpenI_Learning#1207

国产训练全功能GPU能不能上一些,比如 沐曦 MXC500

4 months ago

eatcosmos opened issue zeizei/OpenI_Learning#1034

[建议] 超大数据集导入方案

1 year ago

eatcosmos commented on issue OpenI/MSAdapter#449

为什么不直接用pytorch,而要迁移一遍呢?

> 可以从两个方面思考:生态上,助力国产AI框架的生态建设;性能上,PyTorch并不能在Ascend上获得加速,MindSpore可以结合Ascend发挥最佳性能。 就是主要迁移的是接口名称,里面实现都是mindspore的?

1 year ago

eatcosmos reopened issue OpenI/MSAdapter#450

能开发一个浏览器插件吗?功能是直接转换pytorch官网的案例到mindspore吗?

1 year ago

eatcosmos closed issue OpenI/MSAdapter#450

能开发一个浏览器插件吗?功能是直接转换pytorch官网的案例到mindspore吗?

1 year ago

eatcosmos opened issue OpenI/MSAdapter#450

通开发一个插件,直接转换pytorch官网的案例到mindspore吗?

1 year ago

eatcosmos opened issue OpenI/MSAdapter#449

为什么不直接用pytorch,而要迁移一遍呢?

1 year ago

eatcosmos commented on issue Datawhale/d2l#3

本次组队学习活动或教程的相关建议

内容增强建议 AIGC和大模型是重要方向,希望可以增加一个关系图,阐释d2l各个章节和AIGC、大模型的关系 这样就方便用学习的知识可以过度到AIGC、大模型的学习中,可以吸引更多人来学习

1 year ago

eatcosmos commented on issue Datawhale/d2l#2

【《动手学深度学习》组队学习】学习笔记

jupyter lab切换kernel为conda虚拟环境d2l https://openi.pcl.ac.cn/eatcosmos/d2l/src/branch/master/venv_d2l.sh

1 year ago

eatcosmos opened issue zeizei/OpenI_Learning#440

【功能建议】调试/训练 任务支持动态加载数据集

1 year ago

eatcosmos opened issue zeizei/OpenI_Learning#276

【数据集】建议放宽数据集的上传,考虑百度云之类的云盘转存和全局数据集

1 year ago