Sichen Zhao

Results 4 issues of Sichen Zhao

Now Deepspeed do not support CosineAnnealingLR scheduler. So i want to support it by meself. question is how to develop a custom scheduler, Are there any tutorials available?

enhancement

# Bug Report webui docker images do not support relative path. ## Description for xample, i want to start webui at localhost:8080/webui/, does the image parameter support the relative path...

enhancement
good first issue
help wanted
non-core

Hi, is there a plan to implement the AI driven flame diagram interpreter feature in Grafana Pyroscope in the open source version? link: https://pyroscope.io/blog/ai-powered-flamegraph-interpreter/

hi, I have trained a LLM model with 4 nodes (8 gpus per node), but when I load the checkpoint with 16 nodes, I get the follows error: `deepspeed.runtime.zero.utils.ZeRORuntimeException: The...

enhancement