bytedance/flux

Issues

• [QUESTION] Not supported on A6000?

  #46 opened 25 days ago by Zhuohao-Li
  3

• [QUESTION] Why is GemmRS result on hopper nondeterministic?

  #47 opened 19 days ago by umiswing
  0

• [BUG] Can't find nccl when building from source

  #28 opened 4 months ago by KnowingNothing
  5

• [QUESTION] Can Gemm_V3 be used in SM80?

  #38 opened a month ago by ginowu
  6

• [BUG] incorrect shape output from AGKernel.gather()

  #44 opened a month ago by 152334H
  0

• [QUESTION]some questions about allgather+gemm

  #42 opened a month ago by ChrisRanger
  1

• [QUESTION] Can flux run on RTX 4090?

  #43 opened a month ago by qinghon
  0

• [BUG] Failing to install byte-flux from pypi

  #30 opened 3 months ago by tlrmchlsmth
  8

• [QUESTION] How does flux handle hardware resoureces competition?

  #39 opened 3 months ago by chenhongyu2048
  2

• [QUESTION] Why flux gemm_rs is not faster than torch?

  #34 opened 3 months ago by hxdtest
  5

• [BUG] Illegal memory with multi-node

  #40 opened 2 months ago by YJHMITWEB
  1

• [QUESTION] How to use nvshmem?

  #33 opened 3 months ago by chenhongyu2048
  8

• [QUESTION]is there a plan to support int8?

  #31 opened 3 months ago by Rainlin007
  1

• Are there any difficulties in implementing gemm-allreduce?

  #20 opened 4 months ago by Rainlin007
  2

• [QUESTION] Are you planning on supporting FP8?

  #27 opened 3 months ago by MustafaFayez
  3

• [BUG] `no_nvlink` branch failed to compile

  #32 opened 3 months ago by lucifer1004
  6

• [ENHANCEMENT] support for gpu A40

  #35 opened 3 months ago by 1926627357
  3

• [QUESTION] The gemm time on GPU of different rank under tp8 is very different , and cause low performance

  #36 opened 3 months ago by Rainlin007
  8

• [QUESTION] Why is ring mode fixed to `All2All` in `src/all_gather/ths_op/all_gather_types.h`?

  #37 opened 3 months ago by lucifer1004
  2

• [BUG] Illegal memory access when fuse_reduction=False

  #10 opened 3 months ago by tlrmchlsmth
  5

• [QUESTION]How to run examples in pynvshmem

  #29 opened 3 months ago by TonyWu199
  1

• [BUG] Incorrect results from flux.AGKernel for some problem shapes

  #17 opened 4 months ago by tlrmchlsmth
  15

• [BUG] Exception: not supported device NVIDIA H100 80GB HBM3

  #14 opened 4 months ago by wenscarl
  2

• [BUG] RuntimeError: Could not retrieve or create the backend 2 for device type cuda

  #11 opened 5 months ago by tlrmchlsmth
  14

• [BUG] gemm and reduce-scatter are not overlapped

  #7 opened 5 months ago by wenscarl
  9

• [BUG] Illegal memory access in GemmRS when passing fuse_reduction=True and dtype=bfloat16

  #8 opened 5 months ago by tlrmchlsmth
  4