binmakeswell comments

Results 359 comments of


                                            binmakeswell

[BUG]: an instance of 'c10::CUDAErrorc10::CUDAError' initialization error

Hi @gouchangjiang Did you solve it? This issue was closed due to inactivity. Thanks.

[FEATURE]: Clear Instruction for Checkpoint

用户反馈：目前ColossalAI在如何保存（和加载）混合并行模式下（zero3+tensor+流水线）的优化器和lr调度器参数还没有一个完整的example，这个特性对宕机后从最后一个checkpoint重启训练很重要，可不可以麻烦工作人员补充一下demo。

[FEATURE]: Clear Instruction for Checkpoint

We have completed most of the related Checkpoint development and are doing the final polishing and refinement. Thanks.

[PROPOSAL]: Improve Automation for Better Development Experience

Thanks for the awsome work!

[BUG]: program stucks when using rpc_run and ColoInitContext together.

This issue was closed due to this is a wrong usage. Thanks.

[BUG]: Timeout

We have updated a lot. This issue was closed due to inactivity. Thanks.

[FEATURE]: Support for U-Net in pipeline parallel

We have updated a lot. This issue was closed due to inactivity. Thanks.

[BUG]: How can i save checkpoint trained by the way of GeminiDDP

We have updated a lot. This issue was closed due to inactivity. Thanks.

[BUG]: Dreambooth training example fails with PyTorch nightly build

We have updated a lot. This issue was closed due to inactivity. Thanks.

[BUG]: ERROR:torch.distributed.elastic.multiprocessing.api:failed (exitcode: -9) local_rank: 7 (pid: 2860777) of binary:

Hi @Tron1994 We have updated a lot. You can check our new example. https://github.com/hpcaitech/ColossalAI/tree/main/examples/language/gpt This issue was closed due to inactivity. Thanks.