Antian Liang comments

Results 5 comments of


                                            Antian Liang

F external/org_tensorflow/tensorflow/compiler/xla/python/xla.cc:227] Check failed: stream_device != nullptr (0 vs. nullptr)

我感觉这个问题主要是因为devicemesh.py中有这么一行代码： ``` update_jax_platform("cpu") ``` 到这benchmark的sync函数调用 `jax.local_devices()[0]`,的时候不是用的gpu而是用的cpu

[Bug] Segment fault when using alpa to parallelize llama with jax 0.4.6 environment

A lot of changes have been made in version 0.4 and version 0.3 of jax

How to build debug-version Alpa-modified jaxlib

"""To build with debug information, add the flag --bazel_options='--copt=/Z7'.""" FROM https://jax.readthedocs.io/en/latest/developer.html#id1 不过这个是基于最新的openxla的回答了，不确定这个0.3.22的能不能用 btw，可以加个vx交流一下吗老哥

Problem in building Alpa-modified Jaxlib.

this bug caused by wrong version of libnccl i solved it by reinstalling a right ver libnccl and recreating a new python env based on this libnccl

Problem in building Alpa-modified Jaxlib.

yeah python == 3.8.13 gcc == 7.5.0 nccl == libnccl.so.2.8.4